diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,38374 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9998174182946868,
+  "eval_steps": 500,
+  "global_step": 5476,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003651634106262553,
+      "grad_norm": 858.7462158203125,
+      "learning_rate": 3.6496350364963505e-08,
+      "loss": 4.4937,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007303268212525105,
+      "grad_norm": 6394.86376953125,
+      "learning_rate": 7.299270072992701e-08,
+      "loss": 5.9233,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010954902318787657,
+      "grad_norm": 3305.432861328125,
+      "learning_rate": 1.0948905109489053e-07,
+      "loss": 5.4385,
+      "step": 3
+    },
+    {
+      "epoch": 0.001460653642505021,
+      "grad_norm": 1724.0235595703125,
+      "learning_rate": 1.4598540145985402e-07,
+      "loss": 4.5796,
+      "step": 4
+    },
+    {
+      "epoch": 0.0018258170531312763,
+      "grad_norm": 18848.08203125,
+      "learning_rate": 1.8248175182481753e-07,
+      "loss": 7.1191,
+      "step": 5
+    },
+    {
+      "epoch": 0.0021909804637575314,
+      "grad_norm": 5539.15234375,
+      "learning_rate": 2.1897810218978106e-07,
+      "loss": 6.7446,
+      "step": 6
+    },
+    {
+      "epoch": 0.0025561438743837866,
+      "grad_norm": 2666.61962890625,
+      "learning_rate": 2.5547445255474454e-07,
+      "loss": 8.3125,
+      "step": 7
+    },
+    {
+      "epoch": 0.002921307285010042,
+      "grad_norm": 3787.522705078125,
+      "learning_rate": 2.9197080291970804e-07,
+      "loss": 5.8618,
+      "step": 8
+    },
+    {
+      "epoch": 0.0032864706956362974,
+      "grad_norm": 72566.375,
+      "learning_rate": 3.284671532846716e-07,
+      "loss": 6.7295,
+      "step": 9
+    },
+    {
+      "epoch": 0.0036516341062625525,
+      "grad_norm": 30215.857421875,
+      "learning_rate": 3.6496350364963505e-07,
+      "loss": 7.3945,
+      "step": 10
+    },
+    {
+      "epoch": 0.004016797516888808,
+      "grad_norm": 26900.92578125,
+      "learning_rate": 4.0145985401459856e-07,
+      "loss": 7.6553,
+      "step": 11
+    },
+    {
+      "epoch": 0.004381960927515063,
+      "grad_norm": 3489.6513671875,
+      "learning_rate": 4.379562043795621e-07,
+      "loss": 8.3828,
+      "step": 12
+    },
+    {
+      "epoch": 0.004747124338141318,
+      "grad_norm": 808.1627197265625,
+      "learning_rate": 4.7445255474452557e-07,
+      "loss": 5.2695,
+      "step": 13
+    },
+    {
+      "epoch": 0.005112287748767573,
+      "grad_norm": 5560.9794921875,
+      "learning_rate": 5.109489051094891e-07,
+      "loss": 4.6694,
+      "step": 14
+    },
+    {
+      "epoch": 0.005477451159393828,
+      "grad_norm": 3059.473388671875,
+      "learning_rate": 5.474452554744526e-07,
+      "loss": 6.4053,
+      "step": 15
+    },
+    {
+      "epoch": 0.005842614570020084,
+      "grad_norm": 44171.578125,
+      "learning_rate": 5.839416058394161e-07,
+      "loss": 6.0503,
+      "step": 16
+    },
+    {
+      "epoch": 0.0062077779806463396,
+      "grad_norm": 6308.86474609375,
+      "learning_rate": 6.204379562043796e-07,
+      "loss": 5.3037,
+      "step": 17
+    },
+    {
+      "epoch": 0.006572941391272595,
+      "grad_norm": 1026.9315185546875,
+      "learning_rate": 6.569343065693432e-07,
+      "loss": 6.7432,
+      "step": 18
+    },
+    {
+      "epoch": 0.00693810480189885,
+      "grad_norm": 7379.29296875,
+      "learning_rate": 6.934306569343066e-07,
+      "loss": 5.8125,
+      "step": 19
+    },
+    {
+      "epoch": 0.007303268212525105,
+      "grad_norm": 1578.2115478515625,
+      "learning_rate": 7.299270072992701e-07,
+      "loss": 7.7974,
+      "step": 20
+    },
+    {
+      "epoch": 0.00766843162315136,
+      "grad_norm": 1759.99365234375,
+      "learning_rate": 7.664233576642337e-07,
+      "loss": 5.6411,
+      "step": 21
+    },
+    {
+      "epoch": 0.008033595033777615,
+      "grad_norm": 6159.49169921875,
+      "learning_rate": 8.029197080291971e-07,
+      "loss": 6.9321,
+      "step": 22
+    },
+    {
+      "epoch": 0.00839875844440387,
+      "grad_norm": 2101.160888671875,
+      "learning_rate": 8.394160583941606e-07,
+      "loss": 6.7178,
+      "step": 23
+    },
+    {
+      "epoch": 0.008763921855030126,
+      "grad_norm": 31658.865234375,
+      "learning_rate": 8.759124087591242e-07,
+      "loss": 5.3345,
+      "step": 24
+    },
+    {
+      "epoch": 0.009129085265656381,
+      "grad_norm": 7087.1318359375,
+      "learning_rate": 9.124087591240876e-07,
+      "loss": 6.7251,
+      "step": 25
+    },
+    {
+      "epoch": 0.009494248676282636,
+      "grad_norm": 18886.189453125,
+      "learning_rate": 9.489051094890511e-07,
+      "loss": 6.3369,
+      "step": 26
+    },
+    {
+      "epoch": 0.009859412086908891,
+      "grad_norm": 5172.552734375,
+      "learning_rate": 9.854014598540146e-07,
+      "loss": 6.625,
+      "step": 27
+    },
+    {
+      "epoch": 0.010224575497535146,
+      "grad_norm": 3320.644775390625,
+      "learning_rate": 1.0218978102189781e-06,
+      "loss": 6.4692,
+      "step": 28
+    },
+    {
+      "epoch": 0.010589738908161402,
+      "grad_norm": 5261.62646484375,
+      "learning_rate": 1.0583941605839416e-06,
+      "loss": 7.0078,
+      "step": 29
+    },
+    {
+      "epoch": 0.010954902318787657,
+      "grad_norm": 14572.537109375,
+      "learning_rate": 1.0948905109489052e-06,
+      "loss": 6.2578,
+      "step": 30
+    },
+    {
+      "epoch": 0.011320065729413912,
+      "grad_norm": 2529.656982421875,
+      "learning_rate": 1.1313868613138687e-06,
+      "loss": 6.7266,
+      "step": 31
+    },
+    {
+      "epoch": 0.011685229140040169,
+      "grad_norm": 6337.70458984375,
+      "learning_rate": 1.1678832116788322e-06,
+      "loss": 5.8765,
+      "step": 32
+    },
+    {
+      "epoch": 0.012050392550666424,
+      "grad_norm": 1270.5396728515625,
+      "learning_rate": 1.2043795620437959e-06,
+      "loss": 6.3232,
+      "step": 33
+    },
+    {
+      "epoch": 0.012415555961292679,
+      "grad_norm": 2410.99853515625,
+      "learning_rate": 1.2408759124087592e-06,
+      "loss": 6.2515,
+      "step": 34
+    },
+    {
+      "epoch": 0.012780719371918934,
+      "grad_norm": 1112.746826171875,
+      "learning_rate": 1.2773722627737229e-06,
+      "loss": 7.2603,
+      "step": 35
+    },
+    {
+      "epoch": 0.01314588278254519,
+      "grad_norm": 2323.461669921875,
+      "learning_rate": 1.3138686131386864e-06,
+      "loss": 6.1316,
+      "step": 36
+    },
+    {
+      "epoch": 0.013511046193171445,
+      "grad_norm": 6492.537109375,
+      "learning_rate": 1.3503649635036497e-06,
+      "loss": 5.7026,
+      "step": 37
+    },
+    {
+      "epoch": 0.0138762096037977,
+      "grad_norm": 1916.9757080078125,
+      "learning_rate": 1.3868613138686132e-06,
+      "loss": 6.5596,
+      "step": 38
+    },
+    {
+      "epoch": 0.014241373014423955,
+      "grad_norm": 980.5525512695312,
+      "learning_rate": 1.4233576642335767e-06,
+      "loss": 4.6865,
+      "step": 39
+    },
+    {
+      "epoch": 0.01460653642505021,
+      "grad_norm": 3851.416259765625,
+      "learning_rate": 1.4598540145985402e-06,
+      "loss": 5.9668,
+      "step": 40
+    },
+    {
+      "epoch": 0.014971699835676465,
+      "grad_norm": 1468.562744140625,
+      "learning_rate": 1.496350364963504e-06,
+      "loss": 6.6636,
+      "step": 41
+    },
+    {
+      "epoch": 0.01533686324630272,
+      "grad_norm": 2646.891357421875,
+      "learning_rate": 1.5328467153284674e-06,
+      "loss": 5.6963,
+      "step": 42
+    },
+    {
+      "epoch": 0.015702026656928977,
+      "grad_norm": 10316.486328125,
+      "learning_rate": 1.5693430656934307e-06,
+      "loss": 6.1025,
+      "step": 43
+    },
+    {
+      "epoch": 0.01606719006755523,
+      "grad_norm": 1718.9337158203125,
+      "learning_rate": 1.6058394160583942e-06,
+      "loss": 6.4917,
+      "step": 44
+    },
+    {
+      "epoch": 0.016432353478181488,
+      "grad_norm": 1402.13623046875,
+      "learning_rate": 1.6423357664233577e-06,
+      "loss": 4.6167,
+      "step": 45
+    },
+    {
+      "epoch": 0.01679751688880774,
+      "grad_norm": 514.71923828125,
+      "learning_rate": 1.6788321167883212e-06,
+      "loss": 5.6113,
+      "step": 46
+    },
+    {
+      "epoch": 0.017162680299433998,
+      "grad_norm": 7258.53076171875,
+      "learning_rate": 1.715328467153285e-06,
+      "loss": 5.1318,
+      "step": 47
+    },
+    {
+      "epoch": 0.01752784371006025,
+      "grad_norm": 53655.25,
+      "learning_rate": 1.7518248175182485e-06,
+      "loss": 6.5986,
+      "step": 48
+    },
+    {
+      "epoch": 0.01789300712068651,
+      "grad_norm": 6116.984375,
+      "learning_rate": 1.788321167883212e-06,
+      "loss": 5.8286,
+      "step": 49
+    },
+    {
+      "epoch": 0.018258170531312762,
+      "grad_norm": 1429.343017578125,
+      "learning_rate": 1.8248175182481753e-06,
+      "loss": 6.2109,
+      "step": 50
+    },
+    {
+      "epoch": 0.01862333394193902,
+      "grad_norm": 3730.193115234375,
+      "learning_rate": 1.8613138686131388e-06,
+      "loss": 6.8057,
+      "step": 51
+    },
+    {
+      "epoch": 0.018988497352565272,
+      "grad_norm": 42216.3984375,
+      "learning_rate": 1.8978102189781023e-06,
+      "loss": 4.1436,
+      "step": 52
+    },
+    {
+      "epoch": 0.01935366076319153,
+      "grad_norm": 3568.0703125,
+      "learning_rate": 1.934306569343066e-06,
+      "loss": 6.5127,
+      "step": 53
+    },
+    {
+      "epoch": 0.019718824173817782,
+      "grad_norm": 242.3411407470703,
+      "learning_rate": 1.9708029197080293e-06,
+      "loss": 4.0176,
+      "step": 54
+    },
+    {
+      "epoch": 0.02008398758444404,
+      "grad_norm": 7677.52392578125,
+      "learning_rate": 2.007299270072993e-06,
+      "loss": 6.2373,
+      "step": 55
+    },
+    {
+      "epoch": 0.020449150995070293,
+      "grad_norm": 2871.701904296875,
+      "learning_rate": 2.0437956204379563e-06,
+      "loss": 5.6187,
+      "step": 56
+    },
+    {
+      "epoch": 0.02081431440569655,
+      "grad_norm": 1617.5478515625,
+      "learning_rate": 2.08029197080292e-06,
+      "loss": 6.1074,
+      "step": 57
+    },
+    {
+      "epoch": 0.021179477816322803,
+      "grad_norm": 2647.792724609375,
+      "learning_rate": 2.1167883211678833e-06,
+      "loss": 5.9727,
+      "step": 58
+    },
+    {
+      "epoch": 0.02154464122694906,
+      "grad_norm": 2567.927490234375,
+      "learning_rate": 2.1532846715328466e-06,
+      "loss": 6.4531,
+      "step": 59
+    },
+    {
+      "epoch": 0.021909804637575313,
+      "grad_norm": 549.2553100585938,
+      "learning_rate": 2.1897810218978103e-06,
+      "loss": 5.0977,
+      "step": 60
+    },
+    {
+      "epoch": 0.02227496804820157,
+      "grad_norm": 7288.61474609375,
+      "learning_rate": 2.226277372262774e-06,
+      "loss": 6.0161,
+      "step": 61
+    },
+    {
+      "epoch": 0.022640131458827824,
+      "grad_norm": 18750.427734375,
+      "learning_rate": 2.2627737226277373e-06,
+      "loss": 5.6802,
+      "step": 62
+    },
+    {
+      "epoch": 0.02300529486945408,
+      "grad_norm": 1172.6300048828125,
+      "learning_rate": 2.299270072992701e-06,
+      "loss": 6.1104,
+      "step": 63
+    },
+    {
+      "epoch": 0.023370458280080338,
+      "grad_norm": 2294.61328125,
+      "learning_rate": 2.3357664233576643e-06,
+      "loss": 5.3442,
+      "step": 64
+    },
+    {
+      "epoch": 0.02373562169070659,
+      "grad_norm": 10097.84375,
+      "learning_rate": 2.3722627737226276e-06,
+      "loss": 5.3574,
+      "step": 65
+    },
+    {
+      "epoch": 0.024100785101332848,
+      "grad_norm": 6148.13916015625,
+      "learning_rate": 2.4087591240875918e-06,
+      "loss": 5.4424,
+      "step": 66
+    },
+    {
+      "epoch": 0.0244659485119591,
+      "grad_norm": 1161.4757080078125,
+      "learning_rate": 2.445255474452555e-06,
+      "loss": 5.6011,
+      "step": 67
+    },
+    {
+      "epoch": 0.024831111922585358,
+      "grad_norm": 2529.951904296875,
+      "learning_rate": 2.4817518248175183e-06,
+      "loss": 5.1353,
+      "step": 68
+    },
+    {
+      "epoch": 0.02519627533321161,
+      "grad_norm": 5371.56982421875,
+      "learning_rate": 2.518248175182482e-06,
+      "loss": 3.9297,
+      "step": 69
+    },
+    {
+      "epoch": 0.02556143874383787,
+      "grad_norm": 3334.050537109375,
+      "learning_rate": 2.5547445255474458e-06,
+      "loss": 4.1289,
+      "step": 70
+    },
+    {
+      "epoch": 0.025926602154464122,
+      "grad_norm": 8525.9794921875,
+      "learning_rate": 2.591240875912409e-06,
+      "loss": 5.3735,
+      "step": 71
+    },
+    {
+      "epoch": 0.02629176556509038,
+      "grad_norm": 4757.6689453125,
+      "learning_rate": 2.627737226277373e-06,
+      "loss": 5.7793,
+      "step": 72
+    },
+    {
+      "epoch": 0.026656928975716632,
+      "grad_norm": 6516.1943359375,
+      "learning_rate": 2.664233576642336e-06,
+      "loss": 5.9304,
+      "step": 73
+    },
+    {
+      "epoch": 0.02702209238634289,
+      "grad_norm": 7296.853515625,
+      "learning_rate": 2.7007299270072994e-06,
+      "loss": 4.9775,
+      "step": 74
+    },
+    {
+      "epoch": 0.027387255796969143,
+      "grad_norm": 816.5313720703125,
+      "learning_rate": 2.737226277372263e-06,
+      "loss": 4.5146,
+      "step": 75
+    },
+    {
+      "epoch": 0.0277524192075954,
+      "grad_norm": 1695.573974609375,
+      "learning_rate": 2.7737226277372264e-06,
+      "loss": 5.3511,
+      "step": 76
+    },
+    {
+      "epoch": 0.028117582618221653,
+      "grad_norm": 1111.10009765625,
+      "learning_rate": 2.81021897810219e-06,
+      "loss": 4.2915,
+      "step": 77
+    },
+    {
+      "epoch": 0.02848274602884791,
+      "grad_norm": 6980.4052734375,
+      "learning_rate": 2.8467153284671534e-06,
+      "loss": 3.7622,
+      "step": 78
+    },
+    {
+      "epoch": 0.028847909439474163,
+      "grad_norm": 2465.1806640625,
+      "learning_rate": 2.8832116788321167e-06,
+      "loss": 6.0605,
+      "step": 79
+    },
+    {
+      "epoch": 0.02921307285010042,
+      "grad_norm": 2196.582763671875,
+      "learning_rate": 2.9197080291970804e-06,
+      "loss": 5.7114,
+      "step": 80
+    },
+    {
+      "epoch": 0.029578236260726674,
+      "grad_norm": 893.1383666992188,
+      "learning_rate": 2.956204379562044e-06,
+      "loss": 5.2002,
+      "step": 81
+    },
+    {
+      "epoch": 0.02994339967135293,
+      "grad_norm": 3686.89306640625,
+      "learning_rate": 2.992700729927008e-06,
+      "loss": 5.29,
+      "step": 82
+    },
+    {
+      "epoch": 0.030308563081979184,
+      "grad_norm": 9257.904296875,
+      "learning_rate": 3.029197080291971e-06,
+      "loss": 4.4707,
+      "step": 83
+    },
+    {
+      "epoch": 0.03067372649260544,
+      "grad_norm": 4997.966796875,
+      "learning_rate": 3.065693430656935e-06,
+      "loss": 5.4282,
+      "step": 84
+    },
+    {
+      "epoch": 0.031038889903231698,
+      "grad_norm": 2574.991943359375,
+      "learning_rate": 3.102189781021898e-06,
+      "loss": 4.5264,
+      "step": 85
+    },
+    {
+      "epoch": 0.031404053313857955,
+      "grad_norm": 1940.85546875,
+      "learning_rate": 3.1386861313868614e-06,
+      "loss": 4.8984,
+      "step": 86
+    },
+    {
+      "epoch": 0.031769216724484205,
+      "grad_norm": 1634.5382080078125,
+      "learning_rate": 3.175182481751825e-06,
+      "loss": 4.0796,
+      "step": 87
+    },
+    {
+      "epoch": 0.03213438013511046,
+      "grad_norm": 4987.83154296875,
+      "learning_rate": 3.2116788321167884e-06,
+      "loss": 4.8687,
+      "step": 88
+    },
+    {
+      "epoch": 0.03249954354573672,
+      "grad_norm": 32678.6171875,
+      "learning_rate": 3.248175182481752e-06,
+      "loss": 5.0845,
+      "step": 89
+    },
+    {
+      "epoch": 0.032864706956362975,
+      "grad_norm": 941.7997436523438,
+      "learning_rate": 3.2846715328467155e-06,
+      "loss": 4.0479,
+      "step": 90
+    },
+    {
+      "epoch": 0.033229870366989225,
+      "grad_norm": 6600.39501953125,
+      "learning_rate": 3.3211678832116788e-06,
+      "loss": 4.0234,
+      "step": 91
+    },
+    {
+      "epoch": 0.03359503377761548,
+      "grad_norm": 195.313232421875,
+      "learning_rate": 3.3576642335766425e-06,
+      "loss": 4.543,
+      "step": 92
+    },
+    {
+      "epoch": 0.03396019718824174,
+      "grad_norm": 2158.8779296875,
+      "learning_rate": 3.3941605839416058e-06,
+      "loss": 4.282,
+      "step": 93
+    },
+    {
+      "epoch": 0.034325360598867996,
+      "grad_norm": 2006.3287353515625,
+      "learning_rate": 3.43065693430657e-06,
+      "loss": 4.5869,
+      "step": 94
+    },
+    {
+      "epoch": 0.034690524009494246,
+      "grad_norm": 5730.62646484375,
+      "learning_rate": 3.467153284671533e-06,
+      "loss": 4.3423,
+      "step": 95
+    },
+    {
+      "epoch": 0.0350556874201205,
+      "grad_norm": 6067.22216796875,
+      "learning_rate": 3.503649635036497e-06,
+      "loss": 4.561,
+      "step": 96
+    },
+    {
+      "epoch": 0.03542085083074676,
+      "grad_norm": 7373.00537109375,
+      "learning_rate": 3.54014598540146e-06,
+      "loss": 4.3584,
+      "step": 97
+    },
+    {
+      "epoch": 0.03578601424137302,
+      "grad_norm": 1633.809814453125,
+      "learning_rate": 3.576642335766424e-06,
+      "loss": 4.9429,
+      "step": 98
+    },
+    {
+      "epoch": 0.03615117765199927,
+      "grad_norm": 1958.9189453125,
+      "learning_rate": 3.6131386861313872e-06,
+      "loss": 3.9888,
+      "step": 99
+    },
+    {
+      "epoch": 0.036516341062625524,
+      "grad_norm": 587.3916625976562,
+      "learning_rate": 3.6496350364963505e-06,
+      "loss": 3.7556,
+      "step": 100
+    },
+    {
+      "epoch": 0.03688150447325178,
+      "grad_norm": 1480.251220703125,
+      "learning_rate": 3.6861313868613142e-06,
+      "loss": 4.6995,
+      "step": 101
+    },
+    {
+      "epoch": 0.03724666788387804,
+      "grad_norm": 339.09271240234375,
+      "learning_rate": 3.7226277372262775e-06,
+      "loss": 4.5959,
+      "step": 102
+    },
+    {
+      "epoch": 0.03761183129450429,
+      "grad_norm": 370.29736328125,
+      "learning_rate": 3.7591240875912412e-06,
+      "loss": 3.8401,
+      "step": 103
+    },
+    {
+      "epoch": 0.037976994705130544,
+      "grad_norm": 2126.548583984375,
+      "learning_rate": 3.7956204379562045e-06,
+      "loss": 4.8696,
+      "step": 104
+    },
+    {
+      "epoch": 0.0383421581157568,
+      "grad_norm": 357.5569763183594,
+      "learning_rate": 3.832116788321168e-06,
+      "loss": 4.5408,
+      "step": 105
+    },
+    {
+      "epoch": 0.03870732152638306,
+      "grad_norm": 375.277099609375,
+      "learning_rate": 3.868613138686132e-06,
+      "loss": 4.5688,
+      "step": 106
+    },
+    {
+      "epoch": 0.039072484937009315,
+      "grad_norm": 387.4185485839844,
+      "learning_rate": 3.905109489051096e-06,
+      "loss": 3.8911,
+      "step": 107
+    },
+    {
+      "epoch": 0.039437648347635565,
+      "grad_norm": 732.7127075195312,
+      "learning_rate": 3.9416058394160585e-06,
+      "loss": 4.6123,
+      "step": 108
+    },
+    {
+      "epoch": 0.03980281175826182,
+      "grad_norm": 3045.466064453125,
+      "learning_rate": 3.978102189781022e-06,
+      "loss": 3.9941,
+      "step": 109
+    },
+    {
+      "epoch": 0.04016797516888808,
+      "grad_norm": 6941.19677734375,
+      "learning_rate": 4.014598540145986e-06,
+      "loss": 3.981,
+      "step": 110
+    },
+    {
+      "epoch": 0.040533138579514336,
+      "grad_norm": 1226.5081787109375,
+      "learning_rate": 4.05109489051095e-06,
+      "loss": 3.7681,
+      "step": 111
+    },
+    {
+      "epoch": 0.040898301990140586,
+      "grad_norm": 4508.16748046875,
+      "learning_rate": 4.0875912408759126e-06,
+      "loss": 3.689,
+      "step": 112
+    },
+    {
+      "epoch": 0.04126346540076684,
+      "grad_norm": 8658.3134765625,
+      "learning_rate": 4.124087591240876e-06,
+      "loss": 3.3323,
+      "step": 113
+    },
+    {
+      "epoch": 0.0416286288113931,
+      "grad_norm": 732.7115478515625,
+      "learning_rate": 4.16058394160584e-06,
+      "loss": 3.7373,
+      "step": 114
+    },
+    {
+      "epoch": 0.041993792222019356,
+      "grad_norm": 3595.828369140625,
+      "learning_rate": 4.197080291970803e-06,
+      "loss": 3.7979,
+      "step": 115
+    },
+    {
+      "epoch": 0.042358955632645606,
+      "grad_norm": 558.604736328125,
+      "learning_rate": 4.233576642335767e-06,
+      "loss": 3.6646,
+      "step": 116
+    },
+    {
+      "epoch": 0.04272411904327186,
+      "grad_norm": 1674.7657470703125,
+      "learning_rate": 4.27007299270073e-06,
+      "loss": 4.2749,
+      "step": 117
+    },
+    {
+      "epoch": 0.04308928245389812,
+      "grad_norm": 1014.714599609375,
+      "learning_rate": 4.306569343065693e-06,
+      "loss": 3.411,
+      "step": 118
+    },
+    {
+      "epoch": 0.04345444586452438,
+      "grad_norm": 2597.962158203125,
+      "learning_rate": 4.343065693430658e-06,
+      "loss": 3.8379,
+      "step": 119
+    },
+    {
+      "epoch": 0.04381960927515063,
+      "grad_norm": 3630.7177734375,
+      "learning_rate": 4.379562043795621e-06,
+      "loss": 4.8774,
+      "step": 120
+    },
+    {
+      "epoch": 0.044184772685776884,
+      "grad_norm": 175.34295654296875,
+      "learning_rate": 4.416058394160584e-06,
+      "loss": 3.1763,
+      "step": 121
+    },
+    {
+      "epoch": 0.04454993609640314,
+      "grad_norm": 3908.3720703125,
+      "learning_rate": 4.452554744525548e-06,
+      "loss": 3.3569,
+      "step": 122
+    },
+    {
+      "epoch": 0.0449150995070294,
+      "grad_norm": 4444.5869140625,
+      "learning_rate": 4.489051094890512e-06,
+      "loss": 3.8149,
+      "step": 123
+    },
+    {
+      "epoch": 0.04528026291765565,
+      "grad_norm": 207.31356811523438,
+      "learning_rate": 4.525547445255475e-06,
+      "loss": 2.6179,
+      "step": 124
+    },
+    {
+      "epoch": 0.045645426328281904,
+      "grad_norm": 516.9945068359375,
+      "learning_rate": 4.562043795620438e-06,
+      "loss": 4.0015,
+      "step": 125
+    },
+    {
+      "epoch": 0.04601058973890816,
+      "grad_norm": 4471.39404296875,
+      "learning_rate": 4.598540145985402e-06,
+      "loss": 4.3091,
+      "step": 126
+    },
+    {
+      "epoch": 0.04637575314953442,
+      "grad_norm": 10257.8251953125,
+      "learning_rate": 4.635036496350365e-06,
+      "loss": 3.6528,
+      "step": 127
+    },
+    {
+      "epoch": 0.046740916560160675,
+      "grad_norm": 117.92402648925781,
+      "learning_rate": 4.671532846715329e-06,
+      "loss": 3.4553,
+      "step": 128
+    },
+    {
+      "epoch": 0.047106079970786925,
+      "grad_norm": 3125.854248046875,
+      "learning_rate": 4.708029197080292e-06,
+      "loss": 2.8669,
+      "step": 129
+    },
+    {
+      "epoch": 0.04747124338141318,
+      "grad_norm": 199.3535614013672,
+      "learning_rate": 4.744525547445255e-06,
+      "loss": 2.9316,
+      "step": 130
+    },
+    {
+      "epoch": 0.04783640679203944,
+      "grad_norm": 3418.947265625,
+      "learning_rate": 4.78102189781022e-06,
+      "loss": 4.1367,
+      "step": 131
+    },
+    {
+      "epoch": 0.048201570202665696,
+      "grad_norm": 653.3822631835938,
+      "learning_rate": 4.8175182481751835e-06,
+      "loss": 3.3528,
+      "step": 132
+    },
+    {
+      "epoch": 0.048566733613291946,
+      "grad_norm": 99.99283599853516,
+      "learning_rate": 4.854014598540146e-06,
+      "loss": 2.6455,
+      "step": 133
+    },
+    {
+      "epoch": 0.0489318970239182,
+      "grad_norm": 433.3922424316406,
+      "learning_rate": 4.89051094890511e-06,
+      "loss": 4.5752,
+      "step": 134
+    },
+    {
+      "epoch": 0.04929706043454446,
+      "grad_norm": 746.265380859375,
+      "learning_rate": 4.927007299270074e-06,
+      "loss": 3.9375,
+      "step": 135
+    },
+    {
+      "epoch": 0.049662223845170717,
+      "grad_norm": 750.6600952148438,
+      "learning_rate": 4.963503649635037e-06,
+      "loss": 3.1877,
+      "step": 136
+    },
+    {
+      "epoch": 0.050027387255796966,
+      "grad_norm": 1102.3570556640625,
+      "learning_rate": 5e-06,
+      "loss": 3.3564,
+      "step": 137
+    },
+    {
+      "epoch": 0.05039255066642322,
+      "grad_norm": 2158.283203125,
+      "learning_rate": 5.036496350364964e-06,
+      "loss": 3.2808,
+      "step": 138
+    },
+    {
+      "epoch": 0.05075771407704948,
+      "grad_norm": 308.4197998046875,
+      "learning_rate": 5.072992700729927e-06,
+      "loss": 2.7959,
+      "step": 139
+    },
+    {
+      "epoch": 0.05112287748767574,
+      "grad_norm": 478.9572448730469,
+      "learning_rate": 5.1094890510948916e-06,
+      "loss": 3.1453,
+      "step": 140
+    },
+    {
+      "epoch": 0.05148804089830199,
+      "grad_norm": 494.27197265625,
+      "learning_rate": 5.1459854014598544e-06,
+      "loss": 3.3398,
+      "step": 141
+    },
+    {
+      "epoch": 0.051853204308928244,
+      "grad_norm": 1912.637451171875,
+      "learning_rate": 5.182481751824818e-06,
+      "loss": 3.259,
+      "step": 142
+    },
+    {
+      "epoch": 0.0522183677195545,
+      "grad_norm": 466.3883361816406,
+      "learning_rate": 5.218978102189781e-06,
+      "loss": 3.5239,
+      "step": 143
+    },
+    {
+      "epoch": 0.05258353113018076,
+      "grad_norm": 2344.509765625,
+      "learning_rate": 5.255474452554746e-06,
+      "loss": 3.2029,
+      "step": 144
+    },
+    {
+      "epoch": 0.05294869454080701,
+      "grad_norm": 310.60968017578125,
+      "learning_rate": 5.2919708029197084e-06,
+      "loss": 3.3289,
+      "step": 145
+    },
+    {
+      "epoch": 0.053313857951433265,
+      "grad_norm": 1433.7972412109375,
+      "learning_rate": 5.328467153284672e-06,
+      "loss": 3.1902,
+      "step": 146
+    },
+    {
+      "epoch": 0.05367902136205952,
+      "grad_norm": 3818.810791015625,
+      "learning_rate": 5.364963503649635e-06,
+      "loss": 3.5647,
+      "step": 147
+    },
+    {
+      "epoch": 0.05404418477268578,
+      "grad_norm": 580.8521118164062,
+      "learning_rate": 5.401459854014599e-06,
+      "loss": 3.002,
+      "step": 148
+    },
+    {
+      "epoch": 0.054409348183312035,
+      "grad_norm": 1983.5205078125,
+      "learning_rate": 5.437956204379562e-06,
+      "loss": 3.064,
+      "step": 149
+    },
+    {
+      "epoch": 0.054774511593938285,
+      "grad_norm": 105.86579132080078,
+      "learning_rate": 5.474452554744526e-06,
+      "loss": 2.6968,
+      "step": 150
+    },
+    {
+      "epoch": 0.05513967500456454,
+      "grad_norm": 3389.255615234375,
+      "learning_rate": 5.51094890510949e-06,
+      "loss": 3.416,
+      "step": 151
+    },
+    {
+      "epoch": 0.0555048384151908,
+      "grad_norm": 594.318115234375,
+      "learning_rate": 5.547445255474453e-06,
+      "loss": 2.9917,
+      "step": 152
+    },
+    {
+      "epoch": 0.055870001825817056,
+      "grad_norm": 309.3802185058594,
+      "learning_rate": 5.5839416058394165e-06,
+      "loss": 3.2024,
+      "step": 153
+    },
+    {
+      "epoch": 0.056235165236443306,
+      "grad_norm": 15191.001953125,
+      "learning_rate": 5.62043795620438e-06,
+      "loss": 3.1177,
+      "step": 154
+    },
+    {
+      "epoch": 0.05660032864706956,
+      "grad_norm": 617.5128784179688,
+      "learning_rate": 5.656934306569344e-06,
+      "loss": 2.7318,
+      "step": 155
+    },
+    {
+      "epoch": 0.05696549205769582,
+      "grad_norm": 2553.682861328125,
+      "learning_rate": 5.693430656934307e-06,
+      "loss": 2.7981,
+      "step": 156
+    },
+    {
+      "epoch": 0.05733065546832208,
+      "grad_norm": 434.7556457519531,
+      "learning_rate": 5.7299270072992705e-06,
+      "loss": 3.3635,
+      "step": 157
+    },
+    {
+      "epoch": 0.05769581887894833,
+      "grad_norm": 760.1705932617188,
+      "learning_rate": 5.766423357664233e-06,
+      "loss": 3.2495,
+      "step": 158
+    },
+    {
+      "epoch": 0.058060982289574584,
+      "grad_norm": 1057.6082763671875,
+      "learning_rate": 5.802919708029198e-06,
+      "loss": 2.9773,
+      "step": 159
+    },
+    {
+      "epoch": 0.05842614570020084,
+      "grad_norm": 645.4927368164062,
+      "learning_rate": 5.839416058394161e-06,
+      "loss": 2.7349,
+      "step": 160
+    },
+    {
+      "epoch": 0.0587913091108271,
+      "grad_norm": 1277.3204345703125,
+      "learning_rate": 5.8759124087591245e-06,
+      "loss": 2.5996,
+      "step": 161
+    },
+    {
+      "epoch": 0.05915647252145335,
+      "grad_norm": 52.24332809448242,
+      "learning_rate": 5.912408759124088e-06,
+      "loss": 2.1094,
+      "step": 162
+    },
+    {
+      "epoch": 0.059521635932079604,
+      "grad_norm": 2303.613037109375,
+      "learning_rate": 5.948905109489051e-06,
+      "loss": 2.9954,
+      "step": 163
+    },
+    {
+      "epoch": 0.05988679934270586,
+      "grad_norm": 906.8886108398438,
+      "learning_rate": 5.985401459854016e-06,
+      "loss": 2.9514,
+      "step": 164
+    },
+    {
+      "epoch": 0.06025196275333212,
+      "grad_norm": 430.33221435546875,
+      "learning_rate": 6.0218978102189786e-06,
+      "loss": 2.9661,
+      "step": 165
+    },
+    {
+      "epoch": 0.06061712616395837,
+      "grad_norm": 459.38922119140625,
+      "learning_rate": 6.058394160583942e-06,
+      "loss": 3.4292,
+      "step": 166
+    },
+    {
+      "epoch": 0.060982289574584625,
+      "grad_norm": 453.9920349121094,
+      "learning_rate": 6.094890510948905e-06,
+      "loss": 3.5171,
+      "step": 167
+    },
+    {
+      "epoch": 0.06134745298521088,
+      "grad_norm": 381.0705261230469,
+      "learning_rate": 6.13138686131387e-06,
+      "loss": 2.5144,
+      "step": 168
+    },
+    {
+      "epoch": 0.06171261639583714,
+      "grad_norm": 126.2828598022461,
+      "learning_rate": 6.1678832116788326e-06,
+      "loss": 2.073,
+      "step": 169
+    },
+    {
+      "epoch": 0.062077779806463396,
+      "grad_norm": 211.9131622314453,
+      "learning_rate": 6.204379562043796e-06,
+      "loss": 2.4634,
+      "step": 170
+    },
+    {
+      "epoch": 0.062442943217089646,
+      "grad_norm": 1790.0267333984375,
+      "learning_rate": 6.240875912408759e-06,
+      "loss": 2.6631,
+      "step": 171
+    },
+    {
+      "epoch": 0.06280810662771591,
+      "grad_norm": 6149.90673828125,
+      "learning_rate": 6.277372262773723e-06,
+      "loss": 2.8137,
+      "step": 172
+    },
+    {
+      "epoch": 0.06317327003834215,
+      "grad_norm": 430.6115417480469,
+      "learning_rate": 6.313868613138686e-06,
+      "loss": 2.9197,
+      "step": 173
+    },
+    {
+      "epoch": 0.06353843344896841,
+      "grad_norm": 2553.958740234375,
+      "learning_rate": 6.35036496350365e-06,
+      "loss": 2.7192,
+      "step": 174
+    },
+    {
+      "epoch": 0.06390359685959467,
+      "grad_norm": 653.2872924804688,
+      "learning_rate": 6.386861313868614e-06,
+      "loss": 2.3171,
+      "step": 175
+    },
+    {
+      "epoch": 0.06426876027022092,
+      "grad_norm": 709.2286987304688,
+      "learning_rate": 6.423357664233577e-06,
+      "loss": 2.6941,
+      "step": 176
+    },
+    {
+      "epoch": 0.06463392368084718,
+      "grad_norm": 6361.58740234375,
+      "learning_rate": 6.4598540145985415e-06,
+      "loss": 3.2048,
+      "step": 177
+    },
+    {
+      "epoch": 0.06499908709147344,
+      "grad_norm": 511.78814697265625,
+      "learning_rate": 6.496350364963504e-06,
+      "loss": 2.3953,
+      "step": 178
+    },
+    {
+      "epoch": 0.0653642505020997,
+      "grad_norm": 2117.15576171875,
+      "learning_rate": 6.532846715328468e-06,
+      "loss": 2.8042,
+      "step": 179
+    },
+    {
+      "epoch": 0.06572941391272595,
+      "grad_norm": 448.7147521972656,
+      "learning_rate": 6.569343065693431e-06,
+      "loss": 2.3252,
+      "step": 180
+    },
+    {
+      "epoch": 0.0660945773233522,
+      "grad_norm": 373.40496826171875,
+      "learning_rate": 6.605839416058395e-06,
+      "loss": 2.2825,
+      "step": 181
+    },
+    {
+      "epoch": 0.06645974073397845,
+      "grad_norm": 2423.2509765625,
+      "learning_rate": 6.6423357664233575e-06,
+      "loss": 2.665,
+      "step": 182
+    },
+    {
+      "epoch": 0.06682490414460471,
+      "grad_norm": 810.9012451171875,
+      "learning_rate": 6.678832116788322e-06,
+      "loss": 2.3025,
+      "step": 183
+    },
+    {
+      "epoch": 0.06719006755523096,
+      "grad_norm": 237.67294311523438,
+      "learning_rate": 6.715328467153285e-06,
+      "loss": 2.5515,
+      "step": 184
+    },
+    {
+      "epoch": 0.06755523096585722,
+      "grad_norm": 2281.54345703125,
+      "learning_rate": 6.751824817518249e-06,
+      "loss": 2.478,
+      "step": 185
+    },
+    {
+      "epoch": 0.06792039437648348,
+      "grad_norm": 8467.810546875,
+      "learning_rate": 6.7883211678832115e-06,
+      "loss": 2.3552,
+      "step": 186
+    },
+    {
+      "epoch": 0.06828555778710974,
+      "grad_norm": 262.3053283691406,
+      "learning_rate": 6.824817518248176e-06,
+      "loss": 2.1997,
+      "step": 187
+    },
+    {
+      "epoch": 0.06865072119773599,
+      "grad_norm": 36.84046173095703,
+      "learning_rate": 6.86131386861314e-06,
+      "loss": 1.7383,
+      "step": 188
+    },
+    {
+      "epoch": 0.06901588460836225,
+      "grad_norm": 715.3591918945312,
+      "learning_rate": 6.897810218978103e-06,
+      "loss": 2.1892,
+      "step": 189
+    },
+    {
+      "epoch": 0.06938104801898849,
+      "grad_norm": 1096.4197998046875,
+      "learning_rate": 6.934306569343066e-06,
+      "loss": 2.3547,
+      "step": 190
+    },
+    {
+      "epoch": 0.06974621142961475,
+      "grad_norm": 3342.317138671875,
+      "learning_rate": 6.970802919708029e-06,
+      "loss": 2.7402,
+      "step": 191
+    },
+    {
+      "epoch": 0.070111374840241,
+      "grad_norm": 579.5726318359375,
+      "learning_rate": 7.007299270072994e-06,
+      "loss": 2.5376,
+      "step": 192
+    },
+    {
+      "epoch": 0.07047653825086726,
+      "grad_norm": 4662.072265625,
+      "learning_rate": 7.043795620437957e-06,
+      "loss": 2.4165,
+      "step": 193
+    },
+    {
+      "epoch": 0.07084170166149352,
+      "grad_norm": 74.3611068725586,
+      "learning_rate": 7.08029197080292e-06,
+      "loss": 1.8992,
+      "step": 194
+    },
+    {
+      "epoch": 0.07120686507211978,
+      "grad_norm": 873.2982788085938,
+      "learning_rate": 7.116788321167883e-06,
+      "loss": 2.2776,
+      "step": 195
+    },
+    {
+      "epoch": 0.07157202848274603,
+      "grad_norm": 104.76461791992188,
+      "learning_rate": 7.153284671532848e-06,
+      "loss": 2.1392,
+      "step": 196
+    },
+    {
+      "epoch": 0.07193719189337229,
+      "grad_norm": 13084.2880859375,
+      "learning_rate": 7.189781021897811e-06,
+      "loss": 2.0376,
+      "step": 197
+    },
+    {
+      "epoch": 0.07230235530399853,
+      "grad_norm": 1207.53173828125,
+      "learning_rate": 7.2262773722627744e-06,
+      "loss": 2.4053,
+      "step": 198
+    },
+    {
+      "epoch": 0.07266751871462479,
+      "grad_norm": 453.8113708496094,
+      "learning_rate": 7.262773722627737e-06,
+      "loss": 2.1384,
+      "step": 199
+    },
+    {
+      "epoch": 0.07303268212525105,
+      "grad_norm": 80.14839935302734,
+      "learning_rate": 7.299270072992701e-06,
+      "loss": 1.9971,
+      "step": 200
+    },
+    {
+      "epoch": 0.0733978455358773,
+      "grad_norm": 101.94886016845703,
+      "learning_rate": 7.335766423357666e-06,
+      "loss": 1.7673,
+      "step": 201
+    },
+    {
+      "epoch": 0.07376300894650356,
+      "grad_norm": 333.60888671875,
+      "learning_rate": 7.3722627737226285e-06,
+      "loss": 1.9768,
+      "step": 202
+    },
+    {
+      "epoch": 0.07412817235712982,
+      "grad_norm": 759.494873046875,
+      "learning_rate": 7.408759124087592e-06,
+      "loss": 2.28,
+      "step": 203
+    },
+    {
+      "epoch": 0.07449333576775607,
+      "grad_norm": 364.55377197265625,
+      "learning_rate": 7.445255474452555e-06,
+      "loss": 1.7869,
+      "step": 204
+    },
+    {
+      "epoch": 0.07485849917838233,
+      "grad_norm": 45.348201751708984,
+      "learning_rate": 7.481751824817519e-06,
+      "loss": 1.7795,
+      "step": 205
+    },
+    {
+      "epoch": 0.07522366258900857,
+      "grad_norm": 94.23865509033203,
+      "learning_rate": 7.5182481751824825e-06,
+      "loss": 1.8171,
+      "step": 206
+    },
+    {
+      "epoch": 0.07558882599963483,
+      "grad_norm": 20.518783569335938,
+      "learning_rate": 7.554744525547446e-06,
+      "loss": 1.5852,
+      "step": 207
+    },
+    {
+      "epoch": 0.07595398941026109,
+      "grad_norm": 39.20490646362305,
+      "learning_rate": 7.591240875912409e-06,
+      "loss": 1.7422,
+      "step": 208
+    },
+    {
+      "epoch": 0.07631915282088735,
+      "grad_norm": 27.00390625,
+      "learning_rate": 7.627737226277373e-06,
+      "loss": 1.6777,
+      "step": 209
+    },
+    {
+      "epoch": 0.0766843162315136,
+      "grad_norm": 83.09174346923828,
+      "learning_rate": 7.664233576642336e-06,
+      "loss": 1.5854,
+      "step": 210
+    },
+    {
+      "epoch": 0.07704947964213986,
+      "grad_norm": 34.0961799621582,
+      "learning_rate": 7.7007299270073e-06,
+      "loss": 1.7612,
+      "step": 211
+    },
+    {
+      "epoch": 0.07741464305276612,
+      "grad_norm": 16.32282829284668,
+      "learning_rate": 7.737226277372264e-06,
+      "loss": 1.4873,
+      "step": 212
+    },
+    {
+      "epoch": 0.07777980646339237,
+      "grad_norm": 23.388710021972656,
+      "learning_rate": 7.773722627737227e-06,
+      "loss": 1.6042,
+      "step": 213
+    },
+    {
+      "epoch": 0.07814496987401863,
+      "grad_norm": 14.058931350708008,
+      "learning_rate": 7.810218978102191e-06,
+      "loss": 1.4866,
+      "step": 214
+    },
+    {
+      "epoch": 0.07851013328464487,
+      "grad_norm": 44.53443908691406,
+      "learning_rate": 7.846715328467154e-06,
+      "loss": 1.5852,
+      "step": 215
+    },
+    {
+      "epoch": 0.07887529669527113,
+      "grad_norm": 19.07469940185547,
+      "learning_rate": 7.883211678832117e-06,
+      "loss": 1.5669,
+      "step": 216
+    },
+    {
+      "epoch": 0.07924046010589739,
+      "grad_norm": 25.231599807739258,
+      "learning_rate": 7.91970802919708e-06,
+      "loss": 1.6279,
+      "step": 217
+    },
+    {
+      "epoch": 0.07960562351652364,
+      "grad_norm": 39.810150146484375,
+      "learning_rate": 7.956204379562045e-06,
+      "loss": 1.6566,
+      "step": 218
+    },
+    {
+      "epoch": 0.0799707869271499,
+      "grad_norm": 29.961633682250977,
+      "learning_rate": 7.992700729927007e-06,
+      "loss": 1.6455,
+      "step": 219
+    },
+    {
+      "epoch": 0.08033595033777616,
+      "grad_norm": 28.15206527709961,
+      "learning_rate": 8.029197080291972e-06,
+      "loss": 1.595,
+      "step": 220
+    },
+    {
+      "epoch": 0.08070111374840241,
+      "grad_norm": 19.63481903076172,
+      "learning_rate": 8.065693430656935e-06,
+      "loss": 1.4966,
+      "step": 221
+    },
+    {
+      "epoch": 0.08106627715902867,
+      "grad_norm": 16.381669998168945,
+      "learning_rate": 8.1021897810219e-06,
+      "loss": 1.5696,
+      "step": 222
+    },
+    {
+      "epoch": 0.08143144056965491,
+      "grad_norm": 17.529966354370117,
+      "learning_rate": 8.138686131386862e-06,
+      "loss": 1.5007,
+      "step": 223
+    },
+    {
+      "epoch": 0.08179660398028117,
+      "grad_norm": 14.417710304260254,
+      "learning_rate": 8.175182481751825e-06,
+      "loss": 1.4875,
+      "step": 224
+    },
+    {
+      "epoch": 0.08216176739090743,
+      "grad_norm": 12.957341194152832,
+      "learning_rate": 8.21167883211679e-06,
+      "loss": 1.5208,
+      "step": 225
+    },
+    {
+      "epoch": 0.08252693080153368,
+      "grad_norm": 57.61381912231445,
+      "learning_rate": 8.248175182481753e-06,
+      "loss": 1.4958,
+      "step": 226
+    },
+    {
+      "epoch": 0.08289209421215994,
+      "grad_norm": 11.83564281463623,
+      "learning_rate": 8.284671532846717e-06,
+      "loss": 1.4526,
+      "step": 227
+    },
+    {
+      "epoch": 0.0832572576227862,
+      "grad_norm": 11.580336570739746,
+      "learning_rate": 8.32116788321168e-06,
+      "loss": 1.4099,
+      "step": 228
+    },
+    {
+      "epoch": 0.08362242103341246,
+      "grad_norm": 13.171910285949707,
+      "learning_rate": 8.357664233576643e-06,
+      "loss": 1.4248,
+      "step": 229
+    },
+    {
+      "epoch": 0.08398758444403871,
+      "grad_norm": 19.589122772216797,
+      "learning_rate": 8.394160583941606e-06,
+      "loss": 1.541,
+      "step": 230
+    },
+    {
+      "epoch": 0.08435274785466497,
+      "grad_norm": 18.963899612426758,
+      "learning_rate": 8.43065693430657e-06,
+      "loss": 1.4572,
+      "step": 231
+    },
+    {
+      "epoch": 0.08471791126529121,
+      "grad_norm": 27.862354278564453,
+      "learning_rate": 8.467153284671533e-06,
+      "loss": 1.5496,
+      "step": 232
+    },
+    {
+      "epoch": 0.08508307467591747,
+      "grad_norm": 13.316584587097168,
+      "learning_rate": 8.503649635036498e-06,
+      "loss": 1.418,
+      "step": 233
+    },
+    {
+      "epoch": 0.08544823808654373,
+      "grad_norm": 20.581497192382812,
+      "learning_rate": 8.54014598540146e-06,
+      "loss": 1.4546,
+      "step": 234
+    },
+    {
+      "epoch": 0.08581340149716998,
+      "grad_norm": 16.90022087097168,
+      "learning_rate": 8.576642335766423e-06,
+      "loss": 1.4329,
+      "step": 235
+    },
+    {
+      "epoch": 0.08617856490779624,
+      "grad_norm": 6.15281343460083,
+      "learning_rate": 8.613138686131386e-06,
+      "loss": 1.3525,
+      "step": 236
+    },
+    {
+      "epoch": 0.0865437283184225,
+      "grad_norm": 10.208907127380371,
+      "learning_rate": 8.649635036496351e-06,
+      "loss": 1.3865,
+      "step": 237
+    },
+    {
+      "epoch": 0.08690889172904875,
+      "grad_norm": 12.612737655639648,
+      "learning_rate": 8.686131386861315e-06,
+      "loss": 1.3468,
+      "step": 238
+    },
+    {
+      "epoch": 0.08727405513967501,
+      "grad_norm": 7.821756362915039,
+      "learning_rate": 8.722627737226278e-06,
+      "loss": 1.4331,
+      "step": 239
+    },
+    {
+      "epoch": 0.08763921855030125,
+      "grad_norm": 12.956961631774902,
+      "learning_rate": 8.759124087591241e-06,
+      "loss": 1.4187,
+      "step": 240
+    },
+    {
+      "epoch": 0.08800438196092751,
+      "grad_norm": 11.274295806884766,
+      "learning_rate": 8.795620437956204e-06,
+      "loss": 1.4182,
+      "step": 241
+    },
+    {
+      "epoch": 0.08836954537155377,
+      "grad_norm": 10.392095565795898,
+      "learning_rate": 8.832116788321169e-06,
+      "loss": 1.4207,
+      "step": 242
+    },
+    {
+      "epoch": 0.08873470878218002,
+      "grad_norm": 6.782304286956787,
+      "learning_rate": 8.868613138686132e-06,
+      "loss": 1.3278,
+      "step": 243
+    },
+    {
+      "epoch": 0.08909987219280628,
+      "grad_norm": 5.543254852294922,
+      "learning_rate": 8.905109489051096e-06,
+      "loss": 1.353,
+      "step": 244
+    },
+    {
+      "epoch": 0.08946503560343254,
+      "grad_norm": 4.561470031738281,
+      "learning_rate": 8.941605839416059e-06,
+      "loss": 1.3635,
+      "step": 245
+    },
+    {
+      "epoch": 0.0898301990140588,
+      "grad_norm": 23.20775604248047,
+      "learning_rate": 8.978102189781024e-06,
+      "loss": 1.467,
+      "step": 246
+    },
+    {
+      "epoch": 0.09019536242468505,
+      "grad_norm": 4.162003040313721,
+      "learning_rate": 9.014598540145986e-06,
+      "loss": 1.3699,
+      "step": 247
+    },
+    {
+      "epoch": 0.0905605258353113,
+      "grad_norm": 8.721848487854004,
+      "learning_rate": 9.05109489051095e-06,
+      "loss": 1.4294,
+      "step": 248
+    },
+    {
+      "epoch": 0.09092568924593755,
+      "grad_norm": 14.426276206970215,
+      "learning_rate": 9.087591240875912e-06,
+      "loss": 1.3287,
+      "step": 249
+    },
+    {
+      "epoch": 0.09129085265656381,
+      "grad_norm": 10.680879592895508,
+      "learning_rate": 9.124087591240877e-06,
+      "loss": 1.3496,
+      "step": 250
+    },
+    {
+      "epoch": 0.09165601606719007,
+      "grad_norm": 13.638522148132324,
+      "learning_rate": 9.160583941605841e-06,
+      "loss": 1.3894,
+      "step": 251
+    },
+    {
+      "epoch": 0.09202117947781632,
+      "grad_norm": 5.008195877075195,
+      "learning_rate": 9.197080291970804e-06,
+      "loss": 1.3525,
+      "step": 252
+    },
+    {
+      "epoch": 0.09238634288844258,
+      "grad_norm": 9.899373054504395,
+      "learning_rate": 9.233576642335767e-06,
+      "loss": 1.3335,
+      "step": 253
+    },
+    {
+      "epoch": 0.09275150629906884,
+      "grad_norm": 4.918022155761719,
+      "learning_rate": 9.27007299270073e-06,
+      "loss": 1.3196,
+      "step": 254
+    },
+    {
+      "epoch": 0.0931166697096951,
+      "grad_norm": 4.3167595863342285,
+      "learning_rate": 9.306569343065694e-06,
+      "loss": 1.3231,
+      "step": 255
+    },
+    {
+      "epoch": 0.09348183312032135,
+      "grad_norm": 3.6625378131866455,
+      "learning_rate": 9.343065693430657e-06,
+      "loss": 1.3268,
+      "step": 256
+    },
+    {
+      "epoch": 0.0938469965309476,
+      "grad_norm": 9.615107536315918,
+      "learning_rate": 9.379562043795622e-06,
+      "loss": 1.3452,
+      "step": 257
+    },
+    {
+      "epoch": 0.09421215994157385,
+      "grad_norm": 4.574172019958496,
+      "learning_rate": 9.416058394160585e-06,
+      "loss": 1.3042,
+      "step": 258
+    },
+    {
+      "epoch": 0.09457732335220011,
+      "grad_norm": 6.877755641937256,
+      "learning_rate": 9.452554744525548e-06,
+      "loss": 1.3171,
+      "step": 259
+    },
+    {
+      "epoch": 0.09494248676282636,
+      "grad_norm": 4.867911338806152,
+      "learning_rate": 9.48905109489051e-06,
+      "loss": 1.3306,
+      "step": 260
+    },
+    {
+      "epoch": 0.09530765017345262,
+      "grad_norm": 6.512596607208252,
+      "learning_rate": 9.525547445255475e-06,
+      "loss": 1.2874,
+      "step": 261
+    },
+    {
+      "epoch": 0.09567281358407888,
+      "grad_norm": 3.648726463317871,
+      "learning_rate": 9.56204379562044e-06,
+      "loss": 1.3174,
+      "step": 262
+    },
+    {
+      "epoch": 0.09603797699470513,
+      "grad_norm": 5.357989311218262,
+      "learning_rate": 9.598540145985402e-06,
+      "loss": 1.3412,
+      "step": 263
+    },
+    {
+      "epoch": 0.09640314040533139,
+      "grad_norm": 10.861003875732422,
+      "learning_rate": 9.635036496350367e-06,
+      "loss": 1.296,
+      "step": 264
+    },
+    {
+      "epoch": 0.09676830381595763,
+      "grad_norm": 12.365477561950684,
+      "learning_rate": 9.67153284671533e-06,
+      "loss": 1.3271,
+      "step": 265
+    },
+    {
+      "epoch": 0.09713346722658389,
+      "grad_norm": 7.079929351806641,
+      "learning_rate": 9.708029197080293e-06,
+      "loss": 1.2974,
+      "step": 266
+    },
+    {
+      "epoch": 0.09749863063721015,
+      "grad_norm": 4.8558430671691895,
+      "learning_rate": 9.744525547445256e-06,
+      "loss": 1.2659,
+      "step": 267
+    },
+    {
+      "epoch": 0.0978637940478364,
+      "grad_norm": 3.611086130142212,
+      "learning_rate": 9.78102189781022e-06,
+      "loss": 1.3125,
+      "step": 268
+    },
+    {
+      "epoch": 0.09822895745846266,
+      "grad_norm": 4.1291584968566895,
+      "learning_rate": 9.817518248175183e-06,
+      "loss": 1.3298,
+      "step": 269
+    },
+    {
+      "epoch": 0.09859412086908892,
+      "grad_norm": 6.496428966522217,
+      "learning_rate": 9.854014598540148e-06,
+      "loss": 1.262,
+      "step": 270
+    },
+    {
+      "epoch": 0.09895928427971518,
+      "grad_norm": 5.551502227783203,
+      "learning_rate": 9.89051094890511e-06,
+      "loss": 1.2892,
+      "step": 271
+    },
+    {
+      "epoch": 0.09932444769034143,
+      "grad_norm": 2.4452531337738037,
+      "learning_rate": 9.927007299270073e-06,
+      "loss": 1.3044,
+      "step": 272
+    },
+    {
+      "epoch": 0.09968961110096769,
+      "grad_norm": 8.159255981445312,
+      "learning_rate": 9.963503649635036e-06,
+      "loss": 1.314,
+      "step": 273
+    },
+    {
+      "epoch": 0.10005477451159393,
+      "grad_norm": 3.99063777923584,
+      "learning_rate": 1e-05,
+      "loss": 1.2576,
+      "step": 274
+    },
+    {
+      "epoch": 0.10041993792222019,
+      "grad_norm": 1.5462865829467773,
+      "learning_rate": 1.0036496350364964e-05,
+      "loss": 1.2864,
+      "step": 275
+    },
+    {
+      "epoch": 0.10078510133284645,
+      "grad_norm": 3.204332113265991,
+      "learning_rate": 1.0072992700729928e-05,
+      "loss": 1.2542,
+      "step": 276
+    },
+    {
+      "epoch": 0.1011502647434727,
+      "grad_norm": 2.300490140914917,
+      "learning_rate": 1.0109489051094891e-05,
+      "loss": 1.3162,
+      "step": 277
+    },
+    {
+      "epoch": 0.10151542815409896,
+      "grad_norm": 4.768906593322754,
+      "learning_rate": 1.0145985401459854e-05,
+      "loss": 1.2646,
+      "step": 278
+    },
+    {
+      "epoch": 0.10188059156472522,
+      "grad_norm": 4.017237186431885,
+      "learning_rate": 1.0182481751824817e-05,
+      "loss": 1.2429,
+      "step": 279
+    },
+    {
+      "epoch": 0.10224575497535147,
+      "grad_norm": 4.3871917724609375,
+      "learning_rate": 1.0218978102189783e-05,
+      "loss": 1.2776,
+      "step": 280
+    },
+    {
+      "epoch": 0.10261091838597773,
+      "grad_norm": 3.5851235389709473,
+      "learning_rate": 1.0255474452554746e-05,
+      "loss": 1.3145,
+      "step": 281
+    },
+    {
+      "epoch": 0.10297608179660397,
+      "grad_norm": 2.6969120502471924,
+      "learning_rate": 1.0291970802919709e-05,
+      "loss": 1.2322,
+      "step": 282
+    },
+    {
+      "epoch": 0.10334124520723023,
+      "grad_norm": 2.2774250507354736,
+      "learning_rate": 1.0328467153284672e-05,
+      "loss": 1.2477,
+      "step": 283
+    },
+    {
+      "epoch": 0.10370640861785649,
+      "grad_norm": 2.5099644660949707,
+      "learning_rate": 1.0364963503649636e-05,
+      "loss": 1.2175,
+      "step": 284
+    },
+    {
+      "epoch": 0.10407157202848275,
+      "grad_norm": 5.502248764038086,
+      "learning_rate": 1.04014598540146e-05,
+      "loss": 1.2748,
+      "step": 285
+    },
+    {
+      "epoch": 0.104436735439109,
+      "grad_norm": 2.743541717529297,
+      "learning_rate": 1.0437956204379562e-05,
+      "loss": 1.2744,
+      "step": 286
+    },
+    {
+      "epoch": 0.10480189884973526,
+      "grad_norm": 3.1978044509887695,
+      "learning_rate": 1.0474452554744528e-05,
+      "loss": 1.2834,
+      "step": 287
+    },
+    {
+      "epoch": 0.10516706226036152,
+      "grad_norm": 1.869937539100647,
+      "learning_rate": 1.0510948905109491e-05,
+      "loss": 1.3057,
+      "step": 288
+    },
+    {
+      "epoch": 0.10553222567098777,
+      "grad_norm": 2.7147998809814453,
+      "learning_rate": 1.0547445255474454e-05,
+      "loss": 1.2305,
+      "step": 289
+    },
+    {
+      "epoch": 0.10589738908161402,
+      "grad_norm": 1.5068933963775635,
+      "learning_rate": 1.0583941605839417e-05,
+      "loss": 1.2383,
+      "step": 290
+    },
+    {
+      "epoch": 0.10626255249224027,
+      "grad_norm": 2.4968197345733643,
+      "learning_rate": 1.0620437956204381e-05,
+      "loss": 1.2522,
+      "step": 291
+    },
+    {
+      "epoch": 0.10662771590286653,
+      "grad_norm": 2.359579086303711,
+      "learning_rate": 1.0656934306569344e-05,
+      "loss": 1.2708,
+      "step": 292
+    },
+    {
+      "epoch": 0.10699287931349279,
+      "grad_norm": 1.5981340408325195,
+      "learning_rate": 1.0693430656934307e-05,
+      "loss": 1.2152,
+      "step": 293
+    },
+    {
+      "epoch": 0.10735804272411904,
+      "grad_norm": 2.902353048324585,
+      "learning_rate": 1.072992700729927e-05,
+      "loss": 1.2087,
+      "step": 294
+    },
+    {
+      "epoch": 0.1077232061347453,
+      "grad_norm": 1.5850776433944702,
+      "learning_rate": 1.0766423357664235e-05,
+      "loss": 1.2211,
+      "step": 295
+    },
+    {
+      "epoch": 0.10808836954537156,
+      "grad_norm": 1.5473744869232178,
+      "learning_rate": 1.0802919708029198e-05,
+      "loss": 1.2305,
+      "step": 296
+    },
+    {
+      "epoch": 0.10845353295599781,
+      "grad_norm": 2.7115235328674316,
+      "learning_rate": 1.083941605839416e-05,
+      "loss": 1.1915,
+      "step": 297
+    },
+    {
+      "epoch": 0.10881869636662407,
+      "grad_norm": 1.5588639974594116,
+      "learning_rate": 1.0875912408759123e-05,
+      "loss": 1.1577,
+      "step": 298
+    },
+    {
+      "epoch": 0.10918385977725031,
+      "grad_norm": 1.6646748781204224,
+      "learning_rate": 1.091240875912409e-05,
+      "loss": 1.1941,
+      "step": 299
+    },
+    {
+      "epoch": 0.10954902318787657,
+      "grad_norm": 1.1305725574493408,
+      "learning_rate": 1.0948905109489052e-05,
+      "loss": 1.2269,
+      "step": 300
+    },
+    {
+      "epoch": 0.10991418659850283,
+      "grad_norm": 2.397463083267212,
+      "learning_rate": 1.0985401459854015e-05,
+      "loss": 1.2149,
+      "step": 301
+    },
+    {
+      "epoch": 0.11027935000912908,
+      "grad_norm": 1.4366761445999146,
+      "learning_rate": 1.102189781021898e-05,
+      "loss": 1.1731,
+      "step": 302
+    },
+    {
+      "epoch": 0.11064451341975534,
+      "grad_norm": 1.9387191534042358,
+      "learning_rate": 1.1058394160583943e-05,
+      "loss": 1.2324,
+      "step": 303
+    },
+    {
+      "epoch": 0.1110096768303816,
+      "grad_norm": 1.9468061923980713,
+      "learning_rate": 1.1094890510948906e-05,
+      "loss": 1.2415,
+      "step": 304
+    },
+    {
+      "epoch": 0.11137484024100786,
+      "grad_norm": 1.364353895187378,
+      "learning_rate": 1.1131386861313868e-05,
+      "loss": 1.2037,
+      "step": 305
+    },
+    {
+      "epoch": 0.11174000365163411,
+      "grad_norm": 1.462504267692566,
+      "learning_rate": 1.1167883211678833e-05,
+      "loss": 1.1635,
+      "step": 306
+    },
+    {
+      "epoch": 0.11210516706226036,
+      "grad_norm": 2.742152214050293,
+      "learning_rate": 1.1204379562043798e-05,
+      "loss": 1.209,
+      "step": 307
+    },
+    {
+      "epoch": 0.11247033047288661,
+      "grad_norm": 1.291460633277893,
+      "learning_rate": 1.124087591240876e-05,
+      "loss": 1.2004,
+      "step": 308
+    },
+    {
+      "epoch": 0.11283549388351287,
+      "grad_norm": 1.4753674268722534,
+      "learning_rate": 1.1277372262773723e-05,
+      "loss": 1.1941,
+      "step": 309
+    },
+    {
+      "epoch": 0.11320065729413913,
+      "grad_norm": 1.2697490453720093,
+      "learning_rate": 1.1313868613138688e-05,
+      "loss": 1.1705,
+      "step": 310
+    },
+    {
+      "epoch": 0.11356582070476538,
+      "grad_norm": 1.685782790184021,
+      "learning_rate": 1.135036496350365e-05,
+      "loss": 1.1661,
+      "step": 311
+    },
+    {
+      "epoch": 0.11393098411539164,
+      "grad_norm": 1.2983890771865845,
+      "learning_rate": 1.1386861313868614e-05,
+      "loss": 1.1893,
+      "step": 312
+    },
+    {
+      "epoch": 0.1142961475260179,
+      "grad_norm": 1.7487248182296753,
+      "learning_rate": 1.1423357664233578e-05,
+      "loss": 1.1729,
+      "step": 313
+    },
+    {
+      "epoch": 0.11466131093664415,
+      "grad_norm": 1.314300775527954,
+      "learning_rate": 1.1459854014598541e-05,
+      "loss": 1.1914,
+      "step": 314
+    },
+    {
+      "epoch": 0.11502647434727041,
+      "grad_norm": 1.6196850538253784,
+      "learning_rate": 1.1496350364963504e-05,
+      "loss": 1.1984,
+      "step": 315
+    },
+    {
+      "epoch": 0.11539163775789665,
+      "grad_norm": 1.3013982772827148,
+      "learning_rate": 1.1532846715328467e-05,
+      "loss": 1.1758,
+      "step": 316
+    },
+    {
+      "epoch": 0.11575680116852291,
+      "grad_norm": 1.7583945989608765,
+      "learning_rate": 1.1569343065693433e-05,
+      "loss": 1.1852,
+      "step": 317
+    },
+    {
+      "epoch": 0.11612196457914917,
+      "grad_norm": 1.5669349431991577,
+      "learning_rate": 1.1605839416058396e-05,
+      "loss": 1.1742,
+      "step": 318
+    },
+    {
+      "epoch": 0.11648712798977542,
+      "grad_norm": 1.3974933624267578,
+      "learning_rate": 1.1642335766423359e-05,
+      "loss": 1.1686,
+      "step": 319
+    },
+    {
+      "epoch": 0.11685229140040168,
+      "grad_norm": 1.367573618888855,
+      "learning_rate": 1.1678832116788322e-05,
+      "loss": 1.2064,
+      "step": 320
+    },
+    {
+      "epoch": 0.11721745481102794,
+      "grad_norm": 1.2772947549819946,
+      "learning_rate": 1.1715328467153286e-05,
+      "loss": 1.1616,
+      "step": 321
+    },
+    {
+      "epoch": 0.1175826182216542,
+      "grad_norm": 1.2221980094909668,
+      "learning_rate": 1.1751824817518249e-05,
+      "loss": 1.1732,
+      "step": 322
+    },
+    {
+      "epoch": 0.11794778163228045,
+      "grad_norm": 1.249281883239746,
+      "learning_rate": 1.1788321167883212e-05,
+      "loss": 1.1816,
+      "step": 323
+    },
+    {
+      "epoch": 0.1183129450429067,
+      "grad_norm": 1.674682378768921,
+      "learning_rate": 1.1824817518248176e-05,
+      "loss": 1.1648,
+      "step": 324
+    },
+    {
+      "epoch": 0.11867810845353295,
+      "grad_norm": 1.277409315109253,
+      "learning_rate": 1.186131386861314e-05,
+      "loss": 1.1906,
+      "step": 325
+    },
+    {
+      "epoch": 0.11904327186415921,
+      "grad_norm": 1.164598822593689,
+      "learning_rate": 1.1897810218978102e-05,
+      "loss": 1.1603,
+      "step": 326
+    },
+    {
+      "epoch": 0.11940843527478547,
+      "grad_norm": 1.5280879735946655,
+      "learning_rate": 1.1934306569343067e-05,
+      "loss": 1.1841,
+      "step": 327
+    },
+    {
+      "epoch": 0.11977359868541172,
+      "grad_norm": 1.3797391653060913,
+      "learning_rate": 1.1970802919708031e-05,
+      "loss": 1.1871,
+      "step": 328
+    },
+    {
+      "epoch": 0.12013876209603798,
+      "grad_norm": 1.0593525171279907,
+      "learning_rate": 1.2007299270072994e-05,
+      "loss": 1.1467,
+      "step": 329
+    },
+    {
+      "epoch": 0.12050392550666424,
+      "grad_norm": 1.0836522579193115,
+      "learning_rate": 1.2043795620437957e-05,
+      "loss": 1.1177,
+      "step": 330
+    },
+    {
+      "epoch": 0.1208690889172905,
+      "grad_norm": 0.9108336567878723,
+      "learning_rate": 1.208029197080292e-05,
+      "loss": 1.1685,
+      "step": 331
+    },
+    {
+      "epoch": 0.12123425232791674,
+      "grad_norm": 1.3441603183746338,
+      "learning_rate": 1.2116788321167885e-05,
+      "loss": 1.1779,
+      "step": 332
+    },
+    {
+      "epoch": 0.12159941573854299,
+      "grad_norm": 1.3265718221664429,
+      "learning_rate": 1.2153284671532847e-05,
+      "loss": 1.1953,
+      "step": 333
+    },
+    {
+      "epoch": 0.12196457914916925,
+      "grad_norm": 1.2688337564468384,
+      "learning_rate": 1.218978102189781e-05,
+      "loss": 1.165,
+      "step": 334
+    },
+    {
+      "epoch": 0.1223297425597955,
+      "grad_norm": 1.2843410968780518,
+      "learning_rate": 1.2226277372262773e-05,
+      "loss": 1.1614,
+      "step": 335
+    },
+    {
+      "epoch": 0.12269490597042176,
+      "grad_norm": 1.04667329788208,
+      "learning_rate": 1.226277372262774e-05,
+      "loss": 1.1503,
+      "step": 336
+    },
+    {
+      "epoch": 0.12306006938104802,
+      "grad_norm": 1.1412992477416992,
+      "learning_rate": 1.2299270072992702e-05,
+      "loss": 1.1467,
+      "step": 337
+    },
+    {
+      "epoch": 0.12342523279167428,
+      "grad_norm": 1.4156609773635864,
+      "learning_rate": 1.2335766423357665e-05,
+      "loss": 1.1447,
+      "step": 338
+    },
+    {
+      "epoch": 0.12379039620230053,
+      "grad_norm": 1.8062715530395508,
+      "learning_rate": 1.237226277372263e-05,
+      "loss": 1.1702,
+      "step": 339
+    },
+    {
+      "epoch": 0.12415555961292679,
+      "grad_norm": 1.2953407764434814,
+      "learning_rate": 1.2408759124087593e-05,
+      "loss": 1.1246,
+      "step": 340
+    },
+    {
+      "epoch": 0.12452072302355303,
+      "grad_norm": 1.685192346572876,
+      "learning_rate": 1.2445255474452555e-05,
+      "loss": 1.1724,
+      "step": 341
+    },
+    {
+      "epoch": 0.12488588643417929,
+      "grad_norm": 0.9900994896888733,
+      "learning_rate": 1.2481751824817518e-05,
+      "loss": 1.127,
+      "step": 342
+    },
+    {
+      "epoch": 0.12525104984480556,
+      "grad_norm": 1.2946479320526123,
+      "learning_rate": 1.2518248175182483e-05,
+      "loss": 1.1466,
+      "step": 343
+    },
+    {
+      "epoch": 0.12561621325543182,
+      "grad_norm": 1.1135441064834595,
+      "learning_rate": 1.2554744525547446e-05,
+      "loss": 1.1188,
+      "step": 344
+    },
+    {
+      "epoch": 0.12598137666605805,
+      "grad_norm": 1.2259373664855957,
+      "learning_rate": 1.2591240875912409e-05,
+      "loss": 1.1257,
+      "step": 345
+    },
+    {
+      "epoch": 0.1263465400766843,
+      "grad_norm": 1.0781917572021484,
+      "learning_rate": 1.2627737226277371e-05,
+      "loss": 1.1167,
+      "step": 346
+    },
+    {
+      "epoch": 0.12671170348731056,
+      "grad_norm": 1.1509932279586792,
+      "learning_rate": 1.2664233576642338e-05,
+      "loss": 1.1675,
+      "step": 347
+    },
+    {
+      "epoch": 0.12707686689793682,
+      "grad_norm": 1.7564834356307983,
+      "learning_rate": 1.27007299270073e-05,
+      "loss": 1.1327,
+      "step": 348
+    },
+    {
+      "epoch": 0.12744203030856308,
+      "grad_norm": 1.068116307258606,
+      "learning_rate": 1.2737226277372263e-05,
+      "loss": 1.1166,
+      "step": 349
+    },
+    {
+      "epoch": 0.12780719371918933,
+      "grad_norm": 1.2166095972061157,
+      "learning_rate": 1.2773722627737228e-05,
+      "loss": 1.1189,
+      "step": 350
+    },
+    {
+      "epoch": 0.1281723571298156,
+      "grad_norm": 1.385166049003601,
+      "learning_rate": 1.2810218978102191e-05,
+      "loss": 1.1252,
+      "step": 351
+    },
+    {
+      "epoch": 0.12853752054044185,
+      "grad_norm": 0.9813831448554993,
+      "learning_rate": 1.2846715328467154e-05,
+      "loss": 1.1289,
+      "step": 352
+    },
+    {
+      "epoch": 0.1289026839510681,
+      "grad_norm": 1.0689971446990967,
+      "learning_rate": 1.2883211678832117e-05,
+      "loss": 1.0798,
+      "step": 353
+    },
+    {
+      "epoch": 0.12926784736169436,
+      "grad_norm": 1.1747773885726929,
+      "learning_rate": 1.2919708029197083e-05,
+      "loss": 1.1304,
+      "step": 354
+    },
+    {
+      "epoch": 0.12963301077232062,
+      "grad_norm": 0.9878880381584167,
+      "learning_rate": 1.2956204379562046e-05,
+      "loss": 1.12,
+      "step": 355
+    },
+    {
+      "epoch": 0.12999817418294687,
+      "grad_norm": 1.2414008378982544,
+      "learning_rate": 1.2992700729927009e-05,
+      "loss": 1.1359,
+      "step": 356
+    },
+    {
+      "epoch": 0.13036333759357313,
+      "grad_norm": 1.0068449974060059,
+      "learning_rate": 1.3029197080291972e-05,
+      "loss": 1.1094,
+      "step": 357
+    },
+    {
+      "epoch": 0.1307285010041994,
+      "grad_norm": 1.1552882194519043,
+      "learning_rate": 1.3065693430656936e-05,
+      "loss": 1.1625,
+      "step": 358
+    },
+    {
+      "epoch": 0.13109366441482564,
+      "grad_norm": 1.1237143278121948,
+      "learning_rate": 1.3102189781021899e-05,
+      "loss": 1.1239,
+      "step": 359
+    },
+    {
+      "epoch": 0.1314588278254519,
+      "grad_norm": 1.3414310216903687,
+      "learning_rate": 1.3138686131386862e-05,
+      "loss": 1.1241,
+      "step": 360
+    },
+    {
+      "epoch": 0.13182399123607816,
+      "grad_norm": 1.0340030193328857,
+      "learning_rate": 1.3175182481751825e-05,
+      "loss": 1.1146,
+      "step": 361
+    },
+    {
+      "epoch": 0.1321891546467044,
+      "grad_norm": 1.00321626663208,
+      "learning_rate": 1.321167883211679e-05,
+      "loss": 1.1484,
+      "step": 362
+    },
+    {
+      "epoch": 0.13255431805733064,
+      "grad_norm": 1.6498593091964722,
+      "learning_rate": 1.3248175182481752e-05,
+      "loss": 1.1301,
+      "step": 363
+    },
+    {
+      "epoch": 0.1329194814679569,
+      "grad_norm": 1.1106163263320923,
+      "learning_rate": 1.3284671532846715e-05,
+      "loss": 1.1046,
+      "step": 364
+    },
+    {
+      "epoch": 0.13328464487858316,
+      "grad_norm": 1.888108253479004,
+      "learning_rate": 1.3321167883211681e-05,
+      "loss": 1.0774,
+      "step": 365
+    },
+    {
+      "epoch": 0.13364980828920942,
+      "grad_norm": 1.613932728767395,
+      "learning_rate": 1.3357664233576644e-05,
+      "loss": 1.1235,
+      "step": 366
+    },
+    {
+      "epoch": 0.13401497169983567,
+      "grad_norm": 1.3553868532180786,
+      "learning_rate": 1.3394160583941607e-05,
+      "loss": 1.1141,
+      "step": 367
+    },
+    {
+      "epoch": 0.13438013511046193,
+      "grad_norm": 1.1510350704193115,
+      "learning_rate": 1.343065693430657e-05,
+      "loss": 1.0662,
+      "step": 368
+    },
+    {
+      "epoch": 0.13474529852108819,
+      "grad_norm": 1.0214452743530273,
+      "learning_rate": 1.3467153284671534e-05,
+      "loss": 1.1061,
+      "step": 369
+    },
+    {
+      "epoch": 0.13511046193171444,
+      "grad_norm": 1.2430360317230225,
+      "learning_rate": 1.3503649635036497e-05,
+      "loss": 1.0872,
+      "step": 370
+    },
+    {
+      "epoch": 0.1354756253423407,
+      "grad_norm": 0.8896999359130859,
+      "learning_rate": 1.354014598540146e-05,
+      "loss": 1.0865,
+      "step": 371
+    },
+    {
+      "epoch": 0.13584078875296696,
+      "grad_norm": 1.4086633920669556,
+      "learning_rate": 1.3576642335766423e-05,
+      "loss": 1.1355,
+      "step": 372
+    },
+    {
+      "epoch": 0.1362059521635932,
+      "grad_norm": 1.149206519126892,
+      "learning_rate": 1.361313868613139e-05,
+      "loss": 1.0841,
+      "step": 373
+    },
+    {
+      "epoch": 0.13657111557421947,
+      "grad_norm": 1.3880298137664795,
+      "learning_rate": 1.3649635036496352e-05,
+      "loss": 1.1204,
+      "step": 374
+    },
+    {
+      "epoch": 0.13693627898484573,
+      "grad_norm": 1.1958924531936646,
+      "learning_rate": 1.3686131386861315e-05,
+      "loss": 1.0629,
+      "step": 375
+    },
+    {
+      "epoch": 0.13730144239547198,
+      "grad_norm": 1.2898317575454712,
+      "learning_rate": 1.372262773722628e-05,
+      "loss": 1.1003,
+      "step": 376
+    },
+    {
+      "epoch": 0.13766660580609824,
+      "grad_norm": 1.1109548807144165,
+      "learning_rate": 1.3759124087591242e-05,
+      "loss": 1.1147,
+      "step": 377
+    },
+    {
+      "epoch": 0.1380317692167245,
+      "grad_norm": 1.3594517707824707,
+      "learning_rate": 1.3795620437956205e-05,
+      "loss": 1.1285,
+      "step": 378
+    },
+    {
+      "epoch": 0.13839693262735073,
+      "grad_norm": 0.9975214004516602,
+      "learning_rate": 1.3832116788321168e-05,
+      "loss": 1.0709,
+      "step": 379
+    },
+    {
+      "epoch": 0.13876209603797698,
+      "grad_norm": 1.0834999084472656,
+      "learning_rate": 1.3868613138686133e-05,
+      "loss": 1.0236,
+      "step": 380
+    },
+    {
+      "epoch": 0.13912725944860324,
+      "grad_norm": 1.2681293487548828,
+      "learning_rate": 1.3905109489051096e-05,
+      "loss": 1.0898,
+      "step": 381
+    },
+    {
+      "epoch": 0.1394924228592295,
+      "grad_norm": 1.375718593597412,
+      "learning_rate": 1.3941605839416059e-05,
+      "loss": 1.1216,
+      "step": 382
+    },
+    {
+      "epoch": 0.13985758626985575,
+      "grad_norm": 0.9331570863723755,
+      "learning_rate": 1.3978102189781021e-05,
+      "loss": 1.0582,
+      "step": 383
+    },
+    {
+      "epoch": 0.140222749680482,
+      "grad_norm": 1.582811713218689,
+      "learning_rate": 1.4014598540145988e-05,
+      "loss": 1.0657,
+      "step": 384
+    },
+    {
+      "epoch": 0.14058791309110827,
+      "grad_norm": 1.0264136791229248,
+      "learning_rate": 1.405109489051095e-05,
+      "loss": 1.0863,
+      "step": 385
+    },
+    {
+      "epoch": 0.14095307650173453,
+      "grad_norm": 0.8130198121070862,
+      "learning_rate": 1.4087591240875913e-05,
+      "loss": 1.0668,
+      "step": 386
+    },
+    {
+      "epoch": 0.14131823991236078,
+      "grad_norm": 1.1384094953536987,
+      "learning_rate": 1.4124087591240878e-05,
+      "loss": 1.12,
+      "step": 387
+    },
+    {
+      "epoch": 0.14168340332298704,
+      "grad_norm": 0.7725839018821716,
+      "learning_rate": 1.416058394160584e-05,
+      "loss": 1.0846,
+      "step": 388
+    },
+    {
+      "epoch": 0.1420485667336133,
+      "grad_norm": 0.966789186000824,
+      "learning_rate": 1.4197080291970804e-05,
+      "loss": 1.0905,
+      "step": 389
+    },
+    {
+      "epoch": 0.14241373014423955,
+      "grad_norm": 2.1811580657958984,
+      "learning_rate": 1.4233576642335767e-05,
+      "loss": 1.129,
+      "step": 390
+    },
+    {
+      "epoch": 0.1427788935548658,
+      "grad_norm": 1.425172209739685,
+      "learning_rate": 1.4270072992700733e-05,
+      "loss": 1.0992,
+      "step": 391
+    },
+    {
+      "epoch": 0.14314405696549207,
+      "grad_norm": 0.9881778359413147,
+      "learning_rate": 1.4306569343065696e-05,
+      "loss": 1.0516,
+      "step": 392
+    },
+    {
+      "epoch": 0.14350922037611832,
+      "grad_norm": 0.9002101421356201,
+      "learning_rate": 1.4343065693430659e-05,
+      "loss": 1.0531,
+      "step": 393
+    },
+    {
+      "epoch": 0.14387438378674458,
+      "grad_norm": 1.1981589794158936,
+      "learning_rate": 1.4379562043795621e-05,
+      "loss": 1.0902,
+      "step": 394
+    },
+    {
+      "epoch": 0.1442395471973708,
+      "grad_norm": 1.5116416215896606,
+      "learning_rate": 1.4416058394160586e-05,
+      "loss": 1.1179,
+      "step": 395
+    },
+    {
+      "epoch": 0.14460471060799707,
+      "grad_norm": 0.9920293688774109,
+      "learning_rate": 1.4452554744525549e-05,
+      "loss": 1.0645,
+      "step": 396
+    },
+    {
+      "epoch": 0.14496987401862332,
+      "grad_norm": 1.0004264116287231,
+      "learning_rate": 1.4489051094890512e-05,
+      "loss": 1.0924,
+      "step": 397
+    },
+    {
+      "epoch": 0.14533503742924958,
+      "grad_norm": 0.9262351393699646,
+      "learning_rate": 1.4525547445255475e-05,
+      "loss": 1.0723,
+      "step": 398
+    },
+    {
+      "epoch": 0.14570020083987584,
+      "grad_norm": 1.1073780059814453,
+      "learning_rate": 1.456204379562044e-05,
+      "loss": 1.048,
+      "step": 399
+    },
+    {
+      "epoch": 0.1460653642505021,
+      "grad_norm": 1.540772557258606,
+      "learning_rate": 1.4598540145985402e-05,
+      "loss": 1.0638,
+      "step": 400
+    },
+    {
+      "epoch": 0.14643052766112835,
+      "grad_norm": 1.0480707883834839,
+      "learning_rate": 1.4635036496350365e-05,
+      "loss": 1.0188,
+      "step": 401
+    },
+    {
+      "epoch": 0.1467956910717546,
+      "grad_norm": 0.9333097338676453,
+      "learning_rate": 1.4671532846715331e-05,
+      "loss": 1.0465,
+      "step": 402
+    },
+    {
+      "epoch": 0.14716085448238087,
+      "grad_norm": 1.100213646888733,
+      "learning_rate": 1.4708029197080294e-05,
+      "loss": 1.0583,
+      "step": 403
+    },
+    {
+      "epoch": 0.14752601789300712,
+      "grad_norm": 1.2130063772201538,
+      "learning_rate": 1.4744525547445257e-05,
+      "loss": 1.0928,
+      "step": 404
+    },
+    {
+      "epoch": 0.14789118130363338,
+      "grad_norm": 1.29470956325531,
+      "learning_rate": 1.478102189781022e-05,
+      "loss": 1.0491,
+      "step": 405
+    },
+    {
+      "epoch": 0.14825634471425964,
+      "grad_norm": 1.0126339197158813,
+      "learning_rate": 1.4817518248175184e-05,
+      "loss": 1.08,
+      "step": 406
+    },
+    {
+      "epoch": 0.1486215081248859,
+      "grad_norm": 1.1530481576919556,
+      "learning_rate": 1.4854014598540147e-05,
+      "loss": 1.0546,
+      "step": 407
+    },
+    {
+      "epoch": 0.14898667153551215,
+      "grad_norm": 0.9986507892608643,
+      "learning_rate": 1.489051094890511e-05,
+      "loss": 1.0809,
+      "step": 408
+    },
+    {
+      "epoch": 0.1493518349461384,
+      "grad_norm": 0.8909332156181335,
+      "learning_rate": 1.4927007299270073e-05,
+      "loss": 1.073,
+      "step": 409
+    },
+    {
+      "epoch": 0.14971699835676466,
+      "grad_norm": 1.0050809383392334,
+      "learning_rate": 1.4963503649635038e-05,
+      "loss": 1.0765,
+      "step": 410
+    },
+    {
+      "epoch": 0.15008216176739092,
+      "grad_norm": 1.157792329788208,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1008,
+      "step": 411
+    },
+    {
+      "epoch": 0.15044732517801715,
+      "grad_norm": 0.9719278812408447,
+      "learning_rate": 1.5036496350364965e-05,
+      "loss": 1.0743,
+      "step": 412
+    },
+    {
+      "epoch": 0.1508124885886434,
+      "grad_norm": 1.02153742313385,
+      "learning_rate": 1.507299270072993e-05,
+      "loss": 1.0713,
+      "step": 413
+    },
+    {
+      "epoch": 0.15117765199926966,
+      "grad_norm": 1.0733529329299927,
+      "learning_rate": 1.5109489051094892e-05,
+      "loss": 1.0425,
+      "step": 414
+    },
+    {
+      "epoch": 0.15154281540989592,
+      "grad_norm": 1.4467898607254028,
+      "learning_rate": 1.5145985401459855e-05,
+      "loss": 1.0861,
+      "step": 415
+    },
+    {
+      "epoch": 0.15190797882052218,
+      "grad_norm": 1.0643540620803833,
+      "learning_rate": 1.5182481751824818e-05,
+      "loss": 1.0444,
+      "step": 416
+    },
+    {
+      "epoch": 0.15227314223114843,
+      "grad_norm": 0.9217526912689209,
+      "learning_rate": 1.5218978102189783e-05,
+      "loss": 1.0559,
+      "step": 417
+    },
+    {
+      "epoch": 0.1526383056417747,
+      "grad_norm": 0.9161515831947327,
+      "learning_rate": 1.5255474452554746e-05,
+      "loss": 1.0404,
+      "step": 418
+    },
+    {
+      "epoch": 0.15300346905240095,
+      "grad_norm": 1.1920078992843628,
+      "learning_rate": 1.529197080291971e-05,
+      "loss": 1.0243,
+      "step": 419
+    },
+    {
+      "epoch": 0.1533686324630272,
+      "grad_norm": 1.1917927265167236,
+      "learning_rate": 1.5328467153284673e-05,
+      "loss": 1.0822,
+      "step": 420
+    },
+    {
+      "epoch": 0.15373379587365346,
+      "grad_norm": 0.8970569968223572,
+      "learning_rate": 1.5364963503649638e-05,
+      "loss": 1.0408,
+      "step": 421
+    },
+    {
+      "epoch": 0.15409895928427972,
+      "grad_norm": 1.2498112916946411,
+      "learning_rate": 1.54014598540146e-05,
+      "loss": 1.0618,
+      "step": 422
+    },
+    {
+      "epoch": 0.15446412269490598,
+      "grad_norm": 0.9885013699531555,
+      "learning_rate": 1.5437956204379563e-05,
+      "loss": 1.0234,
+      "step": 423
+    },
+    {
+      "epoch": 0.15482928610553223,
+      "grad_norm": 1.7543072700500488,
+      "learning_rate": 1.5474452554744528e-05,
+      "loss": 1.0657,
+      "step": 424
+    },
+    {
+      "epoch": 0.1551944495161585,
+      "grad_norm": 1.1781865358352661,
+      "learning_rate": 1.5510948905109492e-05,
+      "loss": 1.066,
+      "step": 425
+    },
+    {
+      "epoch": 0.15555961292678475,
+      "grad_norm": 1.1695396900177002,
+      "learning_rate": 1.5547445255474454e-05,
+      "loss": 1.0443,
+      "step": 426
+    },
+    {
+      "epoch": 0.155924776337411,
+      "grad_norm": 1.2203136682510376,
+      "learning_rate": 1.5583941605839418e-05,
+      "loss": 1.0295,
+      "step": 427
+    },
+    {
+      "epoch": 0.15628993974803726,
+      "grad_norm": 1.3869448900222778,
+      "learning_rate": 1.5620437956204383e-05,
+      "loss": 1.0327,
+      "step": 428
+    },
+    {
+      "epoch": 0.1566551031586635,
+      "grad_norm": 0.9070996642112732,
+      "learning_rate": 1.5656934306569344e-05,
+      "loss": 1.0391,
+      "step": 429
+    },
+    {
+      "epoch": 0.15702026656928975,
+      "grad_norm": 1.054938554763794,
+      "learning_rate": 1.569343065693431e-05,
+      "loss": 1.0491,
+      "step": 430
+    },
+    {
+      "epoch": 0.157385429979916,
+      "grad_norm": 1.1217361688613892,
+      "learning_rate": 1.572992700729927e-05,
+      "loss": 1.0445,
+      "step": 431
+    },
+    {
+      "epoch": 0.15775059339054226,
+      "grad_norm": 0.9165348410606384,
+      "learning_rate": 1.5766423357664234e-05,
+      "loss": 1.0095,
+      "step": 432
+    },
+    {
+      "epoch": 0.15811575680116852,
+      "grad_norm": 0.9862052798271179,
+      "learning_rate": 1.58029197080292e-05,
+      "loss": 1.0411,
+      "step": 433
+    },
+    {
+      "epoch": 0.15848092021179477,
+      "grad_norm": 1.22489595413208,
+      "learning_rate": 1.583941605839416e-05,
+      "loss": 0.9929,
+      "step": 434
+    },
+    {
+      "epoch": 0.15884608362242103,
+      "grad_norm": 1.3611830472946167,
+      "learning_rate": 1.5875912408759125e-05,
+      "loss": 1.0294,
+      "step": 435
+    },
+    {
+      "epoch": 0.1592112470330473,
+      "grad_norm": 1.1147099733352661,
+      "learning_rate": 1.591240875912409e-05,
+      "loss": 1.026,
+      "step": 436
+    },
+    {
+      "epoch": 0.15957641044367354,
+      "grad_norm": 1.1320207118988037,
+      "learning_rate": 1.5948905109489054e-05,
+      "loss": 1.0403,
+      "step": 437
+    },
+    {
+      "epoch": 0.1599415738542998,
+      "grad_norm": 1.1242649555206299,
+      "learning_rate": 1.5985401459854015e-05,
+      "loss": 1.011,
+      "step": 438
+    },
+    {
+      "epoch": 0.16030673726492606,
+      "grad_norm": 1.227778673171997,
+      "learning_rate": 1.602189781021898e-05,
+      "loss": 1.0278,
+      "step": 439
+    },
+    {
+      "epoch": 0.16067190067555231,
+      "grad_norm": 1.0648647546768188,
+      "learning_rate": 1.6058394160583944e-05,
+      "loss": 1.0559,
+      "step": 440
+    },
+    {
+      "epoch": 0.16103706408617857,
+      "grad_norm": 0.7608394622802734,
+      "learning_rate": 1.6094890510948905e-05,
+      "loss": 0.9849,
+      "step": 441
+    },
+    {
+      "epoch": 0.16140222749680483,
+      "grad_norm": 1.2371690273284912,
+      "learning_rate": 1.613138686131387e-05,
+      "loss": 1.0323,
+      "step": 442
+    },
+    {
+      "epoch": 0.16176739090743109,
+      "grad_norm": 1.255880355834961,
+      "learning_rate": 1.6167883211678834e-05,
+      "loss": 1.0261,
+      "step": 443
+    },
+    {
+      "epoch": 0.16213255431805734,
+      "grad_norm": 1.570586919784546,
+      "learning_rate": 1.62043795620438e-05,
+      "loss": 1.0492,
+      "step": 444
+    },
+    {
+      "epoch": 0.1624977177286836,
+      "grad_norm": 1.0332871675491333,
+      "learning_rate": 1.624087591240876e-05,
+      "loss": 1.0372,
+      "step": 445
+    },
+    {
+      "epoch": 0.16286288113930983,
+      "grad_norm": 0.8999532461166382,
+      "learning_rate": 1.6277372262773725e-05,
+      "loss": 1.0211,
+      "step": 446
+    },
+    {
+      "epoch": 0.16322804454993609,
+      "grad_norm": 1.821873664855957,
+      "learning_rate": 1.631386861313869e-05,
+      "loss": 1.0408,
+      "step": 447
+    },
+    {
+      "epoch": 0.16359320796056234,
+      "grad_norm": 0.7166489958763123,
+      "learning_rate": 1.635036496350365e-05,
+      "loss": 1.0221,
+      "step": 448
+    },
+    {
+      "epoch": 0.1639583713711886,
+      "grad_norm": 1.1623930931091309,
+      "learning_rate": 1.6386861313868615e-05,
+      "loss": 1.007,
+      "step": 449
+    },
+    {
+      "epoch": 0.16432353478181486,
+      "grad_norm": 0.8305017352104187,
+      "learning_rate": 1.642335766423358e-05,
+      "loss": 0.9558,
+      "step": 450
+    },
+    {
+      "epoch": 0.1646886981924411,
+      "grad_norm": 1.2439148426055908,
+      "learning_rate": 1.645985401459854e-05,
+      "loss": 1.0458,
+      "step": 451
+    },
+    {
+      "epoch": 0.16505386160306737,
+      "grad_norm": 1.6258620023727417,
+      "learning_rate": 1.6496350364963505e-05,
+      "loss": 1.0782,
+      "step": 452
+    },
+    {
+      "epoch": 0.16541902501369363,
+      "grad_norm": 2.255648374557495,
+      "learning_rate": 1.6532846715328466e-05,
+      "loss": 1.0084,
+      "step": 453
+    },
+    {
+      "epoch": 0.16578418842431988,
+      "grad_norm": 1.1743186712265015,
+      "learning_rate": 1.6569343065693434e-05,
+      "loss": 0.9736,
+      "step": 454
+    },
+    {
+      "epoch": 0.16614935183494614,
+      "grad_norm": 1.2281731367111206,
+      "learning_rate": 1.6605839416058395e-05,
+      "loss": 0.9695,
+      "step": 455
+    },
+    {
+      "epoch": 0.1665145152455724,
+      "grad_norm": 1.1789402961730957,
+      "learning_rate": 1.664233576642336e-05,
+      "loss": 1.0486,
+      "step": 456
+    },
+    {
+      "epoch": 0.16687967865619865,
+      "grad_norm": 0.9462199211120605,
+      "learning_rate": 1.667883211678832e-05,
+      "loss": 0.9995,
+      "step": 457
+    },
+    {
+      "epoch": 0.1672448420668249,
+      "grad_norm": 0.989033579826355,
+      "learning_rate": 1.6715328467153286e-05,
+      "loss": 1.0385,
+      "step": 458
+    },
+    {
+      "epoch": 0.16761000547745117,
+      "grad_norm": 1.0912344455718994,
+      "learning_rate": 1.675182481751825e-05,
+      "loss": 0.9836,
+      "step": 459
+    },
+    {
+      "epoch": 0.16797516888807743,
+      "grad_norm": 1.019728660583496,
+      "learning_rate": 1.678832116788321e-05,
+      "loss": 1.0382,
+      "step": 460
+    },
+    {
+      "epoch": 0.16834033229870368,
+      "grad_norm": 1.0688890218734741,
+      "learning_rate": 1.6824817518248176e-05,
+      "loss": 0.9849,
+      "step": 461
+    },
+    {
+      "epoch": 0.16870549570932994,
+      "grad_norm": 0.9562466144561768,
+      "learning_rate": 1.686131386861314e-05,
+      "loss": 1.0332,
+      "step": 462
+    },
+    {
+      "epoch": 0.16907065911995617,
+      "grad_norm": 0.9949936270713806,
+      "learning_rate": 1.6897810218978102e-05,
+      "loss": 1.0527,
+      "step": 463
+    },
+    {
+      "epoch": 0.16943582253058243,
+      "grad_norm": 1.2476692199707031,
+      "learning_rate": 1.6934306569343066e-05,
+      "loss": 1.0553,
+      "step": 464
+    },
+    {
+      "epoch": 0.16980098594120868,
+      "grad_norm": 1.048659324645996,
+      "learning_rate": 1.697080291970803e-05,
+      "loss": 1.0164,
+      "step": 465
+    },
+    {
+      "epoch": 0.17016614935183494,
+      "grad_norm": 0.9506887793540955,
+      "learning_rate": 1.7007299270072995e-05,
+      "loss": 1.022,
+      "step": 466
+    },
+    {
+      "epoch": 0.1705313127624612,
+      "grad_norm": 0.8908308148384094,
+      "learning_rate": 1.7043795620437957e-05,
+      "loss": 0.9797,
+      "step": 467
+    },
+    {
+      "epoch": 0.17089647617308745,
+      "grad_norm": 1.17626953125,
+      "learning_rate": 1.708029197080292e-05,
+      "loss": 0.9773,
+      "step": 468
+    },
+    {
+      "epoch": 0.1712616395837137,
+      "grad_norm": 1.0561892986297607,
+      "learning_rate": 1.7116788321167886e-05,
+      "loss": 0.9957,
+      "step": 469
+    },
+    {
+      "epoch": 0.17162680299433997,
+      "grad_norm": 1.4765760898590088,
+      "learning_rate": 1.7153284671532847e-05,
+      "loss": 1.0071,
+      "step": 470
+    },
+    {
+      "epoch": 0.17199196640496622,
+      "grad_norm": 1.372825264930725,
+      "learning_rate": 1.718978102189781e-05,
+      "loss": 1.0347,
+      "step": 471
+    },
+    {
+      "epoch": 0.17235712981559248,
+      "grad_norm": 1.0499844551086426,
+      "learning_rate": 1.7226277372262773e-05,
+      "loss": 0.9867,
+      "step": 472
+    },
+    {
+      "epoch": 0.17272229322621874,
+      "grad_norm": 1.3431679010391235,
+      "learning_rate": 1.726277372262774e-05,
+      "loss": 1.0115,
+      "step": 473
+    },
+    {
+      "epoch": 0.173087456636845,
+      "grad_norm": 0.8785029649734497,
+      "learning_rate": 1.7299270072992702e-05,
+      "loss": 1.0327,
+      "step": 474
+    },
+    {
+      "epoch": 0.17345262004747125,
+      "grad_norm": 1.3840019702911377,
+      "learning_rate": 1.7335766423357666e-05,
+      "loss": 0.9919,
+      "step": 475
+    },
+    {
+      "epoch": 0.1738177834580975,
+      "grad_norm": 0.955368161201477,
+      "learning_rate": 1.737226277372263e-05,
+      "loss": 0.9966,
+      "step": 476
+    },
+    {
+      "epoch": 0.17418294686872376,
+      "grad_norm": 1.3200948238372803,
+      "learning_rate": 1.7408759124087592e-05,
+      "loss": 1.0052,
+      "step": 477
+    },
+    {
+      "epoch": 0.17454811027935002,
+      "grad_norm": 0.9315057396888733,
+      "learning_rate": 1.7445255474452557e-05,
+      "loss": 0.9813,
+      "step": 478
+    },
+    {
+      "epoch": 0.17491327368997625,
+      "grad_norm": 1.0480930805206299,
+      "learning_rate": 1.7481751824817518e-05,
+      "loss": 0.9749,
+      "step": 479
+    },
+    {
+      "epoch": 0.1752784371006025,
+      "grad_norm": 1.000348687171936,
+      "learning_rate": 1.7518248175182482e-05,
+      "loss": 1.0067,
+      "step": 480
+    },
+    {
+      "epoch": 0.17564360051122876,
+      "grad_norm": 1.0230541229248047,
+      "learning_rate": 1.7554744525547447e-05,
+      "loss": 0.994,
+      "step": 481
+    },
+    {
+      "epoch": 0.17600876392185502,
+      "grad_norm": 1.2165991067886353,
+      "learning_rate": 1.7591240875912408e-05,
+      "loss": 1.0295,
+      "step": 482
+    },
+    {
+      "epoch": 0.17637392733248128,
+      "grad_norm": 1.2207070589065552,
+      "learning_rate": 1.7627737226277373e-05,
+      "loss": 0.9905,
+      "step": 483
+    },
+    {
+      "epoch": 0.17673909074310754,
+      "grad_norm": 1.2231794595718384,
+      "learning_rate": 1.7664233576642337e-05,
+      "loss": 1.0133,
+      "step": 484
+    },
+    {
+      "epoch": 0.1771042541537338,
+      "grad_norm": 1.200690746307373,
+      "learning_rate": 1.7700729927007302e-05,
+      "loss": 0.983,
+      "step": 485
+    },
+    {
+      "epoch": 0.17746941756436005,
+      "grad_norm": 1.118323564529419,
+      "learning_rate": 1.7737226277372263e-05,
+      "loss": 1.0201,
+      "step": 486
+    },
+    {
+      "epoch": 0.1778345809749863,
+      "grad_norm": 1.030706524848938,
+      "learning_rate": 1.7773722627737228e-05,
+      "loss": 1.0175,
+      "step": 487
+    },
+    {
+      "epoch": 0.17819974438561256,
+      "grad_norm": 0.7909834980964661,
+      "learning_rate": 1.7810218978102192e-05,
+      "loss": 0.9784,
+      "step": 488
+    },
+    {
+      "epoch": 0.17856490779623882,
+      "grad_norm": 1.075653076171875,
+      "learning_rate": 1.7846715328467153e-05,
+      "loss": 0.9595,
+      "step": 489
+    },
+    {
+      "epoch": 0.17893007120686508,
+      "grad_norm": 1.1155372858047485,
+      "learning_rate": 1.7883211678832118e-05,
+      "loss": 0.9739,
+      "step": 490
+    },
+    {
+      "epoch": 0.17929523461749133,
+      "grad_norm": 1.4763505458831787,
+      "learning_rate": 1.7919708029197082e-05,
+      "loss": 0.9624,
+      "step": 491
+    },
+    {
+      "epoch": 0.1796603980281176,
+      "grad_norm": 1.2181990146636963,
+      "learning_rate": 1.7956204379562047e-05,
+      "loss": 0.9927,
+      "step": 492
+    },
+    {
+      "epoch": 0.18002556143874385,
+      "grad_norm": 1.0169252157211304,
+      "learning_rate": 1.7992700729927008e-05,
+      "loss": 1.007,
+      "step": 493
+    },
+    {
+      "epoch": 0.1803907248493701,
+      "grad_norm": 1.4996839761734009,
+      "learning_rate": 1.8029197080291973e-05,
+      "loss": 0.998,
+      "step": 494
+    },
+    {
+      "epoch": 0.18075588825999636,
+      "grad_norm": 0.8153586387634277,
+      "learning_rate": 1.8065693430656937e-05,
+      "loss": 0.9808,
+      "step": 495
+    },
+    {
+      "epoch": 0.1811210516706226,
+      "grad_norm": 1.0656777620315552,
+      "learning_rate": 1.81021897810219e-05,
+      "loss": 0.9603,
+      "step": 496
+    },
+    {
+      "epoch": 0.18148621508124885,
+      "grad_norm": 1.1432812213897705,
+      "learning_rate": 1.8138686131386863e-05,
+      "loss": 0.9727,
+      "step": 497
+    },
+    {
+      "epoch": 0.1818513784918751,
+      "grad_norm": 0.8203110694885254,
+      "learning_rate": 1.8175182481751824e-05,
+      "loss": 1.0028,
+      "step": 498
+    },
+    {
+      "epoch": 0.18221654190250136,
+      "grad_norm": 0.9081830978393555,
+      "learning_rate": 1.821167883211679e-05,
+      "loss": 0.9829,
+      "step": 499
+    },
+    {
+      "epoch": 0.18258170531312762,
+      "grad_norm": 1.2304878234863281,
+      "learning_rate": 1.8248175182481753e-05,
+      "loss": 1.0103,
+      "step": 500
+    },
+    {
+      "epoch": 0.18294686872375387,
+      "grad_norm": 1.2801042795181274,
+      "learning_rate": 1.8284671532846715e-05,
+      "loss": 1.0187,
+      "step": 501
+    },
+    {
+      "epoch": 0.18331203213438013,
+      "grad_norm": 1.001322865486145,
+      "learning_rate": 1.8321167883211683e-05,
+      "loss": 0.959,
+      "step": 502
+    },
+    {
+      "epoch": 0.1836771955450064,
+      "grad_norm": 1.000063419342041,
+      "learning_rate": 1.8357664233576644e-05,
+      "loss": 0.993,
+      "step": 503
+    },
+    {
+      "epoch": 0.18404235895563265,
+      "grad_norm": 1.2404290437698364,
+      "learning_rate": 1.8394160583941608e-05,
+      "loss": 0.9767,
+      "step": 504
+    },
+    {
+      "epoch": 0.1844075223662589,
+      "grad_norm": 0.9571080803871155,
+      "learning_rate": 1.843065693430657e-05,
+      "loss": 0.9707,
+      "step": 505
+    },
+    {
+      "epoch": 0.18477268577688516,
+      "grad_norm": 0.9870139360427856,
+      "learning_rate": 1.8467153284671534e-05,
+      "loss": 0.9684,
+      "step": 506
+    },
+    {
+      "epoch": 0.18513784918751142,
+      "grad_norm": 0.9874576330184937,
+      "learning_rate": 1.85036496350365e-05,
+      "loss": 0.9814,
+      "step": 507
+    },
+    {
+      "epoch": 0.18550301259813767,
+      "grad_norm": 1.3045789003372192,
+      "learning_rate": 1.854014598540146e-05,
+      "loss": 0.979,
+      "step": 508
+    },
+    {
+      "epoch": 0.18586817600876393,
+      "grad_norm": 1.4654738903045654,
+      "learning_rate": 1.8576642335766424e-05,
+      "loss": 0.9468,
+      "step": 509
+    },
+    {
+      "epoch": 0.1862333394193902,
+      "grad_norm": 1.3370999097824097,
+      "learning_rate": 1.861313868613139e-05,
+      "loss": 0.9634,
+      "step": 510
+    },
+    {
+      "epoch": 0.18659850283001644,
+      "grad_norm": 0.9240619540214539,
+      "learning_rate": 1.8649635036496353e-05,
+      "loss": 0.9556,
+      "step": 511
+    },
+    {
+      "epoch": 0.1869636662406427,
+      "grad_norm": 0.9380460977554321,
+      "learning_rate": 1.8686131386861315e-05,
+      "loss": 0.9839,
+      "step": 512
+    },
+    {
+      "epoch": 0.18732882965126893,
+      "grad_norm": 0.990183413028717,
+      "learning_rate": 1.872262773722628e-05,
+      "loss": 0.9829,
+      "step": 513
+    },
+    {
+      "epoch": 0.1876939930618952,
+      "grad_norm": 0.9844232201576233,
+      "learning_rate": 1.8759124087591244e-05,
+      "loss": 0.9624,
+      "step": 514
+    },
+    {
+      "epoch": 0.18805915647252144,
+      "grad_norm": 1.3360226154327393,
+      "learning_rate": 1.8795620437956205e-05,
+      "loss": 0.9667,
+      "step": 515
+    },
+    {
+      "epoch": 0.1884243198831477,
+      "grad_norm": 1.0527745485305786,
+      "learning_rate": 1.883211678832117e-05,
+      "loss": 0.9873,
+      "step": 516
+    },
+    {
+      "epoch": 0.18878948329377396,
+      "grad_norm": 0.7709869742393494,
+      "learning_rate": 1.8868613138686134e-05,
+      "loss": 0.9957,
+      "step": 517
+    },
+    {
+      "epoch": 0.18915464670440021,
+      "grad_norm": 0.9885143637657166,
+      "learning_rate": 1.8905109489051095e-05,
+      "loss": 0.9785,
+      "step": 518
+    },
+    {
+      "epoch": 0.18951981011502647,
+      "grad_norm": 0.8128122091293335,
+      "learning_rate": 1.894160583941606e-05,
+      "loss": 0.9491,
+      "step": 519
+    },
+    {
+      "epoch": 0.18988497352565273,
+      "grad_norm": 0.9787136912345886,
+      "learning_rate": 1.897810218978102e-05,
+      "loss": 0.936,
+      "step": 520
+    },
+    {
+      "epoch": 0.19025013693627899,
+      "grad_norm": 1.0835014581680298,
+      "learning_rate": 1.901459854014599e-05,
+      "loss": 0.9873,
+      "step": 521
+    },
+    {
+      "epoch": 0.19061530034690524,
+      "grad_norm": 1.2721729278564453,
+      "learning_rate": 1.905109489051095e-05,
+      "loss": 0.9996,
+      "step": 522
+    },
+    {
+      "epoch": 0.1909804637575315,
+      "grad_norm": 1.0898629426956177,
+      "learning_rate": 1.9087591240875915e-05,
+      "loss": 0.9877,
+      "step": 523
+    },
+    {
+      "epoch": 0.19134562716815776,
+      "grad_norm": 1.5092837810516357,
+      "learning_rate": 1.912408759124088e-05,
+      "loss": 0.9923,
+      "step": 524
+    },
+    {
+      "epoch": 0.191710790578784,
+      "grad_norm": 1.787390947341919,
+      "learning_rate": 1.916058394160584e-05,
+      "loss": 0.9877,
+      "step": 525
+    },
+    {
+      "epoch": 0.19207595398941027,
+      "grad_norm": 1.2582017183303833,
+      "learning_rate": 1.9197080291970805e-05,
+      "loss": 0.9637,
+      "step": 526
+    },
+    {
+      "epoch": 0.19244111740003653,
+      "grad_norm": 1.2805448770523071,
+      "learning_rate": 1.9233576642335766e-05,
+      "loss": 0.9714,
+      "step": 527
+    },
+    {
+      "epoch": 0.19280628081066278,
+      "grad_norm": 1.203843116760254,
+      "learning_rate": 1.9270072992700734e-05,
+      "loss": 0.9453,
+      "step": 528
+    },
+    {
+      "epoch": 0.19317144422128904,
+      "grad_norm": 1.221866250038147,
+      "learning_rate": 1.9306569343065695e-05,
+      "loss": 0.9307,
+      "step": 529
+    },
+    {
+      "epoch": 0.19353660763191527,
+      "grad_norm": 1.4659773111343384,
+      "learning_rate": 1.934306569343066e-05,
+      "loss": 0.9092,
+      "step": 530
+    },
+    {
+      "epoch": 0.19390177104254153,
+      "grad_norm": 0.8587760329246521,
+      "learning_rate": 1.937956204379562e-05,
+      "loss": 0.9265,
+      "step": 531
+    },
+    {
+      "epoch": 0.19426693445316778,
+      "grad_norm": 0.7869857549667358,
+      "learning_rate": 1.9416058394160586e-05,
+      "loss": 0.9355,
+      "step": 532
+    },
+    {
+      "epoch": 0.19463209786379404,
+      "grad_norm": 0.9829152226448059,
+      "learning_rate": 1.945255474452555e-05,
+      "loss": 0.9971,
+      "step": 533
+    },
+    {
+      "epoch": 0.1949972612744203,
+      "grad_norm": 0.9914855360984802,
+      "learning_rate": 1.948905109489051e-05,
+      "loss": 0.9432,
+      "step": 534
+    },
+    {
+      "epoch": 0.19536242468504655,
+      "grad_norm": 1.4708786010742188,
+      "learning_rate": 1.9525547445255476e-05,
+      "loss": 0.9668,
+      "step": 535
+    },
+    {
+      "epoch": 0.1957275880956728,
+      "grad_norm": 1.1222752332687378,
+      "learning_rate": 1.956204379562044e-05,
+      "loss": 0.9619,
+      "step": 536
+    },
+    {
+      "epoch": 0.19609275150629907,
+      "grad_norm": 1.9274342060089111,
+      "learning_rate": 1.95985401459854e-05,
+      "loss": 1.0125,
+      "step": 537
+    },
+    {
+      "epoch": 0.19645791491692532,
+      "grad_norm": 0.9744441509246826,
+      "learning_rate": 1.9635036496350366e-05,
+      "loss": 0.9624,
+      "step": 538
+    },
+    {
+      "epoch": 0.19682307832755158,
+      "grad_norm": 1.1587960720062256,
+      "learning_rate": 1.967153284671533e-05,
+      "loss": 0.9648,
+      "step": 539
+    },
+    {
+      "epoch": 0.19718824173817784,
+      "grad_norm": 1.7533305883407593,
+      "learning_rate": 1.9708029197080295e-05,
+      "loss": 0.9557,
+      "step": 540
+    },
+    {
+      "epoch": 0.1975534051488041,
+      "grad_norm": 0.9296097159385681,
+      "learning_rate": 1.9744525547445256e-05,
+      "loss": 0.9346,
+      "step": 541
+    },
+    {
+      "epoch": 0.19791856855943035,
+      "grad_norm": 1.1713056564331055,
+      "learning_rate": 1.978102189781022e-05,
+      "loss": 0.9485,
+      "step": 542
+    },
+    {
+      "epoch": 0.1982837319700566,
+      "grad_norm": 1.076978325843811,
+      "learning_rate": 1.9817518248175186e-05,
+      "loss": 0.9531,
+      "step": 543
+    },
+    {
+      "epoch": 0.19864889538068287,
+      "grad_norm": 1.4528071880340576,
+      "learning_rate": 1.9854014598540147e-05,
+      "loss": 0.9745,
+      "step": 544
+    },
+    {
+      "epoch": 0.19901405879130912,
+      "grad_norm": 1.236617922782898,
+      "learning_rate": 1.989051094890511e-05,
+      "loss": 0.931,
+      "step": 545
+    },
+    {
+      "epoch": 0.19937922220193538,
+      "grad_norm": 0.9910083413124084,
+      "learning_rate": 1.9927007299270073e-05,
+      "loss": 0.994,
+      "step": 546
+    },
+    {
+      "epoch": 0.1997443856125616,
+      "grad_norm": 0.9027132987976074,
+      "learning_rate": 1.9963503649635037e-05,
+      "loss": 0.9836,
+      "step": 547
+    },
+    {
+      "epoch": 0.20010954902318787,
+      "grad_norm": 1.0734672546386719,
+      "learning_rate": 2e-05,
+      "loss": 0.9387,
+      "step": 548
+    },
+    {
+      "epoch": 0.20047471243381412,
+      "grad_norm": 1.1241096258163452,
+      "learning_rate": 2.0036496350364966e-05,
+      "loss": 1.0135,
+      "step": 549
+    },
+    {
+      "epoch": 0.20083987584444038,
+      "grad_norm": 0.925300121307373,
+      "learning_rate": 2.0072992700729927e-05,
+      "loss": 0.9351,
+      "step": 550
+    },
+    {
+      "epoch": 0.20120503925506664,
+      "grad_norm": 0.923231303691864,
+      "learning_rate": 2.0109489051094892e-05,
+      "loss": 0.9557,
+      "step": 551
+    },
+    {
+      "epoch": 0.2015702026656929,
+      "grad_norm": 1.1165447235107422,
+      "learning_rate": 2.0145985401459857e-05,
+      "loss": 0.9553,
+      "step": 552
+    },
+    {
+      "epoch": 0.20193536607631915,
+      "grad_norm": 0.8924594521522522,
+      "learning_rate": 2.0182481751824818e-05,
+      "loss": 0.9332,
+      "step": 553
+    },
+    {
+      "epoch": 0.2023005294869454,
+      "grad_norm": 1.1063874959945679,
+      "learning_rate": 2.0218978102189782e-05,
+      "loss": 0.9163,
+      "step": 554
+    },
+    {
+      "epoch": 0.20266569289757166,
+      "grad_norm": 1.6629401445388794,
+      "learning_rate": 2.0255474452554743e-05,
+      "loss": 0.97,
+      "step": 555
+    },
+    {
+      "epoch": 0.20303085630819792,
+      "grad_norm": 1.107111930847168,
+      "learning_rate": 2.0291970802919708e-05,
+      "loss": 0.9749,
+      "step": 556
+    },
+    {
+      "epoch": 0.20339601971882418,
+      "grad_norm": 1.0518925189971924,
+      "learning_rate": 2.0328467153284676e-05,
+      "loss": 0.923,
+      "step": 557
+    },
+    {
+      "epoch": 0.20376118312945043,
+      "grad_norm": 1.3342797756195068,
+      "learning_rate": 2.0364963503649634e-05,
+      "loss": 0.9469,
+      "step": 558
+    },
+    {
+      "epoch": 0.2041263465400767,
+      "grad_norm": 0.731576681137085,
+      "learning_rate": 2.04014598540146e-05,
+      "loss": 0.9586,
+      "step": 559
+    },
+    {
+      "epoch": 0.20449150995070295,
+      "grad_norm": 1.0917059183120728,
+      "learning_rate": 2.0437956204379566e-05,
+      "loss": 0.9136,
+      "step": 560
+    },
+    {
+      "epoch": 0.2048566733613292,
+      "grad_norm": 1.082904577255249,
+      "learning_rate": 2.0474452554744527e-05,
+      "loss": 0.9521,
+      "step": 561
+    },
+    {
+      "epoch": 0.20522183677195546,
+      "grad_norm": 1.0768007040023804,
+      "learning_rate": 2.0510948905109492e-05,
+      "loss": 0.9315,
+      "step": 562
+    },
+    {
+      "epoch": 0.2055870001825817,
+      "grad_norm": 0.7562431693077087,
+      "learning_rate": 2.0547445255474457e-05,
+      "loss": 0.9064,
+      "step": 563
+    },
+    {
+      "epoch": 0.20595216359320795,
+      "grad_norm": 1.1797900199890137,
+      "learning_rate": 2.0583941605839418e-05,
+      "loss": 0.999,
+      "step": 564
+    },
+    {
+      "epoch": 0.2063173270038342,
+      "grad_norm": 0.9053616523742676,
+      "learning_rate": 2.0620437956204382e-05,
+      "loss": 0.9337,
+      "step": 565
+    },
+    {
+      "epoch": 0.20668249041446046,
+      "grad_norm": 1.3852813243865967,
+      "learning_rate": 2.0656934306569343e-05,
+      "loss": 0.941,
+      "step": 566
+    },
+    {
+      "epoch": 0.20704765382508672,
+      "grad_norm": 0.6356310844421387,
+      "learning_rate": 2.0693430656934308e-05,
+      "loss": 0.9259,
+      "step": 567
+    },
+    {
+      "epoch": 0.20741281723571298,
+      "grad_norm": 0.9954673647880554,
+      "learning_rate": 2.0729927007299273e-05,
+      "loss": 0.9258,
+      "step": 568
+    },
+    {
+      "epoch": 0.20777798064633923,
+      "grad_norm": 1.2683064937591553,
+      "learning_rate": 2.0766423357664234e-05,
+      "loss": 0.975,
+      "step": 569
+    },
+    {
+      "epoch": 0.2081431440569655,
+      "grad_norm": 0.8792716264724731,
+      "learning_rate": 2.08029197080292e-05,
+      "loss": 0.9497,
+      "step": 570
+    },
+    {
+      "epoch": 0.20850830746759175,
+      "grad_norm": 0.9924480319023132,
+      "learning_rate": 2.0839416058394163e-05,
+      "loss": 0.9425,
+      "step": 571
+    },
+    {
+      "epoch": 0.208873470878218,
+      "grad_norm": 0.9855673909187317,
+      "learning_rate": 2.0875912408759124e-05,
+      "loss": 0.9583,
+      "step": 572
+    },
+    {
+      "epoch": 0.20923863428884426,
+      "grad_norm": 0.9010187983512878,
+      "learning_rate": 2.091240875912409e-05,
+      "loss": 0.9147,
+      "step": 573
+    },
+    {
+      "epoch": 0.20960379769947052,
+      "grad_norm": 1.3942712545394897,
+      "learning_rate": 2.0948905109489057e-05,
+      "loss": 0.9437,
+      "step": 574
+    },
+    {
+      "epoch": 0.20996896111009677,
+      "grad_norm": 1.0104106664657593,
+      "learning_rate": 2.0985401459854014e-05,
+      "loss": 0.9528,
+      "step": 575
+    },
+    {
+      "epoch": 0.21033412452072303,
+      "grad_norm": 0.8935278654098511,
+      "learning_rate": 2.1021897810218982e-05,
+      "loss": 0.9368,
+      "step": 576
+    },
+    {
+      "epoch": 0.2106992879313493,
+      "grad_norm": 1.1053450107574463,
+      "learning_rate": 2.105839416058394e-05,
+      "loss": 0.9342,
+      "step": 577
+    },
+    {
+      "epoch": 0.21106445134197555,
+      "grad_norm": 0.8636622428894043,
+      "learning_rate": 2.1094890510948908e-05,
+      "loss": 0.9568,
+      "step": 578
+    },
+    {
+      "epoch": 0.2114296147526018,
+      "grad_norm": 1.6515229940414429,
+      "learning_rate": 2.1131386861313873e-05,
+      "loss": 0.9174,
+      "step": 579
+    },
+    {
+      "epoch": 0.21179477816322803,
+      "grad_norm": 0.9530323147773743,
+      "learning_rate": 2.1167883211678834e-05,
+      "loss": 0.9052,
+      "step": 580
+    },
+    {
+      "epoch": 0.2121599415738543,
+      "grad_norm": 0.7677631974220276,
+      "learning_rate": 2.12043795620438e-05,
+      "loss": 0.9427,
+      "step": 581
+    },
+    {
+      "epoch": 0.21252510498448055,
+      "grad_norm": 1.109664797782898,
+      "learning_rate": 2.1240875912408763e-05,
+      "loss": 0.938,
+      "step": 582
+    },
+    {
+      "epoch": 0.2128902683951068,
+      "grad_norm": 0.9017308950424194,
+      "learning_rate": 2.1277372262773724e-05,
+      "loss": 0.9496,
+      "step": 583
+    },
+    {
+      "epoch": 0.21325543180573306,
+      "grad_norm": 0.8534948825836182,
+      "learning_rate": 2.131386861313869e-05,
+      "loss": 0.8927,
+      "step": 584
+    },
+    {
+      "epoch": 0.21362059521635932,
+      "grad_norm": 0.9706099629402161,
+      "learning_rate": 2.135036496350365e-05,
+      "loss": 0.9166,
+      "step": 585
+    },
+    {
+      "epoch": 0.21398575862698557,
+      "grad_norm": 1.0126210451126099,
+      "learning_rate": 2.1386861313868614e-05,
+      "loss": 0.9696,
+      "step": 586
+    },
+    {
+      "epoch": 0.21435092203761183,
+      "grad_norm": 1.4508851766586304,
+      "learning_rate": 2.142335766423358e-05,
+      "loss": 0.9596,
+      "step": 587
+    },
+    {
+      "epoch": 0.2147160854482381,
+      "grad_norm": 0.8295338153839111,
+      "learning_rate": 2.145985401459854e-05,
+      "loss": 0.897,
+      "step": 588
+    },
+    {
+      "epoch": 0.21508124885886434,
+      "grad_norm": 1.030462384223938,
+      "learning_rate": 2.1496350364963505e-05,
+      "loss": 0.9136,
+      "step": 589
+    },
+    {
+      "epoch": 0.2154464122694906,
+      "grad_norm": 1.3747875690460205,
+      "learning_rate": 2.153284671532847e-05,
+      "loss": 0.9393,
+      "step": 590
+    },
+    {
+      "epoch": 0.21581157568011686,
+      "grad_norm": 0.7962733507156372,
+      "learning_rate": 2.156934306569343e-05,
+      "loss": 0.9292,
+      "step": 591
+    },
+    {
+      "epoch": 0.21617673909074311,
+      "grad_norm": 1.0487091541290283,
+      "learning_rate": 2.1605839416058395e-05,
+      "loss": 0.9493,
+      "step": 592
+    },
+    {
+      "epoch": 0.21654190250136937,
+      "grad_norm": 0.9827638864517212,
+      "learning_rate": 2.1642335766423363e-05,
+      "loss": 0.9614,
+      "step": 593
+    },
+    {
+      "epoch": 0.21690706591199563,
+      "grad_norm": 0.9397901296615601,
+      "learning_rate": 2.167883211678832e-05,
+      "loss": 0.9199,
+      "step": 594
+    },
+    {
+      "epoch": 0.21727222932262188,
+      "grad_norm": 1.07180655002594,
+      "learning_rate": 2.171532846715329e-05,
+      "loss": 0.8998,
+      "step": 595
+    },
+    {
+      "epoch": 0.21763739273324814,
+      "grad_norm": 0.9420148134231567,
+      "learning_rate": 2.1751824817518246e-05,
+      "loss": 0.9264,
+      "step": 596
+    },
+    {
+      "epoch": 0.21800255614387437,
+      "grad_norm": 0.8558598756790161,
+      "learning_rate": 2.1788321167883214e-05,
+      "loss": 0.9282,
+      "step": 597
+    },
+    {
+      "epoch": 0.21836771955450063,
+      "grad_norm": 0.8148871660232544,
+      "learning_rate": 2.182481751824818e-05,
+      "loss": 0.9066,
+      "step": 598
+    },
+    {
+      "epoch": 0.21873288296512688,
+      "grad_norm": 1.2567858695983887,
+      "learning_rate": 2.186131386861314e-05,
+      "loss": 0.9285,
+      "step": 599
+    },
+    {
+      "epoch": 0.21909804637575314,
+      "grad_norm": 1.0364936590194702,
+      "learning_rate": 2.1897810218978105e-05,
+      "loss": 0.9396,
+      "step": 600
+    },
+    {
+      "epoch": 0.2194632097863794,
+      "grad_norm": 1.0940117835998535,
+      "learning_rate": 2.193430656934307e-05,
+      "loss": 0.9309,
+      "step": 601
+    },
+    {
+      "epoch": 0.21982837319700566,
+      "grad_norm": 1.0114707946777344,
+      "learning_rate": 2.197080291970803e-05,
+      "loss": 0.927,
+      "step": 602
+    },
+    {
+      "epoch": 0.2201935366076319,
+      "grad_norm": 1.1807681322097778,
+      "learning_rate": 2.2007299270072995e-05,
+      "loss": 0.9542,
+      "step": 603
+    },
+    {
+      "epoch": 0.22055870001825817,
+      "grad_norm": 1.4283320903778076,
+      "learning_rate": 2.204379562043796e-05,
+      "loss": 0.9482,
+      "step": 604
+    },
+    {
+      "epoch": 0.22092386342888443,
+      "grad_norm": 1.064704418182373,
+      "learning_rate": 2.208029197080292e-05,
+      "loss": 0.8962,
+      "step": 605
+    },
+    {
+      "epoch": 0.22128902683951068,
+      "grad_norm": 1.1571245193481445,
+      "learning_rate": 2.2116788321167885e-05,
+      "loss": 0.976,
+      "step": 606
+    },
+    {
+      "epoch": 0.22165419025013694,
+      "grad_norm": 1.1410144567489624,
+      "learning_rate": 2.2153284671532847e-05,
+      "loss": 0.918,
+      "step": 607
+    },
+    {
+      "epoch": 0.2220193536607632,
+      "grad_norm": 1.0465383529663086,
+      "learning_rate": 2.218978102189781e-05,
+      "loss": 0.923,
+      "step": 608
+    },
+    {
+      "epoch": 0.22238451707138945,
+      "grad_norm": 1.131148338317871,
+      "learning_rate": 2.2226277372262776e-05,
+      "loss": 0.9437,
+      "step": 609
+    },
+    {
+      "epoch": 0.2227496804820157,
+      "grad_norm": 0.7514604926109314,
+      "learning_rate": 2.2262773722627737e-05,
+      "loss": 0.9127,
+      "step": 610
+    },
+    {
+      "epoch": 0.22311484389264197,
+      "grad_norm": 1.0829296112060547,
+      "learning_rate": 2.22992700729927e-05,
+      "loss": 0.9159,
+      "step": 611
+    },
+    {
+      "epoch": 0.22348000730326822,
+      "grad_norm": 1.2224297523498535,
+      "learning_rate": 2.2335766423357666e-05,
+      "loss": 0.9083,
+      "step": 612
+    },
+    {
+      "epoch": 0.22384517071389448,
+      "grad_norm": 1.065311312675476,
+      "learning_rate": 2.2372262773722627e-05,
+      "loss": 0.9169,
+      "step": 613
+    },
+    {
+      "epoch": 0.2242103341245207,
+      "grad_norm": 0.8168937563896179,
+      "learning_rate": 2.2408759124087595e-05,
+      "loss": 0.8895,
+      "step": 614
+    },
+    {
+      "epoch": 0.22457549753514697,
+      "grad_norm": 1.3119653463363647,
+      "learning_rate": 2.244525547445256e-05,
+      "loss": 0.9395,
+      "step": 615
+    },
+    {
+      "epoch": 0.22494066094577322,
+      "grad_norm": 0.9576928615570068,
+      "learning_rate": 2.248175182481752e-05,
+      "loss": 0.95,
+      "step": 616
+    },
+    {
+      "epoch": 0.22530582435639948,
+      "grad_norm": 1.0545897483825684,
+      "learning_rate": 2.2518248175182485e-05,
+      "loss": 0.9564,
+      "step": 617
+    },
+    {
+      "epoch": 0.22567098776702574,
+      "grad_norm": 1.0971013307571411,
+      "learning_rate": 2.2554744525547447e-05,
+      "loss": 0.9091,
+      "step": 618
+    },
+    {
+      "epoch": 0.226036151177652,
+      "grad_norm": 1.3408724069595337,
+      "learning_rate": 2.259124087591241e-05,
+      "loss": 0.9359,
+      "step": 619
+    },
+    {
+      "epoch": 0.22640131458827825,
+      "grad_norm": 0.7940633893013,
+      "learning_rate": 2.2627737226277376e-05,
+      "loss": 0.9159,
+      "step": 620
+    },
+    {
+      "epoch": 0.2267664779989045,
+      "grad_norm": 0.7596388459205627,
+      "learning_rate": 2.2664233576642337e-05,
+      "loss": 0.9062,
+      "step": 621
+    },
+    {
+      "epoch": 0.22713164140953077,
+      "grad_norm": 1.1658217906951904,
+      "learning_rate": 2.27007299270073e-05,
+      "loss": 0.933,
+      "step": 622
+    },
+    {
+      "epoch": 0.22749680482015702,
+      "grad_norm": 1.2590558528900146,
+      "learning_rate": 2.2737226277372266e-05,
+      "loss": 0.9424,
+      "step": 623
+    },
+    {
+      "epoch": 0.22786196823078328,
+      "grad_norm": 0.8723631501197815,
+      "learning_rate": 2.2773722627737227e-05,
+      "loss": 0.9197,
+      "step": 624
+    },
+    {
+      "epoch": 0.22822713164140954,
+      "grad_norm": 0.9985167980194092,
+      "learning_rate": 2.2810218978102192e-05,
+      "loss": 0.9287,
+      "step": 625
+    },
+    {
+      "epoch": 0.2285922950520358,
+      "grad_norm": 1.1019034385681152,
+      "learning_rate": 2.2846715328467156e-05,
+      "loss": 0.911,
+      "step": 626
+    },
+    {
+      "epoch": 0.22895745846266205,
+      "grad_norm": 1.0437474250793457,
+      "learning_rate": 2.2883211678832117e-05,
+      "loss": 0.9119,
+      "step": 627
+    },
+    {
+      "epoch": 0.2293226218732883,
+      "grad_norm": 0.864844799041748,
+      "learning_rate": 2.2919708029197082e-05,
+      "loss": 0.915,
+      "step": 628
+    },
+    {
+      "epoch": 0.22968778528391456,
+      "grad_norm": 0.829508900642395,
+      "learning_rate": 2.2956204379562043e-05,
+      "loss": 0.9077,
+      "step": 629
+    },
+    {
+      "epoch": 0.23005294869454082,
+      "grad_norm": 0.850080668926239,
+      "learning_rate": 2.2992700729927008e-05,
+      "loss": 0.9023,
+      "step": 630
+    },
+    {
+      "epoch": 0.23041811210516705,
+      "grad_norm": 0.8959830403327942,
+      "learning_rate": 2.3029197080291972e-05,
+      "loss": 0.9038,
+      "step": 631
+    },
+    {
+      "epoch": 0.2307832755157933,
+      "grad_norm": 1.3409028053283691,
+      "learning_rate": 2.3065693430656934e-05,
+      "loss": 0.9269,
+      "step": 632
+    },
+    {
+      "epoch": 0.23114843892641956,
+      "grad_norm": 1.0184214115142822,
+      "learning_rate": 2.31021897810219e-05,
+      "loss": 0.9413,
+      "step": 633
+    },
+    {
+      "epoch": 0.23151360233704582,
+      "grad_norm": 1.1121234893798828,
+      "learning_rate": 2.3138686131386866e-05,
+      "loss": 0.9196,
+      "step": 634
+    },
+    {
+      "epoch": 0.23187876574767208,
+      "grad_norm": 0.9996039271354675,
+      "learning_rate": 2.3175182481751827e-05,
+      "loss": 0.9335,
+      "step": 635
+    },
+    {
+      "epoch": 0.23224392915829833,
+      "grad_norm": 1.0234628915786743,
+      "learning_rate": 2.3211678832116792e-05,
+      "loss": 0.9364,
+      "step": 636
+    },
+    {
+      "epoch": 0.2326090925689246,
+      "grad_norm": 1.0570037364959717,
+      "learning_rate": 2.3248175182481756e-05,
+      "loss": 0.8986,
+      "step": 637
+    },
+    {
+      "epoch": 0.23297425597955085,
+      "grad_norm": 0.6851513981819153,
+      "learning_rate": 2.3284671532846718e-05,
+      "loss": 0.8804,
+      "step": 638
+    },
+    {
+      "epoch": 0.2333394193901771,
+      "grad_norm": 1.0824224948883057,
+      "learning_rate": 2.3321167883211682e-05,
+      "loss": 0.8817,
+      "step": 639
+    },
+    {
+      "epoch": 0.23370458280080336,
+      "grad_norm": 1.2433457374572754,
+      "learning_rate": 2.3357664233576643e-05,
+      "loss": 0.921,
+      "step": 640
+    },
+    {
+      "epoch": 0.23406974621142962,
+      "grad_norm": 0.9732486605644226,
+      "learning_rate": 2.3394160583941608e-05,
+      "loss": 0.8934,
+      "step": 641
+    },
+    {
+      "epoch": 0.23443490962205588,
+      "grad_norm": 1.0137306451797485,
+      "learning_rate": 2.3430656934306572e-05,
+      "loss": 0.9434,
+      "step": 642
+    },
+    {
+      "epoch": 0.23480007303268213,
+      "grad_norm": 0.9960076808929443,
+      "learning_rate": 2.3467153284671534e-05,
+      "loss": 0.8903,
+      "step": 643
+    },
+    {
+      "epoch": 0.2351652364433084,
+      "grad_norm": 0.8179641962051392,
+      "learning_rate": 2.3503649635036498e-05,
+      "loss": 0.9148,
+      "step": 644
+    },
+    {
+      "epoch": 0.23553039985393465,
+      "grad_norm": 1.8424533605575562,
+      "learning_rate": 2.3540145985401463e-05,
+      "loss": 0.9077,
+      "step": 645
+    },
+    {
+      "epoch": 0.2358955632645609,
+      "grad_norm": 1.0837489366531372,
+      "learning_rate": 2.3576642335766424e-05,
+      "loss": 0.9108,
+      "step": 646
+    },
+    {
+      "epoch": 0.23626072667518713,
+      "grad_norm": 0.9237055778503418,
+      "learning_rate": 2.361313868613139e-05,
+      "loss": 0.9159,
+      "step": 647
+    },
+    {
+      "epoch": 0.2366258900858134,
+      "grad_norm": 1.215301513671875,
+      "learning_rate": 2.3649635036496353e-05,
+      "loss": 0.9172,
+      "step": 648
+    },
+    {
+      "epoch": 0.23699105349643965,
+      "grad_norm": 1.474677562713623,
+      "learning_rate": 2.3686131386861314e-05,
+      "loss": 0.9196,
+      "step": 649
+    },
+    {
+      "epoch": 0.2373562169070659,
+      "grad_norm": 0.8014124035835266,
+      "learning_rate": 2.372262773722628e-05,
+      "loss": 0.9036,
+      "step": 650
+    },
+    {
+      "epoch": 0.23772138031769216,
+      "grad_norm": 0.9400027394294739,
+      "learning_rate": 2.375912408759124e-05,
+      "loss": 0.8754,
+      "step": 651
+    },
+    {
+      "epoch": 0.23808654372831842,
+      "grad_norm": 1.0393285751342773,
+      "learning_rate": 2.3795620437956204e-05,
+      "loss": 0.9025,
+      "step": 652
+    },
+    {
+      "epoch": 0.23845170713894467,
+      "grad_norm": 1.153727650642395,
+      "learning_rate": 2.3832116788321172e-05,
+      "loss": 0.9031,
+      "step": 653
+    },
+    {
+      "epoch": 0.23881687054957093,
+      "grad_norm": 1.0982905626296997,
+      "learning_rate": 2.3868613138686134e-05,
+      "loss": 0.8793,
+      "step": 654
+    },
+    {
+      "epoch": 0.2391820339601972,
+      "grad_norm": 0.9338880181312561,
+      "learning_rate": 2.3905109489051098e-05,
+      "loss": 0.8768,
+      "step": 655
+    },
+    {
+      "epoch": 0.23954719737082344,
+      "grad_norm": 0.9995929598808289,
+      "learning_rate": 2.3941605839416063e-05,
+      "loss": 0.8976,
+      "step": 656
+    },
+    {
+      "epoch": 0.2399123607814497,
+      "grad_norm": 1.0906516313552856,
+      "learning_rate": 2.3978102189781024e-05,
+      "loss": 0.9115,
+      "step": 657
+    },
+    {
+      "epoch": 0.24027752419207596,
+      "grad_norm": 1.2404183149337769,
+      "learning_rate": 2.401459854014599e-05,
+      "loss": 0.8925,
+      "step": 658
+    },
+    {
+      "epoch": 0.24064268760270222,
+      "grad_norm": 0.8696293234825134,
+      "learning_rate": 2.405109489051095e-05,
+      "loss": 0.8966,
+      "step": 659
+    },
+    {
+      "epoch": 0.24100785101332847,
+      "grad_norm": 0.8563190698623657,
+      "learning_rate": 2.4087591240875914e-05,
+      "loss": 0.892,
+      "step": 660
+    },
+    {
+      "epoch": 0.24137301442395473,
+      "grad_norm": 0.8124135732650757,
+      "learning_rate": 2.412408759124088e-05,
+      "loss": 0.9028,
+      "step": 661
+    },
+    {
+      "epoch": 0.241738177834581,
+      "grad_norm": 0.9765145778656006,
+      "learning_rate": 2.416058394160584e-05,
+      "loss": 0.9186,
+      "step": 662
+    },
+    {
+      "epoch": 0.24210334124520724,
+      "grad_norm": 1.0169612169265747,
+      "learning_rate": 2.4197080291970805e-05,
+      "loss": 0.8901,
+      "step": 663
+    },
+    {
+      "epoch": 0.24246850465583347,
+      "grad_norm": 0.8080570697784424,
+      "learning_rate": 2.423357664233577e-05,
+      "loss": 0.8871,
+      "step": 664
+    },
+    {
+      "epoch": 0.24283366806645973,
+      "grad_norm": 1.1883140802383423,
+      "learning_rate": 2.427007299270073e-05,
+      "loss": 0.9015,
+      "step": 665
+    },
+    {
+      "epoch": 0.24319883147708599,
+      "grad_norm": 0.9028333425521851,
+      "learning_rate": 2.4306569343065695e-05,
+      "loss": 0.8638,
+      "step": 666
+    },
+    {
+      "epoch": 0.24356399488771224,
+      "grad_norm": 0.7207622528076172,
+      "learning_rate": 2.434306569343066e-05,
+      "loss": 0.8868,
+      "step": 667
+    },
+    {
+      "epoch": 0.2439291582983385,
+      "grad_norm": 1.3190016746520996,
+      "learning_rate": 2.437956204379562e-05,
+      "loss": 0.9341,
+      "step": 668
+    },
+    {
+      "epoch": 0.24429432170896476,
+      "grad_norm": 1.3500337600708008,
+      "learning_rate": 2.4416058394160585e-05,
+      "loss": 0.9249,
+      "step": 669
+    },
+    {
+      "epoch": 0.244659485119591,
+      "grad_norm": 1.042157769203186,
+      "learning_rate": 2.4452554744525546e-05,
+      "loss": 0.8414,
+      "step": 670
+    },
+    {
+      "epoch": 0.24502464853021727,
+      "grad_norm": 1.0846127271652222,
+      "learning_rate": 2.448905109489051e-05,
+      "loss": 0.8883,
+      "step": 671
+    },
+    {
+      "epoch": 0.24538981194084353,
+      "grad_norm": 1.4956647157669067,
+      "learning_rate": 2.452554744525548e-05,
+      "loss": 0.8802,
+      "step": 672
+    },
+    {
+      "epoch": 0.24575497535146978,
+      "grad_norm": 1.2371827363967896,
+      "learning_rate": 2.4562043795620437e-05,
+      "loss": 0.91,
+      "step": 673
+    },
+    {
+      "epoch": 0.24612013876209604,
+      "grad_norm": 1.2039473056793213,
+      "learning_rate": 2.4598540145985405e-05,
+      "loss": 0.8591,
+      "step": 674
+    },
+    {
+      "epoch": 0.2464853021727223,
+      "grad_norm": 0.9505608081817627,
+      "learning_rate": 2.463503649635037e-05,
+      "loss": 0.9138,
+      "step": 675
+    },
+    {
+      "epoch": 0.24685046558334856,
+      "grad_norm": 1.2837300300598145,
+      "learning_rate": 2.467153284671533e-05,
+      "loss": 0.9059,
+      "step": 676
+    },
+    {
+      "epoch": 0.2472156289939748,
+      "grad_norm": 1.084855079650879,
+      "learning_rate": 2.4708029197080295e-05,
+      "loss": 0.9113,
+      "step": 677
+    },
+    {
+      "epoch": 0.24758079240460107,
+      "grad_norm": 0.7835924029350281,
+      "learning_rate": 2.474452554744526e-05,
+      "loss": 0.8977,
+      "step": 678
+    },
+    {
+      "epoch": 0.24794595581522733,
+      "grad_norm": 0.8942123651504517,
+      "learning_rate": 2.478102189781022e-05,
+      "loss": 0.8754,
+      "step": 679
+    },
+    {
+      "epoch": 0.24831111922585358,
+      "grad_norm": 0.9250959753990173,
+      "learning_rate": 2.4817518248175185e-05,
+      "loss": 0.8692,
+      "step": 680
+    },
+    {
+      "epoch": 0.2486762826364798,
+      "grad_norm": 1.7833359241485596,
+      "learning_rate": 2.4854014598540146e-05,
+      "loss": 0.9188,
+      "step": 681
+    },
+    {
+      "epoch": 0.24904144604710607,
+      "grad_norm": 1.1274373531341553,
+      "learning_rate": 2.489051094890511e-05,
+      "loss": 0.8969,
+      "step": 682
+    },
+    {
+      "epoch": 0.24940660945773233,
+      "grad_norm": 0.9775779247283936,
+      "learning_rate": 2.4927007299270075e-05,
+      "loss": 0.8766,
+      "step": 683
+    },
+    {
+      "epoch": 0.24977177286835858,
+      "grad_norm": 0.8695776462554932,
+      "learning_rate": 2.4963503649635037e-05,
+      "loss": 0.9164,
+      "step": 684
+    },
+    {
+      "epoch": 0.25013693627898487,
+      "grad_norm": 0.6799452900886536,
+      "learning_rate": 2.5e-05,
+      "loss": 0.8776,
+      "step": 685
+    },
+    {
+      "epoch": 0.2505020996896111,
+      "grad_norm": 0.9650082588195801,
+      "learning_rate": 2.5036496350364966e-05,
+      "loss": 0.9207,
+      "step": 686
+    },
+    {
+      "epoch": 0.2508672631002374,
+      "grad_norm": 1.0292118787765503,
+      "learning_rate": 2.5072992700729927e-05,
+      "loss": 0.9307,
+      "step": 687
+    },
+    {
+      "epoch": 0.25123242651086364,
+      "grad_norm": 1.9373090267181396,
+      "learning_rate": 2.510948905109489e-05,
+      "loss": 0.8939,
+      "step": 688
+    },
+    {
+      "epoch": 0.25159758992148984,
+      "grad_norm": 1.2657979726791382,
+      "learning_rate": 2.514598540145986e-05,
+      "loss": 0.8967,
+      "step": 689
+    },
+    {
+      "epoch": 0.2519627533321161,
+      "grad_norm": 0.9416601657867432,
+      "learning_rate": 2.5182481751824817e-05,
+      "loss": 0.8701,
+      "step": 690
+    },
+    {
+      "epoch": 0.25232791674274235,
+      "grad_norm": 1.0587490797042847,
+      "learning_rate": 2.5218978102189785e-05,
+      "loss": 0.9169,
+      "step": 691
+    },
+    {
+      "epoch": 0.2526930801533686,
+      "grad_norm": 0.8510803580284119,
+      "learning_rate": 2.5255474452554743e-05,
+      "loss": 0.9108,
+      "step": 692
+    },
+    {
+      "epoch": 0.25305824356399487,
+      "grad_norm": 0.8927858471870422,
+      "learning_rate": 2.529197080291971e-05,
+      "loss": 0.8817,
+      "step": 693
+    },
+    {
+      "epoch": 0.2534234069746211,
+      "grad_norm": 1.1475845575332642,
+      "learning_rate": 2.5328467153284675e-05,
+      "loss": 0.8895,
+      "step": 694
+    },
+    {
+      "epoch": 0.2537885703852474,
+      "grad_norm": 1.0652403831481934,
+      "learning_rate": 2.5364963503649637e-05,
+      "loss": 0.8901,
+      "step": 695
+    },
+    {
+      "epoch": 0.25415373379587364,
+      "grad_norm": 0.7411143779754639,
+      "learning_rate": 2.54014598540146e-05,
+      "loss": 0.8696,
+      "step": 696
+    },
+    {
+      "epoch": 0.2545188972064999,
+      "grad_norm": 0.870008111000061,
+      "learning_rate": 2.5437956204379566e-05,
+      "loss": 0.8942,
+      "step": 697
+    },
+    {
+      "epoch": 0.25488406061712615,
+      "grad_norm": 0.9908129572868347,
+      "learning_rate": 2.5474452554744527e-05,
+      "loss": 0.8719,
+      "step": 698
+    },
+    {
+      "epoch": 0.2552492240277524,
+      "grad_norm": 2.048208475112915,
+      "learning_rate": 2.551094890510949e-05,
+      "loss": 0.9009,
+      "step": 699
+    },
+    {
+      "epoch": 0.25561438743837867,
+      "grad_norm": 0.8549103140830994,
+      "learning_rate": 2.5547445255474456e-05,
+      "loss": 0.8794,
+      "step": 700
+    },
+    {
+      "epoch": 0.2559795508490049,
+      "grad_norm": 1.0178098678588867,
+      "learning_rate": 2.5583941605839417e-05,
+      "loss": 0.8398,
+      "step": 701
+    },
+    {
+      "epoch": 0.2563447142596312,
+      "grad_norm": 0.9158284068107605,
+      "learning_rate": 2.5620437956204382e-05,
+      "loss": 0.8771,
+      "step": 702
+    },
+    {
+      "epoch": 0.25670987767025744,
+      "grad_norm": 0.9023451805114746,
+      "learning_rate": 2.5656934306569343e-05,
+      "loss": 0.907,
+      "step": 703
+    },
+    {
+      "epoch": 0.2570750410808837,
+      "grad_norm": 1.163403868675232,
+      "learning_rate": 2.5693430656934308e-05,
+      "loss": 0.8936,
+      "step": 704
+    },
+    {
+      "epoch": 0.25744020449150995,
+      "grad_norm": 1.1924737691879272,
+      "learning_rate": 2.5729927007299272e-05,
+      "loss": 0.9314,
+      "step": 705
+    },
+    {
+      "epoch": 0.2578053679021362,
+      "grad_norm": 1.0279207229614258,
+      "learning_rate": 2.5766423357664233e-05,
+      "loss": 0.8962,
+      "step": 706
+    },
+    {
+      "epoch": 0.25817053131276246,
+      "grad_norm": 1.5878379344940186,
+      "learning_rate": 2.5802919708029198e-05,
+      "loss": 0.9159,
+      "step": 707
+    },
+    {
+      "epoch": 0.2585356947233887,
+      "grad_norm": 0.899364709854126,
+      "learning_rate": 2.5839416058394166e-05,
+      "loss": 0.8892,
+      "step": 708
+    },
+    {
+      "epoch": 0.258900858134015,
+      "grad_norm": 1.147362470626831,
+      "learning_rate": 2.5875912408759124e-05,
+      "loss": 0.8831,
+      "step": 709
+    },
+    {
+      "epoch": 0.25926602154464123,
+      "grad_norm": 1.191892385482788,
+      "learning_rate": 2.591240875912409e-05,
+      "loss": 0.9058,
+      "step": 710
+    },
+    {
+      "epoch": 0.2596311849552675,
+      "grad_norm": 0.8700737357139587,
+      "learning_rate": 2.5948905109489056e-05,
+      "loss": 0.8944,
+      "step": 711
+    },
+    {
+      "epoch": 0.25999634836589375,
+      "grad_norm": 0.8057035803794861,
+      "learning_rate": 2.5985401459854017e-05,
+      "loss": 0.9044,
+      "step": 712
+    },
+    {
+      "epoch": 0.26036151177652,
+      "grad_norm": 1.1704809665679932,
+      "learning_rate": 2.6021897810218982e-05,
+      "loss": 0.8702,
+      "step": 713
+    },
+    {
+      "epoch": 0.26072667518714626,
+      "grad_norm": 1.2193748950958252,
+      "learning_rate": 2.6058394160583943e-05,
+      "loss": 0.8966,
+      "step": 714
+    },
+    {
+      "epoch": 0.2610918385977725,
+      "grad_norm": 1.676171898841858,
+      "learning_rate": 2.6094890510948908e-05,
+      "loss": 0.8789,
+      "step": 715
+    },
+    {
+      "epoch": 0.2614570020083988,
+      "grad_norm": 1.167014241218567,
+      "learning_rate": 2.6131386861313872e-05,
+      "loss": 0.9003,
+      "step": 716
+    },
+    {
+      "epoch": 0.26182216541902503,
+      "grad_norm": 1.1899787187576294,
+      "learning_rate": 2.6167883211678833e-05,
+      "loss": 0.8865,
+      "step": 717
+    },
+    {
+      "epoch": 0.2621873288296513,
+      "grad_norm": 1.2309730052947998,
+      "learning_rate": 2.6204379562043798e-05,
+      "loss": 0.8909,
+      "step": 718
+    },
+    {
+      "epoch": 0.26255249224027755,
+      "grad_norm": 1.043740153312683,
+      "learning_rate": 2.6240875912408762e-05,
+      "loss": 0.8604,
+      "step": 719
+    },
+    {
+      "epoch": 0.2629176556509038,
+      "grad_norm": 0.9882286190986633,
+      "learning_rate": 2.6277372262773724e-05,
+      "loss": 0.8354,
+      "step": 720
+    },
+    {
+      "epoch": 0.26328281906153006,
+      "grad_norm": 1.0266289710998535,
+      "learning_rate": 2.6313868613138688e-05,
+      "loss": 0.8959,
+      "step": 721
+    },
+    {
+      "epoch": 0.2636479824721563,
+      "grad_norm": 1.2218525409698486,
+      "learning_rate": 2.635036496350365e-05,
+      "loss": 0.9207,
+      "step": 722
+    },
+    {
+      "epoch": 0.2640131458827825,
+      "grad_norm": 1.0779708623886108,
+      "learning_rate": 2.6386861313868614e-05,
+      "loss": 0.8759,
+      "step": 723
+    },
+    {
+      "epoch": 0.2643783092934088,
+      "grad_norm": 1.7302335500717163,
+      "learning_rate": 2.642335766423358e-05,
+      "loss": 0.8741,
+      "step": 724
+    },
+    {
+      "epoch": 0.26474347270403503,
+      "grad_norm": 1.0769745111465454,
+      "learning_rate": 2.645985401459854e-05,
+      "loss": 0.865,
+      "step": 725
+    },
+    {
+      "epoch": 0.2651086361146613,
+      "grad_norm": 1.138060212135315,
+      "learning_rate": 2.6496350364963504e-05,
+      "loss": 0.8702,
+      "step": 726
+    },
+    {
+      "epoch": 0.26547379952528755,
+      "grad_norm": 0.9084373116493225,
+      "learning_rate": 2.6532846715328472e-05,
+      "loss": 0.8986,
+      "step": 727
+    },
+    {
+      "epoch": 0.2658389629359138,
+      "grad_norm": 0.7882066369056702,
+      "learning_rate": 2.656934306569343e-05,
+      "loss": 0.8773,
+      "step": 728
+    },
+    {
+      "epoch": 0.26620412634654006,
+      "grad_norm": 0.9498522281646729,
+      "learning_rate": 2.6605839416058398e-05,
+      "loss": 0.8704,
+      "step": 729
+    },
+    {
+      "epoch": 0.2665692897571663,
+      "grad_norm": 0.8272795081138611,
+      "learning_rate": 2.6642335766423363e-05,
+      "loss": 0.8909,
+      "step": 730
+    },
+    {
+      "epoch": 0.2669344531677926,
+      "grad_norm": 0.9073951244354248,
+      "learning_rate": 2.6678832116788324e-05,
+      "loss": 0.8344,
+      "step": 731
+    },
+    {
+      "epoch": 0.26729961657841883,
+      "grad_norm": 1.3860666751861572,
+      "learning_rate": 2.6715328467153288e-05,
+      "loss": 0.8937,
+      "step": 732
+    },
+    {
+      "epoch": 0.2676647799890451,
+      "grad_norm": 0.8217914700508118,
+      "learning_rate": 2.675182481751825e-05,
+      "loss": 0.8906,
+      "step": 733
+    },
+    {
+      "epoch": 0.26802994339967134,
+      "grad_norm": 0.9651821851730347,
+      "learning_rate": 2.6788321167883214e-05,
+      "loss": 0.9131,
+      "step": 734
+    },
+    {
+      "epoch": 0.2683951068102976,
+      "grad_norm": 0.8943136930465698,
+      "learning_rate": 2.682481751824818e-05,
+      "loss": 0.887,
+      "step": 735
+    },
+    {
+      "epoch": 0.26876027022092386,
+      "grad_norm": 0.8883320093154907,
+      "learning_rate": 2.686131386861314e-05,
+      "loss": 0.8606,
+      "step": 736
+    },
+    {
+      "epoch": 0.2691254336315501,
+      "grad_norm": 1.6274850368499756,
+      "learning_rate": 2.6897810218978104e-05,
+      "loss": 0.8727,
+      "step": 737
+    },
+    {
+      "epoch": 0.26949059704217637,
+      "grad_norm": 0.9518415927886963,
+      "learning_rate": 2.693430656934307e-05,
+      "loss": 0.8876,
+      "step": 738
+    },
+    {
+      "epoch": 0.26985576045280263,
+      "grad_norm": 0.7219151854515076,
+      "learning_rate": 2.697080291970803e-05,
+      "loss": 0.8413,
+      "step": 739
+    },
+    {
+      "epoch": 0.2702209238634289,
+      "grad_norm": 0.9340730905532837,
+      "learning_rate": 2.7007299270072995e-05,
+      "loss": 0.8448,
+      "step": 740
+    },
+    {
+      "epoch": 0.27058608727405514,
+      "grad_norm": 0.9046608209609985,
+      "learning_rate": 2.704379562043796e-05,
+      "loss": 0.8837,
+      "step": 741
+    },
+    {
+      "epoch": 0.2709512506846814,
+      "grad_norm": 1.0275894403457642,
+      "learning_rate": 2.708029197080292e-05,
+      "loss": 0.8754,
+      "step": 742
+    },
+    {
+      "epoch": 0.27131641409530766,
+      "grad_norm": 0.8216059803962708,
+      "learning_rate": 2.7116788321167885e-05,
+      "loss": 0.8639,
+      "step": 743
+    },
+    {
+      "epoch": 0.2716815775059339,
+      "grad_norm": 1.0811870098114014,
+      "learning_rate": 2.7153284671532846e-05,
+      "loss": 0.8712,
+      "step": 744
+    },
+    {
+      "epoch": 0.27204674091656017,
+      "grad_norm": 0.9865444302558899,
+      "learning_rate": 2.718978102189781e-05,
+      "loss": 0.8939,
+      "step": 745
+    },
+    {
+      "epoch": 0.2724119043271864,
+      "grad_norm": 1.233981966972351,
+      "learning_rate": 2.722627737226278e-05,
+      "loss": 0.8718,
+      "step": 746
+    },
+    {
+      "epoch": 0.2727770677378127,
+      "grad_norm": 0.8024433851242065,
+      "learning_rate": 2.7262773722627736e-05,
+      "loss": 0.8561,
+      "step": 747
+    },
+    {
+      "epoch": 0.27314223114843894,
+      "grad_norm": 0.9161458611488342,
+      "learning_rate": 2.7299270072992704e-05,
+      "loss": 0.8787,
+      "step": 748
+    },
+    {
+      "epoch": 0.2735073945590652,
+      "grad_norm": 0.9415838122367859,
+      "learning_rate": 2.733576642335767e-05,
+      "loss": 0.8699,
+      "step": 749
+    },
+    {
+      "epoch": 0.27387255796969145,
+      "grad_norm": 1.0449683666229248,
+      "learning_rate": 2.737226277372263e-05,
+      "loss": 0.859,
+      "step": 750
+    },
+    {
+      "epoch": 0.2742377213803177,
+      "grad_norm": 0.8127017617225647,
+      "learning_rate": 2.7408759124087595e-05,
+      "loss": 0.8369,
+      "step": 751
+    },
+    {
+      "epoch": 0.27460288479094397,
+      "grad_norm": 0.7522885203361511,
+      "learning_rate": 2.744525547445256e-05,
+      "loss": 0.8561,
+      "step": 752
+    },
+    {
+      "epoch": 0.2749680482015702,
+      "grad_norm": 1.016823410987854,
+      "learning_rate": 2.748175182481752e-05,
+      "loss": 0.8719,
+      "step": 753
+    },
+    {
+      "epoch": 0.2753332116121965,
+      "grad_norm": 1.4573849439620972,
+      "learning_rate": 2.7518248175182485e-05,
+      "loss": 0.8424,
+      "step": 754
+    },
+    {
+      "epoch": 0.27569837502282274,
+      "grad_norm": 0.8680856227874756,
+      "learning_rate": 2.7554744525547446e-05,
+      "loss": 0.8884,
+      "step": 755
+    },
+    {
+      "epoch": 0.276063538433449,
+      "grad_norm": 1.0039207935333252,
+      "learning_rate": 2.759124087591241e-05,
+      "loss": 0.8804,
+      "step": 756
+    },
+    {
+      "epoch": 0.2764287018440752,
+      "grad_norm": 1.1460931301116943,
+      "learning_rate": 2.7627737226277375e-05,
+      "loss": 0.8829,
+      "step": 757
+    },
+    {
+      "epoch": 0.27679386525470145,
+      "grad_norm": 0.9026143550872803,
+      "learning_rate": 2.7664233576642336e-05,
+      "loss": 0.8467,
+      "step": 758
+    },
+    {
+      "epoch": 0.2771590286653277,
+      "grad_norm": 1.1995723247528076,
+      "learning_rate": 2.77007299270073e-05,
+      "loss": 0.8937,
+      "step": 759
+    },
+    {
+      "epoch": 0.27752419207595397,
+      "grad_norm": 0.8786799311637878,
+      "learning_rate": 2.7737226277372266e-05,
+      "loss": 0.8345,
+      "step": 760
+    },
+    {
+      "epoch": 0.2778893554865802,
+      "grad_norm": 1.2147480249404907,
+      "learning_rate": 2.7773722627737227e-05,
+      "loss": 0.8653,
+      "step": 761
+    },
+    {
+      "epoch": 0.2782545188972065,
+      "grad_norm": 1.0637128353118896,
+      "learning_rate": 2.781021897810219e-05,
+      "loss": 0.8877,
+      "step": 762
+    },
+    {
+      "epoch": 0.27861968230783274,
+      "grad_norm": 1.7768861055374146,
+      "learning_rate": 2.784671532846716e-05,
+      "loss": 0.8447,
+      "step": 763
+    },
+    {
+      "epoch": 0.278984845718459,
+      "grad_norm": 0.7212632894515991,
+      "learning_rate": 2.7883211678832117e-05,
+      "loss": 0.8678,
+      "step": 764
+    },
+    {
+      "epoch": 0.27935000912908525,
+      "grad_norm": 1.1126927137374878,
+      "learning_rate": 2.7919708029197085e-05,
+      "loss": 0.9049,
+      "step": 765
+    },
+    {
+      "epoch": 0.2797151725397115,
+      "grad_norm": 0.9064791798591614,
+      "learning_rate": 2.7956204379562043e-05,
+      "loss": 0.8501,
+      "step": 766
+    },
+    {
+      "epoch": 0.28008033595033777,
+      "grad_norm": 0.9776251912117004,
+      "learning_rate": 2.799270072992701e-05,
+      "loss": 0.8862,
+      "step": 767
+    },
+    {
+      "epoch": 0.280445499360964,
+      "grad_norm": 1.2621984481811523,
+      "learning_rate": 2.8029197080291975e-05,
+      "loss": 0.8602,
+      "step": 768
+    },
+    {
+      "epoch": 0.2808106627715903,
+      "grad_norm": 0.816348671913147,
+      "learning_rate": 2.8065693430656936e-05,
+      "loss": 0.8169,
+      "step": 769
+    },
+    {
+      "epoch": 0.28117582618221654,
+      "grad_norm": 1.1793899536132812,
+      "learning_rate": 2.81021897810219e-05,
+      "loss": 0.8347,
+      "step": 770
+    },
+    {
+      "epoch": 0.2815409895928428,
+      "grad_norm": 0.8727941513061523,
+      "learning_rate": 2.8138686131386866e-05,
+      "loss": 0.8544,
+      "step": 771
+    },
+    {
+      "epoch": 0.28190615300346905,
+      "grad_norm": 1.2146086692810059,
+      "learning_rate": 2.8175182481751827e-05,
+      "loss": 0.8943,
+      "step": 772
+    },
+    {
+      "epoch": 0.2822713164140953,
+      "grad_norm": 0.974018931388855,
+      "learning_rate": 2.821167883211679e-05,
+      "loss": 0.8328,
+      "step": 773
+    },
+    {
+      "epoch": 0.28263647982472156,
+      "grad_norm": 1.0131561756134033,
+      "learning_rate": 2.8248175182481756e-05,
+      "loss": 0.8574,
+      "step": 774
+    },
+    {
+      "epoch": 0.2830016432353478,
+      "grad_norm": 1.2922457456588745,
+      "learning_rate": 2.8284671532846717e-05,
+      "loss": 0.8866,
+      "step": 775
+    },
+    {
+      "epoch": 0.2833668066459741,
+      "grad_norm": 1.044386863708496,
+      "learning_rate": 2.832116788321168e-05,
+      "loss": 0.8892,
+      "step": 776
+    },
+    {
+      "epoch": 0.28373197005660034,
+      "grad_norm": 0.9163618087768555,
+      "learning_rate": 2.8357664233576643e-05,
+      "loss": 0.8468,
+      "step": 777
+    },
+    {
+      "epoch": 0.2840971334672266,
+      "grad_norm": 0.8878577947616577,
+      "learning_rate": 2.8394160583941607e-05,
+      "loss": 0.8496,
+      "step": 778
+    },
+    {
+      "epoch": 0.28446229687785285,
+      "grad_norm": 1.042579174041748,
+      "learning_rate": 2.8430656934306572e-05,
+      "loss": 0.8405,
+      "step": 779
+    },
+    {
+      "epoch": 0.2848274602884791,
+      "grad_norm": 1.7790285348892212,
+      "learning_rate": 2.8467153284671533e-05,
+      "loss": 0.8541,
+      "step": 780
+    },
+    {
+      "epoch": 0.28519262369910536,
+      "grad_norm": 1.053520679473877,
+      "learning_rate": 2.8503649635036498e-05,
+      "loss": 0.8567,
+      "step": 781
+    },
+    {
+      "epoch": 0.2855577871097316,
+      "grad_norm": 0.8160289525985718,
+      "learning_rate": 2.8540145985401466e-05,
+      "loss": 0.8691,
+      "step": 782
+    },
+    {
+      "epoch": 0.2859229505203579,
+      "grad_norm": 0.8448201417922974,
+      "learning_rate": 2.8576642335766423e-05,
+      "loss": 0.8516,
+      "step": 783
+    },
+    {
+      "epoch": 0.28628811393098413,
+      "grad_norm": 1.0939788818359375,
+      "learning_rate": 2.861313868613139e-05,
+      "loss": 0.8488,
+      "step": 784
+    },
+    {
+      "epoch": 0.2866532773416104,
+      "grad_norm": 0.9668917059898376,
+      "learning_rate": 2.8649635036496356e-05,
+      "loss": 0.8357,
+      "step": 785
+    },
+    {
+      "epoch": 0.28701844075223665,
+      "grad_norm": 1.2287737131118774,
+      "learning_rate": 2.8686131386861317e-05,
+      "loss": 0.8245,
+      "step": 786
+    },
+    {
+      "epoch": 0.2873836041628629,
+      "grad_norm": 0.9395943284034729,
+      "learning_rate": 2.872262773722628e-05,
+      "loss": 0.8193,
+      "step": 787
+    },
+    {
+      "epoch": 0.28774876757348916,
+      "grad_norm": 0.8268449902534485,
+      "learning_rate": 2.8759124087591243e-05,
+      "loss": 0.8689,
+      "step": 788
+    },
+    {
+      "epoch": 0.2881139309841154,
+      "grad_norm": 1.159216284751892,
+      "learning_rate": 2.8795620437956207e-05,
+      "loss": 0.8751,
+      "step": 789
+    },
+    {
+      "epoch": 0.2884790943947416,
+      "grad_norm": 0.7923019528388977,
+      "learning_rate": 2.8832116788321172e-05,
+      "loss": 0.873,
+      "step": 790
+    },
+    {
+      "epoch": 0.2888442578053679,
+      "grad_norm": 0.8604425191879272,
+      "learning_rate": 2.8868613138686133e-05,
+      "loss": 0.8672,
+      "step": 791
+    },
+    {
+      "epoch": 0.28920942121599413,
+      "grad_norm": 1.0001811981201172,
+      "learning_rate": 2.8905109489051098e-05,
+      "loss": 0.827,
+      "step": 792
+    },
+    {
+      "epoch": 0.2895745846266204,
+      "grad_norm": 0.8650938272476196,
+      "learning_rate": 2.8941605839416062e-05,
+      "loss": 0.8575,
+      "step": 793
+    },
+    {
+      "epoch": 0.28993974803724665,
+      "grad_norm": 0.9869515895843506,
+      "learning_rate": 2.8978102189781023e-05,
+      "loss": 0.8517,
+      "step": 794
+    },
+    {
+      "epoch": 0.2903049114478729,
+      "grad_norm": 1.1096405982971191,
+      "learning_rate": 2.9014598540145988e-05,
+      "loss": 0.8355,
+      "step": 795
+    },
+    {
+      "epoch": 0.29067007485849916,
+      "grad_norm": 1.2804101705551147,
+      "learning_rate": 2.905109489051095e-05,
+      "loss": 0.8984,
+      "step": 796
+    },
+    {
+      "epoch": 0.2910352382691254,
+      "grad_norm": 1.031632423400879,
+      "learning_rate": 2.9087591240875914e-05,
+      "loss": 0.8362,
+      "step": 797
+    },
+    {
+      "epoch": 0.2914004016797517,
+      "grad_norm": 1.4014887809753418,
+      "learning_rate": 2.912408759124088e-05,
+      "loss": 0.8417,
+      "step": 798
+    },
+    {
+      "epoch": 0.29176556509037793,
+      "grad_norm": 0.8653863072395325,
+      "learning_rate": 2.916058394160584e-05,
+      "loss": 0.8671,
+      "step": 799
+    },
+    {
+      "epoch": 0.2921307285010042,
+      "grad_norm": 1.3845256567001343,
+      "learning_rate": 2.9197080291970804e-05,
+      "loss": 0.8916,
+      "step": 800
+    },
+    {
+      "epoch": 0.29249589191163045,
+      "grad_norm": 0.8273415565490723,
+      "learning_rate": 2.923357664233577e-05,
+      "loss": 0.8458,
+      "step": 801
+    },
+    {
+      "epoch": 0.2928610553222567,
+      "grad_norm": 0.8840750455856323,
+      "learning_rate": 2.927007299270073e-05,
+      "loss": 0.8491,
+      "step": 802
+    },
+    {
+      "epoch": 0.29322621873288296,
+      "grad_norm": 0.972186267375946,
+      "learning_rate": 2.9306569343065698e-05,
+      "loss": 0.83,
+      "step": 803
+    },
+    {
+      "epoch": 0.2935913821435092,
+      "grad_norm": 1.119389533996582,
+      "learning_rate": 2.9343065693430662e-05,
+      "loss": 0.8501,
+      "step": 804
+    },
+    {
+      "epoch": 0.2939565455541355,
+      "grad_norm": 0.8941442966461182,
+      "learning_rate": 2.9379562043795624e-05,
+      "loss": 0.8246,
+      "step": 805
+    },
+    {
+      "epoch": 0.29432170896476173,
+      "grad_norm": 0.853405237197876,
+      "learning_rate": 2.9416058394160588e-05,
+      "loss": 0.8748,
+      "step": 806
+    },
+    {
+      "epoch": 0.294686872375388,
+      "grad_norm": 1.1405614614486694,
+      "learning_rate": 2.945255474452555e-05,
+      "loss": 0.8303,
+      "step": 807
+    },
+    {
+      "epoch": 0.29505203578601424,
+      "grad_norm": 1.0696759223937988,
+      "learning_rate": 2.9489051094890514e-05,
+      "loss": 0.8627,
+      "step": 808
+    },
+    {
+      "epoch": 0.2954171991966405,
+      "grad_norm": 1.1660170555114746,
+      "learning_rate": 2.952554744525548e-05,
+      "loss": 0.8439,
+      "step": 809
+    },
+    {
+      "epoch": 0.29578236260726676,
+      "grad_norm": 1.073717713356018,
+      "learning_rate": 2.956204379562044e-05,
+      "loss": 0.8474,
+      "step": 810
+    },
+    {
+      "epoch": 0.296147526017893,
+      "grad_norm": 0.9319285750389099,
+      "learning_rate": 2.9598540145985404e-05,
+      "loss": 0.844,
+      "step": 811
+    },
+    {
+      "epoch": 0.29651268942851927,
+      "grad_norm": 1.0393621921539307,
+      "learning_rate": 2.963503649635037e-05,
+      "loss": 0.853,
+      "step": 812
+    },
+    {
+      "epoch": 0.29687785283914553,
+      "grad_norm": 1.3637595176696777,
+      "learning_rate": 2.967153284671533e-05,
+      "loss": 0.9049,
+      "step": 813
+    },
+    {
+      "epoch": 0.2972430162497718,
+      "grad_norm": 1.4779207706451416,
+      "learning_rate": 2.9708029197080294e-05,
+      "loss": 0.826,
+      "step": 814
+    },
+    {
+      "epoch": 0.29760817966039804,
+      "grad_norm": 0.811496913433075,
+      "learning_rate": 2.974452554744526e-05,
+      "loss": 0.8352,
+      "step": 815
+    },
+    {
+      "epoch": 0.2979733430710243,
+      "grad_norm": 1.0108301639556885,
+      "learning_rate": 2.978102189781022e-05,
+      "loss": 0.851,
+      "step": 816
+    },
+    {
+      "epoch": 0.29833850648165056,
+      "grad_norm": 1.167640209197998,
+      "learning_rate": 2.9817518248175185e-05,
+      "loss": 0.8551,
+      "step": 817
+    },
+    {
+      "epoch": 0.2987036698922768,
+      "grad_norm": 1.0643702745437622,
+      "learning_rate": 2.9854014598540146e-05,
+      "loss": 0.8496,
+      "step": 818
+    },
+    {
+      "epoch": 0.29906883330290307,
+      "grad_norm": 0.986891508102417,
+      "learning_rate": 2.989051094890511e-05,
+      "loss": 0.8589,
+      "step": 819
+    },
+    {
+      "epoch": 0.2994339967135293,
+      "grad_norm": 1.0008232593536377,
+      "learning_rate": 2.9927007299270075e-05,
+      "loss": 0.8278,
+      "step": 820
+    },
+    {
+      "epoch": 0.2997991601241556,
+      "grad_norm": 1.130242109298706,
+      "learning_rate": 2.9963503649635036e-05,
+      "loss": 0.8522,
+      "step": 821
+    },
+    {
+      "epoch": 0.30016432353478184,
+      "grad_norm": 1.1431291103363037,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.8372,
+      "step": 822
+    },
+    {
+      "epoch": 0.3005294869454081,
+      "grad_norm": 1.3566477298736572,
+      "learning_rate": 3.003649635036497e-05,
+      "loss": 0.8336,
+      "step": 823
+    },
+    {
+      "epoch": 0.3008946503560343,
+      "grad_norm": 0.8847786784172058,
+      "learning_rate": 3.007299270072993e-05,
+      "loss": 0.8261,
+      "step": 824
+    },
+    {
+      "epoch": 0.30125981376666056,
+      "grad_norm": 1.1674613952636719,
+      "learning_rate": 3.0109489051094894e-05,
+      "loss": 0.8665,
+      "step": 825
+    },
+    {
+      "epoch": 0.3016249771772868,
+      "grad_norm": 1.2328945398330688,
+      "learning_rate": 3.014598540145986e-05,
+      "loss": 0.8342,
+      "step": 826
+    },
+    {
+      "epoch": 0.30199014058791307,
+      "grad_norm": 1.004921317100525,
+      "learning_rate": 3.018248175182482e-05,
+      "loss": 0.8542,
+      "step": 827
+    },
+    {
+      "epoch": 0.3023553039985393,
+      "grad_norm": 0.9154124855995178,
+      "learning_rate": 3.0218978102189785e-05,
+      "loss": 0.8376,
+      "step": 828
+    },
+    {
+      "epoch": 0.3027204674091656,
+      "grad_norm": 0.9534822106361389,
+      "learning_rate": 3.0255474452554746e-05,
+      "loss": 0.8693,
+      "step": 829
+    },
+    {
+      "epoch": 0.30308563081979184,
+      "grad_norm": 0.9566801190376282,
+      "learning_rate": 3.029197080291971e-05,
+      "loss": 0.8574,
+      "step": 830
+    },
+    {
+      "epoch": 0.3034507942304181,
+      "grad_norm": 0.9584648013114929,
+      "learning_rate": 3.0328467153284675e-05,
+      "loss": 0.8534,
+      "step": 831
+    },
+    {
+      "epoch": 0.30381595764104435,
+      "grad_norm": 1.7552622556686401,
+      "learning_rate": 3.0364963503649636e-05,
+      "loss": 0.8157,
+      "step": 832
+    },
+    {
+      "epoch": 0.3041811210516706,
+      "grad_norm": 0.8994453549385071,
+      "learning_rate": 3.04014598540146e-05,
+      "loss": 0.8274,
+      "step": 833
+    },
+    {
+      "epoch": 0.30454628446229687,
+      "grad_norm": 0.6974275708198547,
+      "learning_rate": 3.0437956204379565e-05,
+      "loss": 0.804,
+      "step": 834
+    },
+    {
+      "epoch": 0.3049114478729231,
+      "grad_norm": 0.868697464466095,
+      "learning_rate": 3.0474452554744527e-05,
+      "loss": 0.8411,
+      "step": 835
+    },
+    {
+      "epoch": 0.3052766112835494,
+      "grad_norm": 0.9820177555084229,
+      "learning_rate": 3.051094890510949e-05,
+      "loss": 0.8455,
+      "step": 836
+    },
+    {
+      "epoch": 0.30564177469417564,
+      "grad_norm": 0.7563433647155762,
+      "learning_rate": 3.0547445255474456e-05,
+      "loss": 0.8085,
+      "step": 837
+    },
+    {
+      "epoch": 0.3060069381048019,
+      "grad_norm": 1.0138208866119385,
+      "learning_rate": 3.058394160583942e-05,
+      "loss": 0.8531,
+      "step": 838
+    },
+    {
+      "epoch": 0.30637210151542815,
+      "grad_norm": 0.8455662727355957,
+      "learning_rate": 3.0620437956204385e-05,
+      "loss": 0.849,
+      "step": 839
+    },
+    {
+      "epoch": 0.3067372649260544,
+      "grad_norm": 1.0412776470184326,
+      "learning_rate": 3.0656934306569346e-05,
+      "loss": 0.817,
+      "step": 840
+    },
+    {
+      "epoch": 0.30710242833668067,
+      "grad_norm": 0.8895344734191895,
+      "learning_rate": 3.069343065693431e-05,
+      "loss": 0.8472,
+      "step": 841
+    },
+    {
+      "epoch": 0.3074675917473069,
+      "grad_norm": 1.1234147548675537,
+      "learning_rate": 3.0729927007299275e-05,
+      "loss": 0.819,
+      "step": 842
+    },
+    {
+      "epoch": 0.3078327551579332,
+      "grad_norm": 1.1289489269256592,
+      "learning_rate": 3.0766423357664236e-05,
+      "loss": 0.8502,
+      "step": 843
+    },
+    {
+      "epoch": 0.30819791856855944,
+      "grad_norm": 0.849317729473114,
+      "learning_rate": 3.08029197080292e-05,
+      "loss": 0.8062,
+      "step": 844
+    },
+    {
+      "epoch": 0.3085630819791857,
+      "grad_norm": 0.9962231516838074,
+      "learning_rate": 3.0839416058394165e-05,
+      "loss": 0.8438,
+      "step": 845
+    },
+    {
+      "epoch": 0.30892824538981195,
+      "grad_norm": 1.1971230506896973,
+      "learning_rate": 3.0875912408759127e-05,
+      "loss": 0.8685,
+      "step": 846
+    },
+    {
+      "epoch": 0.3092934088004382,
+      "grad_norm": 1.1716772317886353,
+      "learning_rate": 3.091240875912409e-05,
+      "loss": 0.8779,
+      "step": 847
+    },
+    {
+      "epoch": 0.30965857221106446,
+      "grad_norm": 0.777720034122467,
+      "learning_rate": 3.0948905109489056e-05,
+      "loss": 0.8011,
+      "step": 848
+    },
+    {
+      "epoch": 0.3100237356216907,
+      "grad_norm": 1.3869441747665405,
+      "learning_rate": 3.098540145985402e-05,
+      "loss": 0.8271,
+      "step": 849
+    },
+    {
+      "epoch": 0.310388899032317,
+      "grad_norm": 1.0939347743988037,
+      "learning_rate": 3.1021897810218985e-05,
+      "loss": 0.8478,
+      "step": 850
+    },
+    {
+      "epoch": 0.31075406244294324,
+      "grad_norm": 0.7109103202819824,
+      "learning_rate": 3.105839416058394e-05,
+      "loss": 0.8158,
+      "step": 851
+    },
+    {
+      "epoch": 0.3111192258535695,
+      "grad_norm": 1.3487030267715454,
+      "learning_rate": 3.109489051094891e-05,
+      "loss": 0.8251,
+      "step": 852
+    },
+    {
+      "epoch": 0.31148438926419575,
+      "grad_norm": 1.0756982564926147,
+      "learning_rate": 3.1131386861313875e-05,
+      "loss": 0.8522,
+      "step": 853
+    },
+    {
+      "epoch": 0.311849552674822,
+      "grad_norm": 1.2139660120010376,
+      "learning_rate": 3.1167883211678836e-05,
+      "loss": 0.8373,
+      "step": 854
+    },
+    {
+      "epoch": 0.31221471608544826,
+      "grad_norm": 1.1184643507003784,
+      "learning_rate": 3.12043795620438e-05,
+      "loss": 0.8976,
+      "step": 855
+    },
+    {
+      "epoch": 0.3125798794960745,
+      "grad_norm": 0.9936229586601257,
+      "learning_rate": 3.1240875912408765e-05,
+      "loss": 0.8284,
+      "step": 856
+    },
+    {
+      "epoch": 0.3129450429067007,
+      "grad_norm": 0.8024868369102478,
+      "learning_rate": 3.127737226277373e-05,
+      "loss": 0.8251,
+      "step": 857
+    },
+    {
+      "epoch": 0.313310206317327,
+      "grad_norm": 0.8595249652862549,
+      "learning_rate": 3.131386861313869e-05,
+      "loss": 0.7964,
+      "step": 858
+    },
+    {
+      "epoch": 0.31367536972795323,
+      "grad_norm": 0.880524218082428,
+      "learning_rate": 3.135036496350365e-05,
+      "loss": 0.8198,
+      "step": 859
+    },
+    {
+      "epoch": 0.3140405331385795,
+      "grad_norm": 1.0284209251403809,
+      "learning_rate": 3.138686131386862e-05,
+      "loss": 0.8179,
+      "step": 860
+    },
+    {
+      "epoch": 0.31440569654920575,
+      "grad_norm": 1.0328218936920166,
+      "learning_rate": 3.142335766423358e-05,
+      "loss": 0.8708,
+      "step": 861
+    },
+    {
+      "epoch": 0.314770859959832,
+      "grad_norm": 0.9691762328147888,
+      "learning_rate": 3.145985401459854e-05,
+      "loss": 0.7798,
+      "step": 862
+    },
+    {
+      "epoch": 0.31513602337045826,
+      "grad_norm": 1.3719781637191772,
+      "learning_rate": 3.149635036496351e-05,
+      "loss": 0.7811,
+      "step": 863
+    },
+    {
+      "epoch": 0.3155011867810845,
+      "grad_norm": 0.784748911857605,
+      "learning_rate": 3.153284671532847e-05,
+      "loss": 0.8466,
+      "step": 864
+    },
+    {
+      "epoch": 0.3158663501917108,
+      "grad_norm": 1.022679090499878,
+      "learning_rate": 3.156934306569343e-05,
+      "loss": 0.8276,
+      "step": 865
+    },
+    {
+      "epoch": 0.31623151360233703,
+      "grad_norm": 0.9829046726226807,
+      "learning_rate": 3.16058394160584e-05,
+      "loss": 0.8263,
+      "step": 866
+    },
+    {
+      "epoch": 0.3165966770129633,
+      "grad_norm": 0.8736450672149658,
+      "learning_rate": 3.1642335766423365e-05,
+      "loss": 0.8392,
+      "step": 867
+    },
+    {
+      "epoch": 0.31696184042358955,
+      "grad_norm": 0.9354644417762756,
+      "learning_rate": 3.167883211678832e-05,
+      "loss": 0.8011,
+      "step": 868
+    },
+    {
+      "epoch": 0.3173270038342158,
+      "grad_norm": 1.0501866340637207,
+      "learning_rate": 3.171532846715329e-05,
+      "loss": 0.8572,
+      "step": 869
+    },
+    {
+      "epoch": 0.31769216724484206,
+      "grad_norm": 0.9882240295410156,
+      "learning_rate": 3.175182481751825e-05,
+      "loss": 0.8113,
+      "step": 870
+    },
+    {
+      "epoch": 0.3180573306554683,
+      "grad_norm": 0.8495408296585083,
+      "learning_rate": 3.178832116788322e-05,
+      "loss": 0.8181,
+      "step": 871
+    },
+    {
+      "epoch": 0.3184224940660946,
+      "grad_norm": 1.4235453605651855,
+      "learning_rate": 3.182481751824818e-05,
+      "loss": 0.8707,
+      "step": 872
+    },
+    {
+      "epoch": 0.31878765747672083,
+      "grad_norm": 1.308279037475586,
+      "learning_rate": 3.186131386861314e-05,
+      "loss": 0.8494,
+      "step": 873
+    },
+    {
+      "epoch": 0.3191528208873471,
+      "grad_norm": 1.1783095598220825,
+      "learning_rate": 3.189781021897811e-05,
+      "loss": 0.8304,
+      "step": 874
+    },
+    {
+      "epoch": 0.31951798429797335,
+      "grad_norm": 0.6439776420593262,
+      "learning_rate": 3.193430656934307e-05,
+      "loss": 0.8052,
+      "step": 875
+    },
+    {
+      "epoch": 0.3198831477085996,
+      "grad_norm": 0.74483323097229,
+      "learning_rate": 3.197080291970803e-05,
+      "loss": 0.8425,
+      "step": 876
+    },
+    {
+      "epoch": 0.32024831111922586,
+      "grad_norm": 1.0718963146209717,
+      "learning_rate": 3.2007299270073e-05,
+      "loss": 0.8292,
+      "step": 877
+    },
+    {
+      "epoch": 0.3206134745298521,
+      "grad_norm": 1.2955678701400757,
+      "learning_rate": 3.204379562043796e-05,
+      "loss": 0.8169,
+      "step": 878
+    },
+    {
+      "epoch": 0.3209786379404784,
+      "grad_norm": 1.3649272918701172,
+      "learning_rate": 3.208029197080292e-05,
+      "loss": 0.8435,
+      "step": 879
+    },
+    {
+      "epoch": 0.32134380135110463,
+      "grad_norm": 0.8959698677062988,
+      "learning_rate": 3.211678832116789e-05,
+      "loss": 0.8048,
+      "step": 880
+    },
+    {
+      "epoch": 0.3217089647617309,
+      "grad_norm": 1.2295029163360596,
+      "learning_rate": 3.215328467153285e-05,
+      "loss": 0.833,
+      "step": 881
+    },
+    {
+      "epoch": 0.32207412817235714,
+      "grad_norm": 1.0183993577957153,
+      "learning_rate": 3.218978102189781e-05,
+      "loss": 0.8398,
+      "step": 882
+    },
+    {
+      "epoch": 0.3224392915829834,
+      "grad_norm": 0.7791547179222107,
+      "learning_rate": 3.222627737226278e-05,
+      "loss": 0.8392,
+      "step": 883
+    },
+    {
+      "epoch": 0.32280445499360966,
+      "grad_norm": 2.3649449348449707,
+      "learning_rate": 3.226277372262774e-05,
+      "loss": 0.8372,
+      "step": 884
+    },
+    {
+      "epoch": 0.3231696184042359,
+      "grad_norm": 1.179885983467102,
+      "learning_rate": 3.22992700729927e-05,
+      "loss": 0.8252,
+      "step": 885
+    },
+    {
+      "epoch": 0.32353478181486217,
+      "grad_norm": 1.0518529415130615,
+      "learning_rate": 3.233576642335767e-05,
+      "loss": 0.8553,
+      "step": 886
+    },
+    {
+      "epoch": 0.32389994522548843,
+      "grad_norm": 0.9633598923683167,
+      "learning_rate": 3.237226277372263e-05,
+      "loss": 0.8253,
+      "step": 887
+    },
+    {
+      "epoch": 0.3242651086361147,
+      "grad_norm": 1.2298461198806763,
+      "learning_rate": 3.24087591240876e-05,
+      "loss": 0.8173,
+      "step": 888
+    },
+    {
+      "epoch": 0.32463027204674094,
+      "grad_norm": 0.9368358254432678,
+      "learning_rate": 3.244525547445256e-05,
+      "loss": 0.8084,
+      "step": 889
+    },
+    {
+      "epoch": 0.3249954354573672,
+      "grad_norm": 0.8894047141075134,
+      "learning_rate": 3.248175182481752e-05,
+      "loss": 0.8243,
+      "step": 890
+    },
+    {
+      "epoch": 0.3253605988679934,
+      "grad_norm": 0.9334021806716919,
+      "learning_rate": 3.251824817518249e-05,
+      "loss": 0.8069,
+      "step": 891
+    },
+    {
+      "epoch": 0.32572576227861966,
+      "grad_norm": 1.008568525314331,
+      "learning_rate": 3.255474452554745e-05,
+      "loss": 0.8408,
+      "step": 892
+    },
+    {
+      "epoch": 0.3260909256892459,
+      "grad_norm": 1.0324256420135498,
+      "learning_rate": 3.259124087591241e-05,
+      "loss": 0.8558,
+      "step": 893
+    },
+    {
+      "epoch": 0.32645608909987217,
+      "grad_norm": 2.0214602947235107,
+      "learning_rate": 3.262773722627738e-05,
+      "loss": 0.7908,
+      "step": 894
+    },
+    {
+      "epoch": 0.32682125251049843,
+      "grad_norm": 0.6830151677131653,
+      "learning_rate": 3.266423357664234e-05,
+      "loss": 0.8373,
+      "step": 895
+    },
+    {
+      "epoch": 0.3271864159211247,
+      "grad_norm": 1.074393391609192,
+      "learning_rate": 3.27007299270073e-05,
+      "loss": 0.8124,
+      "step": 896
+    },
+    {
+      "epoch": 0.32755157933175094,
+      "grad_norm": 0.845180869102478,
+      "learning_rate": 3.273722627737227e-05,
+      "loss": 0.832,
+      "step": 897
+    },
+    {
+      "epoch": 0.3279167427423772,
+      "grad_norm": 1.0447711944580078,
+      "learning_rate": 3.277372262773723e-05,
+      "loss": 0.8257,
+      "step": 898
+    },
+    {
+      "epoch": 0.32828190615300346,
+      "grad_norm": 0.8940461277961731,
+      "learning_rate": 3.281021897810219e-05,
+      "loss": 0.8353,
+      "step": 899
+    },
+    {
+      "epoch": 0.3286470695636297,
+      "grad_norm": 1.0841431617736816,
+      "learning_rate": 3.284671532846716e-05,
+      "loss": 0.8226,
+      "step": 900
+    },
+    {
+      "epoch": 0.32901223297425597,
+      "grad_norm": 0.9496214389801025,
+      "learning_rate": 3.288321167883212e-05,
+      "loss": 0.8202,
+      "step": 901
+    },
+    {
+      "epoch": 0.3293773963848822,
+      "grad_norm": 0.8948093056678772,
+      "learning_rate": 3.291970802919708e-05,
+      "loss": 0.8251,
+      "step": 902
+    },
+    {
+      "epoch": 0.3297425597955085,
+      "grad_norm": 1.312648892402649,
+      "learning_rate": 3.295620437956204e-05,
+      "loss": 0.8773,
+      "step": 903
+    },
+    {
+      "epoch": 0.33010772320613474,
+      "grad_norm": 1.1946377754211426,
+      "learning_rate": 3.299270072992701e-05,
+      "loss": 0.8223,
+      "step": 904
+    },
+    {
+      "epoch": 0.330472886616761,
+      "grad_norm": 0.9710689187049866,
+      "learning_rate": 3.302919708029197e-05,
+      "loss": 0.8355,
+      "step": 905
+    },
+    {
+      "epoch": 0.33083805002738725,
+      "grad_norm": 0.8350407481193542,
+      "learning_rate": 3.306569343065693e-05,
+      "loss": 0.8265,
+      "step": 906
+    },
+    {
+      "epoch": 0.3312032134380135,
+      "grad_norm": 1.2623652219772339,
+      "learning_rate": 3.31021897810219e-05,
+      "loss": 0.8221,
+      "step": 907
+    },
+    {
+      "epoch": 0.33156837684863977,
+      "grad_norm": 0.8959325551986694,
+      "learning_rate": 3.313868613138687e-05,
+      "loss": 0.8311,
+      "step": 908
+    },
+    {
+      "epoch": 0.331933540259266,
+      "grad_norm": 0.9143708944320679,
+      "learning_rate": 3.317518248175183e-05,
+      "loss": 0.8245,
+      "step": 909
+    },
+    {
+      "epoch": 0.3322987036698923,
+      "grad_norm": 1.5401510000228882,
+      "learning_rate": 3.321167883211679e-05,
+      "loss": 0.8101,
+      "step": 910
+    },
+    {
+      "epoch": 0.33266386708051854,
+      "grad_norm": 0.6910380721092224,
+      "learning_rate": 3.324817518248176e-05,
+      "loss": 0.8,
+      "step": 911
+    },
+    {
+      "epoch": 0.3330290304911448,
+      "grad_norm": 2.472301483154297,
+      "learning_rate": 3.328467153284672e-05,
+      "loss": 0.8077,
+      "step": 912
+    },
+    {
+      "epoch": 0.33339419390177105,
+      "grad_norm": 1.2135626077651978,
+      "learning_rate": 3.332116788321168e-05,
+      "loss": 0.8113,
+      "step": 913
+    },
+    {
+      "epoch": 0.3337593573123973,
+      "grad_norm": 1.107142448425293,
+      "learning_rate": 3.335766423357664e-05,
+      "loss": 0.843,
+      "step": 914
+    },
+    {
+      "epoch": 0.33412452072302357,
+      "grad_norm": 1.1342405080795288,
+      "learning_rate": 3.339416058394161e-05,
+      "loss": 0.8278,
+      "step": 915
+    },
+    {
+      "epoch": 0.3344896841336498,
+      "grad_norm": 0.953411877155304,
+      "learning_rate": 3.343065693430657e-05,
+      "loss": 0.8277,
+      "step": 916
+    },
+    {
+      "epoch": 0.3348548475442761,
+      "grad_norm": 0.9323768019676208,
+      "learning_rate": 3.346715328467153e-05,
+      "loss": 0.8081,
+      "step": 917
+    },
+    {
+      "epoch": 0.33522001095490234,
+      "grad_norm": 0.9772979021072388,
+      "learning_rate": 3.35036496350365e-05,
+      "loss": 0.8453,
+      "step": 918
+    },
+    {
+      "epoch": 0.3355851743655286,
+      "grad_norm": 1.1281118392944336,
+      "learning_rate": 3.354014598540146e-05,
+      "loss": 0.8509,
+      "step": 919
+    },
+    {
+      "epoch": 0.33595033777615485,
+      "grad_norm": 0.9218109846115112,
+      "learning_rate": 3.357664233576642e-05,
+      "loss": 0.8181,
+      "step": 920
+    },
+    {
+      "epoch": 0.3363155011867811,
+      "grad_norm": 1.230379581451416,
+      "learning_rate": 3.361313868613139e-05,
+      "loss": 0.8436,
+      "step": 921
+    },
+    {
+      "epoch": 0.33668066459740736,
+      "grad_norm": 1.3046516180038452,
+      "learning_rate": 3.364963503649635e-05,
+      "loss": 0.8668,
+      "step": 922
+    },
+    {
+      "epoch": 0.3370458280080336,
+      "grad_norm": 0.9936146140098572,
+      "learning_rate": 3.368613138686131e-05,
+      "loss": 0.8121,
+      "step": 923
+    },
+    {
+      "epoch": 0.3374109914186599,
+      "grad_norm": 0.7893944382667542,
+      "learning_rate": 3.372262773722628e-05,
+      "loss": 0.8416,
+      "step": 924
+    },
+    {
+      "epoch": 0.3377761548292861,
+      "grad_norm": 0.9127933382987976,
+      "learning_rate": 3.375912408759124e-05,
+      "loss": 0.8004,
+      "step": 925
+    },
+    {
+      "epoch": 0.33814131823991234,
+      "grad_norm": 0.8317862749099731,
+      "learning_rate": 3.3795620437956204e-05,
+      "loss": 0.8025,
+      "step": 926
+    },
+    {
+      "epoch": 0.3385064816505386,
+      "grad_norm": 0.8821642994880676,
+      "learning_rate": 3.383211678832117e-05,
+      "loss": 0.8265,
+      "step": 927
+    },
+    {
+      "epoch": 0.33887164506116485,
+      "grad_norm": 1.2264710664749146,
+      "learning_rate": 3.386861313868613e-05,
+      "loss": 0.8625,
+      "step": 928
+    },
+    {
+      "epoch": 0.3392368084717911,
+      "grad_norm": 0.8349049687385559,
+      "learning_rate": 3.39051094890511e-05,
+      "loss": 0.8036,
+      "step": 929
+    },
+    {
+      "epoch": 0.33960197188241736,
+      "grad_norm": 0.8627322316169739,
+      "learning_rate": 3.394160583941606e-05,
+      "loss": 0.8652,
+      "step": 930
+    },
+    {
+      "epoch": 0.3399671352930436,
+      "grad_norm": 1.1609947681427002,
+      "learning_rate": 3.397810218978102e-05,
+      "loss": 0.8444,
+      "step": 931
+    },
+    {
+      "epoch": 0.3403322987036699,
+      "grad_norm": 1.0138678550720215,
+      "learning_rate": 3.401459854014599e-05,
+      "loss": 0.833,
+      "step": 932
+    },
+    {
+      "epoch": 0.34069746211429613,
+      "grad_norm": 1.2408621311187744,
+      "learning_rate": 3.405109489051095e-05,
+      "loss": 0.833,
+      "step": 933
+    },
+    {
+      "epoch": 0.3410626255249224,
+      "grad_norm": 0.9205710291862488,
+      "learning_rate": 3.408759124087591e-05,
+      "loss": 0.8199,
+      "step": 934
+    },
+    {
+      "epoch": 0.34142778893554865,
+      "grad_norm": 1.0878050327301025,
+      "learning_rate": 3.412408759124088e-05,
+      "loss": 0.8083,
+      "step": 935
+    },
+    {
+      "epoch": 0.3417929523461749,
+      "grad_norm": 0.8882326483726501,
+      "learning_rate": 3.416058394160584e-05,
+      "loss": 0.8295,
+      "step": 936
+    },
+    {
+      "epoch": 0.34215811575680116,
+      "grad_norm": 1.1191176176071167,
+      "learning_rate": 3.4197080291970804e-05,
+      "loss": 0.811,
+      "step": 937
+    },
+    {
+      "epoch": 0.3425232791674274,
+      "grad_norm": 0.829477846622467,
+      "learning_rate": 3.423357664233577e-05,
+      "loss": 0.8364,
+      "step": 938
+    },
+    {
+      "epoch": 0.3428884425780537,
+      "grad_norm": 0.9906582236289978,
+      "learning_rate": 3.427007299270073e-05,
+      "loss": 0.8206,
+      "step": 939
+    },
+    {
+      "epoch": 0.34325360598867993,
+      "grad_norm": 1.1381598711013794,
+      "learning_rate": 3.4306569343065694e-05,
+      "loss": 0.8435,
+      "step": 940
+    },
+    {
+      "epoch": 0.3436187693993062,
+      "grad_norm": 0.9499815702438354,
+      "learning_rate": 3.434306569343066e-05,
+      "loss": 0.7904,
+      "step": 941
+    },
+    {
+      "epoch": 0.34398393280993245,
+      "grad_norm": 1.2010533809661865,
+      "learning_rate": 3.437956204379562e-05,
+      "loss": 0.8521,
+      "step": 942
+    },
+    {
+      "epoch": 0.3443490962205587,
+      "grad_norm": 0.9433616399765015,
+      "learning_rate": 3.4416058394160584e-05,
+      "loss": 0.8716,
+      "step": 943
+    },
+    {
+      "epoch": 0.34471425963118496,
+      "grad_norm": 0.9508865475654602,
+      "learning_rate": 3.4452554744525545e-05,
+      "loss": 0.8239,
+      "step": 944
+    },
+    {
+      "epoch": 0.3450794230418112,
+      "grad_norm": 0.9482638239860535,
+      "learning_rate": 3.448905109489051e-05,
+      "loss": 0.839,
+      "step": 945
+    },
+    {
+      "epoch": 0.3454445864524375,
+      "grad_norm": 0.8817934393882751,
+      "learning_rate": 3.452554744525548e-05,
+      "loss": 0.8252,
+      "step": 946
+    },
+    {
+      "epoch": 0.34580974986306373,
+      "grad_norm": 1.4641749858856201,
+      "learning_rate": 3.4562043795620436e-05,
+      "loss": 0.8228,
+      "step": 947
+    },
+    {
+      "epoch": 0.34617491327369,
+      "grad_norm": 1.001059889793396,
+      "learning_rate": 3.4598540145985404e-05,
+      "loss": 0.8246,
+      "step": 948
+    },
+    {
+      "epoch": 0.34654007668431625,
+      "grad_norm": 1.1823747158050537,
+      "learning_rate": 3.463503649635037e-05,
+      "loss": 0.8475,
+      "step": 949
+    },
+    {
+      "epoch": 0.3469052400949425,
+      "grad_norm": 0.9262009263038635,
+      "learning_rate": 3.467153284671533e-05,
+      "loss": 0.8162,
+      "step": 950
+    },
+    {
+      "epoch": 0.34727040350556876,
+      "grad_norm": 0.6685298681259155,
+      "learning_rate": 3.4708029197080294e-05,
+      "loss": 0.8395,
+      "step": 951
+    },
+    {
+      "epoch": 0.347635566916195,
+      "grad_norm": 1.6057016849517822,
+      "learning_rate": 3.474452554744526e-05,
+      "loss": 0.8174,
+      "step": 952
+    },
+    {
+      "epoch": 0.3480007303268213,
+      "grad_norm": 0.9343808889389038,
+      "learning_rate": 3.478102189781022e-05,
+      "loss": 0.8411,
+      "step": 953
+    },
+    {
+      "epoch": 0.34836589373744753,
+      "grad_norm": 0.9032874703407288,
+      "learning_rate": 3.4817518248175184e-05,
+      "loss": 0.801,
+      "step": 954
+    },
+    {
+      "epoch": 0.3487310571480738,
+      "grad_norm": 0.9119698405265808,
+      "learning_rate": 3.4854014598540145e-05,
+      "loss": 0.8065,
+      "step": 955
+    },
+    {
+      "epoch": 0.34909622055870004,
+      "grad_norm": 1.0663105249404907,
+      "learning_rate": 3.4890510948905113e-05,
+      "loss": 0.7919,
+      "step": 956
+    },
+    {
+      "epoch": 0.3494613839693263,
+      "grad_norm": 1.0270241498947144,
+      "learning_rate": 3.4927007299270075e-05,
+      "loss": 0.8098,
+      "step": 957
+    },
+    {
+      "epoch": 0.3498265473799525,
+      "grad_norm": 1.1155675649642944,
+      "learning_rate": 3.4963503649635036e-05,
+      "loss": 0.8612,
+      "step": 958
+    },
+    {
+      "epoch": 0.35019171079057876,
+      "grad_norm": 1.2419427633285522,
+      "learning_rate": 3.5000000000000004e-05,
+      "loss": 0.8745,
+      "step": 959
+    },
+    {
+      "epoch": 0.350556874201205,
+      "grad_norm": 0.8625814914703369,
+      "learning_rate": 3.5036496350364965e-05,
+      "loss": 0.8186,
+      "step": 960
+    },
+    {
+      "epoch": 0.3509220376118313,
+      "grad_norm": 0.964387834072113,
+      "learning_rate": 3.5072992700729926e-05,
+      "loss": 0.8287,
+      "step": 961
+    },
+    {
+      "epoch": 0.35128720102245753,
+      "grad_norm": 0.7243093252182007,
+      "learning_rate": 3.5109489051094894e-05,
+      "loss": 0.7982,
+      "step": 962
+    },
+    {
+      "epoch": 0.3516523644330838,
+      "grad_norm": 1.322411060333252,
+      "learning_rate": 3.514598540145986e-05,
+      "loss": 0.8702,
+      "step": 963
+    },
+    {
+      "epoch": 0.35201752784371004,
+      "grad_norm": 0.9760127663612366,
+      "learning_rate": 3.5182481751824816e-05,
+      "loss": 0.8479,
+      "step": 964
+    },
+    {
+      "epoch": 0.3523826912543363,
+      "grad_norm": 1.1912221908569336,
+      "learning_rate": 3.5218978102189784e-05,
+      "loss": 0.8054,
+      "step": 965
+    },
+    {
+      "epoch": 0.35274785466496256,
+      "grad_norm": 1.126135230064392,
+      "learning_rate": 3.5255474452554745e-05,
+      "loss": 0.8484,
+      "step": 966
+    },
+    {
+      "epoch": 0.3531130180755888,
+      "grad_norm": 1.1270838975906372,
+      "learning_rate": 3.5291970802919713e-05,
+      "loss": 0.8258,
+      "step": 967
+    },
+    {
+      "epoch": 0.35347818148621507,
+      "grad_norm": 0.8599242568016052,
+      "learning_rate": 3.5328467153284675e-05,
+      "loss": 0.8273,
+      "step": 968
+    },
+    {
+      "epoch": 0.3538433448968413,
+      "grad_norm": 1.205792784690857,
+      "learning_rate": 3.5364963503649636e-05,
+      "loss": 0.8451,
+      "step": 969
+    },
+    {
+      "epoch": 0.3542085083074676,
+      "grad_norm": 1.0273261070251465,
+      "learning_rate": 3.5401459854014604e-05,
+      "loss": 0.8134,
+      "step": 970
+    },
+    {
+      "epoch": 0.35457367171809384,
+      "grad_norm": 1.1758073568344116,
+      "learning_rate": 3.5437956204379565e-05,
+      "loss": 0.8381,
+      "step": 971
+    },
+    {
+      "epoch": 0.3549388351287201,
+      "grad_norm": 1.1069955825805664,
+      "learning_rate": 3.5474452554744526e-05,
+      "loss": 0.8257,
+      "step": 972
+    },
+    {
+      "epoch": 0.35530399853934636,
+      "grad_norm": 1.0383074283599854,
+      "learning_rate": 3.5510948905109494e-05,
+      "loss": 0.8691,
+      "step": 973
+    },
+    {
+      "epoch": 0.3556691619499726,
+      "grad_norm": 1.0610301494598389,
+      "learning_rate": 3.5547445255474455e-05,
+      "loss": 0.8293,
+      "step": 974
+    },
+    {
+      "epoch": 0.35603432536059887,
+      "grad_norm": 1.1219865083694458,
+      "learning_rate": 3.5583941605839416e-05,
+      "loss": 0.8011,
+      "step": 975
+    },
+    {
+      "epoch": 0.3563994887712251,
+      "grad_norm": 1.0254887342453003,
+      "learning_rate": 3.5620437956204384e-05,
+      "loss": 0.8262,
+      "step": 976
+    },
+    {
+      "epoch": 0.3567646521818514,
+      "grad_norm": 1.057141900062561,
+      "learning_rate": 3.5656934306569346e-05,
+      "loss": 0.8442,
+      "step": 977
+    },
+    {
+      "epoch": 0.35712981559247764,
+      "grad_norm": 0.9983533024787903,
+      "learning_rate": 3.569343065693431e-05,
+      "loss": 0.809,
+      "step": 978
+    },
+    {
+      "epoch": 0.3574949790031039,
+      "grad_norm": 1.0989717245101929,
+      "learning_rate": 3.5729927007299275e-05,
+      "loss": 0.8024,
+      "step": 979
+    },
+    {
+      "epoch": 0.35786014241373015,
+      "grad_norm": 0.8879622220993042,
+      "learning_rate": 3.5766423357664236e-05,
+      "loss": 0.8134,
+      "step": 980
+    },
+    {
+      "epoch": 0.3582253058243564,
+      "grad_norm": 0.9703028798103333,
+      "learning_rate": 3.58029197080292e-05,
+      "loss": 0.8459,
+      "step": 981
+    },
+    {
+      "epoch": 0.35859046923498267,
+      "grad_norm": 0.9390744566917419,
+      "learning_rate": 3.5839416058394165e-05,
+      "loss": 0.8263,
+      "step": 982
+    },
+    {
+      "epoch": 0.3589556326456089,
+      "grad_norm": 0.9232451915740967,
+      "learning_rate": 3.5875912408759126e-05,
+      "loss": 0.7977,
+      "step": 983
+    },
+    {
+      "epoch": 0.3593207960562352,
+      "grad_norm": 0.9906948208808899,
+      "learning_rate": 3.5912408759124094e-05,
+      "loss": 0.7958,
+      "step": 984
+    },
+    {
+      "epoch": 0.35968595946686144,
+      "grad_norm": 1.122028112411499,
+      "learning_rate": 3.5948905109489055e-05,
+      "loss": 0.8314,
+      "step": 985
+    },
+    {
+      "epoch": 0.3600511228774877,
+      "grad_norm": 1.034443736076355,
+      "learning_rate": 3.5985401459854016e-05,
+      "loss": 0.7949,
+      "step": 986
+    },
+    {
+      "epoch": 0.36041628628811395,
+      "grad_norm": 1.1605156660079956,
+      "learning_rate": 3.6021897810218984e-05,
+      "loss": 0.804,
+      "step": 987
+    },
+    {
+      "epoch": 0.3607814496987402,
+      "grad_norm": 0.9331724047660828,
+      "learning_rate": 3.6058394160583946e-05,
+      "loss": 0.7858,
+      "step": 988
+    },
+    {
+      "epoch": 0.36114661310936647,
+      "grad_norm": 0.9008059501647949,
+      "learning_rate": 3.609489051094891e-05,
+      "loss": 0.8015,
+      "step": 989
+    },
+    {
+      "epoch": 0.3615117765199927,
+      "grad_norm": 1.032978892326355,
+      "learning_rate": 3.6131386861313875e-05,
+      "loss": 0.8217,
+      "step": 990
+    },
+    {
+      "epoch": 0.361876939930619,
+      "grad_norm": 0.9546392560005188,
+      "learning_rate": 3.6167883211678836e-05,
+      "loss": 0.8102,
+      "step": 991
+    },
+    {
+      "epoch": 0.3622421033412452,
+      "grad_norm": 1.1435415744781494,
+      "learning_rate": 3.62043795620438e-05,
+      "loss": 0.8077,
+      "step": 992
+    },
+    {
+      "epoch": 0.36260726675187144,
+      "grad_norm": 1.009759545326233,
+      "learning_rate": 3.6240875912408765e-05,
+      "loss": 0.802,
+      "step": 993
+    },
+    {
+      "epoch": 0.3629724301624977,
+      "grad_norm": 0.9189592599868774,
+      "learning_rate": 3.6277372262773726e-05,
+      "loss": 0.8082,
+      "step": 994
+    },
+    {
+      "epoch": 0.36333759357312395,
+      "grad_norm": 0.9083291292190552,
+      "learning_rate": 3.631386861313869e-05,
+      "loss": 0.8265,
+      "step": 995
+    },
+    {
+      "epoch": 0.3637027569837502,
+      "grad_norm": 1.0146288871765137,
+      "learning_rate": 3.635036496350365e-05,
+      "loss": 0.7642,
+      "step": 996
+    },
+    {
+      "epoch": 0.36406792039437647,
+      "grad_norm": 1.2165806293487549,
+      "learning_rate": 3.6386861313868616e-05,
+      "loss": 0.8153,
+      "step": 997
+    },
+    {
+      "epoch": 0.3644330838050027,
+      "grad_norm": 1.1724755764007568,
+      "learning_rate": 3.642335766423358e-05,
+      "loss": 0.78,
+      "step": 998
+    },
+    {
+      "epoch": 0.364798247215629,
+      "grad_norm": 1.15825355052948,
+      "learning_rate": 3.645985401459854e-05,
+      "loss": 0.8104,
+      "step": 999
+    },
+    {
+      "epoch": 0.36516341062625524,
+      "grad_norm": 1.1032559871673584,
+      "learning_rate": 3.649635036496351e-05,
+      "loss": 0.8246,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3655285740368815,
+      "grad_norm": 0.9770479798316956,
+      "learning_rate": 3.6532846715328475e-05,
+      "loss": 0.7734,
+      "step": 1001
+    },
+    {
+      "epoch": 0.36589373744750775,
+      "grad_norm": 1.4258418083190918,
+      "learning_rate": 3.656934306569343e-05,
+      "loss": 0.7863,
+      "step": 1002
+    },
+    {
+      "epoch": 0.366258900858134,
+      "grad_norm": 1.1266810894012451,
+      "learning_rate": 3.66058394160584e-05,
+      "loss": 0.8093,
+      "step": 1003
+    },
+    {
+      "epoch": 0.36662406426876026,
+      "grad_norm": 1.1982899904251099,
+      "learning_rate": 3.6642335766423365e-05,
+      "loss": 0.8171,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3669892276793865,
+      "grad_norm": 1.5895270109176636,
+      "learning_rate": 3.6678832116788326e-05,
+      "loss": 0.8324,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3673543910900128,
+      "grad_norm": 1.1550204753875732,
+      "learning_rate": 3.671532846715329e-05,
+      "loss": 0.8187,
+      "step": 1006
+    },
+    {
+      "epoch": 0.36771955450063903,
+      "grad_norm": 3.356076955795288,
+      "learning_rate": 3.675182481751825e-05,
+      "loss": 0.7982,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3680847179112653,
+      "grad_norm": 1.1151169538497925,
+      "learning_rate": 3.6788321167883217e-05,
+      "loss": 0.8,
+      "step": 1008
+    },
+    {
+      "epoch": 0.36844988132189155,
+      "grad_norm": 1.7109216451644897,
+      "learning_rate": 3.682481751824818e-05,
+      "loss": 0.8066,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3688150447325178,
+      "grad_norm": 1.1339987516403198,
+      "learning_rate": 3.686131386861314e-05,
+      "loss": 0.8025,
+      "step": 1010
+    },
+    {
+      "epoch": 0.36918020814314406,
+      "grad_norm": 1.3794639110565186,
+      "learning_rate": 3.689781021897811e-05,
+      "loss": 0.7701,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3695453715537703,
+      "grad_norm": 2.926950454711914,
+      "learning_rate": 3.693430656934307e-05,
+      "loss": 0.8243,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3699105349643966,
+      "grad_norm": 1.6819735765457153,
+      "learning_rate": 3.697080291970803e-05,
+      "loss": 0.8391,
+      "step": 1013
+    },
+    {
+      "epoch": 0.37027569837502283,
+      "grad_norm": 1.31744384765625,
+      "learning_rate": 3.7007299270073e-05,
+      "loss": 0.7911,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3706408617856491,
+      "grad_norm": 1.478202223777771,
+      "learning_rate": 3.704379562043796e-05,
+      "loss": 0.8079,
+      "step": 1015
+    },
+    {
+      "epoch": 0.37100602519627535,
+      "grad_norm": 1.2199921607971191,
+      "learning_rate": 3.708029197080292e-05,
+      "loss": 0.8317,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3713711886069016,
+      "grad_norm": 2.6387546062469482,
+      "learning_rate": 3.711678832116789e-05,
+      "loss": 0.7961,
+      "step": 1017
+    },
+    {
+      "epoch": 0.37173635201752786,
+      "grad_norm": 1.2078938484191895,
+      "learning_rate": 3.715328467153285e-05,
+      "loss": 0.7894,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3721015154281541,
+      "grad_norm": 1.1073907613754272,
+      "learning_rate": 3.718978102189781e-05,
+      "loss": 0.8289,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3724666788387804,
+      "grad_norm": 1.514863133430481,
+      "learning_rate": 3.722627737226278e-05,
+      "loss": 0.7914,
+      "step": 1020
+    },
+    {
+      "epoch": 0.37283184224940663,
+      "grad_norm": 1.3350275754928589,
+      "learning_rate": 3.726277372262774e-05,
+      "loss": 0.7986,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3731970056600329,
+      "grad_norm": 1.2805231809616089,
+      "learning_rate": 3.729927007299271e-05,
+      "loss": 0.8132,
+      "step": 1022
+    },
+    {
+      "epoch": 0.37356216907065914,
+      "grad_norm": 1.068396806716919,
+      "learning_rate": 3.733576642335767e-05,
+      "loss": 0.8007,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3739273324812854,
+      "grad_norm": 0.9894810914993286,
+      "learning_rate": 3.737226277372263e-05,
+      "loss": 0.8082,
+      "step": 1024
+    },
+    {
+      "epoch": 0.3742924958919116,
+      "grad_norm": 1.133439302444458,
+      "learning_rate": 3.74087591240876e-05,
+      "loss": 0.7847,
+      "step": 1025
+    },
+    {
+      "epoch": 0.37465765930253786,
+      "grad_norm": 0.7983406782150269,
+      "learning_rate": 3.744525547445256e-05,
+      "loss": 0.7599,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3750228227131641,
+      "grad_norm": 0.8845236301422119,
+      "learning_rate": 3.748175182481752e-05,
+      "loss": 0.7507,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3753879861237904,
+      "grad_norm": 1.125713586807251,
+      "learning_rate": 3.751824817518249e-05,
+      "loss": 0.7647,
+      "step": 1028
+    },
+    {
+      "epoch": 0.37575314953441663,
+      "grad_norm": 1.0719150304794312,
+      "learning_rate": 3.755474452554745e-05,
+      "loss": 0.8115,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3761183129450429,
+      "grad_norm": 0.8788213133811951,
+      "learning_rate": 3.759124087591241e-05,
+      "loss": 0.7986,
+      "step": 1030
+    },
+    {
+      "epoch": 0.37648347635566914,
+      "grad_norm": 0.8796796798706055,
+      "learning_rate": 3.762773722627738e-05,
+      "loss": 0.8007,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3768486397662954,
+      "grad_norm": 1.013904333114624,
+      "learning_rate": 3.766423357664234e-05,
+      "loss": 0.8134,
+      "step": 1032
+    },
+    {
+      "epoch": 0.37721380317692166,
+      "grad_norm": 1.1152280569076538,
+      "learning_rate": 3.77007299270073e-05,
+      "loss": 0.7745,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3775789665875479,
+      "grad_norm": 0.913651168346405,
+      "learning_rate": 3.773722627737227e-05,
+      "loss": 0.8302,
+      "step": 1034
+    },
+    {
+      "epoch": 0.37794412999817417,
+      "grad_norm": 1.1180498600006104,
+      "learning_rate": 3.777372262773723e-05,
+      "loss": 0.796,
+      "step": 1035
+    },
+    {
+      "epoch": 0.37830929340880043,
+      "grad_norm": 1.4855592250823975,
+      "learning_rate": 3.781021897810219e-05,
+      "loss": 0.8248,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3786744568194267,
+      "grad_norm": 0.6914875507354736,
+      "learning_rate": 3.784671532846716e-05,
+      "loss": 0.807,
+      "step": 1037
+    },
+    {
+      "epoch": 0.37903962023005294,
+      "grad_norm": 1.0839126110076904,
+      "learning_rate": 3.788321167883212e-05,
+      "loss": 0.828,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3794047836406792,
+      "grad_norm": 1.0782427787780762,
+      "learning_rate": 3.791970802919709e-05,
+      "loss": 0.772,
+      "step": 1039
+    },
+    {
+      "epoch": 0.37976994705130546,
+      "grad_norm": 1.0255452394485474,
+      "learning_rate": 3.795620437956204e-05,
+      "loss": 0.8265,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3801351104619317,
+      "grad_norm": 0.9261432886123657,
+      "learning_rate": 3.799270072992701e-05,
+      "loss": 0.812,
+      "step": 1041
+    },
+    {
+      "epoch": 0.38050027387255797,
+      "grad_norm": 1.079363465309143,
+      "learning_rate": 3.802919708029198e-05,
+      "loss": 0.7696,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3808654372831842,
+      "grad_norm": 0.9970255494117737,
+      "learning_rate": 3.806569343065694e-05,
+      "loss": 0.7926,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3812306006938105,
+      "grad_norm": 1.0042568445205688,
+      "learning_rate": 3.81021897810219e-05,
+      "loss": 0.7717,
+      "step": 1044
+    },
+    {
+      "epoch": 0.38159576410443674,
+      "grad_norm": 0.8755733966827393,
+      "learning_rate": 3.813868613138687e-05,
+      "loss": 0.8062,
+      "step": 1045
+    },
+    {
+      "epoch": 0.381960927515063,
+      "grad_norm": 0.9101842045783997,
+      "learning_rate": 3.817518248175183e-05,
+      "loss": 0.8138,
+      "step": 1046
+    },
+    {
+      "epoch": 0.38232609092568925,
+      "grad_norm": 0.9083014726638794,
+      "learning_rate": 3.821167883211679e-05,
+      "loss": 0.8264,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3826912543363155,
+      "grad_norm": 0.8410999774932861,
+      "learning_rate": 3.824817518248176e-05,
+      "loss": 0.8135,
+      "step": 1048
+    },
+    {
+      "epoch": 0.38305641774694177,
+      "grad_norm": 0.8920347094535828,
+      "learning_rate": 3.828467153284672e-05,
+      "loss": 0.7825,
+      "step": 1049
+    },
+    {
+      "epoch": 0.383421581157568,
+      "grad_norm": 1.1922773122787476,
+      "learning_rate": 3.832116788321168e-05,
+      "loss": 0.8097,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3837867445681943,
+      "grad_norm": 1.257304072380066,
+      "learning_rate": 3.835766423357664e-05,
+      "loss": 0.8099,
+      "step": 1051
+    },
+    {
+      "epoch": 0.38415190797882054,
+      "grad_norm": 1.5688021183013916,
+      "learning_rate": 3.839416058394161e-05,
+      "loss": 0.8091,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3845170713894468,
+      "grad_norm": 1.091172695159912,
+      "learning_rate": 3.843065693430657e-05,
+      "loss": 0.8044,
+      "step": 1053
+    },
+    {
+      "epoch": 0.38488223480007305,
+      "grad_norm": 0.9299972057342529,
+      "learning_rate": 3.846715328467153e-05,
+      "loss": 0.8152,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3852473982106993,
+      "grad_norm": 0.9777531027793884,
+      "learning_rate": 3.85036496350365e-05,
+      "loss": 0.7755,
+      "step": 1055
+    },
+    {
+      "epoch": 0.38561256162132557,
+      "grad_norm": 0.9919531345367432,
+      "learning_rate": 3.854014598540147e-05,
+      "loss": 0.8351,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3859777250319518,
+      "grad_norm": 0.8127278685569763,
+      "learning_rate": 3.857664233576642e-05,
+      "loss": 0.7995,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3863428884425781,
+      "grad_norm": 0.7582124471664429,
+      "learning_rate": 3.861313868613139e-05,
+      "loss": 0.8107,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3867080518532043,
+      "grad_norm": 1.0248321294784546,
+      "learning_rate": 3.864963503649636e-05,
+      "loss": 0.7926,
+      "step": 1059
+    },
+    {
+      "epoch": 0.38707321526383054,
+      "grad_norm": 0.9556481242179871,
+      "learning_rate": 3.868613138686132e-05,
+      "loss": 0.801,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3874383786744568,
+      "grad_norm": 0.9816288352012634,
+      "learning_rate": 3.872262773722628e-05,
+      "loss": 0.8016,
+      "step": 1061
+    },
+    {
+      "epoch": 0.38780354208508305,
+      "grad_norm": 0.8992881178855896,
+      "learning_rate": 3.875912408759124e-05,
+      "loss": 0.7817,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3881687054957093,
+      "grad_norm": 1.2685658931732178,
+      "learning_rate": 3.879562043795621e-05,
+      "loss": 0.8052,
+      "step": 1063
+    },
+    {
+      "epoch": 0.38853386890633557,
+      "grad_norm": 1.1701579093933105,
+      "learning_rate": 3.883211678832117e-05,
+      "loss": 0.7946,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3888990323169618,
+      "grad_norm": 1.1810351610183716,
+      "learning_rate": 3.886861313868613e-05,
+      "loss": 0.8519,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3892641957275881,
+      "grad_norm": 0.8648379445075989,
+      "learning_rate": 3.89051094890511e-05,
+      "loss": 0.8398,
+      "step": 1066
+    },
+    {
+      "epoch": 0.38962935913821434,
+      "grad_norm": 0.9467642903327942,
+      "learning_rate": 3.894160583941606e-05,
+      "loss": 0.7784,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3899945225488406,
+      "grad_norm": 0.9347090721130371,
+      "learning_rate": 3.897810218978102e-05,
+      "loss": 0.7626,
+      "step": 1068
+    },
+    {
+      "epoch": 0.39035968595946685,
+      "grad_norm": 1.0562098026275635,
+      "learning_rate": 3.901459854014599e-05,
+      "loss": 0.7798,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3907248493700931,
+      "grad_norm": 0.8048838973045349,
+      "learning_rate": 3.905109489051095e-05,
+      "loss": 0.7616,
+      "step": 1070
+    },
+    {
+      "epoch": 0.39109001278071936,
+      "grad_norm": 1.2834361791610718,
+      "learning_rate": 3.908759124087591e-05,
+      "loss": 0.7972,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3914551761913456,
+      "grad_norm": 0.891139030456543,
+      "learning_rate": 3.912408759124088e-05,
+      "loss": 0.7899,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3918203396019719,
+      "grad_norm": 1.0749610662460327,
+      "learning_rate": 3.916058394160584e-05,
+      "loss": 0.8253,
+      "step": 1073
+    },
+    {
+      "epoch": 0.39218550301259814,
+      "grad_norm": 0.7955366969108582,
+      "learning_rate": 3.91970802919708e-05,
+      "loss": 0.7832,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3925506664232244,
+      "grad_norm": 1.152219533920288,
+      "learning_rate": 3.923357664233577e-05,
+      "loss": 0.7932,
+      "step": 1075
+    },
+    {
+      "epoch": 0.39291582983385065,
+      "grad_norm": 1.2308257818222046,
+      "learning_rate": 3.927007299270073e-05,
+      "loss": 0.7938,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3932809932444769,
+      "grad_norm": 0.9965977072715759,
+      "learning_rate": 3.93065693430657e-05,
+      "loss": 0.8392,
+      "step": 1077
+    },
+    {
+      "epoch": 0.39364615665510316,
+      "grad_norm": 0.9761756062507629,
+      "learning_rate": 3.934306569343066e-05,
+      "loss": 0.7936,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3940113200657294,
+      "grad_norm": 1.0288221836090088,
+      "learning_rate": 3.937956204379562e-05,
+      "loss": 0.816,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3943764834763557,
+      "grad_norm": 1.1133888959884644,
+      "learning_rate": 3.941605839416059e-05,
+      "loss": 0.822,
+      "step": 1080
+    },
+    {
+      "epoch": 0.39474164688698193,
+      "grad_norm": 0.9675469994544983,
+      "learning_rate": 3.945255474452555e-05,
+      "loss": 0.7877,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3951068102976082,
+      "grad_norm": 1.2215927839279175,
+      "learning_rate": 3.948905109489051e-05,
+      "loss": 0.8262,
+      "step": 1082
+    },
+    {
+      "epoch": 0.39547197370823445,
+      "grad_norm": 1.0631349086761475,
+      "learning_rate": 3.952554744525548e-05,
+      "loss": 0.7987,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3958371371188607,
+      "grad_norm": 1.1623072624206543,
+      "learning_rate": 3.956204379562044e-05,
+      "loss": 0.7977,
+      "step": 1084
+    },
+    {
+      "epoch": 0.39620230052948696,
+      "grad_norm": 0.8911407589912415,
+      "learning_rate": 3.95985401459854e-05,
+      "loss": 0.7855,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3965674639401132,
+      "grad_norm": 1.10833740234375,
+      "learning_rate": 3.963503649635037e-05,
+      "loss": 0.8246,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3969326273507395,
+      "grad_norm": 1.0649516582489014,
+      "learning_rate": 3.967153284671533e-05,
+      "loss": 0.7784,
+      "step": 1087
+    },
+    {
+      "epoch": 0.39729779076136573,
+      "grad_norm": 0.9672961235046387,
+      "learning_rate": 3.9708029197080294e-05,
+      "loss": 0.8223,
+      "step": 1088
+    },
+    {
+      "epoch": 0.397662954171992,
+      "grad_norm": 1.1473028659820557,
+      "learning_rate": 3.974452554744526e-05,
+      "loss": 0.7705,
+      "step": 1089
+    },
+    {
+      "epoch": 0.39802811758261825,
+      "grad_norm": 0.9718714952468872,
+      "learning_rate": 3.978102189781022e-05,
+      "loss": 0.8162,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3983932809932445,
+      "grad_norm": 1.0650386810302734,
+      "learning_rate": 3.9817518248175184e-05,
+      "loss": 0.7981,
+      "step": 1091
+    },
+    {
+      "epoch": 0.39875844440387076,
+      "grad_norm": 1.0443297624588013,
+      "learning_rate": 3.9854014598540145e-05,
+      "loss": 0.8149,
+      "step": 1092
+    },
+    {
+      "epoch": 0.39912360781449696,
+      "grad_norm": 1.0627777576446533,
+      "learning_rate": 3.989051094890511e-05,
+      "loss": 0.8289,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3994887712251232,
+      "grad_norm": 1.1238811016082764,
+      "learning_rate": 3.9927007299270074e-05,
+      "loss": 0.7831,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3998539346357495,
+      "grad_norm": 0.9035513401031494,
+      "learning_rate": 3.9963503649635035e-05,
+      "loss": 0.771,
+      "step": 1095
+    },
+    {
+      "epoch": 0.40021909804637573,
+      "grad_norm": 1.0562816858291626,
+      "learning_rate": 4e-05,
+      "loss": 0.7944,
+      "step": 1096
+    },
+    {
+      "epoch": 0.400584261457002,
+      "grad_norm": 0.821844220161438,
+      "learning_rate": 3.999999485540128e-05,
+      "loss": 0.8033,
+      "step": 1097
+    },
+    {
+      "epoch": 0.40094942486762825,
+      "grad_norm": 0.959084689617157,
+      "learning_rate": 3.999997942160775e-05,
+      "loss": 0.7932,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4013145882782545,
+      "grad_norm": 0.7904175519943237,
+      "learning_rate": 3.9999953698627355e-05,
+      "loss": 0.787,
+      "step": 1099
+    },
+    {
+      "epoch": 0.40167975168888076,
+      "grad_norm": 0.7241582870483398,
+      "learning_rate": 3.9999917686473335e-05,
+      "loss": 0.7588,
+      "step": 1100
+    },
+    {
+      "epoch": 0.402044915099507,
+      "grad_norm": 0.9488881826400757,
+      "learning_rate": 3.9999871385164215e-05,
+      "loss": 0.7584,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4024100785101333,
+      "grad_norm": 0.968736469745636,
+      "learning_rate": 3.9999814794723805e-05,
+      "loss": 0.8066,
+      "step": 1102
+    },
+    {
+      "epoch": 0.40277524192075953,
+      "grad_norm": 0.7286505103111267,
+      "learning_rate": 3.999974791518123e-05,
+      "loss": 0.8229,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4031404053313858,
+      "grad_norm": 1.0029511451721191,
+      "learning_rate": 3.999967074657089e-05,
+      "loss": 0.7956,
+      "step": 1104
+    },
+    {
+      "epoch": 0.40350556874201204,
+      "grad_norm": 1.0827014446258545,
+      "learning_rate": 3.9999583288932495e-05,
+      "loss": 0.764,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4038707321526383,
+      "grad_norm": 0.9567440152168274,
+      "learning_rate": 3.999948554231102e-05,
+      "loss": 0.7845,
+      "step": 1106
+    },
+    {
+      "epoch": 0.40423589556326456,
+      "grad_norm": 0.6938320994377136,
+      "learning_rate": 3.9999377506756765e-05,
+      "loss": 0.7732,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4046010589738908,
+      "grad_norm": 0.9644039869308472,
+      "learning_rate": 3.9999259182325315e-05,
+      "loss": 0.7891,
+      "step": 1108
+    },
+    {
+      "epoch": 0.40496622238451707,
+      "grad_norm": 1.138069748878479,
+      "learning_rate": 3.999913056907753e-05,
+      "loss": 0.7866,
+      "step": 1109
+    },
+    {
+      "epoch": 0.40533138579514333,
+      "grad_norm": 0.8856624364852905,
+      "learning_rate": 3.999899166707959e-05,
+      "loss": 0.7845,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4056965492057696,
+      "grad_norm": 1.0707448720932007,
+      "learning_rate": 3.999884247640293e-05,
+      "loss": 0.7736,
+      "step": 1111
+    },
+    {
+      "epoch": 0.40606171261639584,
+      "grad_norm": 1.0259274244308472,
+      "learning_rate": 3.999868299712434e-05,
+      "loss": 0.8112,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4064268760270221,
+      "grad_norm": 1.0142202377319336,
+      "learning_rate": 3.999851322932583e-05,
+      "loss": 0.7892,
+      "step": 1113
+    },
+    {
+      "epoch": 0.40679203943764836,
+      "grad_norm": 0.7966458797454834,
+      "learning_rate": 3.9998333173094764e-05,
+      "loss": 0.797,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4071572028482746,
+      "grad_norm": 1.1583501100540161,
+      "learning_rate": 3.999814282852375e-05,
+      "loss": 0.7563,
+      "step": 1115
+    },
+    {
+      "epoch": 0.40752236625890087,
+      "grad_norm": 1.2791121006011963,
+      "learning_rate": 3.9997942195710744e-05,
+      "loss": 0.7943,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4078875296695271,
+      "grad_norm": 0.9490485191345215,
+      "learning_rate": 3.999773127475894e-05,
+      "loss": 0.7856,
+      "step": 1117
+    },
+    {
+      "epoch": 0.4082526930801534,
+      "grad_norm": 1.0569360256195068,
+      "learning_rate": 3.9997510065776843e-05,
+      "loss": 0.792,
+      "step": 1118
+    },
+    {
+      "epoch": 0.40861785649077964,
+      "grad_norm": 1.1657289266586304,
+      "learning_rate": 3.9997278568878275e-05,
+      "loss": 0.755,
+      "step": 1119
+    },
+    {
+      "epoch": 0.4089830199014059,
+      "grad_norm": 0.8284951448440552,
+      "learning_rate": 3.9997036784182325e-05,
+      "loss": 0.7788,
+      "step": 1120
+    },
+    {
+      "epoch": 0.40934818331203215,
+      "grad_norm": 0.819286048412323,
+      "learning_rate": 3.999678471181338e-05,
+      "loss": 0.7683,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4097133467226584,
+      "grad_norm": 1.0198330879211426,
+      "learning_rate": 3.999652235190112e-05,
+      "loss": 0.802,
+      "step": 1122
+    },
+    {
+      "epoch": 0.41007851013328467,
+      "grad_norm": 1.0188407897949219,
+      "learning_rate": 3.999624970458053e-05,
+      "loss": 0.7964,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4104436735439109,
+      "grad_norm": 0.8926658034324646,
+      "learning_rate": 3.999596676999185e-05,
+      "loss": 0.785,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4108088369545372,
+      "grad_norm": 1.0475916862487793,
+      "learning_rate": 3.999567354828067e-05,
+      "loss": 0.8413,
+      "step": 1125
+    },
+    {
+      "epoch": 0.4111740003651634,
+      "grad_norm": 1.21036958694458,
+      "learning_rate": 3.9995370039597826e-05,
+      "loss": 0.7864,
+      "step": 1126
+    },
+    {
+      "epoch": 0.41153916377578964,
+      "grad_norm": 0.8281807899475098,
+      "learning_rate": 3.9995056244099444e-05,
+      "loss": 0.7778,
+      "step": 1127
+    },
+    {
+      "epoch": 0.4119043271864159,
+      "grad_norm": 1.1479681730270386,
+      "learning_rate": 3.9994732161946986e-05,
+      "loss": 0.7848,
+      "step": 1128
+    },
+    {
+      "epoch": 0.41226949059704215,
+      "grad_norm": 0.6831890940666199,
+      "learning_rate": 3.9994397793307175e-05,
+      "loss": 0.7786,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4126346540076684,
+      "grad_norm": 0.9135376214981079,
+      "learning_rate": 3.999405313835202e-05,
+      "loss": 0.7648,
+      "step": 1130
+    },
+    {
+      "epoch": 0.41299981741829467,
+      "grad_norm": 0.7735843658447266,
+      "learning_rate": 3.999369819725884e-05,
+      "loss": 0.793,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4133649808289209,
+      "grad_norm": 0.8167039752006531,
+      "learning_rate": 3.999333297021023e-05,
+      "loss": 0.7635,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4137301442395472,
+      "grad_norm": 0.9002150297164917,
+      "learning_rate": 3.999295745739409e-05,
+      "loss": 0.7849,
+      "step": 1133
+    },
+    {
+      "epoch": 0.41409530765017344,
+      "grad_norm": 1.089339017868042,
+      "learning_rate": 3.999257165900361e-05,
+      "loss": 0.7899,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4144604710607997,
+      "grad_norm": 1.2341511249542236,
+      "learning_rate": 3.999217557523725e-05,
+      "loss": 0.7859,
+      "step": 1135
+    },
+    {
+      "epoch": 0.41482563447142595,
+      "grad_norm": 1.0575354099273682,
+      "learning_rate": 3.9991769206298805e-05,
+      "loss": 0.787,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4151907978820522,
+      "grad_norm": 1.286300539970398,
+      "learning_rate": 3.999135255239732e-05,
+      "loss": 0.8298,
+      "step": 1137
+    },
+    {
+      "epoch": 0.41555596129267847,
+      "grad_norm": 0.8028886318206787,
+      "learning_rate": 3.999092561374715e-05,
+      "loss": 0.7725,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4159211247033047,
+      "grad_norm": 1.2141560316085815,
+      "learning_rate": 3.999048839056793e-05,
+      "loss": 0.83,
+      "step": 1139
+    },
+    {
+      "epoch": 0.416286288113931,
+      "grad_norm": 0.7901906371116638,
+      "learning_rate": 3.999004088308462e-05,
+      "loss": 0.8021,
+      "step": 1140
+    },
+    {
+      "epoch": 0.41665145152455724,
+      "grad_norm": 1.0362855195999146,
+      "learning_rate": 3.998958309152741e-05,
+      "loss": 0.7839,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4170166149351835,
+      "grad_norm": 1.0257155895233154,
+      "learning_rate": 3.998911501613184e-05,
+      "loss": 0.8022,
+      "step": 1142
+    },
+    {
+      "epoch": 0.41738177834580975,
+      "grad_norm": 0.9644251465797424,
+      "learning_rate": 3.99886366571387e-05,
+      "loss": 0.7898,
+      "step": 1143
+    },
+    {
+      "epoch": 0.417746941756436,
+      "grad_norm": 0.9577879905700684,
+      "learning_rate": 3.99881480147941e-05,
+      "loss": 0.765,
+      "step": 1144
+    },
+    {
+      "epoch": 0.41811210516706226,
+      "grad_norm": 0.9823921918869019,
+      "learning_rate": 3.998764908934942e-05,
+      "loss": 0.8112,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4184772685776885,
+      "grad_norm": 1.0017061233520508,
+      "learning_rate": 3.998713988106134e-05,
+      "loss": 0.8097,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4188424319883148,
+      "grad_norm": 0.8555564880371094,
+      "learning_rate": 3.9986620390191815e-05,
+      "loss": 0.7759,
+      "step": 1147
+    },
+    {
+      "epoch": 0.41920759539894104,
+      "grad_norm": 1.3428617715835571,
+      "learning_rate": 3.998609061700812e-05,
+      "loss": 0.7601,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4195727588095673,
+      "grad_norm": 1.0684926509857178,
+      "learning_rate": 3.998555056178279e-05,
+      "loss": 0.8042,
+      "step": 1149
+    },
+    {
+      "epoch": 0.41993792222019355,
+      "grad_norm": 0.8180867433547974,
+      "learning_rate": 3.998500022479367e-05,
+      "loss": 0.7991,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4203030856308198,
+      "grad_norm": 0.9196853041648865,
+      "learning_rate": 3.998443960632388e-05,
+      "loss": 0.7787,
+      "step": 1151
+    },
+    {
+      "epoch": 0.42066824904144606,
+      "grad_norm": 0.7535399794578552,
+      "learning_rate": 3.998386870666185e-05,
+      "loss": 0.793,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4210334124520723,
+      "grad_norm": 0.9972962141036987,
+      "learning_rate": 3.9983287526101256e-05,
+      "loss": 0.7958,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4213985758626986,
+      "grad_norm": 1.0572669506072998,
+      "learning_rate": 3.9982696064941116e-05,
+      "loss": 0.7844,
+      "step": 1154
+    },
+    {
+      "epoch": 0.42176373927332483,
+      "grad_norm": 1.0186238288879395,
+      "learning_rate": 3.9982094323485706e-05,
+      "loss": 0.7692,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4221289026839511,
+      "grad_norm": 0.851263165473938,
+      "learning_rate": 3.9981482302044604e-05,
+      "loss": 0.7744,
+      "step": 1156
+    },
+    {
+      "epoch": 0.42249406609457735,
+      "grad_norm": 1.0665268898010254,
+      "learning_rate": 3.998086000093266e-05,
+      "loss": 0.7941,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4228592295052036,
+      "grad_norm": 0.9266101717948914,
+      "learning_rate": 3.998022742047002e-05,
+      "loss": 0.802,
+      "step": 1158
+    },
+    {
+      "epoch": 0.42322439291582986,
+      "grad_norm": 0.9939461946487427,
+      "learning_rate": 3.9979584560982144e-05,
+      "loss": 0.793,
+      "step": 1159
+    },
+    {
+      "epoch": 0.42358955632645606,
+      "grad_norm": 1.3110156059265137,
+      "learning_rate": 3.997893142279973e-05,
+      "loss": 0.8345,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4239547197370823,
+      "grad_norm": 1.0854302644729614,
+      "learning_rate": 3.997826800625881e-05,
+      "loss": 0.8214,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4243198831477086,
+      "grad_norm": 0.5960488319396973,
+      "learning_rate": 3.9977594311700676e-05,
+      "loss": 0.7914,
+      "step": 1162
+    },
+    {
+      "epoch": 0.42468504655833483,
+      "grad_norm": 1.0404022932052612,
+      "learning_rate": 3.9976910339471914e-05,
+      "loss": 0.7772,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4250502099689611,
+      "grad_norm": 1.0241742134094238,
+      "learning_rate": 3.9976216089924415e-05,
+      "loss": 0.7376,
+      "step": 1164
+    },
+    {
+      "epoch": 0.42541537337958735,
+      "grad_norm": 1.1148185729980469,
+      "learning_rate": 3.9975511563415336e-05,
+      "loss": 0.8058,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4257805367902136,
+      "grad_norm": 0.7152006030082703,
+      "learning_rate": 3.997479676030711e-05,
+      "loss": 0.7782,
+      "step": 1166
+    },
+    {
+      "epoch": 0.42614570020083986,
+      "grad_norm": 0.9840139746665955,
+      "learning_rate": 3.9974071680967504e-05,
+      "loss": 0.8091,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4265108636114661,
+      "grad_norm": 0.9908798336982727,
+      "learning_rate": 3.9973336325769526e-05,
+      "loss": 0.7751,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4268760270220924,
+      "grad_norm": 1.0988410711288452,
+      "learning_rate": 3.9972590695091476e-05,
+      "loss": 0.7939,
+      "step": 1169
+    },
+    {
+      "epoch": 0.42724119043271863,
+      "grad_norm": 0.8988221883773804,
+      "learning_rate": 3.997183478931698e-05,
+      "loss": 0.7952,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4276063538433449,
+      "grad_norm": 0.8373317718505859,
+      "learning_rate": 3.9971068608834895e-05,
+      "loss": 0.7942,
+      "step": 1171
+    },
+    {
+      "epoch": 0.42797151725397115,
+      "grad_norm": 1.4707140922546387,
+      "learning_rate": 3.9970292154039396e-05,
+      "loss": 0.7861,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4283366806645974,
+      "grad_norm": 1.840456485748291,
+      "learning_rate": 3.9969505425329955e-05,
+      "loss": 0.7849,
+      "step": 1173
+    },
+    {
+      "epoch": 0.42870184407522366,
+      "grad_norm": 0.8191938400268555,
+      "learning_rate": 3.996870842311129e-05,
+      "loss": 0.816,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4290670074858499,
+      "grad_norm": 1.8431330919265747,
+      "learning_rate": 3.9967901147793436e-05,
+      "loss": 0.7844,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4294321708964762,
+      "grad_norm": 0.8188057541847229,
+      "learning_rate": 3.99670835997917e-05,
+      "loss": 0.7913,
+      "step": 1176
+    },
+    {
+      "epoch": 0.42979733430710243,
+      "grad_norm": 0.8033564686775208,
+      "learning_rate": 3.996625577952669e-05,
+      "loss": 0.7655,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4301624977177287,
+      "grad_norm": 1.0804041624069214,
+      "learning_rate": 3.9965417687424274e-05,
+      "loss": 0.8119,
+      "step": 1178
+    },
+    {
+      "epoch": 0.43052766112835494,
+      "grad_norm": 0.9008678793907166,
+      "learning_rate": 3.996456932391562e-05,
+      "loss": 0.8007,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4308928245389812,
+      "grad_norm": 0.7237406373023987,
+      "learning_rate": 3.9963710689437174e-05,
+      "loss": 0.7888,
+      "step": 1180
+    },
+    {
+      "epoch": 0.43125798794960746,
+      "grad_norm": 0.8602893948554993,
+      "learning_rate": 3.996284178443068e-05,
+      "loss": 0.7904,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4316231513602337,
+      "grad_norm": 1.1926499605178833,
+      "learning_rate": 3.996196260934314e-05,
+      "loss": 0.7903,
+      "step": 1182
+    },
+    {
+      "epoch": 0.43198831477085997,
+      "grad_norm": 1.187950849533081,
+      "learning_rate": 3.996107316462686e-05,
+      "loss": 0.7792,
+      "step": 1183
+    },
+    {
+      "epoch": 0.43235347818148623,
+      "grad_norm": 0.7057816386222839,
+      "learning_rate": 3.9960173450739425e-05,
+      "loss": 0.7725,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4327186415921125,
+      "grad_norm": 1.08821702003479,
+      "learning_rate": 3.9959263468143706e-05,
+      "loss": 0.7986,
+      "step": 1185
+    },
+    {
+      "epoch": 0.43308380500273874,
+      "grad_norm": 0.9124771952629089,
+      "learning_rate": 3.995834321730785e-05,
+      "loss": 0.8041,
+      "step": 1186
+    },
+    {
+      "epoch": 0.433448968413365,
+      "grad_norm": 0.9151847958564758,
+      "learning_rate": 3.995741269870528e-05,
+      "loss": 0.8016,
+      "step": 1187
+    },
+    {
+      "epoch": 0.43381413182399126,
+      "grad_norm": 1.1104234457015991,
+      "learning_rate": 3.9956471912814715e-05,
+      "loss": 0.7729,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4341792952346175,
+      "grad_norm": 0.7967671155929565,
+      "learning_rate": 3.9955520860120164e-05,
+      "loss": 0.7914,
+      "step": 1189
+    },
+    {
+      "epoch": 0.43454445864524377,
+      "grad_norm": 0.9998862147331238,
+      "learning_rate": 3.995455954111089e-05,
+      "loss": 0.7782,
+      "step": 1190
+    },
+    {
+      "epoch": 0.43490962205587,
+      "grad_norm": 0.8886136412620544,
+      "learning_rate": 3.995358795628146e-05,
+      "loss": 0.7969,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4352747854664963,
+      "grad_norm": 0.7756946086883545,
+      "learning_rate": 3.995260610613172e-05,
+      "loss": 0.7875,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4356399488771225,
+      "grad_norm": 0.9421478509902954,
+      "learning_rate": 3.995161399116678e-05,
+      "loss": 0.7557,
+      "step": 1193
+    },
+    {
+      "epoch": 0.43600511228774874,
+      "grad_norm": 0.9283790588378906,
+      "learning_rate": 3.9950611611897055e-05,
+      "loss": 0.7792,
+      "step": 1194
+    },
+    {
+      "epoch": 0.436370275698375,
+      "grad_norm": 0.7054854035377502,
+      "learning_rate": 3.994959896883821e-05,
+      "loss": 0.7354,
+      "step": 1195
+    },
+    {
+      "epoch": 0.43673543910900126,
+      "grad_norm": 0.9545449614524841,
+      "learning_rate": 3.994857606251124e-05,
+      "loss": 0.7786,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4371006025196275,
+      "grad_norm": 0.8932951092720032,
+      "learning_rate": 3.994754289344236e-05,
+      "loss": 0.795,
+      "step": 1197
+    },
+    {
+      "epoch": 0.43746576593025377,
+      "grad_norm": 0.7554992437362671,
+      "learning_rate": 3.9946499462163116e-05,
+      "loss": 0.7732,
+      "step": 1198
+    },
+    {
+      "epoch": 0.43783092934088,
+      "grad_norm": 1.0914173126220703,
+      "learning_rate": 3.99454457692103e-05,
+      "loss": 0.795,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4381960927515063,
+      "grad_norm": 0.92984938621521,
+      "learning_rate": 3.9944381815125987e-05,
+      "loss": 0.7977,
+      "step": 1200
+    },
+    {
+      "epoch": 0.43856125616213254,
+      "grad_norm": 0.8287634253501892,
+      "learning_rate": 3.9943307600457563e-05,
+      "loss": 0.7723,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4389264195727588,
+      "grad_norm": 1.0759422779083252,
+      "learning_rate": 3.994222312575764e-05,
+      "loss": 0.7911,
+      "step": 1202
+    },
+    {
+      "epoch": 0.43929158298338505,
+      "grad_norm": 0.7861177921295166,
+      "learning_rate": 3.994112839158416e-05,
+      "loss": 0.7568,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4396567463940113,
+      "grad_norm": 0.8400720357894897,
+      "learning_rate": 3.99400233985003e-05,
+      "loss": 0.7809,
+      "step": 1204
+    },
+    {
+      "epoch": 0.44002190980463757,
+      "grad_norm": 1.083005666732788,
+      "learning_rate": 3.993890814707455e-05,
+      "loss": 0.7705,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4403870732152638,
+      "grad_norm": 1.2543330192565918,
+      "learning_rate": 3.9937782637880665e-05,
+      "loss": 0.8156,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4407522366258901,
+      "grad_norm": 0.8394210338592529,
+      "learning_rate": 3.9936646871497656e-05,
+      "loss": 0.7837,
+      "step": 1207
+    },
+    {
+      "epoch": 0.44111740003651634,
+      "grad_norm": 0.8094637393951416,
+      "learning_rate": 3.9935500848509845e-05,
+      "loss": 0.7788,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4414825634471426,
+      "grad_norm": 0.898333728313446,
+      "learning_rate": 3.993434456950681e-05,
+      "loss": 0.7497,
+      "step": 1209
+    },
+    {
+      "epoch": 0.44184772685776885,
+      "grad_norm": 0.8007587790489197,
+      "learning_rate": 3.9933178035083406e-05,
+      "loss": 0.7402,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4422128902683951,
+      "grad_norm": 0.8324313759803772,
+      "learning_rate": 3.993200124583977e-05,
+      "loss": 0.7768,
+      "step": 1211
+    },
+    {
+      "epoch": 0.44257805367902137,
+      "grad_norm": 1.1105968952178955,
+      "learning_rate": 3.993081420238132e-05,
+      "loss": 0.7723,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4429432170896476,
+      "grad_norm": 0.8521186709403992,
+      "learning_rate": 3.992961690531873e-05,
+      "loss": 0.7604,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4433083805002739,
+      "grad_norm": 0.850679337978363,
+      "learning_rate": 3.992840935526797e-05,
+      "loss": 0.7555,
+      "step": 1214
+    },
+    {
+      "epoch": 0.44367354391090014,
+      "grad_norm": 0.8649604916572571,
+      "learning_rate": 3.992719155285028e-05,
+      "loss": 0.7933,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4440387073215264,
+      "grad_norm": 0.7320381999015808,
+      "learning_rate": 3.992596349869216e-05,
+      "loss": 0.777,
+      "step": 1216
+    },
+    {
+      "epoch": 0.44440387073215265,
+      "grad_norm": 1.2951277494430542,
+      "learning_rate": 3.99247251934254e-05,
+      "loss": 0.8055,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4447690341427789,
+      "grad_norm": 0.9105983972549438,
+      "learning_rate": 3.992347663768705e-05,
+      "loss": 0.7768,
+      "step": 1218
+    },
+    {
+      "epoch": 0.44513419755340516,
+      "grad_norm": 0.9981197118759155,
+      "learning_rate": 3.9922217832119464e-05,
+      "loss": 0.7576,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4454993609640314,
+      "grad_norm": 1.0948731899261475,
+      "learning_rate": 3.992094877737022e-05,
+      "loss": 0.8055,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4458645243746577,
+      "grad_norm": 1.0124883651733398,
+      "learning_rate": 3.991966947409221e-05,
+      "loss": 0.7642,
+      "step": 1221
+    },
+    {
+      "epoch": 0.44622968778528393,
+      "grad_norm": 0.8435676693916321,
+      "learning_rate": 3.991837992294358e-05,
+      "loss": 0.7785,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4465948511959102,
+      "grad_norm": 1.0190502405166626,
+      "learning_rate": 3.991708012458777e-05,
+      "loss": 0.7943,
+      "step": 1223
+    },
+    {
+      "epoch": 0.44696001460653645,
+      "grad_norm": 1.195723056793213,
+      "learning_rate": 3.991577007969344e-05,
+      "loss": 0.8112,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4473251780171627,
+      "grad_norm": 1.138889193534851,
+      "learning_rate": 3.9914449788934584e-05,
+      "loss": 0.7964,
+      "step": 1225
+    },
+    {
+      "epoch": 0.44769034142778896,
+      "grad_norm": 0.848869264125824,
+      "learning_rate": 3.991311925299042e-05,
+      "loss": 0.7592,
+      "step": 1226
+    },
+    {
+      "epoch": 0.44805550483841516,
+      "grad_norm": 1.003063440322876,
+      "learning_rate": 3.991177847254547e-05,
+      "loss": 0.7567,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4484206682490414,
+      "grad_norm": 1.5992316007614136,
+      "learning_rate": 3.991042744828951e-05,
+      "loss": 0.781,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4487858316596677,
+      "grad_norm": 0.8211202621459961,
+      "learning_rate": 3.990906618091758e-05,
+      "loss": 0.7533,
+      "step": 1229
+    },
+    {
+      "epoch": 0.44915099507029393,
+      "grad_norm": 0.915510356426239,
+      "learning_rate": 3.9907694671129996e-05,
+      "loss": 0.7424,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4495161584809202,
+      "grad_norm": 0.9703736305236816,
+      "learning_rate": 3.990631291963236e-05,
+      "loss": 0.7519,
+      "step": 1231
+    },
+    {
+      "epoch": 0.44988132189154645,
+      "grad_norm": 1.0902239084243774,
+      "learning_rate": 3.9904920927135504e-05,
+      "loss": 0.8147,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4502464853021727,
+      "grad_norm": 0.7888534069061279,
+      "learning_rate": 3.9903518694355575e-05,
+      "loss": 0.772,
+      "step": 1233
+    },
+    {
+      "epoch": 0.45061164871279896,
+      "grad_norm": 0.9694254994392395,
+      "learning_rate": 3.990210622201396e-05,
+      "loss": 0.7618,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4509768121234252,
+      "grad_norm": 0.8534974455833435,
+      "learning_rate": 3.9900683510837306e-05,
+      "loss": 0.7538,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4513419755340515,
+      "grad_norm": 0.6274154782295227,
+      "learning_rate": 3.989925056155756e-05,
+      "loss": 0.7614,
+      "step": 1236
+    },
+    {
+      "epoch": 0.45170713894467773,
+      "grad_norm": 0.985271692276001,
+      "learning_rate": 3.9897807374911895e-05,
+      "loss": 0.7385,
+      "step": 1237
+    },
+    {
+      "epoch": 0.452072302355304,
+      "grad_norm": 0.7203537821769714,
+      "learning_rate": 3.9896353951642795e-05,
+      "loss": 0.7429,
+      "step": 1238
+    },
+    {
+      "epoch": 0.45243746576593025,
+      "grad_norm": 0.9454881548881531,
+      "learning_rate": 3.989489029249797e-05,
+      "loss": 0.7766,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4528026291765565,
+      "grad_norm": 0.8015196323394775,
+      "learning_rate": 3.989341639823042e-05,
+      "loss": 0.7528,
+      "step": 1240
+    },
+    {
+      "epoch": 0.45316779258718276,
+      "grad_norm": 0.8929771184921265,
+      "learning_rate": 3.9891932269598414e-05,
+      "loss": 0.7491,
+      "step": 1241
+    },
+    {
+      "epoch": 0.453532955997809,
+      "grad_norm": 0.7127432823181152,
+      "learning_rate": 3.989043790736547e-05,
+      "loss": 0.7749,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4538981194084353,
+      "grad_norm": 1.1709378957748413,
+      "learning_rate": 3.988893331230038e-05,
+      "loss": 0.8119,
+      "step": 1243
+    },
+    {
+      "epoch": 0.45426328281906153,
+      "grad_norm": 0.9406178593635559,
+      "learning_rate": 3.9887418485177175e-05,
+      "loss": 0.7858,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4546284462296878,
+      "grad_norm": 1.2341252565383911,
+      "learning_rate": 3.9885893426775204e-05,
+      "loss": 0.7908,
+      "step": 1245
+    },
+    {
+      "epoch": 0.45499360964031405,
+      "grad_norm": 0.9134424328804016,
+      "learning_rate": 3.988435813787904e-05,
+      "loss": 0.7761,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4553587730509403,
+      "grad_norm": 1.2409087419509888,
+      "learning_rate": 3.988281261927852e-05,
+      "loss": 0.7548,
+      "step": 1247
+    },
+    {
+      "epoch": 0.45572393646156656,
+      "grad_norm": 0.9813235402107239,
+      "learning_rate": 3.9881256871768756e-05,
+      "loss": 0.8041,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4560890998721928,
+      "grad_norm": 0.9483786225318909,
+      "learning_rate": 3.9879690896150114e-05,
+      "loss": 0.7639,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4564542632828191,
+      "grad_norm": 0.9267379641532898,
+      "learning_rate": 3.9878114693228236e-05,
+      "loss": 0.7598,
+      "step": 1250
+    },
+    {
+      "epoch": 0.45681942669344533,
+      "grad_norm": 1.0630885362625122,
+      "learning_rate": 3.9876528263813995e-05,
+      "loss": 0.7939,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4571845901040716,
+      "grad_norm": 0.8597736954689026,
+      "learning_rate": 3.9874931608723566e-05,
+      "loss": 0.7781,
+      "step": 1252
+    },
+    {
+      "epoch": 0.45754975351469784,
+      "grad_norm": 0.8005722761154175,
+      "learning_rate": 3.9873324728778354e-05,
+      "loss": 0.7593,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4579149169253241,
+      "grad_norm": 0.9881361126899719,
+      "learning_rate": 3.9871707624805037e-05,
+      "loss": 0.7797,
+      "step": 1254
+    },
+    {
+      "epoch": 0.45828008033595036,
+      "grad_norm": 0.9582185745239258,
+      "learning_rate": 3.987008029763555e-05,
+      "loss": 0.7805,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4586452437465766,
+      "grad_norm": 0.8981064558029175,
+      "learning_rate": 3.9868442748107076e-05,
+      "loss": 0.7714,
+      "step": 1256
+    },
+    {
+      "epoch": 0.45901040715720287,
+      "grad_norm": 1.3101154565811157,
+      "learning_rate": 3.9866794977062086e-05,
+      "loss": 0.7825,
+      "step": 1257
+    },
+    {
+      "epoch": 0.45937557056782913,
+      "grad_norm": 1.0810472965240479,
+      "learning_rate": 3.986513698534829e-05,
+      "loss": 0.7424,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4597407339784554,
+      "grad_norm": 0.9976481795310974,
+      "learning_rate": 3.9863468773818646e-05,
+      "loss": 0.7919,
+      "step": 1259
+    },
+    {
+      "epoch": 0.46010589738908164,
+      "grad_norm": 1.0877742767333984,
+      "learning_rate": 3.986179034333139e-05,
+      "loss": 0.733,
+      "step": 1260
+    },
+    {
+      "epoch": 0.46047106079970784,
+      "grad_norm": 1.0023266077041626,
+      "learning_rate": 3.986010169475002e-05,
+      "loss": 0.7323,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4608362242103341,
+      "grad_norm": 0.9857776761054993,
+      "learning_rate": 3.985840282894325e-05,
+      "loss": 0.7279,
+      "step": 1262
+    },
+    {
+      "epoch": 0.46120138762096036,
+      "grad_norm": 1.1565684080123901,
+      "learning_rate": 3.9856693746785095e-05,
+      "loss": 0.7472,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4615665510315866,
+      "grad_norm": 1.14054536819458,
+      "learning_rate": 3.9854974449154805e-05,
+      "loss": 0.7546,
+      "step": 1264
+    },
+    {
+      "epoch": 0.46193171444221287,
+      "grad_norm": 1.011195421218872,
+      "learning_rate": 3.985324493693689e-05,
+      "loss": 0.776,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4622968778528391,
+      "grad_norm": 0.7427286505699158,
+      "learning_rate": 3.985150521102113e-05,
+      "loss": 0.7865,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4626620412634654,
+      "grad_norm": 1.1418955326080322,
+      "learning_rate": 3.9849755272302515e-05,
+      "loss": 0.7852,
+      "step": 1267
+    },
+    {
+      "epoch": 0.46302720467409164,
+      "grad_norm": 0.8222767114639282,
+      "learning_rate": 3.984799512168134e-05,
+      "loss": 0.7686,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4633923680847179,
+      "grad_norm": 0.7613946199417114,
+      "learning_rate": 3.9846224760063125e-05,
+      "loss": 0.7412,
+      "step": 1269
+    },
+    {
+      "epoch": 0.46375753149534416,
+      "grad_norm": 0.8206880688667297,
+      "learning_rate": 3.984444418835865e-05,
+      "loss": 0.7761,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4641226949059704,
+      "grad_norm": 0.8956536054611206,
+      "learning_rate": 3.984265340748395e-05,
+      "loss": 0.7845,
+      "step": 1271
+    },
+    {
+      "epoch": 0.46448785831659667,
+      "grad_norm": 1.3047102689743042,
+      "learning_rate": 3.98408524183603e-05,
+      "loss": 0.7668,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4648530217272229,
+      "grad_norm": 1.1105831861495972,
+      "learning_rate": 3.983904122191425e-05,
+      "loss": 0.7783,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4652181851378492,
+      "grad_norm": 0.9006888270378113,
+      "learning_rate": 3.9837219819077584e-05,
+      "loss": 0.7582,
+      "step": 1274
+    },
+    {
+      "epoch": 0.46558334854847544,
+      "grad_norm": 0.736251175403595,
+      "learning_rate": 3.983538821078734e-05,
+      "loss": 0.7581,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4659485119591017,
+      "grad_norm": 1.315656065940857,
+      "learning_rate": 3.98335463979858e-05,
+      "loss": 0.7577,
+      "step": 1276
+    },
+    {
+      "epoch": 0.46631367536972795,
+      "grad_norm": 1.501815676689148,
+      "learning_rate": 3.9831694381620513e-05,
+      "loss": 0.7814,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4666788387803542,
+      "grad_norm": 0.8803642988204956,
+      "learning_rate": 3.982983216264427e-05,
+      "loss": 0.736,
+      "step": 1278
+    },
+    {
+      "epoch": 0.46704400219098047,
+      "grad_norm": 0.748648464679718,
+      "learning_rate": 3.982795974201509e-05,
+      "loss": 0.7902,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4674091656016067,
+      "grad_norm": 0.9274646043777466,
+      "learning_rate": 3.982607712069627e-05,
+      "loss": 0.7483,
+      "step": 1280
+    },
+    {
+      "epoch": 0.467774329012233,
+      "grad_norm": 0.9979590773582458,
+      "learning_rate": 3.982418429965635e-05,
+      "loss": 0.7606,
+      "step": 1281
+    },
+    {
+      "epoch": 0.46813949242285924,
+      "grad_norm": 0.6248117685317993,
+      "learning_rate": 3.982228127986909e-05,
+      "loss": 0.7252,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4685046558334855,
+      "grad_norm": 1.1311684846878052,
+      "learning_rate": 3.9820368062313546e-05,
+      "loss": 0.7732,
+      "step": 1283
+    },
+    {
+      "epoch": 0.46886981924411175,
+      "grad_norm": 1.0071941614151,
+      "learning_rate": 3.981844464797397e-05,
+      "loss": 0.7675,
+      "step": 1284
+    },
+    {
+      "epoch": 0.469234982654738,
+      "grad_norm": 1.47381591796875,
+      "learning_rate": 3.981651103783988e-05,
+      "loss": 0.7737,
+      "step": 1285
+    },
+    {
+      "epoch": 0.46960014606536427,
+      "grad_norm": 1.114651083946228,
+      "learning_rate": 3.9814567232906054e-05,
+      "loss": 0.7441,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4699653094759905,
+      "grad_norm": 0.6732168793678284,
+      "learning_rate": 3.98126132341725e-05,
+      "loss": 0.7352,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4703304728866168,
+      "grad_norm": 0.9219527840614319,
+      "learning_rate": 3.981064904264446e-05,
+      "loss": 0.7621,
+      "step": 1288
+    },
+    {
+      "epoch": 0.47069563629724304,
+      "grad_norm": 1.1353275775909424,
+      "learning_rate": 3.9808674659332445e-05,
+      "loss": 0.7874,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4710607997078693,
+      "grad_norm": 0.9650176167488098,
+      "learning_rate": 3.9806690085252184e-05,
+      "loss": 0.792,
+      "step": 1290
+    },
+    {
+      "epoch": 0.47142596311849555,
+      "grad_norm": 0.731926441192627,
+      "learning_rate": 3.980469532142467e-05,
+      "loss": 0.7357,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4717911265291218,
+      "grad_norm": 1.2001556158065796,
+      "learning_rate": 3.980269036887613e-05,
+      "loss": 0.7553,
+      "step": 1292
+    },
+    {
+      "epoch": 0.47215628993974806,
+      "grad_norm": 1.1558605432510376,
+      "learning_rate": 3.980067522863802e-05,
+      "loss": 0.7749,
+      "step": 1293
+    },
+    {
+      "epoch": 0.47252145335037427,
+      "grad_norm": 0.8277838230133057,
+      "learning_rate": 3.9798649901747064e-05,
+      "loss": 0.778,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4728866167610005,
+      "grad_norm": 1.0746656656265259,
+      "learning_rate": 3.9796614389245205e-05,
+      "loss": 0.797,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4732517801716268,
+      "grad_norm": 0.9039919972419739,
+      "learning_rate": 3.979456869217962e-05,
+      "loss": 0.7693,
+      "step": 1296
+    },
+    {
+      "epoch": 0.47361694358225304,
+      "grad_norm": 1.7959991693496704,
+      "learning_rate": 3.979251281160277e-05,
+      "loss": 0.7856,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4739821069928793,
+      "grad_norm": 0.8877509236335754,
+      "learning_rate": 3.979044674857228e-05,
+      "loss": 0.7679,
+      "step": 1298
+    },
+    {
+      "epoch": 0.47434727040350555,
+      "grad_norm": 3.5038464069366455,
+      "learning_rate": 3.978837050415109e-05,
+      "loss": 0.7628,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4747124338141318,
+      "grad_norm": 0.9983282685279846,
+      "learning_rate": 3.9786284079407325e-05,
+      "loss": 0.7321,
+      "step": 1300
+    },
+    {
+      "epoch": 0.47507759722475806,
+      "grad_norm": 1.103953242301941,
+      "learning_rate": 3.978418747541438e-05,
+      "loss": 0.7543,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4754427606353843,
+      "grad_norm": 1.1223280429840088,
+      "learning_rate": 3.9782080693250875e-05,
+      "loss": 0.7411,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4758079240460106,
+      "grad_norm": 1.0355007648468018,
+      "learning_rate": 3.977996373400066e-05,
+      "loss": 0.7828,
+      "step": 1303
+    },
+    {
+      "epoch": 0.47617308745663683,
+      "grad_norm": 0.9945892691612244,
+      "learning_rate": 3.9777836598752814e-05,
+      "loss": 0.7616,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4765382508672631,
+      "grad_norm": 1.0359525680541992,
+      "learning_rate": 3.977569928860168e-05,
+      "loss": 0.7593,
+      "step": 1305
+    },
+    {
+      "epoch": 0.47690341427788935,
+      "grad_norm": 1.0656094551086426,
+      "learning_rate": 3.977355180464681e-05,
+      "loss": 0.7476,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4772685776885156,
+      "grad_norm": 1.0946440696716309,
+      "learning_rate": 3.9771394147993e-05,
+      "loss": 0.7534,
+      "step": 1307
+    },
+    {
+      "epoch": 0.47763374109914186,
+      "grad_norm": 1.5356392860412598,
+      "learning_rate": 3.976922631975028e-05,
+      "loss": 0.806,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4779989045097681,
+      "grad_norm": 0.8914726376533508,
+      "learning_rate": 3.97670483210339e-05,
+      "loss": 0.7457,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4783640679203944,
+      "grad_norm": 0.9732979536056519,
+      "learning_rate": 3.9764860152964365e-05,
+      "loss": 0.782,
+      "step": 1310
+    },
+    {
+      "epoch": 0.47872923133102063,
+      "grad_norm": 0.9373230934143066,
+      "learning_rate": 3.9762661816667404e-05,
+      "loss": 0.7806,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4790943947416469,
+      "grad_norm": 1.190435528755188,
+      "learning_rate": 3.9760453313273954e-05,
+      "loss": 0.7563,
+      "step": 1312
+    },
+    {
+      "epoch": 0.47945955815227315,
+      "grad_norm": 1.0832659006118774,
+      "learning_rate": 3.9758234643920214e-05,
+      "loss": 0.7213,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4798247215628994,
+      "grad_norm": 0.9625949263572693,
+      "learning_rate": 3.9756005809747604e-05,
+      "loss": 0.7792,
+      "step": 1314
+    },
+    {
+      "epoch": 0.48018988497352566,
+      "grad_norm": 1.1885652542114258,
+      "learning_rate": 3.9753766811902756e-05,
+      "loss": 0.7377,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4805550483841519,
+      "grad_norm": 1.1306437253952026,
+      "learning_rate": 3.975151765153756e-05,
+      "loss": 0.7626,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4809202117947782,
+      "grad_norm": 1.4518258571624756,
+      "learning_rate": 3.9749258329809104e-05,
+      "loss": 0.7622,
+      "step": 1317
+    },
+    {
+      "epoch": 0.48128537520540443,
+      "grad_norm": 1.0818719863891602,
+      "learning_rate": 3.974698884787973e-05,
+      "loss": 0.7561,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4816505386160307,
+      "grad_norm": 0.9876286387443542,
+      "learning_rate": 3.974470920691699e-05,
+      "loss": 0.7512,
+      "step": 1319
+    },
+    {
+      "epoch": 0.48201570202665694,
+      "grad_norm": 2.4548518657684326,
+      "learning_rate": 3.974241940809367e-05,
+      "loss": 0.7738,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4823808654372832,
+      "grad_norm": 0.925119936466217,
+      "learning_rate": 3.9740119452587784e-05,
+      "loss": 0.7352,
+      "step": 1321
+    },
+    {
+      "epoch": 0.48274602884790946,
+      "grad_norm": 0.841683566570282,
+      "learning_rate": 3.9737809341582545e-05,
+      "loss": 0.7432,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4831111922585357,
+      "grad_norm": 1.1326254606246948,
+      "learning_rate": 3.973548907626644e-05,
+      "loss": 0.7402,
+      "step": 1323
+    },
+    {
+      "epoch": 0.483476355669162,
+      "grad_norm": 0.839130699634552,
+      "learning_rate": 3.973315865783314e-05,
+      "loss": 0.7539,
+      "step": 1324
+    },
+    {
+      "epoch": 0.48384151907978823,
+      "grad_norm": 1.288762092590332,
+      "learning_rate": 3.9730818087481554e-05,
+      "loss": 0.7786,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4842066824904145,
+      "grad_norm": 0.9070988893508911,
+      "learning_rate": 3.9728467366415815e-05,
+      "loss": 0.7703,
+      "step": 1326
+    },
+    {
+      "epoch": 0.48457184590104074,
+      "grad_norm": 0.861164391040802,
+      "learning_rate": 3.972610649584526e-05,
+      "loss": 0.7523,
+      "step": 1327
+    },
+    {
+      "epoch": 0.48493700931166694,
+      "grad_norm": 1.061942458152771,
+      "learning_rate": 3.972373547698448e-05,
+      "loss": 0.7517,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4853021727222932,
+      "grad_norm": 1.091475009918213,
+      "learning_rate": 3.9721354311053256e-05,
+      "loss": 0.7523,
+      "step": 1329
+    },
+    {
+      "epoch": 0.48566733613291946,
+      "grad_norm": 0.9926595687866211,
+      "learning_rate": 3.971896299927661e-05,
+      "loss": 0.7897,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4860324995435457,
+      "grad_norm": 1.066672444343567,
+      "learning_rate": 3.971656154288477e-05,
+      "loss": 0.7565,
+      "step": 1331
+    },
+    {
+      "epoch": 0.48639766295417197,
+      "grad_norm": 1.009468674659729,
+      "learning_rate": 3.97141499431132e-05,
+      "loss": 0.7789,
+      "step": 1332
+    },
+    {
+      "epoch": 0.48676282636479823,
+      "grad_norm": 1.015899419784546,
+      "learning_rate": 3.971172820120256e-05,
+      "loss": 0.7394,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4871279897754245,
+      "grad_norm": 0.8125002980232239,
+      "learning_rate": 3.970929631839874e-05,
+      "loss": 0.755,
+      "step": 1334
+    },
+    {
+      "epoch": 0.48749315318605074,
+      "grad_norm": 1.096103549003601,
+      "learning_rate": 3.9706854295952856e-05,
+      "loss": 0.726,
+      "step": 1335
+    },
+    {
+      "epoch": 0.487858316596677,
+      "grad_norm": 0.8944382667541504,
+      "learning_rate": 3.9704402135121214e-05,
+      "loss": 0.7747,
+      "step": 1336
+    },
+    {
+      "epoch": 0.48822348000730326,
+      "grad_norm": 1.324345588684082,
+      "learning_rate": 3.970193983716537e-05,
+      "loss": 0.809,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4885886434179295,
+      "grad_norm": 0.9750800728797913,
+      "learning_rate": 3.9699467403352066e-05,
+      "loss": 0.792,
+      "step": 1338
+    },
+    {
+      "epoch": 0.48895380682855577,
+      "grad_norm": 0.7252906560897827,
+      "learning_rate": 3.9696984834953274e-05,
+      "loss": 0.7506,
+      "step": 1339
+    },
+    {
+      "epoch": 0.489318970239182,
+      "grad_norm": 0.9124673008918762,
+      "learning_rate": 3.969449213324617e-05,
+      "loss": 0.7542,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4896841336498083,
+      "grad_norm": 0.9339630603790283,
+      "learning_rate": 3.969198929951316e-05,
+      "loss": 0.7512,
+      "step": 1341
+    },
+    {
+      "epoch": 0.49004929706043454,
+      "grad_norm": 1.2227842807769775,
+      "learning_rate": 3.9689476335041844e-05,
+      "loss": 0.7772,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4904144604710608,
+      "grad_norm": 0.8912896513938904,
+      "learning_rate": 3.9686953241125045e-05,
+      "loss": 0.762,
+      "step": 1343
+    },
+    {
+      "epoch": 0.49077962388168705,
+      "grad_norm": 0.9073852300643921,
+      "learning_rate": 3.96844200190608e-05,
+      "loss": 0.7357,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4911447872923133,
+      "grad_norm": 0.837660014629364,
+      "learning_rate": 3.968187667015233e-05,
+      "loss": 0.7352,
+      "step": 1345
+    },
+    {
+      "epoch": 0.49150995070293957,
+      "grad_norm": 0.9073949456214905,
+      "learning_rate": 3.9679323195708095e-05,
+      "loss": 0.7387,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4918751141135658,
+      "grad_norm": 0.7531459331512451,
+      "learning_rate": 3.9676759597041765e-05,
+      "loss": 0.7518,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4922402775241921,
+      "grad_norm": 1.2089084386825562,
+      "learning_rate": 3.96741858754722e-05,
+      "loss": 0.7792,
+      "step": 1348
+    },
+    {
+      "epoch": 0.49260544093481834,
+      "grad_norm": 0.8173801302909851,
+      "learning_rate": 3.9671602032323475e-05,
+      "loss": 0.7516,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4929706043454446,
+      "grad_norm": 0.9236201643943787,
+      "learning_rate": 3.9669008068924885e-05,
+      "loss": 0.726,
+      "step": 1350
+    },
+    {
+      "epoch": 0.49333576775607085,
+      "grad_norm": 0.8511644601821899,
+      "learning_rate": 3.9666403986610904e-05,
+      "loss": 0.7853,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4937009311666971,
+      "grad_norm": 1.0044788122177124,
+      "learning_rate": 3.9663789786721235e-05,
+      "loss": 0.7512,
+      "step": 1352
+    },
+    {
+      "epoch": 0.49406609457732337,
+      "grad_norm": 0.9926662445068359,
+      "learning_rate": 3.966116547060078e-05,
+      "loss": 0.7505,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4944312579879496,
+      "grad_norm": 0.8975332975387573,
+      "learning_rate": 3.965853103959965e-05,
+      "loss": 0.7635,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4947964213985759,
+      "grad_norm": 1.0068562030792236,
+      "learning_rate": 3.965588649507314e-05,
+      "loss": 0.7328,
+      "step": 1355
+    },
+    {
+      "epoch": 0.49516158480920214,
+      "grad_norm": 1.42154860496521,
+      "learning_rate": 3.965323183838177e-05,
+      "loss": 0.8043,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4955267482198284,
+      "grad_norm": 1.7872182130813599,
+      "learning_rate": 3.9650567070891256e-05,
+      "loss": 0.7196,
+      "step": 1357
+    },
+    {
+      "epoch": 0.49589191163045465,
+      "grad_norm": 1.0129311084747314,
+      "learning_rate": 3.964789219397252e-05,
+      "loss": 0.7634,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4962570750410809,
+      "grad_norm": 1.1521090269088745,
+      "learning_rate": 3.964520720900167e-05,
+      "loss": 0.7689,
+      "step": 1359
+    },
+    {
+      "epoch": 0.49662223845170717,
+      "grad_norm": 1.0223087072372437,
+      "learning_rate": 3.964251211736002e-05,
+      "loss": 0.76,
+      "step": 1360
+    },
+    {
+      "epoch": 0.49698740186233337,
+      "grad_norm": 1.1527354717254639,
+      "learning_rate": 3.963980692043408e-05,
+      "loss": 0.7615,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4973525652729596,
+      "grad_norm": 1.0977435111999512,
+      "learning_rate": 3.963709161961559e-05,
+      "loss": 0.7582,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4977177286835859,
+      "grad_norm": 1.1074302196502686,
+      "learning_rate": 3.9634366216301445e-05,
+      "loss": 0.761,
+      "step": 1363
+    },
+    {
+      "epoch": 0.49808289209421214,
+      "grad_norm": 0.6354876160621643,
+      "learning_rate": 3.963163071189376e-05,
+      "loss": 0.7312,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4984480555048384,
+      "grad_norm": 0.8839811086654663,
+      "learning_rate": 3.962888510779984e-05,
+      "loss": 0.7709,
+      "step": 1365
+    },
+    {
+      "epoch": 0.49881321891546465,
+      "grad_norm": 1.2112983465194702,
+      "learning_rate": 3.962612940543219e-05,
+      "loss": 0.7876,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4991783823260909,
+      "grad_norm": 1.0086325407028198,
+      "learning_rate": 3.962336360620851e-05,
+      "loss": 0.7424,
+      "step": 1367
+    },
+    {
+      "epoch": 0.49954354573671716,
+      "grad_norm": 0.9076545834541321,
+      "learning_rate": 3.962058771155169e-05,
+      "loss": 0.7189,
+      "step": 1368
+    },
+    {
+      "epoch": 0.4999087091473434,
+      "grad_norm": 0.8234534859657288,
+      "learning_rate": 3.9617801722889815e-05,
+      "loss": 0.7291,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5002738725579697,
+      "grad_norm": 1.1717032194137573,
+      "learning_rate": 3.9615005641656175e-05,
+      "loss": 0.7498,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5006390359685959,
+      "grad_norm": 1.1757103204727173,
+      "learning_rate": 3.961219946928923e-05,
+      "loss": 0.7739,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5010041993792222,
+      "grad_norm": 0.7470300197601318,
+      "learning_rate": 3.960938320723265e-05,
+      "loss": 0.7401,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5013693627898484,
+      "grad_norm": 1.7062366008758545,
+      "learning_rate": 3.960655685693528e-05,
+      "loss": 0.7811,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5017345262004748,
+      "grad_norm": 0.990139365196228,
+      "learning_rate": 3.960372041985117e-05,
+      "loss": 0.7343,
+      "step": 1374
+    },
+    {
+      "epoch": 0.502099689611101,
+      "grad_norm": 1.0472018718719482,
+      "learning_rate": 3.960087389743955e-05,
+      "loss": 0.7577,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5024648530217273,
+      "grad_norm": 0.8663569688796997,
+      "learning_rate": 3.959801729116485e-05,
+      "loss": 0.7581,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5028300164323535,
+      "grad_norm": 1.1145840883255005,
+      "learning_rate": 3.959515060249666e-05,
+      "loss": 0.724,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5031951798429797,
+      "grad_norm": 0.9530529975891113,
+      "learning_rate": 3.959227383290981e-05,
+      "loss": 0.7391,
+      "step": 1378
+    },
+    {
+      "epoch": 0.503560343253606,
+      "grad_norm": 1.080680251121521,
+      "learning_rate": 3.9589386983884245e-05,
+      "loss": 0.7557,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5039255066642322,
+      "grad_norm": 1.1287989616394043,
+      "learning_rate": 3.9586490056905155e-05,
+      "loss": 0.7414,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5042906700748585,
+      "grad_norm": 0.7276008129119873,
+      "learning_rate": 3.958358305346289e-05,
+      "loss": 0.7595,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5046558334854847,
+      "grad_norm": 1.0778876543045044,
+      "learning_rate": 3.958066597505299e-05,
+      "loss": 0.7636,
+      "step": 1382
+    },
+    {
+      "epoch": 0.505020996896111,
+      "grad_norm": 1.0213855504989624,
+      "learning_rate": 3.957773882317615e-05,
+      "loss": 0.7161,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5053861603067372,
+      "grad_norm": 1.043364405632019,
+      "learning_rate": 3.957480159933831e-05,
+      "loss": 0.7676,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5057513237173635,
+      "grad_norm": 1.728406548500061,
+      "learning_rate": 3.957185430505052e-05,
+      "loss": 0.7615,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5061164871279897,
+      "grad_norm": 1.02801513671875,
+      "learning_rate": 3.9568896941829076e-05,
+      "loss": 0.6919,
+      "step": 1386
+    },
+    {
+      "epoch": 0.506481650538616,
+      "grad_norm": 0.7651058435440063,
+      "learning_rate": 3.9565929511195395e-05,
+      "loss": 0.7399,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5068468139492422,
+      "grad_norm": 0.8804731965065002,
+      "learning_rate": 3.9562952014676116e-05,
+      "loss": 0.7604,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5072119773598686,
+      "grad_norm": 1.0950775146484375,
+      "learning_rate": 3.955996445380303e-05,
+      "loss": 0.7845,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5075771407704948,
+      "grad_norm": 0.9987009167671204,
+      "learning_rate": 3.955696683011314e-05,
+      "loss": 0.7397,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5079423041811211,
+      "grad_norm": 0.7001053094863892,
+      "learning_rate": 3.9553959145148585e-05,
+      "loss": 0.7282,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5083074675917473,
+      "grad_norm": 1.055344581604004,
+      "learning_rate": 3.955094140045669e-05,
+      "loss": 0.7471,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5086726310023736,
+      "grad_norm": 1.044423222541809,
+      "learning_rate": 3.954791359758998e-05,
+      "loss": 0.7891,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5090377944129998,
+      "grad_norm": 0.9530107975006104,
+      "learning_rate": 3.9544875738106136e-05,
+      "loss": 0.7474,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5094029578236261,
+      "grad_norm": 1.2218527793884277,
+      "learning_rate": 3.9541827823568016e-05,
+      "loss": 0.7604,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5097681212342523,
+      "grad_norm": 1.285076379776001,
+      "learning_rate": 3.953876985554364e-05,
+      "loss": 0.7841,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5101332846448786,
+      "grad_norm": 0.9849181175231934,
+      "learning_rate": 3.953570183560621e-05,
+      "loss": 0.7719,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5104984480555048,
+      "grad_norm": 1.0370405912399292,
+      "learning_rate": 3.953262376533412e-05,
+      "loss": 0.7454,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5108636114661311,
+      "grad_norm": 0.9289689064025879,
+      "learning_rate": 3.9529535646310876e-05,
+      "loss": 0.7605,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5112287748767573,
+      "grad_norm": 1.2506901025772095,
+      "learning_rate": 3.9526437480125227e-05,
+      "loss": 0.7914,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5115939382873836,
+      "grad_norm": 0.8542978167533875,
+      "learning_rate": 3.952332926837105e-05,
+      "loss": 0.7587,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5119591016980098,
+      "grad_norm": 1.0645527839660645,
+      "learning_rate": 3.9520211012647366e-05,
+      "loss": 0.7507,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5123242651086362,
+      "grad_norm": 1.0783454179763794,
+      "learning_rate": 3.951708271455843e-05,
+      "loss": 0.7578,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5126894285192624,
+      "grad_norm": 0.8975682258605957,
+      "learning_rate": 3.95139443757136e-05,
+      "loss": 0.7546,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5130545919298887,
+      "grad_norm": 1.2350717782974243,
+      "learning_rate": 3.951079599772744e-05,
+      "loss": 0.7856,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5134197553405149,
+      "grad_norm": 1.0456730127334595,
+      "learning_rate": 3.950763758221966e-05,
+      "loss": 0.7516,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5137849187511412,
+      "grad_norm": 0.9096729755401611,
+      "learning_rate": 3.950446913081513e-05,
+      "loss": 0.7571,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5141500821617674,
+      "grad_norm": 0.8735365271568298,
+      "learning_rate": 3.9501290645143905e-05,
+      "loss": 0.7809,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5145152455723937,
+      "grad_norm": 1.090287446975708,
+      "learning_rate": 3.949810212684117e-05,
+      "loss": 0.7471,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5148804089830199,
+      "grad_norm": 0.8822610378265381,
+      "learning_rate": 3.949490357754731e-05,
+      "loss": 0.7285,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5152455723936462,
+      "grad_norm": 0.9613785743713379,
+      "learning_rate": 3.9491694998907835e-05,
+      "loss": 0.7302,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5156107358042724,
+      "grad_norm": 0.9969671368598938,
+      "learning_rate": 3.948847639257344e-05,
+      "loss": 0.7629,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5159758992148986,
+      "grad_norm": 1.1526439189910889,
+      "learning_rate": 3.948524776019997e-05,
+      "loss": 0.765,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5163410626255249,
+      "grad_norm": 0.8454238176345825,
+      "learning_rate": 3.9482009103448415e-05,
+      "loss": 0.7268,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5167062260361511,
+      "grad_norm": 0.8136591911315918,
+      "learning_rate": 3.947876042398494e-05,
+      "loss": 0.7429,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5170713894467774,
+      "grad_norm": 1.1326630115509033,
+      "learning_rate": 3.947550172348087e-05,
+      "loss": 0.7675,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5174365528574036,
+      "grad_norm": 0.917960524559021,
+      "learning_rate": 3.947223300361265e-05,
+      "loss": 0.7319,
+      "step": 1417
+    },
+    {
+      "epoch": 0.51780171626803,
+      "grad_norm": 0.93192058801651,
+      "learning_rate": 3.946895426606194e-05,
+      "loss": 0.7578,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5181668796786562,
+      "grad_norm": 0.9007557034492493,
+      "learning_rate": 3.946566551251549e-05,
+      "loss": 0.7427,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5185320430892825,
+      "grad_norm": 1.0617369413375854,
+      "learning_rate": 3.946236674466524e-05,
+      "loss": 0.7753,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5188972064999087,
+      "grad_norm": 0.7483951449394226,
+      "learning_rate": 3.945905796420828e-05,
+      "loss": 0.7314,
+      "step": 1421
+    },
+    {
+      "epoch": 0.519262369910535,
+      "grad_norm": 0.8350456357002258,
+      "learning_rate": 3.945573917284685e-05,
+      "loss": 0.7306,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5196275333211612,
+      "grad_norm": 0.791534960269928,
+      "learning_rate": 3.945241037228831e-05,
+      "loss": 0.7414,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5199926967317875,
+      "grad_norm": 0.9981648921966553,
+      "learning_rate": 3.944907156424522e-05,
+      "loss": 0.7371,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5203578601424137,
+      "grad_norm": 0.7343462705612183,
+      "learning_rate": 3.9445722750435244e-05,
+      "loss": 0.7511,
+      "step": 1425
+    },
+    {
+      "epoch": 0.52072302355304,
+      "grad_norm": 0.775215208530426,
+      "learning_rate": 3.944236393258123e-05,
+      "loss": 0.7365,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5210881869636662,
+      "grad_norm": 1.1900813579559326,
+      "learning_rate": 3.9438995112411144e-05,
+      "loss": 0.7915,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5214533503742925,
+      "grad_norm": 0.951445460319519,
+      "learning_rate": 3.943561629165811e-05,
+      "loss": 0.7274,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5218185137849187,
+      "grad_norm": 0.9051117897033691,
+      "learning_rate": 3.94322274720604e-05,
+      "loss": 0.7634,
+      "step": 1429
+    },
+    {
+      "epoch": 0.522183677195545,
+      "grad_norm": 0.8373913764953613,
+      "learning_rate": 3.942882865536142e-05,
+      "loss": 0.7477,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5225488406061712,
+      "grad_norm": 1.136154294013977,
+      "learning_rate": 3.942541984330972e-05,
+      "loss": 0.7323,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5229140040167976,
+      "grad_norm": 0.8140801191329956,
+      "learning_rate": 3.942200103765901e-05,
+      "loss": 0.7441,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5232791674274238,
+      "grad_norm": 0.7399319410324097,
+      "learning_rate": 3.941857224016812e-05,
+      "loss": 0.7267,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5236443308380501,
+      "grad_norm": 0.8513920307159424,
+      "learning_rate": 3.941513345260104e-05,
+      "loss": 0.7428,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5240094942486763,
+      "grad_norm": 0.9583787322044373,
+      "learning_rate": 3.941168467672687e-05,
+      "loss": 0.7773,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5243746576593026,
+      "grad_norm": 1.1312326192855835,
+      "learning_rate": 3.940822591431988e-05,
+      "loss": 0.7528,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5247398210699288,
+      "grad_norm": 1.0751779079437256,
+      "learning_rate": 3.940475716715946e-05,
+      "loss": 0.7648,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5251049844805551,
+      "grad_norm": 0.7694438695907593,
+      "learning_rate": 3.9401278437030144e-05,
+      "loss": 0.7466,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5254701478911813,
+      "grad_norm": 1.1514031887054443,
+      "learning_rate": 3.9397789725721594e-05,
+      "loss": 0.7599,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5258353113018076,
+      "grad_norm": 1.1192198991775513,
+      "learning_rate": 3.939429103502862e-05,
+      "loss": 0.7544,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5262004747124338,
+      "grad_norm": 0.6443295478820801,
+      "learning_rate": 3.939078236675115e-05,
+      "loss": 0.7329,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5265656381230601,
+      "grad_norm": 0.7252869606018066,
+      "learning_rate": 3.938726372269425e-05,
+      "loss": 0.7239,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5269308015336863,
+      "grad_norm": 0.8113353848457336,
+      "learning_rate": 3.9383735104668135e-05,
+      "loss": 0.7357,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5272959649443126,
+      "grad_norm": 1.05000638961792,
+      "learning_rate": 3.9380196514488126e-05,
+      "loss": 0.7494,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5276611283549388,
+      "grad_norm": 1.1334178447723389,
+      "learning_rate": 3.937664795397469e-05,
+      "loss": 0.7516,
+      "step": 1445
+    },
+    {
+      "epoch": 0.528026291765565,
+      "grad_norm": 1.180598258972168,
+      "learning_rate": 3.937308942495342e-05,
+      "loss": 0.7391,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5283914551761913,
+      "grad_norm": 0.9825596213340759,
+      "learning_rate": 3.936952092925503e-05,
+      "loss": 0.7528,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5287566185868176,
+      "grad_norm": 1.298930287361145,
+      "learning_rate": 3.9365942468715375e-05,
+      "loss": 0.735,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5291217819974439,
+      "grad_norm": 1.0083543062210083,
+      "learning_rate": 3.936235404517543e-05,
+      "loss": 0.731,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5294869454080701,
+      "grad_norm": 0.8948465585708618,
+      "learning_rate": 3.935875566048129e-05,
+      "loss": 0.7324,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5298521088186964,
+      "grad_norm": 1.0688496828079224,
+      "learning_rate": 3.935514731648418e-05,
+      "loss": 0.7656,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5302172722293226,
+      "grad_norm": 1.411491870880127,
+      "learning_rate": 3.935152901504045e-05,
+      "loss": 0.7747,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5305824356399489,
+      "grad_norm": 0.8527326583862305,
+      "learning_rate": 3.934790075801156e-05,
+      "loss": 0.7478,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5309475990505751,
+      "grad_norm": 0.7693621516227722,
+      "learning_rate": 3.934426254726413e-05,
+      "loss": 0.7347,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5313127624612014,
+      "grad_norm": 0.9457259774208069,
+      "learning_rate": 3.934061438466985e-05,
+      "loss": 0.7649,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5316779258718276,
+      "grad_norm": 0.6615961194038391,
+      "learning_rate": 3.933695627210555e-05,
+      "loss": 0.7629,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5320430892824539,
+      "grad_norm": 0.5765036344528198,
+      "learning_rate": 3.93332882114532e-05,
+      "loss": 0.7267,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5324082526930801,
+      "grad_norm": 0.880248486995697,
+      "learning_rate": 3.9329610204599864e-05,
+      "loss": 0.7397,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5327734161037064,
+      "grad_norm": 0.7687374353408813,
+      "learning_rate": 3.932592225343772e-05,
+      "loss": 0.7455,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5331385795143326,
+      "grad_norm": 0.8867073059082031,
+      "learning_rate": 3.932222435986408e-05,
+      "loss": 0.748,
+      "step": 1460
+    },
+    {
+      "epoch": 0.533503742924959,
+      "grad_norm": 1.0743588209152222,
+      "learning_rate": 3.931851652578137e-05,
+      "loss": 0.764,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5338689063355851,
+      "grad_norm": 0.8158539533615112,
+      "learning_rate": 3.93147987530971e-05,
+      "loss": 0.7408,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5342340697462115,
+      "grad_norm": 0.9781192541122437,
+      "learning_rate": 3.9311071043723927e-05,
+      "loss": 0.7259,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5345992331568377,
+      "grad_norm": 1.0258907079696655,
+      "learning_rate": 3.930733339957961e-05,
+      "loss": 0.7621,
+      "step": 1464
+    },
+    {
+      "epoch": 0.534964396567464,
+      "grad_norm": 1.2095290422439575,
+      "learning_rate": 3.9303585822587014e-05,
+      "loss": 0.748,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5353295599780902,
+      "grad_norm": 1.3170123100280762,
+      "learning_rate": 3.929982831467412e-05,
+      "loss": 0.7411,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5356947233887165,
+      "grad_norm": 0.941737949848175,
+      "learning_rate": 3.9296060877774004e-05,
+      "loss": 0.7373,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5360598867993427,
+      "grad_norm": 1.0546380281448364,
+      "learning_rate": 3.9292283513824873e-05,
+      "loss": 0.7261,
+      "step": 1468
+    },
+    {
+      "epoch": 0.536425050209969,
+      "grad_norm": 0.8461571931838989,
+      "learning_rate": 3.928849622477002e-05,
+      "loss": 0.7213,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5367902136205952,
+      "grad_norm": 0.902228832244873,
+      "learning_rate": 3.928469901255787e-05,
+      "loss": 0.7261,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5371553770312215,
+      "grad_norm": 0.6545456051826477,
+      "learning_rate": 3.928089187914192e-05,
+      "loss": 0.7326,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5375205404418477,
+      "grad_norm": 0.6541972160339355,
+      "learning_rate": 3.927707482648079e-05,
+      "loss": 0.7266,
+      "step": 1472
+    },
+    {
+      "epoch": 0.537885703852474,
+      "grad_norm": 0.862709641456604,
+      "learning_rate": 3.92732478565382e-05,
+      "loss": 0.7932,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5382508672631002,
+      "grad_norm": 1.082711935043335,
+      "learning_rate": 3.926941097128298e-05,
+      "loss": 0.7378,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5386160306737265,
+      "grad_norm": 1.1237872838974,
+      "learning_rate": 3.9265564172689046e-05,
+      "loss": 0.7468,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5389811940843527,
+      "grad_norm": 0.8410653471946716,
+      "learning_rate": 3.926170746273543e-05,
+      "loss": 0.7385,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5393463574949791,
+      "grad_norm": 0.7173160910606384,
+      "learning_rate": 3.925784084340624e-05,
+      "loss": 0.7298,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5397115209056053,
+      "grad_norm": 0.9982025027275085,
+      "learning_rate": 3.9253964316690707e-05,
+      "loss": 0.7587,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5400766843162315,
+      "grad_norm": 0.919310986995697,
+      "learning_rate": 3.925007788458315e-05,
+      "loss": 0.7244,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5404418477268578,
+      "grad_norm": 0.8118817806243896,
+      "learning_rate": 3.924618154908298e-05,
+      "loss": 0.7427,
+      "step": 1480
+    },
+    {
+      "epoch": 0.540807011137484,
+      "grad_norm": 0.8505826592445374,
+      "learning_rate": 3.9242275312194694e-05,
+      "loss": 0.7567,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5411721745481103,
+      "grad_norm": 0.8501365780830383,
+      "learning_rate": 3.923835917592792e-05,
+      "loss": 0.7379,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5415373379587365,
+      "grad_norm": 0.6969407796859741,
+      "learning_rate": 3.923443314229732e-05,
+      "loss": 0.7416,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5419025013693628,
+      "grad_norm": 0.9816417694091797,
+      "learning_rate": 3.9230497213322715e-05,
+      "loss": 0.7499,
+      "step": 1484
+    },
+    {
+      "epoch": 0.542267664779989,
+      "grad_norm": 0.8559290170669556,
+      "learning_rate": 3.922655139102895e-05,
+      "loss": 0.7424,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5426328281906153,
+      "grad_norm": 0.8443220257759094,
+      "learning_rate": 3.922259567744602e-05,
+      "loss": 0.7454,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5429979916012415,
+      "grad_norm": 1.1401548385620117,
+      "learning_rate": 3.9218630074608966e-05,
+      "loss": 0.735,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5433631550118678,
+      "grad_norm": 0.926574170589447,
+      "learning_rate": 3.921465458455793e-05,
+      "loss": 0.7263,
+      "step": 1488
+    },
+    {
+      "epoch": 0.543728318422494,
+      "grad_norm": 0.9735384583473206,
+      "learning_rate": 3.9210669209338144e-05,
+      "loss": 0.735,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5440934818331203,
+      "grad_norm": 0.8426558375358582,
+      "learning_rate": 3.920667395099993e-05,
+      "loss": 0.73,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5444586452437465,
+      "grad_norm": 0.9400650262832642,
+      "learning_rate": 3.920266881159869e-05,
+      "loss": 0.7649,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5448238086543729,
+      "grad_norm": 0.765070915222168,
+      "learning_rate": 3.9198653793194896e-05,
+      "loss": 0.7173,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5451889720649991,
+      "grad_norm": 0.9452735185623169,
+      "learning_rate": 3.919462889785412e-05,
+      "loss": 0.7412,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5455541354756254,
+      "grad_norm": 1.000473976135254,
+      "learning_rate": 3.9190594127647005e-05,
+      "loss": 0.7458,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5459192988862516,
+      "grad_norm": 0.6995605826377869,
+      "learning_rate": 3.918654948464928e-05,
+      "loss": 0.7232,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5462844622968779,
+      "grad_norm": 0.7358512878417969,
+      "learning_rate": 3.918249497094176e-05,
+      "loss": 0.7213,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5466496257075041,
+      "grad_norm": 0.8587970733642578,
+      "learning_rate": 3.917843058861032e-05,
+      "loss": 0.7213,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5470147891181304,
+      "grad_norm": 1.0094513893127441,
+      "learning_rate": 3.9174356339745933e-05,
+      "loss": 0.7267,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5473799525287566,
+      "grad_norm": 1.1970059871673584,
+      "learning_rate": 3.917027222644462e-05,
+      "loss": 0.7336,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5477451159393829,
+      "grad_norm": 0.9547330737113953,
+      "learning_rate": 3.9166178250807504e-05,
+      "loss": 0.7314,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5481102793500091,
+      "grad_norm": 0.8042092323303223,
+      "learning_rate": 3.9162074414940764e-05,
+      "loss": 0.674,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5484754427606354,
+      "grad_norm": 0.9922061562538147,
+      "learning_rate": 3.915796072095567e-05,
+      "loss": 0.7426,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5488406061712616,
+      "grad_norm": 0.9946082234382629,
+      "learning_rate": 3.9153837170968544e-05,
+      "loss": 0.7634,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5492057695818879,
+      "grad_norm": 0.8678110241889954,
+      "learning_rate": 3.914970376710079e-05,
+      "loss": 0.7316,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5495709329925141,
+      "grad_norm": 0.9004887342453003,
+      "learning_rate": 3.914556051147887e-05,
+      "loss": 0.7429,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5499360964031405,
+      "grad_norm": 0.8457182049751282,
+      "learning_rate": 3.914140740623434e-05,
+      "loss": 0.7302,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5503012598137667,
+      "grad_norm": 0.6601680517196655,
+      "learning_rate": 3.9137244453503794e-05,
+      "loss": 0.7152,
+      "step": 1507
+    },
+    {
+      "epoch": 0.550666423224393,
+      "grad_norm": 0.9470217823982239,
+      "learning_rate": 3.9133071655428904e-05,
+      "loss": 0.7292,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5510315866350192,
+      "grad_norm": 1.0125720500946045,
+      "learning_rate": 3.9128889014156415e-05,
+      "loss": 0.7549,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5513967500456455,
+      "grad_norm": 0.8530757427215576,
+      "learning_rate": 3.9124696531838114e-05,
+      "loss": 0.7444,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5517619134562717,
+      "grad_norm": 0.9094074368476868,
+      "learning_rate": 3.9120494210630886e-05,
+      "loss": 0.7416,
+      "step": 1511
+    },
+    {
+      "epoch": 0.552127076866898,
+      "grad_norm": 0.9099465012550354,
+      "learning_rate": 3.911628205269663e-05,
+      "loss": 0.7158,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5524922402775242,
+      "grad_norm": 0.9260149598121643,
+      "learning_rate": 3.911206006020235e-05,
+      "loss": 0.7482,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5528574036881504,
+      "grad_norm": 1.0228321552276611,
+      "learning_rate": 3.910782823532009e-05,
+      "loss": 0.7572,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5532225670987767,
+      "grad_norm": 0.957773745059967,
+      "learning_rate": 3.910358658022696e-05,
+      "loss": 0.7161,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5535877305094029,
+      "grad_norm": 1.198459506034851,
+      "learning_rate": 3.909933509710511e-05,
+      "loss": 0.736,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5539528939200292,
+      "grad_norm": 1.1073055267333984,
+      "learning_rate": 3.909507378814175e-05,
+      "loss": 0.7547,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5543180573306554,
+      "grad_norm": 0.9325324296951294,
+      "learning_rate": 3.909080265552918e-05,
+      "loss": 0.7434,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5546832207412817,
+      "grad_norm": 0.7317970395088196,
+      "learning_rate": 3.90865217014647e-05,
+      "loss": 0.7222,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5550483841519079,
+      "grad_norm": 1.3255256414413452,
+      "learning_rate": 3.90822309281507e-05,
+      "loss": 0.7048,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5554135475625342,
+      "grad_norm": 0.7833756804466248,
+      "learning_rate": 3.9077930337794614e-05,
+      "loss": 0.7237,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5557787109731605,
+      "grad_norm": 0.8257026672363281,
+      "learning_rate": 3.907361993260891e-05,
+      "loss": 0.7181,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5561438743837868,
+      "grad_norm": 0.9362462162971497,
+      "learning_rate": 3.906929971481114e-05,
+      "loss": 0.7233,
+      "step": 1523
+    },
+    {
+      "epoch": 0.556509037794413,
+      "grad_norm": 1.0759377479553223,
+      "learning_rate": 3.906496968662386e-05,
+      "loss": 0.7137,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5568742012050393,
+      "grad_norm": 0.9503883123397827,
+      "learning_rate": 3.906062985027471e-05,
+      "loss": 0.7472,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5572393646156655,
+      "grad_norm": 0.9840250611305237,
+      "learning_rate": 3.905628020799636e-05,
+      "loss": 0.7241,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5576045280262918,
+      "grad_norm": 0.9575956463813782,
+      "learning_rate": 3.905192076202652e-05,
+      "loss": 0.7015,
+      "step": 1527
+    },
+    {
+      "epoch": 0.557969691436918,
+      "grad_norm": 0.8100535273551941,
+      "learning_rate": 3.904755151460795e-05,
+      "loss": 0.7181,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5583348548475443,
+      "grad_norm": 1.001095175743103,
+      "learning_rate": 3.9043172467988464e-05,
+      "loss": 0.7142,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5587000182581705,
+      "grad_norm": 1.096186637878418,
+      "learning_rate": 3.9038783624420894e-05,
+      "loss": 0.749,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5590651816687968,
+      "grad_norm": 3.8753530979156494,
+      "learning_rate": 3.9034384986163126e-05,
+      "loss": 0.7357,
+      "step": 1531
+    },
+    {
+      "epoch": 0.559430345079423,
+      "grad_norm": 0.928016185760498,
+      "learning_rate": 3.902997655547809e-05,
+      "loss": 0.7152,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5597955084900493,
+      "grad_norm": 0.7244706749916077,
+      "learning_rate": 3.9025558334633735e-05,
+      "loss": 0.7244,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5601606719006755,
+      "grad_norm": 0.8810442686080933,
+      "learning_rate": 3.9021130325903076e-05,
+      "loss": 0.6888,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5605258353113018,
+      "grad_norm": 1.0424643754959106,
+      "learning_rate": 3.9016692531564125e-05,
+      "loss": 0.7013,
+      "step": 1535
+    },
+    {
+      "epoch": 0.560890998721928,
+      "grad_norm": 0.968743085861206,
+      "learning_rate": 3.901224495389996e-05,
+      "loss": 0.749,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5612561621325544,
+      "grad_norm": 0.9253996014595032,
+      "learning_rate": 3.900778759519868e-05,
+      "loss": 0.7476,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5616213255431806,
+      "grad_norm": 0.746687114238739,
+      "learning_rate": 3.9003320457753425e-05,
+      "loss": 0.7317,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5619864889538069,
+      "grad_norm": 1.5913606882095337,
+      "learning_rate": 3.8998843543862347e-05,
+      "loss": 0.7412,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5623516523644331,
+      "grad_norm": 1.0188117027282715,
+      "learning_rate": 3.899435685582864e-05,
+      "loss": 0.7125,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5627168157750594,
+      "grad_norm": 1.0400840044021606,
+      "learning_rate": 3.898986039596052e-05,
+      "loss": 0.7058,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5630819791856856,
+      "grad_norm": 0.8724197745323181,
+      "learning_rate": 3.898535416657125e-05,
+      "loss": 0.7518,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5634471425963119,
+      "grad_norm": 0.9361122846603394,
+      "learning_rate": 3.89808381699791e-05,
+      "loss": 0.738,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5638123060069381,
+      "grad_norm": 0.8224091529846191,
+      "learning_rate": 3.8976312408507356e-05,
+      "loss": 0.7749,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5641774694175644,
+      "grad_norm": 0.7766799926757812,
+      "learning_rate": 3.897177688448435e-05,
+      "loss": 0.7444,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5645426328281906,
+      "grad_norm": 2.588979959487915,
+      "learning_rate": 3.8967231600243434e-05,
+      "loss": 0.7341,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5649077962388168,
+      "grad_norm": 0.7987133860588074,
+      "learning_rate": 3.8962676558122965e-05,
+      "loss": 0.7125,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5652729596494431,
+      "grad_norm": 0.8955987095832825,
+      "learning_rate": 3.895811176046633e-05,
+      "loss": 0.7365,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5656381230600693,
+      "grad_norm": 0.7170660495758057,
+      "learning_rate": 3.895353720962193e-05,
+      "loss": 0.7212,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5660032864706956,
+      "grad_norm": 1.0151087045669556,
+      "learning_rate": 3.8948952907943206e-05,
+      "loss": 0.7556,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5663684498813218,
+      "grad_norm": 1.3191031217575073,
+      "learning_rate": 3.8944358857788576e-05,
+      "loss": 0.77,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5667336132919482,
+      "grad_norm": 1.0472934246063232,
+      "learning_rate": 3.893975506152151e-05,
+      "loss": 0.7274,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5670987767025744,
+      "grad_norm": 0.9767879247665405,
+      "learning_rate": 3.8935141521510466e-05,
+      "loss": 0.7597,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5674639401132007,
+      "grad_norm": 0.8284198045730591,
+      "learning_rate": 3.8930518240128926e-05,
+      "loss": 0.7382,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5678291035238269,
+      "grad_norm": 1.0305380821228027,
+      "learning_rate": 3.892588521975539e-05,
+      "loss": 0.75,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5681942669344532,
+      "grad_norm": 1.0287115573883057,
+      "learning_rate": 3.892124246277336e-05,
+      "loss": 0.7576,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5685594303450794,
+      "grad_norm": 0.7308568358421326,
+      "learning_rate": 3.891658997157134e-05,
+      "loss": 0.7251,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5689245937557057,
+      "grad_norm": 0.9927620887756348,
+      "learning_rate": 3.891192774854285e-05,
+      "loss": 0.724,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5692897571663319,
+      "grad_norm": 1.1937997341156006,
+      "learning_rate": 3.890725579608643e-05,
+      "loss": 0.7468,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5696549205769582,
+      "grad_norm": 0.9900869727134705,
+      "learning_rate": 3.89025741166056e-05,
+      "loss": 0.7531,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5700200839875844,
+      "grad_norm": 0.9900295734405518,
+      "learning_rate": 3.8897882712508906e-05,
+      "loss": 0.7347,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5703852473982107,
+      "grad_norm": 0.9560722708702087,
+      "learning_rate": 3.8893181586209883e-05,
+      "loss": 0.759,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5707504108088369,
+      "grad_norm": 0.7202712297439575,
+      "learning_rate": 3.888847074012706e-05,
+      "loss": 0.7329,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5711155742194632,
+      "grad_norm": 1.0042426586151123,
+      "learning_rate": 3.8883750176684e-05,
+      "loss": 0.7676,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5714807376300894,
+      "grad_norm": 1.1066584587097168,
+      "learning_rate": 3.8879019898309215e-05,
+      "loss": 0.7616,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5718459010407158,
+      "grad_norm": 1.1834222078323364,
+      "learning_rate": 3.8874279907436274e-05,
+      "loss": 0.7126,
+      "step": 1566
+    },
+    {
+      "epoch": 0.572211064451342,
+      "grad_norm": 0.8603668808937073,
+      "learning_rate": 3.886953020650369e-05,
+      "loss": 0.7305,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5725762278619683,
+      "grad_norm": 0.7139928340911865,
+      "learning_rate": 3.8864770797955e-05,
+      "loss": 0.7516,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5729413912725945,
+      "grad_norm": 0.8934731483459473,
+      "learning_rate": 3.8860001684238744e-05,
+      "loss": 0.7286,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5733065546832208,
+      "grad_norm": 0.8853731155395508,
+      "learning_rate": 3.885522286780842e-05,
+      "loss": 0.7301,
+      "step": 1570
+    },
+    {
+      "epoch": 0.573671718093847,
+      "grad_norm": 0.9740007519721985,
+      "learning_rate": 3.8850434351122536e-05,
+      "loss": 0.7448,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5740368815044733,
+      "grad_norm": 1.2016581296920776,
+      "learning_rate": 3.884563613664461e-05,
+      "loss": 0.7446,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5744020449150995,
+      "grad_norm": 1.046797513961792,
+      "learning_rate": 3.8840828226843113e-05,
+      "loss": 0.7207,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5747672083257258,
+      "grad_norm": 0.9799802899360657,
+      "learning_rate": 3.8836010624191535e-05,
+      "loss": 0.734,
+      "step": 1574
+    },
+    {
+      "epoch": 0.575132371736352,
+      "grad_norm": 1.2142255306243896,
+      "learning_rate": 3.883118333116833e-05,
+      "loss": 0.7166,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5754975351469783,
+      "grad_norm": 1.001091480255127,
+      "learning_rate": 3.8826346350256943e-05,
+      "loss": 0.7248,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5758626985576045,
+      "grad_norm": 1.177527666091919,
+      "learning_rate": 3.882149968394582e-05,
+      "loss": 0.7349,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5762278619682308,
+      "grad_norm": 1.2621814012527466,
+      "learning_rate": 3.881664333472837e-05,
+      "loss": 0.7143,
+      "step": 1578
+    },
+    {
+      "epoch": 0.576593025378857,
+      "grad_norm": 0.8858799338340759,
+      "learning_rate": 3.8811777305102986e-05,
+      "loss": 0.7115,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5769581887894832,
+      "grad_norm": 0.8787485361099243,
+      "learning_rate": 3.880690159757305e-05,
+      "loss": 0.7211,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5773233522001096,
+      "grad_norm": 0.9342215061187744,
+      "learning_rate": 3.880201621464691e-05,
+      "loss": 0.7169,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5776885156107358,
+      "grad_norm": 1.0358704328536987,
+      "learning_rate": 3.8797121158837914e-05,
+      "loss": 0.7211,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5780536790213621,
+      "grad_norm": 1.1126457452774048,
+      "learning_rate": 3.8792216432664356e-05,
+      "loss": 0.7659,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5784188424319883,
+      "grad_norm": 0.8449137210845947,
+      "learning_rate": 3.878730203864954e-05,
+      "loss": 0.7141,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5787840058426146,
+      "grad_norm": 0.9435628056526184,
+      "learning_rate": 3.87823779793217e-05,
+      "loss": 0.7293,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5791491692532408,
+      "grad_norm": 1.09213125705719,
+      "learning_rate": 3.877744425721408e-05,
+      "loss": 0.7275,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5795143326638671,
+      "grad_norm": 1.1995439529418945,
+      "learning_rate": 3.8772500874864886e-05,
+      "loss": 0.739,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5798794960744933,
+      "grad_norm": 0.8401466012001038,
+      "learning_rate": 3.876754783481729e-05,
+      "loss": 0.7004,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5802446594851196,
+      "grad_norm": 1.449797511100769,
+      "learning_rate": 3.8762585139619415e-05,
+      "loss": 0.7337,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5806098228957458,
+      "grad_norm": 1.1313889026641846,
+      "learning_rate": 3.875761279182439e-05,
+      "loss": 0.6908,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5809749863063721,
+      "grad_norm": 0.8749697804450989,
+      "learning_rate": 3.875263079399028e-05,
+      "loss": 0.749,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5813401497169983,
+      "grad_norm": 0.8486689925193787,
+      "learning_rate": 3.874763914868013e-05,
+      "loss": 0.7147,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5817053131276246,
+      "grad_norm": 1.2051169872283936,
+      "learning_rate": 3.874263785846192e-05,
+      "loss": 0.7404,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5820704765382508,
+      "grad_norm": 1.1990227699279785,
+      "learning_rate": 3.873762692590863e-05,
+      "loss": 0.7632,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5824356399488771,
+      "grad_norm": 0.8588769435882568,
+      "learning_rate": 3.8732606353598185e-05,
+      "loss": 0.7407,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5828008033595033,
+      "grad_norm": 0.9404013752937317,
+      "learning_rate": 3.872757614411346e-05,
+      "loss": 0.729,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5831659667701297,
+      "grad_norm": 1.1304678916931152,
+      "learning_rate": 3.8722536300042305e-05,
+      "loss": 0.781,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5835311301807559,
+      "grad_norm": 1.142740249633789,
+      "learning_rate": 3.871748682397751e-05,
+      "loss": 0.7141,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5838962935913822,
+      "grad_norm": 1.2788124084472656,
+      "learning_rate": 3.871242771851683e-05,
+      "loss": 0.7212,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5842614570020084,
+      "grad_norm": 0.8648350238800049,
+      "learning_rate": 3.870735898626297e-05,
+      "loss": 0.7303,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5846266204126347,
+      "grad_norm": 0.9944093227386475,
+      "learning_rate": 3.8702280629823595e-05,
+      "loss": 0.7536,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5849917838232609,
+      "grad_norm": 0.7696415185928345,
+      "learning_rate": 3.8697192651811305e-05,
+      "loss": 0.7068,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5853569472338872,
+      "grad_norm": 1.4209320545196533,
+      "learning_rate": 3.869209505484367e-05,
+      "loss": 0.7426,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5857221106445134,
+      "grad_norm": 0.8690351843833923,
+      "learning_rate": 3.86869878415432e-05,
+      "loss": 0.7363,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5860872740551397,
+      "grad_norm": 1.0473322868347168,
+      "learning_rate": 3.868187101453734e-05,
+      "loss": 0.7094,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5864524374657659,
+      "grad_norm": 1.0611331462860107,
+      "learning_rate": 3.867674457645851e-05,
+      "loss": 0.7642,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5868176008763922,
+      "grad_norm": 0.8476665616035461,
+      "learning_rate": 3.8671608529944035e-05,
+      "loss": 0.7202,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5871827642870184,
+      "grad_norm": 1.0705032348632812,
+      "learning_rate": 3.866646287763622e-05,
+      "loss": 0.7548,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5875479276976447,
+      "grad_norm": 0.7043454051017761,
+      "learning_rate": 3.86613076221823e-05,
+      "loss": 0.6984,
+      "step": 1609
+    },
+    {
+      "epoch": 0.587913091108271,
+      "grad_norm": 1.0177797079086304,
+      "learning_rate": 3.865614276623443e-05,
+      "loss": 0.7512,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5882782545188973,
+      "grad_norm": 1.0479447841644287,
+      "learning_rate": 3.8650968312449745e-05,
+      "loss": 0.723,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5886434179295235,
+      "grad_norm": 0.7406904101371765,
+      "learning_rate": 3.864578426349027e-05,
+      "loss": 0.7261,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5890085813401497,
+      "grad_norm": 1.4430965185165405,
+      "learning_rate": 3.8640590622023005e-05,
+      "loss": 0.7601,
+      "step": 1613
+    },
+    {
+      "epoch": 0.589373744750776,
+      "grad_norm": 0.7748079299926758,
+      "learning_rate": 3.863538739071986e-05,
+      "loss": 0.7094,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5897389081614022,
+      "grad_norm": 2.9764978885650635,
+      "learning_rate": 3.86301745722577e-05,
+      "loss": 0.7087,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5901040715720285,
+      "grad_norm": 0.768072783946991,
+      "learning_rate": 3.86249521693183e-05,
+      "loss": 0.7376,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5904692349826547,
+      "grad_norm": 0.743763267993927,
+      "learning_rate": 3.861972018458838e-05,
+      "loss": 0.713,
+      "step": 1617
+    },
+    {
+      "epoch": 0.590834398393281,
+      "grad_norm": 0.926245927810669,
+      "learning_rate": 3.861447862075959e-05,
+      "loss": 0.6924,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5911995618039072,
+      "grad_norm": 1.0158087015151978,
+      "learning_rate": 3.86092274805285e-05,
+      "loss": 0.7286,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5915647252145335,
+      "grad_norm": 0.8567479848861694,
+      "learning_rate": 3.8603966766596624e-05,
+      "loss": 0.6996,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5919298886251597,
+      "grad_norm": 0.9987544417381287,
+      "learning_rate": 3.8598696481670364e-05,
+      "loss": 0.7487,
+      "step": 1621
+    },
+    {
+      "epoch": 0.592295052035786,
+      "grad_norm": 0.9453714489936829,
+      "learning_rate": 3.859341662846109e-05,
+      "loss": 0.6998,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5926602154464122,
+      "grad_norm": 0.8023524880409241,
+      "learning_rate": 3.858812720968507e-05,
+      "loss": 0.7198,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5930253788570385,
+      "grad_norm": 1.01613450050354,
+      "learning_rate": 3.858282822806349e-05,
+      "loss": 0.7234,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5933905422676647,
+      "grad_norm": 1.1630140542984009,
+      "learning_rate": 3.857751968632247e-05,
+      "loss": 0.7289,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5937557056782911,
+      "grad_norm": 1.1454566717147827,
+      "learning_rate": 3.857220158719305e-05,
+      "loss": 0.7469,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5941208690889173,
+      "grad_norm": 1.7499412298202515,
+      "learning_rate": 3.8566873933411156e-05,
+      "loss": 0.7163,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5944860324995436,
+      "grad_norm": 1.0432333946228027,
+      "learning_rate": 3.856153672771767e-05,
+      "loss": 0.7432,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5948511959101698,
+      "grad_norm": 0.8729361891746521,
+      "learning_rate": 3.855618997285837e-05,
+      "loss": 0.7484,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5952163593207961,
+      "grad_norm": 0.9249089956283569,
+      "learning_rate": 3.855083367158394e-05,
+      "loss": 0.7289,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5955815227314223,
+      "grad_norm": 1.153282642364502,
+      "learning_rate": 3.854546782664998e-05,
+      "loss": 0.7185,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5959466861420486,
+      "grad_norm": 1.1534794569015503,
+      "learning_rate": 3.854009244081701e-05,
+      "loss": 0.7081,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5963118495526748,
+      "grad_norm": 0.6747300028800964,
+      "learning_rate": 3.8534707516850446e-05,
+      "loss": 0.7314,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5966770129633011,
+      "grad_norm": 1.1119229793548584,
+      "learning_rate": 3.852931305752062e-05,
+      "loss": 0.7388,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5970421763739273,
+      "grad_norm": 1.1446058750152588,
+      "learning_rate": 3.852390906560276e-05,
+      "loss": 0.7186,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5974073397845536,
+      "grad_norm": 1.0130422115325928,
+      "learning_rate": 3.8518495543877e-05,
+      "loss": 0.7465,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5977725031951798,
+      "grad_norm": 1.2683770656585693,
+      "learning_rate": 3.8513072495128385e-05,
+      "loss": 0.7214,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5981376666058061,
+      "grad_norm": 0.9249864816665649,
+      "learning_rate": 3.850763992214686e-05,
+      "loss": 0.7571,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5985028300164323,
+      "grad_norm": 0.9933344721794128,
+      "learning_rate": 3.8502197827727254e-05,
+      "loss": 0.738,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5988679934270587,
+      "grad_norm": 1.3973751068115234,
+      "learning_rate": 3.849674621466931e-05,
+      "loss": 0.7045,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5992331568376849,
+      "grad_norm": 1.2577351331710815,
+      "learning_rate": 3.849128508577767e-05,
+      "loss": 0.7463,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5995983202483112,
+      "grad_norm": 1.1957579851150513,
+      "learning_rate": 3.848581444386187e-05,
+      "loss": 0.7256,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5999634836589374,
+      "grad_norm": 1.1578030586242676,
+      "learning_rate": 3.848033429173632e-05,
+      "loss": 0.7491,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6003286470695637,
+      "grad_norm": 0.8890886306762695,
+      "learning_rate": 3.847484463222035e-05,
+      "loss": 0.6758,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6006938104801899,
+      "grad_norm": 0.6987606883049011,
+      "learning_rate": 3.846934546813816e-05,
+      "loss": 0.7345,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6010589738908162,
+      "grad_norm": 0.7313883304595947,
+      "learning_rate": 3.8463836802318865e-05,
+      "loss": 0.7071,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6014241373014424,
+      "grad_norm": 1.2004591226577759,
+      "learning_rate": 3.8458318637596434e-05,
+      "loss": 0.7462,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6017893007120686,
+      "grad_norm": 1.004254698753357,
+      "learning_rate": 3.845279097680975e-05,
+      "loss": 0.7172,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6021544641226949,
+      "grad_norm": 1.0235750675201416,
+      "learning_rate": 3.844725382280258e-05,
+      "loss": 0.743,
+      "step": 1649
+    },
+    {
+      "epoch": 0.6025196275333211,
+      "grad_norm": 0.9419151544570923,
+      "learning_rate": 3.8441707178423554e-05,
+      "loss": 0.7057,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6028847909439474,
+      "grad_norm": 0.799010157585144,
+      "learning_rate": 3.843615104652621e-05,
+      "loss": 0.7042,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6032499543545736,
+      "grad_norm": 1.102571964263916,
+      "learning_rate": 3.843058542996895e-05,
+      "loss": 0.7648,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6036151177651999,
+      "grad_norm": 0.751781165599823,
+      "learning_rate": 3.842501033161505e-05,
+      "loss": 0.7295,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6039802811758261,
+      "grad_norm": 0.8896080851554871,
+      "learning_rate": 3.8419425754332694e-05,
+      "loss": 0.7025,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6043454445864525,
+      "grad_norm": 1.2246979475021362,
+      "learning_rate": 3.8413831700994905e-05,
+      "loss": 0.7541,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6047106079970787,
+      "grad_norm": 1.4045801162719727,
+      "learning_rate": 3.840822817447961e-05,
+      "loss": 0.7764,
+      "step": 1656
+    },
+    {
+      "epoch": 0.605075771407705,
+      "grad_norm": 1.0411593914031982,
+      "learning_rate": 3.8402615177669604e-05,
+      "loss": 0.7406,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6054409348183312,
+      "grad_norm": 1.3511064052581787,
+      "learning_rate": 3.839699271345253e-05,
+      "loss": 0.7415,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6058060982289575,
+      "grad_norm": 0.8141111135482788,
+      "learning_rate": 3.839136078472093e-05,
+      "loss": 0.7292,
+      "step": 1659
+    },
+    {
+      "epoch": 0.6061712616395837,
+      "grad_norm": 0.959466278553009,
+      "learning_rate": 3.838571939437221e-05,
+      "loss": 0.7207,
+      "step": 1660
+    },
+    {
+      "epoch": 0.60653642505021,
+      "grad_norm": 1.0917295217514038,
+      "learning_rate": 3.838006854530863e-05,
+      "loss": 0.7007,
+      "step": 1661
+    },
+    {
+      "epoch": 0.6069015884608362,
+      "grad_norm": 0.9539211392402649,
+      "learning_rate": 3.837440824043734e-05,
+      "loss": 0.7369,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6072667518714625,
+      "grad_norm": 0.9652641415596008,
+      "learning_rate": 3.8368738482670315e-05,
+      "loss": 0.7266,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6076319152820887,
+      "grad_norm": 1.208406686782837,
+      "learning_rate": 3.8363059274924445e-05,
+      "loss": 0.7199,
+      "step": 1664
+    },
+    {
+      "epoch": 0.607997078692715,
+      "grad_norm": 1.1405893564224243,
+      "learning_rate": 3.835737062012143e-05,
+      "loss": 0.7505,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6083622421033412,
+      "grad_norm": 1.0682623386383057,
+      "learning_rate": 3.8351672521187874e-05,
+      "loss": 0.7272,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6087274055139675,
+      "grad_norm": 0.9367862343788147,
+      "learning_rate": 3.834596498105521e-05,
+      "loss": 0.7195,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6090925689245937,
+      "grad_norm": 0.6378055214881897,
+      "learning_rate": 3.8340248002659745e-05,
+      "loss": 0.7269,
+      "step": 1668
+    },
+    {
+      "epoch": 0.60945773233522,
+      "grad_norm": 1.028985619544983,
+      "learning_rate": 3.8334521588942626e-05,
+      "loss": 0.7161,
+      "step": 1669
+    },
+    {
+      "epoch": 0.6098228957458462,
+      "grad_norm": 1.2288942337036133,
+      "learning_rate": 3.832878574284988e-05,
+      "loss": 0.7233,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6101880591564726,
+      "grad_norm": 0.8505779504776001,
+      "learning_rate": 3.8323040467332344e-05,
+      "loss": 0.7125,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6105532225670988,
+      "grad_norm": 0.9262564778327942,
+      "learning_rate": 3.8317285765345746e-05,
+      "loss": 0.7317,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6109183859777251,
+      "grad_norm": 0.9521602392196655,
+      "learning_rate": 3.831152163985065e-05,
+      "loss": 0.7239,
+      "step": 1673
+    },
+    {
+      "epoch": 0.6112835493883513,
+      "grad_norm": 0.8419995903968811,
+      "learning_rate": 3.830574809381247e-05,
+      "loss": 0.6981,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6116487127989776,
+      "grad_norm": 0.906650185585022,
+      "learning_rate": 3.829996513020146e-05,
+      "loss": 0.7193,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6120138762096038,
+      "grad_norm": 1.6719812154769897,
+      "learning_rate": 3.829417275199272e-05,
+      "loss": 0.7359,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6123790396202301,
+      "grad_norm": 0.9106917977333069,
+      "learning_rate": 3.8288370962166194e-05,
+      "loss": 0.741,
+      "step": 1677
+    },
+    {
+      "epoch": 0.6127442030308563,
+      "grad_norm": 0.7947406768798828,
+      "learning_rate": 3.828255976370668e-05,
+      "loss": 0.7325,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6131093664414826,
+      "grad_norm": 1.6704083681106567,
+      "learning_rate": 3.8276739159603795e-05,
+      "loss": 0.7078,
+      "step": 1679
+    },
+    {
+      "epoch": 0.6134745298521088,
+      "grad_norm": 1.1100096702575684,
+      "learning_rate": 3.827090915285202e-05,
+      "loss": 0.7399,
+      "step": 1680
+    },
+    {
+      "epoch": 0.613839693262735,
+      "grad_norm": 0.9204789996147156,
+      "learning_rate": 3.826506974645065e-05,
+      "loss": 0.7225,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6142048566733613,
+      "grad_norm": 0.7737852334976196,
+      "learning_rate": 3.8259220943403825e-05,
+      "loss": 0.7071,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6145700200839875,
+      "grad_norm": 1.0756714344024658,
+      "learning_rate": 3.825336274672053e-05,
+      "loss": 0.7123,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6149351834946138,
+      "grad_norm": 1.1082093715667725,
+      "learning_rate": 3.824749515941455e-05,
+      "loss": 0.7314,
+      "step": 1684
+    },
+    {
+      "epoch": 0.61530034690524,
+      "grad_norm": 0.9082583785057068,
+      "learning_rate": 3.824161818450454e-05,
+      "loss": 0.7487,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6156655103158664,
+      "grad_norm": 0.7466766834259033,
+      "learning_rate": 3.823573182501397e-05,
+      "loss": 0.7064,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6160306737264926,
+      "grad_norm": 0.8748775124549866,
+      "learning_rate": 3.822983608397113e-05,
+      "loss": 0.7094,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6163958371371189,
+      "grad_norm": 0.8320901989936829,
+      "learning_rate": 3.8223930964409136e-05,
+      "loss": 0.7298,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6167610005477451,
+      "grad_norm": 0.9852972626686096,
+      "learning_rate": 3.821801646936595e-05,
+      "loss": 0.7457,
+      "step": 1689
+    },
+    {
+      "epoch": 0.6171261639583714,
+      "grad_norm": 0.7243798971176147,
+      "learning_rate": 3.821209260188433e-05,
+      "loss": 0.7272,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6174913273689976,
+      "grad_norm": 0.8015673756599426,
+      "learning_rate": 3.8206159365011875e-05,
+      "loss": 0.7139,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6178564907796239,
+      "grad_norm": 0.7652367949485779,
+      "learning_rate": 3.8200216761800986e-05,
+      "loss": 0.7321,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6182216541902501,
+      "grad_norm": 1.6092801094055176,
+      "learning_rate": 3.819426479530891e-05,
+      "loss": 0.7166,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6185868176008764,
+      "grad_norm": 1.0409109592437744,
+      "learning_rate": 3.8188303468597684e-05,
+      "loss": 0.7145,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6189519810115026,
+      "grad_norm": 1.179543137550354,
+      "learning_rate": 3.818233278473417e-05,
+      "loss": 0.7096,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6193171444221289,
+      "grad_norm": 0.6408907175064087,
+      "learning_rate": 3.817635274679006e-05,
+      "loss": 0.7106,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6196823078327551,
+      "grad_norm": 0.9521438479423523,
+      "learning_rate": 3.817036335784183e-05,
+      "loss": 0.7045,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6200474712433814,
+      "grad_norm": 0.9148648977279663,
+      "learning_rate": 3.816436462097079e-05,
+      "loss": 0.7177,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6204126346540076,
+      "grad_norm": 0.8411362767219543,
+      "learning_rate": 3.815835653926303e-05,
+      "loss": 0.7051,
+      "step": 1699
+    },
+    {
+      "epoch": 0.620777798064634,
+      "grad_norm": 0.799384355545044,
+      "learning_rate": 3.8152339115809486e-05,
+      "loss": 0.6688,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6211429614752602,
+      "grad_norm": 0.9203513264656067,
+      "learning_rate": 3.814631235370587e-05,
+      "loss": 0.7238,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6215081248858865,
+      "grad_norm": 0.7966442108154297,
+      "learning_rate": 3.814027625605272e-05,
+      "loss": 0.6946,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6218732882965127,
+      "grad_norm": 0.865376353263855,
+      "learning_rate": 3.8134230825955366e-05,
+      "loss": 0.7072,
+      "step": 1703
+    },
+    {
+      "epoch": 0.622238451707139,
+      "grad_norm": 0.8998957276344299,
+      "learning_rate": 3.812817606652392e-05,
+      "loss": 0.7385,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6226036151177652,
+      "grad_norm": 0.8941604495048523,
+      "learning_rate": 3.812211198087333e-05,
+      "loss": 0.7095,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6229687785283915,
+      "grad_norm": 0.8669780492782593,
+      "learning_rate": 3.8116038572123325e-05,
+      "loss": 0.7166,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6233339419390177,
+      "grad_norm": 1.0933927297592163,
+      "learning_rate": 3.810995584339843e-05,
+      "loss": 0.744,
+      "step": 1707
+    },
+    {
+      "epoch": 0.623699105349644,
+      "grad_norm": 0.6955661773681641,
+      "learning_rate": 3.8103863797827955e-05,
+      "loss": 0.7256,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6240642687602702,
+      "grad_norm": 1.25681471824646,
+      "learning_rate": 3.809776243854602e-05,
+      "loss": 0.7371,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6244294321708965,
+      "grad_norm": 1.0992910861968994,
+      "learning_rate": 3.8091651768691526e-05,
+      "loss": 0.6931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6247945955815227,
+      "grad_norm": 1.6440374851226807,
+      "learning_rate": 3.808553179140817e-05,
+      "loss": 0.704,
+      "step": 1711
+    },
+    {
+      "epoch": 0.625159758992149,
+      "grad_norm": 1.3347744941711426,
+      "learning_rate": 3.807940250984444e-05,
+      "loss": 0.6899,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6255249224027752,
+      "grad_norm": 1.079863429069519,
+      "learning_rate": 3.807326392715359e-05,
+      "loss": 0.7344,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6258900858134014,
+      "grad_norm": 0.8595447540283203,
+      "learning_rate": 3.806711604649369e-05,
+      "loss": 0.7137,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6262552492240278,
+      "grad_norm": 0.827134907245636,
+      "learning_rate": 3.806095887102757e-05,
+      "loss": 0.6954,
+      "step": 1715
+    },
+    {
+      "epoch": 0.626620412634654,
+      "grad_norm": 1.483556866645813,
+      "learning_rate": 3.805479240392286e-05,
+      "loss": 0.7368,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6269855760452803,
+      "grad_norm": 1.2429213523864746,
+      "learning_rate": 3.804861664835195e-05,
+      "loss": 0.7454,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6273507394559065,
+      "grad_norm": 0.8351743817329407,
+      "learning_rate": 3.8042431607492015e-05,
+      "loss": 0.7059,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6277159028665328,
+      "grad_norm": 0.8552349209785461,
+      "learning_rate": 3.8036237284525016e-05,
+      "loss": 0.7266,
+      "step": 1719
+    },
+    {
+      "epoch": 0.628081066277159,
+      "grad_norm": 1.1395808458328247,
+      "learning_rate": 3.8030033682637686e-05,
+      "loss": 0.7244,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6284462296877853,
+      "grad_norm": 1.2721116542816162,
+      "learning_rate": 3.8023820805021524e-05,
+      "loss": 0.722,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6288113930984115,
+      "grad_norm": 0.9128320813179016,
+      "learning_rate": 3.801759865487281e-05,
+      "loss": 0.7298,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6291765565090378,
+      "grad_norm": 1.0605303049087524,
+      "learning_rate": 3.801136723539259e-05,
+      "loss": 0.7627,
+      "step": 1723
+    },
+    {
+      "epoch": 0.629541719919664,
+      "grad_norm": 0.8743678331375122,
+      "learning_rate": 3.8005126549786674e-05,
+      "loss": 0.7209,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6299068833302903,
+      "grad_norm": 0.8217443227767944,
+      "learning_rate": 3.7998876601265654e-05,
+      "loss": 0.7207,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6302720467409165,
+      "grad_norm": 1.1639951467514038,
+      "learning_rate": 3.799261739304487e-05,
+      "loss": 0.683,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6306372101515428,
+      "grad_norm": 0.8091967105865479,
+      "learning_rate": 3.798634892834444e-05,
+      "loss": 0.718,
+      "step": 1727
+    },
+    {
+      "epoch": 0.631002373562169,
+      "grad_norm": 0.9289019703865051,
+      "learning_rate": 3.798007121038923e-05,
+      "loss": 0.7153,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6313675369727954,
+      "grad_norm": 1.0279673337936401,
+      "learning_rate": 3.797378424240888e-05,
+      "loss": 0.7405,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6317327003834216,
+      "grad_norm": 0.9682115912437439,
+      "learning_rate": 3.7967488027637776e-05,
+      "loss": 0.7032,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6320978637940479,
+      "grad_norm": 1.1122698783874512,
+      "learning_rate": 3.796118256931507e-05,
+      "loss": 0.7122,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6324630272046741,
+      "grad_norm": 1.057283639907837,
+      "learning_rate": 3.7954867870684677e-05,
+      "loss": 0.7413,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6328281906153004,
+      "grad_norm": 0.8642354011535645,
+      "learning_rate": 3.794854393499525e-05,
+      "loss": 0.6952,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6331933540259266,
+      "grad_norm": 0.9934948086738586,
+      "learning_rate": 3.7942210765500197e-05,
+      "loss": 0.7555,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6335585174365529,
+      "grad_norm": 0.855786144733429,
+      "learning_rate": 3.7935868365457674e-05,
+      "loss": 0.7121,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6339236808471791,
+      "grad_norm": 1.0481681823730469,
+      "learning_rate": 3.7929516738130606e-05,
+      "loss": 0.6837,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6342888442578054,
+      "grad_norm": 0.8370383381843567,
+      "learning_rate": 3.7923155886786636e-05,
+      "loss": 0.7074,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6346540076684316,
+      "grad_norm": 0.963932454586029,
+      "learning_rate": 3.791678581469818e-05,
+      "loss": 0.7545,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6350191710790579,
+      "grad_norm": 0.819789707660675,
+      "learning_rate": 3.7910406525142374e-05,
+      "loss": 0.7273,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6353843344896841,
+      "grad_norm": 0.7305337190628052,
+      "learning_rate": 3.790401802140111e-05,
+      "loss": 0.7001,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6357494979003104,
+      "grad_norm": 0.7566940188407898,
+      "learning_rate": 3.789762030676103e-05,
+      "loss": 0.7009,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6361146613109366,
+      "grad_norm": 0.914252758026123,
+      "learning_rate": 3.7891213384513476e-05,
+      "loss": 0.7362,
+      "step": 1742
+    },
+    {
+      "epoch": 0.636479824721563,
+      "grad_norm": 1.0864005088806152,
+      "learning_rate": 3.7884797257954565e-05,
+      "loss": 0.6954,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6368449881321891,
+      "grad_norm": 0.9217675924301147,
+      "learning_rate": 3.7878371930385144e-05,
+      "loss": 0.7169,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6372101515428155,
+      "grad_norm": 0.9754314422607422,
+      "learning_rate": 3.787193740511077e-05,
+      "loss": 0.7153,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6375753149534417,
+      "grad_norm": 1.1855268478393555,
+      "learning_rate": 3.786549368544177e-05,
+      "loss": 0.7084,
+      "step": 1746
+    },
+    {
+      "epoch": 0.637940478364068,
+      "grad_norm": 0.854855477809906,
+      "learning_rate": 3.7859040774693156e-05,
+      "loss": 0.7084,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6383056417746942,
+      "grad_norm": 0.9696609973907471,
+      "learning_rate": 3.7852578676184705e-05,
+      "loss": 0.7118,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6386708051853204,
+      "grad_norm": 0.8385400176048279,
+      "learning_rate": 3.784610739324091e-05,
+      "loss": 0.7386,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6390359685959467,
+      "grad_norm": 1.112640142440796,
+      "learning_rate": 3.7839626929190976e-05,
+      "loss": 0.7151,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6394011320065729,
+      "grad_norm": 1.1101984977722168,
+      "learning_rate": 3.783313728736884e-05,
+      "loss": 0.7083,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6397662954171992,
+      "grad_norm": 0.865172803401947,
+      "learning_rate": 3.782663847111318e-05,
+      "loss": 0.705,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6401314588278254,
+      "grad_norm": 1.0278229713439941,
+      "learning_rate": 3.782013048376736e-05,
+      "loss": 0.7145,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6404966222384517,
+      "grad_norm": 0.9916568398475647,
+      "learning_rate": 3.781361332867948e-05,
+      "loss": 0.7386,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6408617856490779,
+      "grad_norm": 1.145369529724121,
+      "learning_rate": 3.7807087009202366e-05,
+      "loss": 0.7388,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6412269490597042,
+      "grad_norm": 0.8243128061294556,
+      "learning_rate": 3.780055152869354e-05,
+      "loss": 0.7056,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6415921124703304,
+      "grad_norm": 1.1970804929733276,
+      "learning_rate": 3.7794006890515235e-05,
+      "loss": 0.7265,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6419572758809567,
+      "grad_norm": 0.7989738583564758,
+      "learning_rate": 3.778745309803442e-05,
+      "loss": 0.7139,
+      "step": 1758
+    },
+    {
+      "epoch": 0.642322439291583,
+      "grad_norm": 0.8789658546447754,
+      "learning_rate": 3.778089015462275e-05,
+      "loss": 0.713,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6426876027022093,
+      "grad_norm": 0.6590216159820557,
+      "learning_rate": 3.77743180636566e-05,
+      "loss": 0.7172,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6430527661128355,
+      "grad_norm": 0.92206871509552,
+      "learning_rate": 3.776773682851705e-05,
+      "loss": 0.7135,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6434179295234618,
+      "grad_norm": 1.0673507452011108,
+      "learning_rate": 3.776114645258987e-05,
+      "loss": 0.7426,
+      "step": 1762
+    },
+    {
+      "epoch": 0.643783092934088,
+      "grad_norm": 0.7360542416572571,
+      "learning_rate": 3.775454693926554e-05,
+      "loss": 0.7217,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6441482563447143,
+      "grad_norm": 1.0386346578598022,
+      "learning_rate": 3.774793829193927e-05,
+      "loss": 0.6968,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6445134197553405,
+      "grad_norm": 1.0472970008850098,
+      "learning_rate": 3.774132051401093e-05,
+      "loss": 0.7058,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6448785831659668,
+      "grad_norm": 0.9268572330474854,
+      "learning_rate": 3.77346936088851e-05,
+      "loss": 0.7255,
+      "step": 1766
+    },
+    {
+      "epoch": 0.645243746576593,
+      "grad_norm": 0.9543839693069458,
+      "learning_rate": 3.772805757997105e-05,
+      "loss": 0.6696,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6456089099872193,
+      "grad_norm": 1.1249008178710938,
+      "learning_rate": 3.7721412430682766e-05,
+      "loss": 0.725,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6459740733978455,
+      "grad_norm": 1.248288631439209,
+      "learning_rate": 3.7714758164438896e-05,
+      "loss": 0.7472,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6463392368084718,
+      "grad_norm": 1.2799824476242065,
+      "learning_rate": 3.7708094784662804e-05,
+      "loss": 0.7057,
+      "step": 1770
+    },
+    {
+      "epoch": 0.646704400219098,
+      "grad_norm": 1.0398547649383545,
+      "learning_rate": 3.7701422294782514e-05,
+      "loss": 0.7277,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6470695636297243,
+      "grad_norm": 0.8737127184867859,
+      "learning_rate": 3.769474069823078e-05,
+      "loss": 0.7192,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6474347270403505,
+      "grad_norm": 0.949883222579956,
+      "learning_rate": 3.7688049998445e-05,
+      "loss": 0.7462,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6477998904509769,
+      "grad_norm": 0.878869354724884,
+      "learning_rate": 3.7681350198867274e-05,
+      "loss": 0.7163,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6481650538616031,
+      "grad_norm": 0.7875892519950867,
+      "learning_rate": 3.767464130294438e-05,
+      "loss": 0.7289,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6485302172722294,
+      "grad_norm": 0.9135648608207703,
+      "learning_rate": 3.7667923314127774e-05,
+      "loss": 0.7115,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6488953806828556,
+      "grad_norm": 0.7511104345321655,
+      "learning_rate": 3.766119623587359e-05,
+      "loss": 0.7015,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6492605440934819,
+      "grad_norm": 1.0761756896972656,
+      "learning_rate": 3.765446007164264e-05,
+      "loss": 0.6735,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6496257075041081,
+      "grad_norm": 0.8264586925506592,
+      "learning_rate": 3.764771482490042e-05,
+      "loss": 0.7061,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6499908709147344,
+      "grad_norm": 1.6037218570709229,
+      "learning_rate": 3.7640960499117076e-05,
+      "loss": 0.702,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6503560343253606,
+      "grad_norm": 0.7423888444900513,
+      "learning_rate": 3.763419709776744e-05,
+      "loss": 0.7206,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6507211977359868,
+      "grad_norm": 0.7805701494216919,
+      "learning_rate": 3.762742462433102e-05,
+      "loss": 0.6962,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6510863611466131,
+      "grad_norm": 0.6788707971572876,
+      "learning_rate": 3.7620643082291976e-05,
+      "loss": 0.681,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6514515245572393,
+      "grad_norm": 1.0078482627868652,
+      "learning_rate": 3.761385247513913e-05,
+      "loss": 0.7579,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6518166879678656,
+      "grad_norm": 1.1813578605651855,
+      "learning_rate": 3.760705280636599e-05,
+      "loss": 0.7061,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6521818513784918,
+      "grad_norm": 1.0172258615493774,
+      "learning_rate": 3.760024407947072e-05,
+      "loss": 0.6906,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6525470147891181,
+      "grad_norm": 1.272959589958191,
+      "learning_rate": 3.759342629795611e-05,
+      "loss": 0.7523,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6529121781997443,
+      "grad_norm": 1.2016862630844116,
+      "learning_rate": 3.758659946532965e-05,
+      "loss": 0.715,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6532773416103707,
+      "grad_norm": 0.8221237063407898,
+      "learning_rate": 3.757976358510348e-05,
+      "loss": 0.7262,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6536425050209969,
+      "grad_norm": 0.9917743802070618,
+      "learning_rate": 3.757291866079437e-05,
+      "loss": 0.6689,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6540076684316232,
+      "grad_norm": 1.1483347415924072,
+      "learning_rate": 3.756606469592377e-05,
+      "loss": 0.6982,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6543728318422494,
+      "grad_norm": 0.7733443975448608,
+      "learning_rate": 3.755920169401777e-05,
+      "loss": 0.7169,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6547379952528757,
+      "grad_norm": 1.0676130056381226,
+      "learning_rate": 3.7552329658607096e-05,
+      "loss": 0.751,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6551031586635019,
+      "grad_norm": 1.0856797695159912,
+      "learning_rate": 3.754544859322715e-05,
+      "loss": 0.7233,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6554683220741282,
+      "grad_norm": 0.9236862659454346,
+      "learning_rate": 3.753855850141795e-05,
+      "loss": 0.7167,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6558334854847544,
+      "grad_norm": 0.7494192123413086,
+      "learning_rate": 3.7531659386724195e-05,
+      "loss": 0.727,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6561986488953807,
+      "grad_norm": 0.850273847579956,
+      "learning_rate": 3.752475125269517e-05,
+      "loss": 0.7317,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6565638123060069,
+      "grad_norm": 1.0726951360702515,
+      "learning_rate": 3.7517834102884865e-05,
+      "loss": 0.7631,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6569289757166332,
+      "grad_norm": 0.916655421257019,
+      "learning_rate": 3.751090794085185e-05,
+      "loss": 0.6746,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6572941391272594,
+      "grad_norm": 1.0388219356536865,
+      "learning_rate": 3.750397277015937e-05,
+      "loss": 0.6825,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6576593025378857,
+      "grad_norm": 1.2060333490371704,
+      "learning_rate": 3.74970285943753e-05,
+      "loss": 0.6987,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6580244659485119,
+      "grad_norm": 1.0017542839050293,
+      "learning_rate": 3.749007541707212e-05,
+      "loss": 0.7143,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6583896293591383,
+      "grad_norm": 1.0995457172393799,
+      "learning_rate": 3.7483113241826974e-05,
+      "loss": 0.7184,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6587547927697645,
+      "grad_norm": 0.8721252679824829,
+      "learning_rate": 3.747614207222162e-05,
+      "loss": 0.6864,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6591199561803908,
+      "grad_norm": 0.9801641702651978,
+      "learning_rate": 3.7469161911842444e-05,
+      "loss": 0.7024,
+      "step": 1805
+    },
+    {
+      "epoch": 0.659485119591017,
+      "grad_norm": 1.2977049350738525,
+      "learning_rate": 3.7462172764280456e-05,
+      "loss": 0.6916,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6598502830016433,
+      "grad_norm": 1.3299815654754639,
+      "learning_rate": 3.745517463313129e-05,
+      "loss": 0.7258,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6602154464122695,
+      "grad_norm": 1.1427744626998901,
+      "learning_rate": 3.7448167521995216e-05,
+      "loss": 0.743,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6605806098228958,
+      "grad_norm": 0.9667388200759888,
+      "learning_rate": 3.7441151434477096e-05,
+      "loss": 0.7051,
+      "step": 1809
+    },
+    {
+      "epoch": 0.660945773233522,
+      "grad_norm": 0.8338121771812439,
+      "learning_rate": 3.743412637418644e-05,
+      "loss": 0.7105,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6613109366441483,
+      "grad_norm": 1.0420618057250977,
+      "learning_rate": 3.742709234473735e-05,
+      "loss": 0.7135,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6616761000547745,
+      "grad_norm": 1.397230625152588,
+      "learning_rate": 3.7420049349748555e-05,
+      "loss": 0.6804,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6620412634654008,
+      "grad_norm": 0.7304776310920715,
+      "learning_rate": 3.7412997392843385e-05,
+      "loss": 0.6901,
+      "step": 1813
+    },
+    {
+      "epoch": 0.662406426876027,
+      "grad_norm": 0.9349809288978577,
+      "learning_rate": 3.7405936477649806e-05,
+      "loss": 0.7291,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6627715902866532,
+      "grad_norm": 0.813755214214325,
+      "learning_rate": 3.739886660780037e-05,
+      "loss": 0.7118,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6631367536972795,
+      "grad_norm": 0.8275956511497498,
+      "learning_rate": 3.739178778693222e-05,
+      "loss": 0.7351,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6635019171079057,
+      "grad_norm": 0.8273405432701111,
+      "learning_rate": 3.7384700018687154e-05,
+      "loss": 0.7005,
+      "step": 1817
+    },
+    {
+      "epoch": 0.663867080518532,
+      "grad_norm": 0.8671954870223999,
+      "learning_rate": 3.737760330671153e-05,
+      "loss": 0.715,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6642322439291583,
+      "grad_norm": 1.006998896598816,
+      "learning_rate": 3.737049765465633e-05,
+      "loss": 0.6922,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6645974073397846,
+      "grad_norm": 1.0910762548446655,
+      "learning_rate": 3.736338306617712e-05,
+      "loss": 0.7117,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6649625707504108,
+      "grad_norm": 0.9704879522323608,
+      "learning_rate": 3.735625954493406e-05,
+      "loss": 0.724,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6653277341610371,
+      "grad_norm": 1.0446968078613281,
+      "learning_rate": 3.734912709459194e-05,
+      "loss": 0.7117,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6656928975716633,
+      "grad_norm": 1.0730209350585938,
+      "learning_rate": 3.7341985718820106e-05,
+      "loss": 0.6985,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6660580609822896,
+      "grad_norm": 0.7722004652023315,
+      "learning_rate": 3.733483542129251e-05,
+      "loss": 0.7067,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6664232243929158,
+      "grad_norm": 0.9696376323699951,
+      "learning_rate": 3.732767620568769e-05,
+      "loss": 0.6758,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6667883878035421,
+      "grad_norm": 0.747405469417572,
+      "learning_rate": 3.732050807568878e-05,
+      "loss": 0.6826,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6671535512141683,
+      "grad_norm": 1.2408199310302734,
+      "learning_rate": 3.731333103498349e-05,
+      "loss": 0.7484,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6675187146247946,
+      "grad_norm": 1.3128819465637207,
+      "learning_rate": 3.730614508726413e-05,
+      "loss": 0.7032,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6678838780354208,
+      "grad_norm": 1.001400113105774,
+      "learning_rate": 3.729895023622756e-05,
+      "loss": 0.7014,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6682490414460471,
+      "grad_norm": 0.7678777575492859,
+      "learning_rate": 3.729174648557528e-05,
+      "loss": 0.6633,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6686142048566733,
+      "grad_norm": 0.8901016712188721,
+      "learning_rate": 3.728453383901329e-05,
+      "loss": 0.7197,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6689793682672996,
+      "grad_norm": 0.8703147172927856,
+      "learning_rate": 3.727731230025224e-05,
+      "loss": 0.7244,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6693445316779258,
+      "grad_norm": 1.0041730403900146,
+      "learning_rate": 3.727008187300729e-05,
+      "loss": 0.7062,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6697096950885522,
+      "grad_norm": 1.2102035284042358,
+      "learning_rate": 3.726284256099823e-05,
+      "loss": 0.7057,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6700748584991784,
+      "grad_norm": 0.9731162786483765,
+      "learning_rate": 3.725559436794939e-05,
+      "loss": 0.6986,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6704400219098047,
+      "grad_norm": 0.7118486166000366,
+      "learning_rate": 3.7248337297589666e-05,
+      "loss": 0.7186,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6708051853204309,
+      "grad_norm": 0.725632905960083,
+      "learning_rate": 3.724107135365254e-05,
+      "loss": 0.7206,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6711703487310572,
+      "grad_norm": 1.1240463256835938,
+      "learning_rate": 3.723379653987604e-05,
+      "loss": 0.7195,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6715355121416834,
+      "grad_norm": 0.9441981315612793,
+      "learning_rate": 3.722651286000277e-05,
+      "loss": 0.6832,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6719006755523097,
+      "grad_norm": 2.023256540298462,
+      "learning_rate": 3.7219220317779886e-05,
+      "loss": 0.7089,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6722658389629359,
+      "grad_norm": 0.8320139050483704,
+      "learning_rate": 3.721191891695912e-05,
+      "loss": 0.6735,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6726310023735622,
+      "grad_norm": 0.827463686466217,
+      "learning_rate": 3.720460866129674e-05,
+      "loss": 0.6902,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6729961657841884,
+      "grad_norm": 0.6844232082366943,
+      "learning_rate": 3.719728955455359e-05,
+      "loss": 0.7166,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6733613291948147,
+      "grad_norm": 1.0527371168136597,
+      "learning_rate": 3.718996160049504e-05,
+      "loss": 0.7418,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6737264926054409,
+      "grad_norm": 0.9820262789726257,
+      "learning_rate": 3.718262480289103e-05,
+      "loss": 0.7046,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6740916560160672,
+      "grad_norm": 0.8342486023902893,
+      "learning_rate": 3.7175279165516064e-05,
+      "loss": 0.6899,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6744568194266934,
+      "grad_norm": 1.2259405851364136,
+      "learning_rate": 3.7167924692149164e-05,
+      "loss": 0.7542,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6748219828373198,
+      "grad_norm": 0.8796677589416504,
+      "learning_rate": 3.7160561386573916e-05,
+      "loss": 0.7241,
+      "step": 1848
+    },
+    {
+      "epoch": 0.675187146247946,
+      "grad_norm": 0.9300341606140137,
+      "learning_rate": 3.7153189252578454e-05,
+      "loss": 0.7022,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6755523096585722,
+      "grad_norm": 0.7318381071090698,
+      "learning_rate": 3.7145808293955427e-05,
+      "loss": 0.7015,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6759174730691985,
+      "grad_norm": 1.1048039197921753,
+      "learning_rate": 3.7138418514502055e-05,
+      "loss": 0.7115,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6762826364798247,
+      "grad_norm": 0.9916225671768188,
+      "learning_rate": 3.7131019918020074e-05,
+      "loss": 0.7175,
+      "step": 1852
+    },
+    {
+      "epoch": 0.676647799890451,
+      "grad_norm": 1.0461235046386719,
+      "learning_rate": 3.712361250831578e-05,
+      "loss": 0.6848,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6770129633010772,
+      "grad_norm": 1.782787561416626,
+      "learning_rate": 3.711619628919997e-05,
+      "loss": 0.6769,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6773781267117035,
+      "grad_norm": 1.2121411561965942,
+      "learning_rate": 3.7108771264488e-05,
+      "loss": 0.707,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6777432901223297,
+      "grad_norm": 0.8399494886398315,
+      "learning_rate": 3.7101337437999746e-05,
+      "loss": 0.6915,
+      "step": 1856
+    },
+    {
+      "epoch": 0.678108453532956,
+      "grad_norm": 0.8495551943778992,
+      "learning_rate": 3.709389481355962e-05,
+      "loss": 0.719,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6784736169435822,
+      "grad_norm": 0.7249609231948853,
+      "learning_rate": 3.708644339499654e-05,
+      "loss": 0.7111,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6788387803542085,
+      "grad_norm": 0.7661548256874084,
+      "learning_rate": 3.7078983186143976e-05,
+      "loss": 0.6877,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6792039437648347,
+      "grad_norm": 0.9450551867485046,
+      "learning_rate": 3.7071514190839895e-05,
+      "loss": 0.7233,
+      "step": 1860
+    },
+    {
+      "epoch": 0.679569107175461,
+      "grad_norm": 0.8381118774414062,
+      "learning_rate": 3.706403641292681e-05,
+      "loss": 0.6978,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6799342705860872,
+      "grad_norm": 0.859592616558075,
+      "learning_rate": 3.705654985625171e-05,
+      "loss": 0.7018,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6802994339967136,
+      "grad_norm": 0.7822218537330627,
+      "learning_rate": 3.704905452466616e-05,
+      "loss": 0.7034,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6806645974073398,
+      "grad_norm": 0.9308319687843323,
+      "learning_rate": 3.704155042202619e-05,
+      "loss": 0.7197,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6810297608179661,
+      "grad_norm": 0.8447214365005493,
+      "learning_rate": 3.703403755219236e-05,
+      "loss": 0.7,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6813949242285923,
+      "grad_norm": 0.8934256434440613,
+      "learning_rate": 3.702651591902974e-05,
+      "loss": 0.7039,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6817600876392186,
+      "grad_norm": 0.8217077255249023,
+      "learning_rate": 3.701898552640792e-05,
+      "loss": 0.7259,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6821252510498448,
+      "grad_norm": 1.7904006242752075,
+      "learning_rate": 3.7011446378200965e-05,
+      "loss": 0.6941,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6824904144604711,
+      "grad_norm": 1.0466225147247314,
+      "learning_rate": 3.700389847828749e-05,
+      "loss": 0.6877,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6828555778710973,
+      "grad_norm": 0.8028392791748047,
+      "learning_rate": 3.699634183055056e-05,
+      "loss": 0.6982,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6832207412817236,
+      "grad_norm": 0.9090667366981506,
+      "learning_rate": 3.6988776438877784e-05,
+      "loss": 0.7179,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6835859046923498,
+      "grad_norm": 0.7476855516433716,
+      "learning_rate": 3.698120230716124e-05,
+      "loss": 0.6968,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6839510681029761,
+      "grad_norm": 1.658331274986267,
+      "learning_rate": 3.697361943929753e-05,
+      "loss": 0.7238,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6843162315136023,
+      "grad_norm": 0.8094294667243958,
+      "learning_rate": 3.696602783918773e-05,
+      "loss": 0.6913,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6846813949242286,
+      "grad_norm": 1.531867265701294,
+      "learning_rate": 3.69584275107374e-05,
+      "loss": 0.7319,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6850465583348548,
+      "grad_norm": 0.8356072306632996,
+      "learning_rate": 3.695081845785663e-05,
+      "loss": 0.6787,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6854117217454812,
+      "grad_norm": 0.961771547794342,
+      "learning_rate": 3.6943200684459944e-05,
+      "loss": 0.671,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6857768851561074,
+      "grad_norm": 0.7966907024383545,
+      "learning_rate": 3.69355741944664e-05,
+      "loss": 0.6841,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6861420485667337,
+      "grad_norm": 0.9229552745819092,
+      "learning_rate": 3.692793899179951e-05,
+      "loss": 0.7295,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6865072119773599,
+      "grad_norm": 0.9129878878593445,
+      "learning_rate": 3.6920295080387295e-05,
+      "loss": 0.7352,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6868723753879862,
+      "grad_norm": 0.9927907586097717,
+      "learning_rate": 3.691264246416222e-05,
+      "loss": 0.6975,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6872375387986124,
+      "grad_norm": 0.7063228487968445,
+      "learning_rate": 3.6904981147061265e-05,
+      "loss": 0.7197,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6876027022092386,
+      "grad_norm": 0.8225526809692383,
+      "learning_rate": 3.689731113302587e-05,
+      "loss": 0.702,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6879678656198649,
+      "grad_norm": 0.9241859912872314,
+      "learning_rate": 3.688963242600193e-05,
+      "loss": 0.7156,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6883330290304911,
+      "grad_norm": 0.970552384853363,
+      "learning_rate": 3.688194502993985e-05,
+      "loss": 0.7197,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6886981924411174,
+      "grad_norm": 0.9685255289077759,
+      "learning_rate": 3.6874248948794494e-05,
+      "loss": 0.7068,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6890633558517436,
+      "grad_norm": 0.8250409364700317,
+      "learning_rate": 3.6866544186525156e-05,
+      "loss": 0.691,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6894285192623699,
+      "grad_norm": 1.1441587209701538,
+      "learning_rate": 3.685883074709566e-05,
+      "loss": 0.7062,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6897936826729961,
+      "grad_norm": 1.142409086227417,
+      "learning_rate": 3.685110863447424e-05,
+      "loss": 0.7308,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6901588460836224,
+      "grad_norm": 0.834210991859436,
+      "learning_rate": 3.684337785263363e-05,
+      "loss": 0.7133,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6905240094942486,
+      "grad_norm": 1.0087356567382812,
+      "learning_rate": 3.6835638405550994e-05,
+      "loss": 0.7025,
+      "step": 1891
+    },
+    {
+      "epoch": 0.690889172904875,
+      "grad_norm": 0.8793954849243164,
+      "learning_rate": 3.6827890297207964e-05,
+      "loss": 0.6901,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6912543363155011,
+      "grad_norm": 1.214178204536438,
+      "learning_rate": 3.682013353159065e-05,
+      "loss": 0.701,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6916194997261275,
+      "grad_norm": 1.5995608568191528,
+      "learning_rate": 3.681236811268957e-05,
+      "loss": 0.723,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6919846631367537,
+      "grad_norm": 0.9377299547195435,
+      "learning_rate": 3.680459404449974e-05,
+      "loss": 0.7001,
+      "step": 1895
+    },
+    {
+      "epoch": 0.69234982654738,
+      "grad_norm": 0.7437262535095215,
+      "learning_rate": 3.67968113310206e-05,
+      "loss": 0.7089,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6927149899580062,
+      "grad_norm": 0.8654033541679382,
+      "learning_rate": 3.6789019976256045e-05,
+      "loss": 0.7159,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6930801533686325,
+      "grad_norm": 1.072096347808838,
+      "learning_rate": 3.678121998421441e-05,
+      "loss": 0.7229,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6934453167792587,
+      "grad_norm": 0.7390715479850769,
+      "learning_rate": 3.6773411358908486e-05,
+      "loss": 0.6949,
+      "step": 1899
+    },
+    {
+      "epoch": 0.693810480189885,
+      "grad_norm": 0.9722973704338074,
+      "learning_rate": 3.676559410435549e-05,
+      "loss": 0.7297,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6941756436005112,
+      "grad_norm": 1.1977391242980957,
+      "learning_rate": 3.6757768224577086e-05,
+      "loss": 0.6913,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6945408070111375,
+      "grad_norm": 0.7919796109199524,
+      "learning_rate": 3.6749933723599385e-05,
+      "loss": 0.6944,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6949059704217637,
+      "grad_norm": 0.9592905044555664,
+      "learning_rate": 3.674209060545291e-05,
+      "loss": 0.7401,
+      "step": 1903
+    },
+    {
+      "epoch": 0.69527113383239,
+      "grad_norm": 0.7853878140449524,
+      "learning_rate": 3.6734238874172644e-05,
+      "loss": 0.7006,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6956362972430162,
+      "grad_norm": 0.7083317041397095,
+      "learning_rate": 3.6726378533797976e-05,
+      "loss": 0.6706,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6960014606536425,
+      "grad_norm": 0.6648118495941162,
+      "learning_rate": 3.6718509588372737e-05,
+      "loss": 0.6747,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6963666240642687,
+      "grad_norm": 1.0176913738250732,
+      "learning_rate": 3.6710632041945195e-05,
+      "loss": 0.6844,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6967317874748951,
+      "grad_norm": 0.7009652853012085,
+      "learning_rate": 3.670274589856802e-05,
+      "loss": 0.7198,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6970969508855213,
+      "grad_norm": 1.0065317153930664,
+      "learning_rate": 3.6694851162298315e-05,
+      "loss": 0.71,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6974621142961476,
+      "grad_norm": 0.9386461973190308,
+      "learning_rate": 3.668694783719762e-05,
+      "loss": 0.6891,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6978272777067738,
+      "grad_norm": 1.139102578163147,
+      "learning_rate": 3.6679035927331855e-05,
+      "loss": 0.6519,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6981924411174001,
+      "grad_norm": 0.8256600499153137,
+      "learning_rate": 3.6671115436771404e-05,
+      "loss": 0.7115,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6985576045280263,
+      "grad_norm": 0.9241785407066345,
+      "learning_rate": 3.666318636959102e-05,
+      "loss": 0.6836,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6989227679386526,
+      "grad_norm": 1.092182993888855,
+      "learning_rate": 3.665524872986991e-05,
+      "loss": 0.7064,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6992879313492788,
+      "grad_norm": 1.0015578269958496,
+      "learning_rate": 3.664730252169166e-05,
+      "loss": 0.6837,
+      "step": 1915
+    },
+    {
+      "epoch": 0.699653094759905,
+      "grad_norm": 0.8351472020149231,
+      "learning_rate": 3.663934774914428e-05,
+      "loss": 0.6622,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7000182581705313,
+      "grad_norm": 1.0238572359085083,
+      "learning_rate": 3.6631384416320176e-05,
+      "loss": 0.665,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7003834215811575,
+      "grad_norm": 0.8432892560958862,
+      "learning_rate": 3.662341252731616e-05,
+      "loss": 0.7377,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7007485849917838,
+      "grad_norm": 0.8084632754325867,
+      "learning_rate": 3.6615432086233466e-05,
+      "loss": 0.6875,
+      "step": 1919
+    },
+    {
+      "epoch": 0.70111374840241,
+      "grad_norm": 0.8538629412651062,
+      "learning_rate": 3.660744309717769e-05,
+      "loss": 0.7096,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7014789118130363,
+      "grad_norm": 0.7508218884468079,
+      "learning_rate": 3.6599445564258855e-05,
+      "loss": 0.7041,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7018440752236625,
+      "grad_norm": 0.8841976523399353,
+      "learning_rate": 3.659143949159138e-05,
+      "loss": 0.7489,
+      "step": 1922
+    },
+    {
+      "epoch": 0.7022092386342889,
+      "grad_norm": 0.6620965003967285,
+      "learning_rate": 3.6583424883294053e-05,
+      "loss": 0.7083,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7025744020449151,
+      "grad_norm": 1.078153371810913,
+      "learning_rate": 3.657540174349007e-05,
+      "loss": 0.703,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7029395654555414,
+      "grad_norm": 0.7211769223213196,
+      "learning_rate": 3.656737007630703e-05,
+      "loss": 0.7263,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7033047288661676,
+      "grad_norm": 1.0884227752685547,
+      "learning_rate": 3.6559329885876896e-05,
+      "loss": 0.7,
+      "step": 1926
+    },
+    {
+      "epoch": 0.7036698922767939,
+      "grad_norm": 3.626338243484497,
+      "learning_rate": 3.6551281176336015e-05,
+      "loss": 0.7142,
+      "step": 1927
+    },
+    {
+      "epoch": 0.7040350556874201,
+      "grad_norm": 0.6280802488327026,
+      "learning_rate": 3.654322395182512e-05,
+      "loss": 0.687,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7044002190980464,
+      "grad_norm": 0.8909886479377747,
+      "learning_rate": 3.653515821648936e-05,
+      "loss": 0.7003,
+      "step": 1929
+    },
+    {
+      "epoch": 0.7047653825086726,
+      "grad_norm": 1.0985990762710571,
+      "learning_rate": 3.6527083974478193e-05,
+      "loss": 0.6906,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7051305459192989,
+      "grad_norm": 0.9323737025260925,
+      "learning_rate": 3.651900122994552e-05,
+      "loss": 0.7028,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7054957093299251,
+      "grad_norm": 1.0363316535949707,
+      "learning_rate": 3.651090998704958e-05,
+      "loss": 0.7089,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7058608727405514,
+      "grad_norm": 0.8372681736946106,
+      "learning_rate": 3.650281024995299e-05,
+      "loss": 0.722,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7062260361511776,
+      "grad_norm": 0.9375368356704712,
+      "learning_rate": 3.649470202282275e-05,
+      "loss": 0.6845,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7065911995618039,
+      "grad_norm": 0.8614623546600342,
+      "learning_rate": 3.648658530983021e-05,
+      "loss": 0.7002,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7069563629724301,
+      "grad_norm": 0.7346733212471008,
+      "learning_rate": 3.6478460115151084e-05,
+      "loss": 0.6824,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7073215263830565,
+      "grad_norm": 0.7642632126808167,
+      "learning_rate": 3.6470326442965475e-05,
+      "loss": 0.7172,
+      "step": 1937
+    },
+    {
+      "epoch": 0.7076866897936827,
+      "grad_norm": 0.8814178705215454,
+      "learning_rate": 3.6462184297457826e-05,
+      "loss": 0.6914,
+      "step": 1938
+    },
+    {
+      "epoch": 0.708051853204309,
+      "grad_norm": 0.9905987977981567,
+      "learning_rate": 3.6454033682816946e-05,
+      "loss": 0.7083,
+      "step": 1939
+    },
+    {
+      "epoch": 0.7084170166149352,
+      "grad_norm": 0.930086076259613,
+      "learning_rate": 3.6445874603235986e-05,
+      "loss": 0.7191,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7087821800255615,
+      "grad_norm": 0.8876291513442993,
+      "learning_rate": 3.643770706291248e-05,
+      "loss": 0.6849,
+      "step": 1941
+    },
+    {
+      "epoch": 0.7091473434361877,
+      "grad_norm": 0.9349446892738342,
+      "learning_rate": 3.642953106604829e-05,
+      "loss": 0.6666,
+      "step": 1942
+    },
+    {
+      "epoch": 0.709512506846814,
+      "grad_norm": 0.8735674619674683,
+      "learning_rate": 3.6421346616849645e-05,
+      "loss": 0.719,
+      "step": 1943
+    },
+    {
+      "epoch": 0.7098776702574402,
+      "grad_norm": 0.7789167761802673,
+      "learning_rate": 3.641315371952711e-05,
+      "loss": 0.719,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7102428336680665,
+      "grad_norm": 0.9366868138313293,
+      "learning_rate": 3.640495237829561e-05,
+      "loss": 0.7332,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7106079970786927,
+      "grad_norm": 0.8824470043182373,
+      "learning_rate": 3.63967425973744e-05,
+      "loss": 0.694,
+      "step": 1946
+    },
+    {
+      "epoch": 0.710973160489319,
+      "grad_norm": 0.8090077638626099,
+      "learning_rate": 3.638852438098708e-05,
+      "loss": 0.6626,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7113383238999452,
+      "grad_norm": 0.7595183849334717,
+      "learning_rate": 3.63802977333616e-05,
+      "loss": 0.6954,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7117034873105714,
+      "grad_norm": 0.6301791071891785,
+      "learning_rate": 3.637206265873024e-05,
+      "loss": 0.7201,
+      "step": 1949
+    },
+    {
+      "epoch": 0.7120686507211977,
+      "grad_norm": 1.1005266904830933,
+      "learning_rate": 3.6363819161329606e-05,
+      "loss": 0.7225,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7124338141318239,
+      "grad_norm": 0.7750469446182251,
+      "learning_rate": 3.6355567245400655e-05,
+      "loss": 0.6826,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7127989775424503,
+      "grad_norm": 0.9254831671714783,
+      "learning_rate": 3.634730691518866e-05,
+      "loss": 0.7193,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7131641409530765,
+      "grad_norm": 1.1714247465133667,
+      "learning_rate": 3.633903817494324e-05,
+      "loss": 0.7113,
+      "step": 1953
+    },
+    {
+      "epoch": 0.7135293043637028,
+      "grad_norm": 1.0585075616836548,
+      "learning_rate": 3.633076102891832e-05,
+      "loss": 0.7272,
+      "step": 1954
+    },
+    {
+      "epoch": 0.713894467774329,
+      "grad_norm": 0.7965645790100098,
+      "learning_rate": 3.632247548137217e-05,
+      "loss": 0.6878,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7142596311849553,
+      "grad_norm": 0.6610022187232971,
+      "learning_rate": 3.631418153656736e-05,
+      "loss": 0.6795,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7146247945955815,
+      "grad_norm": 1.0526721477508545,
+      "learning_rate": 3.630587919877079e-05,
+      "loss": 0.7363,
+      "step": 1957
+    },
+    {
+      "epoch": 0.7149899580062078,
+      "grad_norm": 0.8518906831741333,
+      "learning_rate": 3.6297568472253694e-05,
+      "loss": 0.7076,
+      "step": 1958
+    },
+    {
+      "epoch": 0.715355121416834,
+      "grad_norm": 0.9094949960708618,
+      "learning_rate": 3.628924936129161e-05,
+      "loss": 0.6866,
+      "step": 1959
+    },
+    {
+      "epoch": 0.7157202848274603,
+      "grad_norm": 0.8456898331642151,
+      "learning_rate": 3.628092187016436e-05,
+      "loss": 0.7202,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7160854482380865,
+      "grad_norm": 1.3246712684631348,
+      "learning_rate": 3.627258600315612e-05,
+      "loss": 0.7158,
+      "step": 1961
+    },
+    {
+      "epoch": 0.7164506116487128,
+      "grad_norm": 0.8540048599243164,
+      "learning_rate": 3.626424176455537e-05,
+      "loss": 0.6913,
+      "step": 1962
+    },
+    {
+      "epoch": 0.716815775059339,
+      "grad_norm": 0.7801914811134338,
+      "learning_rate": 3.625588915865487e-05,
+      "loss": 0.6888,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7171809384699653,
+      "grad_norm": 0.9037594199180603,
+      "learning_rate": 3.624752818975171e-05,
+      "loss": 0.6992,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7175461018805915,
+      "grad_norm": 1.1611043214797974,
+      "learning_rate": 3.623915886214726e-05,
+      "loss": 0.687,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7179112652912178,
+      "grad_norm": 0.8219496011734009,
+      "learning_rate": 3.6230781180147225e-05,
+      "loss": 0.7015,
+      "step": 1966
+    },
+    {
+      "epoch": 0.718276428701844,
+      "grad_norm": 0.8429041504859924,
+      "learning_rate": 3.622239514806157e-05,
+      "loss": 0.738,
+      "step": 1967
+    },
+    {
+      "epoch": 0.7186415921124704,
+      "grad_norm": 0.7997975945472717,
+      "learning_rate": 3.621400077020457e-05,
+      "loss": 0.6908,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7190067555230966,
+      "grad_norm": 0.9351162314414978,
+      "learning_rate": 3.62055980508948e-05,
+      "loss": 0.6897,
+      "step": 1969
+    },
+    {
+      "epoch": 0.7193719189337229,
+      "grad_norm": 0.7457789778709412,
+      "learning_rate": 3.619718699445513e-05,
+      "loss": 0.665,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7197370823443491,
+      "grad_norm": 0.8901992440223694,
+      "learning_rate": 3.61887676052127e-05,
+      "loss": 0.6707,
+      "step": 1971
+    },
+    {
+      "epoch": 0.7201022457549754,
+      "grad_norm": 0.6807724833488464,
+      "learning_rate": 3.6180339887498953e-05,
+      "loss": 0.6902,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7204674091656016,
+      "grad_norm": 0.8766950964927673,
+      "learning_rate": 3.617190384564961e-05,
+      "loss": 0.6949,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7208325725762279,
+      "grad_norm": 0.9084246754646301,
+      "learning_rate": 3.616345948400468e-05,
+      "loss": 0.7058,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7211977359868541,
+      "grad_norm": 1.0412424802780151,
+      "learning_rate": 3.615500680690843e-05,
+      "loss": 0.6998,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7215628993974804,
+      "grad_norm": 0.8140221238136292,
+      "learning_rate": 3.614654581870945e-05,
+      "loss": 0.6876,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7219280628081066,
+      "grad_norm": 0.6115467548370361,
+      "learning_rate": 3.613807652376057e-05,
+      "loss": 0.6809,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7222932262187329,
+      "grad_norm": 0.7723403573036194,
+      "learning_rate": 3.6129598926418896e-05,
+      "loss": 0.6692,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7226583896293591,
+      "grad_norm": 0.732070803642273,
+      "learning_rate": 3.6121113031045815e-05,
+      "loss": 0.6859,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7230235530399854,
+      "grad_norm": 0.9471202492713928,
+      "learning_rate": 3.611261884200698e-05,
+      "loss": 0.7017,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7233887164506116,
+      "grad_norm": 0.7620534896850586,
+      "learning_rate": 3.6104116363672304e-05,
+      "loss": 0.6991,
+      "step": 1981
+    },
+    {
+      "epoch": 0.723753879861238,
+      "grad_norm": 0.8713348507881165,
+      "learning_rate": 3.6095605600415985e-05,
+      "loss": 0.6714,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7241190432718642,
+      "grad_norm": 0.8815789818763733,
+      "learning_rate": 3.6087086556616457e-05,
+      "loss": 0.6954,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7244842066824904,
+      "grad_norm": 1.0424094200134277,
+      "learning_rate": 3.607855923665643e-05,
+      "loss": 0.6781,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7248493700931167,
+      "grad_norm": 0.8389812111854553,
+      "learning_rate": 3.607002364492287e-05,
+      "loss": 0.6414,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7252145335037429,
+      "grad_norm": 0.8990022540092468,
+      "learning_rate": 3.6061479785806996e-05,
+      "loss": 0.6917,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7255796969143692,
+      "grad_norm": 1.0698925256729126,
+      "learning_rate": 3.6052927663704276e-05,
+      "loss": 0.7104,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7259448603249954,
+      "grad_norm": 0.9811051487922668,
+      "learning_rate": 3.604436728301443e-05,
+      "loss": 0.7292,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7263100237356217,
+      "grad_norm": 0.7819287776947021,
+      "learning_rate": 3.603579864814145e-05,
+      "loss": 0.7457,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7266751871462479,
+      "grad_norm": 0.7379570603370667,
+      "learning_rate": 3.6027221763493534e-05,
+      "loss": 0.7169,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7270403505568742,
+      "grad_norm": 1.0845247507095337,
+      "learning_rate": 3.6018636633483154e-05,
+      "loss": 0.6904,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7274055139675004,
+      "grad_norm": 1.1536235809326172,
+      "learning_rate": 3.601004326252702e-05,
+      "loss": 0.6995,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7277706773781267,
+      "grad_norm": 0.9554802179336548,
+      "learning_rate": 3.600144165504607e-05,
+      "loss": 0.6824,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7281358407887529,
+      "grad_norm": 1.0458979606628418,
+      "learning_rate": 3.5992831815465476e-05,
+      "loss": 0.7018,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7285010041993792,
+      "grad_norm": 0.8219031095504761,
+      "learning_rate": 3.598421374821468e-05,
+      "loss": 0.6919,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7288661676100054,
+      "grad_norm": 0.7273349761962891,
+      "learning_rate": 3.59755874577273e-05,
+      "loss": 0.6747,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7292313310206318,
+      "grad_norm": 0.8271771669387817,
+      "learning_rate": 3.596695294844124e-05,
+      "loss": 0.6915,
+      "step": 1997
+    },
+    {
+      "epoch": 0.729596494431258,
+      "grad_norm": 0.9043617248535156,
+      "learning_rate": 3.5958310224798605e-05,
+      "loss": 0.7148,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7299616578418843,
+      "grad_norm": 0.9108407497406006,
+      "learning_rate": 3.5949659291245727e-05,
+      "loss": 0.699,
+      "step": 1999
+    },
+    {
+      "epoch": 0.7303268212525105,
+      "grad_norm": 0.9600886106491089,
+      "learning_rate": 3.5941000152233166e-05,
+      "loss": 0.6747,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7306919846631368,
+      "grad_norm": 2.9126079082489014,
+      "learning_rate": 3.5932332812215694e-05,
+      "loss": 0.6936,
+      "step": 2001
+    },
+    {
+      "epoch": 0.731057148073763,
+      "grad_norm": 0.8958035111427307,
+      "learning_rate": 3.5923657275652316e-05,
+      "loss": 0.6804,
+      "step": 2002
+    },
+    {
+      "epoch": 0.7314223114843893,
+      "grad_norm": 1.0890897512435913,
+      "learning_rate": 3.5914973547006244e-05,
+      "loss": 0.7119,
+      "step": 2003
+    },
+    {
+      "epoch": 0.7317874748950155,
+      "grad_norm": 0.9848058223724365,
+      "learning_rate": 3.5906281630744914e-05,
+      "loss": 0.7091,
+      "step": 2004
+    },
+    {
+      "epoch": 0.7321526383056418,
+      "grad_norm": 1.0069018602371216,
+      "learning_rate": 3.589758153133996e-05,
+      "loss": 0.7079,
+      "step": 2005
+    },
+    {
+      "epoch": 0.732517801716268,
+      "grad_norm": 0.8909217119216919,
+      "learning_rate": 3.588887325326725e-05,
+      "loss": 0.7091,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7328829651268943,
+      "grad_norm": 0.9334930181503296,
+      "learning_rate": 3.5880156801006826e-05,
+      "loss": 0.6807,
+      "step": 2007
+    },
+    {
+      "epoch": 0.7332481285375205,
+      "grad_norm": 0.8287849426269531,
+      "learning_rate": 3.587143217904295e-05,
+      "loss": 0.6672,
+      "step": 2008
+    },
+    {
+      "epoch": 0.7336132919481468,
+      "grad_norm": 1.1333422660827637,
+      "learning_rate": 3.58626993918641e-05,
+      "loss": 0.6932,
+      "step": 2009
+    },
+    {
+      "epoch": 0.733978455358773,
+      "grad_norm": 1.2113679647445679,
+      "learning_rate": 3.585395844396295e-05,
+      "loss": 0.6713,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7343436187693994,
+      "grad_norm": 0.7048681378364563,
+      "learning_rate": 3.584520933983636e-05,
+      "loss": 0.6991,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7347087821800256,
+      "grad_norm": 0.9552350044250488,
+      "learning_rate": 3.5836452083985394e-05,
+      "loss": 0.7224,
+      "step": 2012
+    },
+    {
+      "epoch": 0.7350739455906519,
+      "grad_norm": 1.0799976587295532,
+      "learning_rate": 3.58276866809153e-05,
+      "loss": 0.6576,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7354391090012781,
+      "grad_norm": 2.1734097003936768,
+      "learning_rate": 3.581891313513555e-05,
+      "loss": 0.7068,
+      "step": 2014
+    },
+    {
+      "epoch": 0.7358042724119044,
+      "grad_norm": 0.9834950566291809,
+      "learning_rate": 3.581013145115975e-05,
+      "loss": 0.6852,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7361694358225306,
+      "grad_norm": 0.8883312940597534,
+      "learning_rate": 3.580134163350575e-05,
+      "loss": 0.6896,
+      "step": 2016
+    },
+    {
+      "epoch": 0.7365345992331568,
+      "grad_norm": 0.9405683279037476,
+      "learning_rate": 3.5792543686695544e-05,
+      "loss": 0.6899,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7368997626437831,
+      "grad_norm": 0.8160017728805542,
+      "learning_rate": 3.5783737615255326e-05,
+      "loss": 0.7014,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7372649260544093,
+      "grad_norm": 0.8430017828941345,
+      "learning_rate": 3.5774923423715464e-05,
+      "loss": 0.6951,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7376300894650356,
+      "grad_norm": 1.1991838216781616,
+      "learning_rate": 3.576610111661051e-05,
+      "loss": 0.6572,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7379952528756618,
+      "grad_norm": 0.6597726941108704,
+      "learning_rate": 3.5757270698479186e-05,
+      "loss": 0.6934,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7383604162862881,
+      "grad_norm": 1.0859240293502808,
+      "learning_rate": 3.5748432173864394e-05,
+      "loss": 0.7139,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7387255796969143,
+      "grad_norm": 0.8668961524963379,
+      "learning_rate": 3.573958554731319e-05,
+      "loss": 0.6971,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7390907431075406,
+      "grad_norm": 0.9613633155822754,
+      "learning_rate": 3.573073082337681e-05,
+      "loss": 0.7083,
+      "step": 2024
+    },
+    {
+      "epoch": 0.7394559065181668,
+      "grad_norm": 0.6835755109786987,
+      "learning_rate": 3.572186800661065e-05,
+      "loss": 0.7037,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7398210699287932,
+      "grad_norm": 0.9177113771438599,
+      "learning_rate": 3.571299710157429e-05,
+      "loss": 0.6993,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7401862333394194,
+      "grad_norm": 0.9868367314338684,
+      "learning_rate": 3.570411811283144e-05,
+      "loss": 0.7054,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7405513967500457,
+      "grad_norm": 1.4232624769210815,
+      "learning_rate": 3.569523104494999e-05,
+      "loss": 0.6859,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7409165601606719,
+      "grad_norm": 1.0719857215881348,
+      "learning_rate": 3.568633590250198e-05,
+      "loss": 0.7003,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7412817235712982,
+      "grad_norm": 0.705696702003479,
+      "learning_rate": 3.56774326900636e-05,
+      "loss": 0.6842,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7416468869819244,
+      "grad_norm": 1.0737361907958984,
+      "learning_rate": 3.5668521412215194e-05,
+      "loss": 0.7257,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7420120503925507,
+      "grad_norm": 0.9407960772514343,
+      "learning_rate": 3.5659602073541256e-05,
+      "loss": 0.7186,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7423772138031769,
+      "grad_norm": 0.8402684330940247,
+      "learning_rate": 3.565067467863044e-05,
+      "loss": 0.7084,
+      "step": 2033
+    },
+    {
+      "epoch": 0.7427423772138032,
+      "grad_norm": 1.0480576753616333,
+      "learning_rate": 3.564173923207553e-05,
+      "loss": 0.696,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7431075406244294,
+      "grad_norm": 1.7584154605865479,
+      "learning_rate": 3.563279573847344e-05,
+      "loss": 0.7129,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7434727040350557,
+      "grad_norm": 1.2279152870178223,
+      "learning_rate": 3.5623844202425245e-05,
+      "loss": 0.7009,
+      "step": 2036
+    },
+    {
+      "epoch": 0.7438378674456819,
+      "grad_norm": 0.8674067258834839,
+      "learning_rate": 3.5614884628536156e-05,
+      "loss": 0.7026,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7442030308563082,
+      "grad_norm": 0.8557435274124146,
+      "learning_rate": 3.560591702141552e-05,
+      "loss": 0.6932,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7445681942669344,
+      "grad_norm": 0.8920378088951111,
+      "learning_rate": 3.559694138567679e-05,
+      "loss": 0.6843,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7449333576775607,
+      "grad_norm": 1.034100890159607,
+      "learning_rate": 3.558795772593759e-05,
+      "loss": 0.7097,
+      "step": 2040
+    },
+    {
+      "epoch": 0.745298521088187,
+      "grad_norm": 0.7674341201782227,
+      "learning_rate": 3.5578966046819644e-05,
+      "loss": 0.6938,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7456636844988133,
+      "grad_norm": 0.9016082286834717,
+      "learning_rate": 3.556996635294881e-05,
+      "loss": 0.6827,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7460288479094395,
+      "grad_norm": 0.782569408416748,
+      "learning_rate": 3.556095864895508e-05,
+      "loss": 0.6498,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7463940113200658,
+      "grad_norm": 0.9244042038917542,
+      "learning_rate": 3.555194293947254e-05,
+      "loss": 0.698,
+      "step": 2044
+    },
+    {
+      "epoch": 0.746759174730692,
+      "grad_norm": 0.8407185673713684,
+      "learning_rate": 3.554291922913942e-05,
+      "loss": 0.7162,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7471243381413183,
+      "grad_norm": 1.2670552730560303,
+      "learning_rate": 3.553388752259806e-05,
+      "loss": 0.6927,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7474895015519445,
+      "grad_norm": 0.8736078143119812,
+      "learning_rate": 3.5524847824494896e-05,
+      "loss": 0.7036,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7478546649625708,
+      "grad_norm": 0.7163779735565186,
+      "learning_rate": 3.5515800139480505e-05,
+      "loss": 0.7062,
+      "step": 2048
+    },
+    {
+      "epoch": 0.748219828373197,
+      "grad_norm": 0.5790550112724304,
+      "learning_rate": 3.5506744472209556e-05,
+      "loss": 0.6837,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7485849917838232,
+      "grad_norm": 1.0888993740081787,
+      "learning_rate": 3.5497680827340816e-05,
+      "loss": 0.7061,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7489501551944495,
+      "grad_norm": 0.7524549961090088,
+      "learning_rate": 3.5488609209537176e-05,
+      "loss": 0.6826,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7493153186050757,
+      "grad_norm": 1.128859281539917,
+      "learning_rate": 3.547952962346562e-05,
+      "loss": 0.7017,
+      "step": 2052
+    },
+    {
+      "epoch": 0.749680482015702,
+      "grad_norm": 0.9107731580734253,
+      "learning_rate": 3.5470442073797224e-05,
+      "loss": 0.6801,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7500456454263282,
+      "grad_norm": 0.8646042346954346,
+      "learning_rate": 3.5461346565207174e-05,
+      "loss": 0.6924,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7504108088369545,
+      "grad_norm": 0.8979294896125793,
+      "learning_rate": 3.5452243102374737e-05,
+      "loss": 0.6661,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7507759722475807,
+      "grad_norm": 1.0486830472946167,
+      "learning_rate": 3.5443131689983285e-05,
+      "loss": 0.6636,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7511411356582071,
+      "grad_norm": 0.8791133761405945,
+      "learning_rate": 3.543401233272028e-05,
+      "loss": 0.7052,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7515062990688333,
+      "grad_norm": 0.8093293905258179,
+      "learning_rate": 3.5424885035277255e-05,
+      "loss": 0.6693,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7518714624794596,
+      "grad_norm": 0.7917428612709045,
+      "learning_rate": 3.541574980234983e-05,
+      "loss": 0.6754,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7522366258900858,
+      "grad_norm": 0.9093384742736816,
+      "learning_rate": 3.540660663863774e-05,
+      "loss": 0.7178,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7526017893007121,
+      "grad_norm": 0.7561656832695007,
+      "learning_rate": 3.539745554884476e-05,
+      "loss": 0.6684,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7529669527113383,
+      "grad_norm": 0.9925743937492371,
+      "learning_rate": 3.5388296537678765e-05,
+      "loss": 0.7123,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7533321161219646,
+      "grad_norm": 0.7667102813720703,
+      "learning_rate": 3.537912960985169e-05,
+      "loss": 0.6581,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7536972795325908,
+      "grad_norm": 1.0372451543807983,
+      "learning_rate": 3.536995477007955e-05,
+      "loss": 0.6929,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7540624429432171,
+      "grad_norm": 1.0563349723815918,
+      "learning_rate": 3.5360772023082446e-05,
+      "loss": 0.7054,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7544276063538433,
+      "grad_norm": 1.1406291723251343,
+      "learning_rate": 3.535158137358453e-05,
+      "loss": 0.7117,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7547927697644696,
+      "grad_norm": 0.67351233959198,
+      "learning_rate": 3.534238282631401e-05,
+      "loss": 0.703,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7551579331750958,
+      "grad_norm": 0.8350493311882019,
+      "learning_rate": 3.533317638600319e-05,
+      "loss": 0.6952,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7555230965857221,
+      "grad_norm": 0.8542644381523132,
+      "learning_rate": 3.532396205738839e-05,
+      "loss": 0.6871,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7558882599963483,
+      "grad_norm": 0.8126198649406433,
+      "learning_rate": 3.5314739845210027e-05,
+      "loss": 0.6661,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7562534234069747,
+      "grad_norm": 1.0363764762878418,
+      "learning_rate": 3.530550975421255e-05,
+      "loss": 0.7021,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7566185868176009,
+      "grad_norm": 0.8353523015975952,
+      "learning_rate": 3.529627178914448e-05,
+      "loss": 0.7183,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7569837502282272,
+      "grad_norm": 0.7634402513504028,
+      "learning_rate": 3.5287025954758385e-05,
+      "loss": 0.696,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7573489136388534,
+      "grad_norm": 1.201458215713501,
+      "learning_rate": 3.5277772255810855e-05,
+      "loss": 0.6416,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7577140770494797,
+      "grad_norm": 0.7530384659767151,
+      "learning_rate": 3.526851069706256e-05,
+      "loss": 0.6996,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7580792404601059,
+      "grad_norm": 0.8554724454879761,
+      "learning_rate": 3.5259241283278204e-05,
+      "loss": 0.6973,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7584444038707322,
+      "grad_norm": 1.2510430812835693,
+      "learning_rate": 3.5249964019226514e-05,
+      "loss": 0.7042,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7588095672813584,
+      "grad_norm": 0.9825018048286438,
+      "learning_rate": 3.524067890968029e-05,
+      "loss": 0.6988,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7591747306919847,
+      "grad_norm": 0.7476685643196106,
+      "learning_rate": 3.523138595941633e-05,
+      "loss": 0.6993,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7595398941026109,
+      "grad_norm": 0.8344430923461914,
+      "learning_rate": 3.5222085173215495e-05,
+      "loss": 0.6678,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7599050575132372,
+      "grad_norm": 0.9352977275848389,
+      "learning_rate": 3.521277655586266e-05,
+      "loss": 0.7004,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7602702209238634,
+      "grad_norm": 1.1583284139633179,
+      "learning_rate": 3.520346011214674e-05,
+      "loss": 0.6855,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7606353843344897,
+      "grad_norm": 0.9419161081314087,
+      "learning_rate": 3.519413584686067e-05,
+      "loss": 0.687,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7610005477451159,
+      "grad_norm": 1.1509552001953125,
+      "learning_rate": 3.518480376480141e-05,
+      "loss": 0.7069,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7613657111557421,
+      "grad_norm": 0.9876951575279236,
+      "learning_rate": 3.5175463870769935e-05,
+      "loss": 0.6963,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7617308745663685,
+      "grad_norm": 1.2705250978469849,
+      "learning_rate": 3.516611616957125e-05,
+      "loss": 0.7005,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7620960379769947,
+      "grad_norm": 0.8080708980560303,
+      "learning_rate": 3.515676066601438e-05,
+      "loss": 0.6875,
+      "step": 2087
+    },
+    {
+      "epoch": 0.762461201387621,
+      "grad_norm": 0.7331348657608032,
+      "learning_rate": 3.514739736491235e-05,
+      "loss": 0.6668,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7628263647982472,
+      "grad_norm": 0.698076069355011,
+      "learning_rate": 3.51380262710822e-05,
+      "loss": 0.6801,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7631915282088735,
+      "grad_norm": 0.7155959010124207,
+      "learning_rate": 3.512864738934499e-05,
+      "loss": 0.6964,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7635566916194997,
+      "grad_norm": 0.8522117733955383,
+      "learning_rate": 3.511926072452578e-05,
+      "loss": 0.7069,
+      "step": 2091
+    },
+    {
+      "epoch": 0.763921855030126,
+      "grad_norm": 0.8499336838722229,
+      "learning_rate": 3.5109866281453606e-05,
+      "loss": 0.6711,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7642870184407522,
+      "grad_norm": 0.7255849242210388,
+      "learning_rate": 3.510046406496157e-05,
+      "loss": 0.6665,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7646521818513785,
+      "grad_norm": 1.1719845533370972,
+      "learning_rate": 3.50910540798867e-05,
+      "loss": 0.666,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7650173452620047,
+      "grad_norm": 0.9627401232719421,
+      "learning_rate": 3.508163633107008e-05,
+      "loss": 0.7095,
+      "step": 2095
+    },
+    {
+      "epoch": 0.765382508672631,
+      "grad_norm": 0.8184304237365723,
+      "learning_rate": 3.507221082335676e-05,
+      "loss": 0.6829,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7657476720832572,
+      "grad_norm": 0.9856240153312683,
+      "learning_rate": 3.5062777561595776e-05,
+      "loss": 0.7136,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7661128354938835,
+      "grad_norm": 1.0894156694412231,
+      "learning_rate": 3.505333655064017e-05,
+      "loss": 0.6995,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7664779989045097,
+      "grad_norm": 1.1927454471588135,
+      "learning_rate": 3.5043887795346966e-05,
+      "loss": 0.7103,
+      "step": 2099
+    },
+    {
+      "epoch": 0.766843162315136,
+      "grad_norm": 0.9579023718833923,
+      "learning_rate": 3.503443130057715e-05,
+      "loss": 0.6916,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7672083257257623,
+      "grad_norm": 1.1369895935058594,
+      "learning_rate": 3.5024967071195736e-05,
+      "loss": 0.7256,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7675734891363886,
+      "grad_norm": 0.7931919097900391,
+      "learning_rate": 3.501549511207168e-05,
+      "loss": 0.7003,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7679386525470148,
+      "grad_norm": 1.0025581121444702,
+      "learning_rate": 3.500601542807792e-05,
+      "loss": 0.6724,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7683038159576411,
+      "grad_norm": 0.9976342916488647,
+      "learning_rate": 3.499652802409137e-05,
+      "loss": 0.6695,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7686689793682673,
+      "grad_norm": 0.6871845722198486,
+      "learning_rate": 3.4987032904992935e-05,
+      "loss": 0.6914,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7690341427788936,
+      "grad_norm": 1.060022234916687,
+      "learning_rate": 3.497753007566746e-05,
+      "loss": 0.6821,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7693993061895198,
+      "grad_norm": 1.0832650661468506,
+      "learning_rate": 3.4968019541003765e-05,
+      "loss": 0.6801,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7697644696001461,
+      "grad_norm": 0.8646666407585144,
+      "learning_rate": 3.495850130589465e-05,
+      "loss": 0.6887,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7701296330107723,
+      "grad_norm": 0.6750617623329163,
+      "learning_rate": 3.494897537523686e-05,
+      "loss": 0.6979,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7704947964213986,
+      "grad_norm": 1.157203197479248,
+      "learning_rate": 3.493944175393111e-05,
+      "loss": 0.6683,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7708599598320248,
+      "grad_norm": 0.9452328085899353,
+      "learning_rate": 3.492990044688205e-05,
+      "loss": 0.6818,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7712251232426511,
+      "grad_norm": 0.7917987704277039,
+      "learning_rate": 3.4920351458998316e-05,
+      "loss": 0.6987,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7715902866532773,
+      "grad_norm": 0.9351409077644348,
+      "learning_rate": 3.491079479519248e-05,
+      "loss": 0.6997,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7719554500639036,
+      "grad_norm": 0.8905813097953796,
+      "learning_rate": 3.490123046038104e-05,
+      "loss": 0.6809,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7723206134745298,
+      "grad_norm": 0.8470001220703125,
+      "learning_rate": 3.489165845948448e-05,
+      "loss": 0.6904,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7726857768851562,
+      "grad_norm": 1.801979660987854,
+      "learning_rate": 3.488207879742722e-05,
+      "loss": 0.6968,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7730509402957824,
+      "grad_norm": 0.9084315299987793,
+      "learning_rate": 3.487249147913759e-05,
+      "loss": 0.6625,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7734161037064086,
+      "grad_norm": 1.1759934425354004,
+      "learning_rate": 3.4862896509547886e-05,
+      "loss": 0.6912,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7737812671170349,
+      "grad_norm": 0.7581272125244141,
+      "learning_rate": 3.485329389359434e-05,
+      "loss": 0.7298,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7741464305276611,
+      "grad_norm": 0.9350868463516235,
+      "learning_rate": 3.484368363621712e-05,
+      "loss": 0.6889,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7745115939382874,
+      "grad_norm": 1.0078556537628174,
+      "learning_rate": 3.48340657423603e-05,
+      "loss": 0.718,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7748767573489136,
+      "grad_norm": 0.9204308986663818,
+      "learning_rate": 3.482444021697192e-05,
+      "loss": 0.686,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7752419207595399,
+      "grad_norm": 0.7891234755516052,
+      "learning_rate": 3.481480706500391e-05,
+      "loss": 0.6902,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7756070841701661,
+      "grad_norm": 0.8696504831314087,
+      "learning_rate": 3.480516629141214e-05,
+      "loss": 0.7117,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7759722475807924,
+      "grad_norm": 0.9487229585647583,
+      "learning_rate": 3.479551790115642e-05,
+      "loss": 0.6794,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7763374109914186,
+      "grad_norm": 0.8154385089874268,
+      "learning_rate": 3.4785861899200434e-05,
+      "loss": 0.6934,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7767025744020449,
+      "grad_norm": 1.1768670082092285,
+      "learning_rate": 3.477619829051183e-05,
+      "loss": 0.6923,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7770677378126711,
+      "grad_norm": 0.9852544069290161,
+      "learning_rate": 3.476652708006214e-05,
+      "loss": 0.7153,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7774329012232974,
+      "grad_norm": 0.7022532820701599,
+      "learning_rate": 3.4756848272826795e-05,
+      "loss": 0.6879,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7777980646339236,
+      "grad_norm": 1.0566223859786987,
+      "learning_rate": 3.474716187378518e-05,
+      "loss": 0.6808,
+      "step": 2130
+    },
+    {
+      "epoch": 0.77816322804455,
+      "grad_norm": 0.8685449361801147,
+      "learning_rate": 3.4737467887920556e-05,
+      "loss": 0.6829,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7785283914551762,
+      "grad_norm": 0.8262388706207275,
+      "learning_rate": 3.4727766320220064e-05,
+      "loss": 0.6729,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7788935548658025,
+      "grad_norm": 0.6778239607810974,
+      "learning_rate": 3.47180571756748e-05,
+      "loss": 0.6755,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7792587182764287,
+      "grad_norm": 0.7228227853775024,
+      "learning_rate": 3.470834045927971e-05,
+      "loss": 0.7108,
+      "step": 2134
+    },
+    {
+      "epoch": 0.779623881687055,
+      "grad_norm": 0.8152421116828918,
+      "learning_rate": 3.469861617603367e-05,
+      "loss": 0.6943,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7799890450976812,
+      "grad_norm": 0.9563167095184326,
+      "learning_rate": 3.468888433093943e-05,
+      "loss": 0.6711,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7803542085083075,
+      "grad_norm": 0.8263843059539795,
+      "learning_rate": 3.4679144929003624e-05,
+      "loss": 0.6996,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7807193719189337,
+      "grad_norm": 0.9184187650680542,
+      "learning_rate": 3.466939797523679e-05,
+      "loss": 0.6791,
+      "step": 2138
+    },
+    {
+      "epoch": 0.78108453532956,
+      "grad_norm": 1.192243218421936,
+      "learning_rate": 3.465964347465334e-05,
+      "loss": 0.6846,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7814496987401862,
+      "grad_norm": 1.0710678100585938,
+      "learning_rate": 3.464988143227158e-05,
+      "loss": 0.7174,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7818148621508125,
+      "grad_norm": 0.8729625940322876,
+      "learning_rate": 3.4640111853113686e-05,
+      "loss": 0.6285,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7821800255614387,
+      "grad_norm": 1.0296928882598877,
+      "learning_rate": 3.463033474220572e-05,
+      "loss": 0.6873,
+      "step": 2142
+    },
+    {
+      "epoch": 0.782545188972065,
+      "grad_norm": 1.0441279411315918,
+      "learning_rate": 3.46205501045776e-05,
+      "loss": 0.6527,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7829103523826912,
+      "grad_norm": 0.9956523180007935,
+      "learning_rate": 3.461075794526314e-05,
+      "loss": 0.6841,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7832755157933176,
+      "grad_norm": 0.896945059299469,
+      "learning_rate": 3.460095826930001e-05,
+      "loss": 0.7008,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7836406792039438,
+      "grad_norm": 0.9649824500083923,
+      "learning_rate": 3.4591151081729756e-05,
+      "loss": 0.6786,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7840058426145701,
+      "grad_norm": 0.975921630859375,
+      "learning_rate": 3.458133638759777e-05,
+      "loss": 0.696,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7843710060251963,
+      "grad_norm": 0.9992843270301819,
+      "learning_rate": 3.457151419195332e-05,
+      "loss": 0.7158,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7847361694358226,
+      "grad_norm": 1.2938470840454102,
+      "learning_rate": 3.456168449984955e-05,
+      "loss": 0.657,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7851013328464488,
+      "grad_norm": 0.8014959096908569,
+      "learning_rate": 3.4551847316343426e-05,
+      "loss": 0.7029,
+      "step": 2150
+    },
+    {
+      "epoch": 0.785466496257075,
+      "grad_norm": 1.136978030204773,
+      "learning_rate": 3.4542002646495784e-05,
+      "loss": 0.7255,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7858316596677013,
+      "grad_norm": 0.9775698184967041,
+      "learning_rate": 3.453215049537131e-05,
+      "loss": 0.6677,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7861968230783275,
+      "grad_norm": 0.7203195095062256,
+      "learning_rate": 3.452229086803856e-05,
+      "loss": 0.6564,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7865619864889538,
+      "grad_norm": 1.0495548248291016,
+      "learning_rate": 3.451242376956988e-05,
+      "loss": 0.7411,
+      "step": 2154
+    },
+    {
+      "epoch": 0.78692714989958,
+      "grad_norm": 1.1219384670257568,
+      "learning_rate": 3.4502549205041534e-05,
+      "loss": 0.6976,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7872923133102063,
+      "grad_norm": 0.7579952478408813,
+      "learning_rate": 3.449266717953357e-05,
+      "loss": 0.6864,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7876574767208325,
+      "grad_norm": 0.9256203174591064,
+      "learning_rate": 3.44827776981299e-05,
+      "loss": 0.7102,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7880226401314588,
+      "grad_norm": 1.4947396516799927,
+      "learning_rate": 3.447288076591825e-05,
+      "loss": 0.7104,
+      "step": 2158
+    },
+    {
+      "epoch": 0.788387803542085,
+      "grad_norm": 1.0526587963104248,
+      "learning_rate": 3.446297638799022e-05,
+      "loss": 0.6885,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7887529669527114,
+      "grad_norm": 1.2474300861358643,
+      "learning_rate": 3.445306456944119e-05,
+      "loss": 0.6862,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7891181303633376,
+      "grad_norm": 0.6906046867370605,
+      "learning_rate": 3.444314531537041e-05,
+      "loss": 0.6976,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7894832937739639,
+      "grad_norm": 1.101379632949829,
+      "learning_rate": 3.443321863088093e-05,
+      "loss": 0.675,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7898484571845901,
+      "grad_norm": 0.7387844324111938,
+      "learning_rate": 3.4423284521079635e-05,
+      "loss": 0.6929,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7902136205952164,
+      "grad_norm": 1.1480523347854614,
+      "learning_rate": 3.441334299107722e-05,
+      "loss": 0.6868,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7905787840058426,
+      "grad_norm": 0.8580271601676941,
+      "learning_rate": 3.440339404598822e-05,
+      "loss": 0.6732,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7909439474164689,
+      "grad_norm": 1.411083459854126,
+      "learning_rate": 3.4393437690930944e-05,
+      "loss": 0.7169,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7913091108270951,
+      "grad_norm": 1.0092064142227173,
+      "learning_rate": 3.438347393102755e-05,
+      "loss": 0.645,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7916742742377214,
+      "grad_norm": 0.6816556453704834,
+      "learning_rate": 3.4373502771403995e-05,
+      "loss": 0.6823,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7920394376483476,
+      "grad_norm": 0.9781081080436707,
+      "learning_rate": 3.436352421719004e-05,
+      "loss": 0.7077,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7924046010589739,
+      "grad_norm": 0.9609267115592957,
+      "learning_rate": 3.4353538273519244e-05,
+      "loss": 0.6879,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7927697644696001,
+      "grad_norm": 0.986900806427002,
+      "learning_rate": 3.4343544945528975e-05,
+      "loss": 0.6833,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7931349278802264,
+      "grad_norm": 0.7960975766181946,
+      "learning_rate": 3.4333544238360404e-05,
+      "loss": 0.6609,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7935000912908526,
+      "grad_norm": 0.7304735779762268,
+      "learning_rate": 3.432353615715849e-05,
+      "loss": 0.6893,
+      "step": 2173
+    },
+    {
+      "epoch": 0.793865254701479,
+      "grad_norm": 0.8587673902511597,
+      "learning_rate": 3.431352070707199e-05,
+      "loss": 0.6996,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7942304181121052,
+      "grad_norm": 1.1969038248062134,
+      "learning_rate": 3.430349789325346e-05,
+      "loss": 0.685,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7945955815227315,
+      "grad_norm": 0.762546718120575,
+      "learning_rate": 3.429346772085923e-05,
+      "loss": 0.6876,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7949607449333577,
+      "grad_norm": 1.1056230068206787,
+      "learning_rate": 3.42834301950494e-05,
+      "loss": 0.7068,
+      "step": 2177
+    },
+    {
+      "epoch": 0.795325908343984,
+      "grad_norm": 0.8168462514877319,
+      "learning_rate": 3.427338532098791e-05,
+      "loss": 0.6572,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7956910717546102,
+      "grad_norm": 1.0522792339324951,
+      "learning_rate": 3.4263333103842415e-05,
+      "loss": 0.7001,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7960562351652365,
+      "grad_norm": 0.9588885307312012,
+      "learning_rate": 3.42532735487844e-05,
+      "loss": 0.6782,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7964213985758627,
+      "grad_norm": 0.9216328859329224,
+      "learning_rate": 3.424320666098909e-05,
+      "loss": 0.6941,
+      "step": 2181
+    },
+    {
+      "epoch": 0.796786561986489,
+      "grad_norm": 0.9530096650123596,
+      "learning_rate": 3.4233132445635496e-05,
+      "loss": 0.6492,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7971517253971152,
+      "grad_norm": 1.0291130542755127,
+      "learning_rate": 3.42230509079064e-05,
+      "loss": 0.7058,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7975168888077415,
+      "grad_norm": 0.9647759199142456,
+      "learning_rate": 3.421296205298835e-05,
+      "loss": 0.6777,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7978820522183677,
+      "grad_norm": 1.066928744316101,
+      "learning_rate": 3.420286588607165e-05,
+      "loss": 0.683,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7982472156289939,
+      "grad_norm": 1.107342004776001,
+      "learning_rate": 3.4192762412350375e-05,
+      "loss": 0.6785,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7986123790396202,
+      "grad_norm": 0.8849497437477112,
+      "learning_rate": 3.418265163702236e-05,
+      "loss": 0.6871,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7989775424502464,
+      "grad_norm": 0.997079074382782,
+      "learning_rate": 3.41725335652892e-05,
+      "loss": 0.6908,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7993427058608727,
+      "grad_norm": 1.009993553161621,
+      "learning_rate": 3.4162408202356224e-05,
+      "loss": 0.7019,
+      "step": 2189
+    },
+    {
+      "epoch": 0.799707869271499,
+      "grad_norm": 0.8812515735626221,
+      "learning_rate": 3.4152275553432524e-05,
+      "loss": 0.663,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8000730326821253,
+      "grad_norm": 0.9019457697868347,
+      "learning_rate": 3.4142135623730954e-05,
+      "loss": 0.6851,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8004381960927515,
+      "grad_norm": 0.6627798676490784,
+      "learning_rate": 3.413198841846809e-05,
+      "loss": 0.688,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8008033595033778,
+      "grad_norm": 0.8205757141113281,
+      "learning_rate": 3.412183394286427e-05,
+      "loss": 0.6604,
+      "step": 2193
+    },
+    {
+      "epoch": 0.801168522914004,
+      "grad_norm": 0.9085656404495239,
+      "learning_rate": 3.411167220214356e-05,
+      "loss": 0.7146,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8015336863246303,
+      "grad_norm": 0.7780555486679077,
+      "learning_rate": 3.410150320153377e-05,
+      "loss": 0.6484,
+      "step": 2195
+    },
+    {
+      "epoch": 0.8018988497352565,
+      "grad_norm": 0.7425792813301086,
+      "learning_rate": 3.409132694626643e-05,
+      "loss": 0.6873,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8022640131458828,
+      "grad_norm": 0.7671288251876831,
+      "learning_rate": 3.408114344157684e-05,
+      "loss": 0.6995,
+      "step": 2197
+    },
+    {
+      "epoch": 0.802629176556509,
+      "grad_norm": 1.4739527702331543,
+      "learning_rate": 3.407095269270398e-05,
+      "loss": 0.6724,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8029943399671353,
+      "grad_norm": 0.9535475373268127,
+      "learning_rate": 3.40607547048906e-05,
+      "loss": 0.6924,
+      "step": 2199
+    },
+    {
+      "epoch": 0.8033595033777615,
+      "grad_norm": 0.8332328796386719,
+      "learning_rate": 3.405054948338314e-05,
+      "loss": 0.6776,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8037246667883878,
+      "grad_norm": 1.0553945302963257,
+      "learning_rate": 3.404033703343179e-05,
+      "loss": 0.6918,
+      "step": 2201
+    },
+    {
+      "epoch": 0.804089830199014,
+      "grad_norm": 0.998306930065155,
+      "learning_rate": 3.4030117360290436e-05,
+      "loss": 0.6989,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8044549936096403,
+      "grad_norm": 0.9074344038963318,
+      "learning_rate": 3.40198904692167e-05,
+      "loss": 0.688,
+      "step": 2203
+    },
+    {
+      "epoch": 0.8048201570202665,
+      "grad_norm": 1.1983717679977417,
+      "learning_rate": 3.4009656365471895e-05,
+      "loss": 0.7019,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8051853204308929,
+      "grad_norm": 1.010562777519226,
+      "learning_rate": 3.399941505432106e-05,
+      "loss": 0.7024,
+      "step": 2205
+    },
+    {
+      "epoch": 0.8055504838415191,
+      "grad_norm": 1.1205374002456665,
+      "learning_rate": 3.398916654103294e-05,
+      "loss": 0.6799,
+      "step": 2206
+    },
+    {
+      "epoch": 0.8059156472521454,
+      "grad_norm": 0.6950740218162537,
+      "learning_rate": 3.397891083088e-05,
+      "loss": 0.6813,
+      "step": 2207
+    },
+    {
+      "epoch": 0.8062808106627716,
+      "grad_norm": 0.9909610152244568,
+      "learning_rate": 3.396864792913836e-05,
+      "loss": 0.665,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8066459740733979,
+      "grad_norm": 1.141450047492981,
+      "learning_rate": 3.3958377841087894e-05,
+      "loss": 0.6785,
+      "step": 2209
+    },
+    {
+      "epoch": 0.8070111374840241,
+      "grad_norm": 0.7786381840705872,
+      "learning_rate": 3.3948100572012145e-05,
+      "loss": 0.668,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8073763008946504,
+      "grad_norm": 2.428544521331787,
+      "learning_rate": 3.393781612719835e-05,
+      "loss": 0.6535,
+      "step": 2211
+    },
+    {
+      "epoch": 0.8077414643052766,
+      "grad_norm": 1.2587729692459106,
+      "learning_rate": 3.3927524511937446e-05,
+      "loss": 0.6436,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8081066277159029,
+      "grad_norm": 0.8259543776512146,
+      "learning_rate": 3.391722573152406e-05,
+      "loss": 0.7026,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8084717911265291,
+      "grad_norm": 0.7508890628814697,
+      "learning_rate": 3.39069197912565e-05,
+      "loss": 0.6814,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8088369545371554,
+      "grad_norm": 1.0003023147583008,
+      "learning_rate": 3.389660669643676e-05,
+      "loss": 0.7075,
+      "step": 2215
+    },
+    {
+      "epoch": 0.8092021179477816,
+      "grad_norm": 1.233620047569275,
+      "learning_rate": 3.3886286452370505e-05,
+      "loss": 0.7073,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8095672813584079,
+      "grad_norm": 0.9432274699211121,
+      "learning_rate": 3.387595906436709e-05,
+      "loss": 0.6842,
+      "step": 2217
+    },
+    {
+      "epoch": 0.8099324447690341,
+      "grad_norm": 0.9820045232772827,
+      "learning_rate": 3.386562453773955e-05,
+      "loss": 0.6941,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8102976081796603,
+      "grad_norm": 0.9450540542602539,
+      "learning_rate": 3.3855282877804575e-05,
+      "loss": 0.707,
+      "step": 2219
+    },
+    {
+      "epoch": 0.8106627715902867,
+      "grad_norm": 0.8697330951690674,
+      "learning_rate": 3.384493408988254e-05,
+      "loss": 0.7067,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8110279350009129,
+      "grad_norm": 1.2985196113586426,
+      "learning_rate": 3.3834578179297484e-05,
+      "loss": 0.705,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8113930984115392,
+      "grad_norm": 0.7868382334709167,
+      "learning_rate": 3.3824215151377095e-05,
+      "loss": 0.6523,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8117582618221654,
+      "grad_norm": 0.9879230260848999,
+      "learning_rate": 3.381384501145274e-05,
+      "loss": 0.6889,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8121234252327917,
+      "grad_norm": 0.7710115909576416,
+      "learning_rate": 3.380346776485944e-05,
+      "loss": 0.6675,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8124885886434179,
+      "grad_norm": 1.0650300979614258,
+      "learning_rate": 3.379308341693588e-05,
+      "loss": 0.6707,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8128537520540442,
+      "grad_norm": 1.146630048751831,
+      "learning_rate": 3.378269197302438e-05,
+      "loss": 0.6971,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8132189154646704,
+      "grad_norm": 0.9396244287490845,
+      "learning_rate": 3.3772293438470924e-05,
+      "loss": 0.6747,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8135840788752967,
+      "grad_norm": 0.7806180715560913,
+      "learning_rate": 3.376188781862515e-05,
+      "loss": 0.6843,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8139492422859229,
+      "grad_norm": 1.0326999425888062,
+      "learning_rate": 3.375147511884032e-05,
+      "loss": 0.7123,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8143144056965492,
+      "grad_norm": 0.7732445597648621,
+      "learning_rate": 3.374105534447334e-05,
+      "loss": 0.6622,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8146795691071754,
+      "grad_norm": 1.0355019569396973,
+      "learning_rate": 3.37306285008848e-05,
+      "loss": 0.6843,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8150447325178017,
+      "grad_norm": 0.9251493215560913,
+      "learning_rate": 3.372019459343886e-05,
+      "loss": 0.6624,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8154098959284279,
+      "grad_norm": 1.0084545612335205,
+      "learning_rate": 3.370975362750335e-05,
+      "loss": 0.6965,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8157750593390543,
+      "grad_norm": 0.7215114235877991,
+      "learning_rate": 3.369930560844975e-05,
+      "loss": 0.6849,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8161402227496805,
+      "grad_norm": 0.6900755167007446,
+      "learning_rate": 3.368885054165314e-05,
+      "loss": 0.6823,
+      "step": 2235
+    },
+    {
+      "epoch": 0.8165053861603068,
+      "grad_norm": 1.0220706462860107,
+      "learning_rate": 3.367838843249222e-05,
+      "loss": 0.675,
+      "step": 2236
+    },
+    {
+      "epoch": 0.816870549570933,
+      "grad_norm": 0.7549731135368347,
+      "learning_rate": 3.366791928634932e-05,
+      "loss": 0.6699,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8172357129815593,
+      "grad_norm": 1.0896284580230713,
+      "learning_rate": 3.365744310861041e-05,
+      "loss": 0.6907,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8176008763921855,
+      "grad_norm": 1.1018774509429932,
+      "learning_rate": 3.364695990466507e-05,
+      "loss": 0.6816,
+      "step": 2239
+    },
+    {
+      "epoch": 0.8179660398028118,
+      "grad_norm": 1.1370127201080322,
+      "learning_rate": 3.363646967990647e-05,
+      "loss": 0.6642,
+      "step": 2240
+    },
+    {
+      "epoch": 0.818331203213438,
+      "grad_norm": 1.2615329027175903,
+      "learning_rate": 3.3625972439731425e-05,
+      "loss": 0.7133,
+      "step": 2241
+    },
+    {
+      "epoch": 0.8186963666240643,
+      "grad_norm": 0.8279294967651367,
+      "learning_rate": 3.361546818954033e-05,
+      "loss": 0.6337,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8190615300346905,
+      "grad_norm": 0.9214347004890442,
+      "learning_rate": 3.3604956934737206e-05,
+      "loss": 0.6723,
+      "step": 2243
+    },
+    {
+      "epoch": 0.8194266934453168,
+      "grad_norm": 0.9047579169273376,
+      "learning_rate": 3.359443868072967e-05,
+      "loss": 0.6589,
+      "step": 2244
+    },
+    {
+      "epoch": 0.819791856855943,
+      "grad_norm": 1.0666414499282837,
+      "learning_rate": 3.3583913432928945e-05,
+      "loss": 0.6718,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8201570202665693,
+      "grad_norm": 0.9169327020645142,
+      "learning_rate": 3.357338119674985e-05,
+      "loss": 0.675,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8205221836771955,
+      "grad_norm": 0.9106442332267761,
+      "learning_rate": 3.3562841977610796e-05,
+      "loss": 0.6606,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8208873470878219,
+      "grad_norm": 0.6304537057876587,
+      "learning_rate": 3.355229578093378e-05,
+      "loss": 0.6827,
+      "step": 2248
+    },
+    {
+      "epoch": 0.821252510498448,
+      "grad_norm": 0.9223216772079468,
+      "learning_rate": 3.354174261214441e-05,
+      "loss": 0.6802,
+      "step": 2249
+    },
+    {
+      "epoch": 0.8216176739090744,
+      "grad_norm": 1.1750116348266602,
+      "learning_rate": 3.353118247667186e-05,
+      "loss": 0.6851,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8219828373197006,
+      "grad_norm": 1.092160940170288,
+      "learning_rate": 3.35206153799489e-05,
+      "loss": 0.6767,
+      "step": 2251
+    },
+    {
+      "epoch": 0.8223480007303268,
+      "grad_norm": 0.8775092363357544,
+      "learning_rate": 3.351004132741188e-05,
+      "loss": 0.6913,
+      "step": 2252
+    },
+    {
+      "epoch": 0.8227131641409531,
+      "grad_norm": 1.434619665145874,
+      "learning_rate": 3.349946032450071e-05,
+      "loss": 0.7225,
+      "step": 2253
+    },
+    {
+      "epoch": 0.8230783275515793,
+      "grad_norm": 0.6954415440559387,
+      "learning_rate": 3.348887237665891e-05,
+      "loss": 0.6674,
+      "step": 2254
+    },
+    {
+      "epoch": 0.8234434909622056,
+      "grad_norm": 0.7986772656440735,
+      "learning_rate": 3.3478277489333554e-05,
+      "loss": 0.681,
+      "step": 2255
+    },
+    {
+      "epoch": 0.8238086543728318,
+      "grad_norm": 1.0696231126785278,
+      "learning_rate": 3.346767566797527e-05,
+      "loss": 0.6731,
+      "step": 2256
+    },
+    {
+      "epoch": 0.8241738177834581,
+      "grad_norm": 1.0807222127914429,
+      "learning_rate": 3.345706691803828e-05,
+      "loss": 0.7114,
+      "step": 2257
+    },
+    {
+      "epoch": 0.8245389811940843,
+      "grad_norm": 0.8983349800109863,
+      "learning_rate": 3.344645124498036e-05,
+      "loss": 0.7061,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8249041446047106,
+      "grad_norm": 1.091439127922058,
+      "learning_rate": 3.3435828654262844e-05,
+      "loss": 0.712,
+      "step": 2259
+    },
+    {
+      "epoch": 0.8252693080153368,
+      "grad_norm": 0.816214382648468,
+      "learning_rate": 3.3425199151350636e-05,
+      "loss": 0.6725,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8256344714259631,
+      "grad_norm": 0.6547426581382751,
+      "learning_rate": 3.341456274171218e-05,
+      "loss": 0.6605,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8259996348365893,
+      "grad_norm": 0.8819223046302795,
+      "learning_rate": 3.340391943081949e-05,
+      "loss": 0.6647,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8263647982472156,
+      "grad_norm": 0.8542426228523254,
+      "learning_rate": 3.339326922414812e-05,
+      "loss": 0.6961,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8267299616578418,
+      "grad_norm": 1.0729966163635254,
+      "learning_rate": 3.3382612127177166e-05,
+      "loss": 0.6595,
+      "step": 2264
+    },
+    {
+      "epoch": 0.8270951250684682,
+      "grad_norm": 1.184866189956665,
+      "learning_rate": 3.337194814538929e-05,
+      "loss": 0.6446,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8274602884790944,
+      "grad_norm": 0.8264432549476624,
+      "learning_rate": 3.336127728427067e-05,
+      "loss": 0.6735,
+      "step": 2266
+    },
+    {
+      "epoch": 0.8278254518897207,
+      "grad_norm": 0.7857865691184998,
+      "learning_rate": 3.335059954931105e-05,
+      "loss": 0.6556,
+      "step": 2267
+    },
+    {
+      "epoch": 0.8281906153003469,
+      "grad_norm": 0.828025221824646,
+      "learning_rate": 3.333991494600368e-05,
+      "loss": 0.6747,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8285557787109732,
+      "grad_norm": 1.0558624267578125,
+      "learning_rate": 3.332922347984537e-05,
+      "loss": 0.7017,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8289209421215994,
+      "grad_norm": 0.8370004296302795,
+      "learning_rate": 3.331852515633645e-05,
+      "loss": 0.6711,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8292861055322257,
+      "grad_norm": 0.8168753981590271,
+      "learning_rate": 3.330781998098078e-05,
+      "loss": 0.6727,
+      "step": 2271
+    },
+    {
+      "epoch": 0.8296512689428519,
+      "grad_norm": 0.7985580563545227,
+      "learning_rate": 3.3297107959285734e-05,
+      "loss": 0.6577,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8300164323534782,
+      "grad_norm": 1.1410428285598755,
+      "learning_rate": 3.328638909676222e-05,
+      "loss": 0.6677,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8303815957641044,
+      "grad_norm": 1.0028425455093384,
+      "learning_rate": 3.327566339892467e-05,
+      "loss": 0.6589,
+      "step": 2274
+    },
+    {
+      "epoch": 0.8307467591747307,
+      "grad_norm": 0.9950466156005859,
+      "learning_rate": 3.326493087129102e-05,
+      "loss": 0.6871,
+      "step": 2275
+    },
+    {
+      "epoch": 0.8311119225853569,
+      "grad_norm": 0.6724287271499634,
+      "learning_rate": 3.325419151938273e-05,
+      "loss": 0.6873,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8314770859959832,
+      "grad_norm": 0.7970866560935974,
+      "learning_rate": 3.3243445348724756e-05,
+      "loss": 0.6788,
+      "step": 2277
+    },
+    {
+      "epoch": 0.8318422494066094,
+      "grad_norm": 1.309187412261963,
+      "learning_rate": 3.323269236484557e-05,
+      "loss": 0.6971,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8322074128172358,
+      "grad_norm": 0.9678725004196167,
+      "learning_rate": 3.322193257327716e-05,
+      "loss": 0.6932,
+      "step": 2279
+    },
+    {
+      "epoch": 0.832572576227862,
+      "grad_norm": 0.7458254098892212,
+      "learning_rate": 3.321116597955501e-05,
+      "loss": 0.668,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8329377396384883,
+      "grad_norm": 2.006092071533203,
+      "learning_rate": 3.320039258921809e-05,
+      "loss": 0.652,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8333029030491145,
+      "grad_norm": 0.9141222238540649,
+      "learning_rate": 3.318961240780889e-05,
+      "loss": 0.6888,
+      "step": 2282
+    },
+    {
+      "epoch": 0.8336680664597408,
+      "grad_norm": 1.0954430103302002,
+      "learning_rate": 3.317882544087336e-05,
+      "loss": 0.6824,
+      "step": 2283
+    },
+    {
+      "epoch": 0.834033229870367,
+      "grad_norm": 0.8255243301391602,
+      "learning_rate": 3.316803169396098e-05,
+      "loss": 0.6443,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8343983932809932,
+      "grad_norm": 1.5680209398269653,
+      "learning_rate": 3.31572311726247e-05,
+      "loss": 0.674,
+      "step": 2285
+    },
+    {
+      "epoch": 0.8347635566916195,
+      "grad_norm": 1.061949610710144,
+      "learning_rate": 3.3146423882420935e-05,
+      "loss": 0.717,
+      "step": 2286
+    },
+    {
+      "epoch": 0.8351287201022457,
+      "grad_norm": 0.993720531463623,
+      "learning_rate": 3.313560982890963e-05,
+      "loss": 0.6495,
+      "step": 2287
+    },
+    {
+      "epoch": 0.835493883512872,
+      "grad_norm": 0.9374502301216125,
+      "learning_rate": 3.3124789017654154e-05,
+      "loss": 0.6553,
+      "step": 2288
+    },
+    {
+      "epoch": 0.8358590469234982,
+      "grad_norm": 0.6694275140762329,
+      "learning_rate": 3.31139614542214e-05,
+      "loss": 0.6793,
+      "step": 2289
+    },
+    {
+      "epoch": 0.8362242103341245,
+      "grad_norm": 1.1059433221817017,
+      "learning_rate": 3.310312714418171e-05,
+      "loss": 0.705,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8365893737447507,
+      "grad_norm": 0.9983911514282227,
+      "learning_rate": 3.3092286093108894e-05,
+      "loss": 0.6574,
+      "step": 2291
+    },
+    {
+      "epoch": 0.836954537155377,
+      "grad_norm": 0.8774459362030029,
+      "learning_rate": 3.308143830658025e-05,
+      "loss": 0.673,
+      "step": 2292
+    },
+    {
+      "epoch": 0.8373197005660032,
+      "grad_norm": 0.8330726027488708,
+      "learning_rate": 3.307058379017652e-05,
+      "loss": 0.6395,
+      "step": 2293
+    },
+    {
+      "epoch": 0.8376848639766296,
+      "grad_norm": 0.9212683439254761,
+      "learning_rate": 3.305972254948191e-05,
+      "loss": 0.6776,
+      "step": 2294
+    },
+    {
+      "epoch": 0.8380500273872558,
+      "grad_norm": 1.1847418546676636,
+      "learning_rate": 3.304885459008412e-05,
+      "loss": 0.6734,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8384151907978821,
+      "grad_norm": 1.0819242000579834,
+      "learning_rate": 3.303797991757425e-05,
+      "loss": 0.7296,
+      "step": 2296
+    },
+    {
+      "epoch": 0.8387803542085083,
+      "grad_norm": 0.9311473369598389,
+      "learning_rate": 3.3027098537546904e-05,
+      "loss": 0.6743,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8391455176191346,
+      "grad_norm": 1.4701728820800781,
+      "learning_rate": 3.3016210455600094e-05,
+      "loss": 0.6945,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8395106810297608,
+      "grad_norm": 0.6400688886642456,
+      "learning_rate": 3.300531567733532e-05,
+      "loss": 0.6655,
+      "step": 2299
+    },
+    {
+      "epoch": 0.8398758444403871,
+      "grad_norm": 1.135380744934082,
+      "learning_rate": 3.2994414208357496e-05,
+      "loss": 0.6844,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8402410078510133,
+      "grad_norm": 1.25801420211792,
+      "learning_rate": 3.2983506054274995e-05,
+      "loss": 0.6521,
+      "step": 2301
+    },
+    {
+      "epoch": 0.8406061712616396,
+      "grad_norm": 1.4099013805389404,
+      "learning_rate": 3.297259122069963e-05,
+      "loss": 0.7244,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8409713346722658,
+      "grad_norm": 0.9134430885314941,
+      "learning_rate": 3.296166971324664e-05,
+      "loss": 0.7045,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8413364980828921,
+      "grad_norm": 0.8753640651702881,
+      "learning_rate": 3.29507415375347e-05,
+      "loss": 0.6834,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8417016614935183,
+      "grad_norm": 1.0376008749008179,
+      "learning_rate": 3.293980669918592e-05,
+      "loss": 0.6962,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8420668249041446,
+      "grad_norm": 0.8525744080543518,
+      "learning_rate": 3.292886520382583e-05,
+      "loss": 0.6828,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8424319883147708,
+      "grad_norm": 1.1410913467407227,
+      "learning_rate": 3.29179170570834e-05,
+      "loss": 0.6629,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8427971517253972,
+      "grad_norm": 0.9425336718559265,
+      "learning_rate": 3.2906962264591014e-05,
+      "loss": 0.6633,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8431623151360234,
+      "grad_norm": 1.1711310148239136,
+      "learning_rate": 3.2896000831984456e-05,
+      "loss": 0.6816,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8435274785466497,
+      "grad_norm": 0.7658752799034119,
+      "learning_rate": 3.288503276490296e-05,
+      "loss": 0.7057,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8438926419572759,
+      "grad_norm": 0.8208125829696655,
+      "learning_rate": 3.287405806898915e-05,
+      "loss": 0.6657,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8442578053679022,
+      "grad_norm": 0.9931808710098267,
+      "learning_rate": 3.2863076749889064e-05,
+      "loss": 0.6545,
+      "step": 2312
+    },
+    {
+      "epoch": 0.8446229687785284,
+      "grad_norm": 1.2441120147705078,
+      "learning_rate": 3.285208881325216e-05,
+      "loss": 0.6757,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8449881321891547,
+      "grad_norm": 1.0434895753860474,
+      "learning_rate": 3.2841094264731274e-05,
+      "loss": 0.6835,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8453532955997809,
+      "grad_norm": 0.935052752494812,
+      "learning_rate": 3.283009310998268e-05,
+      "loss": 0.6379,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8457184590104072,
+      "grad_norm": 1.2280255556106567,
+      "learning_rate": 3.2819085354666015e-05,
+      "loss": 0.6865,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8460836224210334,
+      "grad_norm": 0.9368286728858948,
+      "learning_rate": 3.280807100444433e-05,
+      "loss": 0.6586,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8464487858316597,
+      "grad_norm": 0.6505674719810486,
+      "learning_rate": 3.279705006498408e-05,
+      "loss": 0.6982,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8468139492422859,
+      "grad_norm": 1.2371175289154053,
+      "learning_rate": 3.278602254195507e-05,
+      "loss": 0.6849,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8471791126529121,
+      "grad_norm": 0.9046905040740967,
+      "learning_rate": 3.277498844103055e-05,
+      "loss": 0.6556,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8475442760635384,
+      "grad_norm": 1.8878294229507446,
+      "learning_rate": 3.276394776788709e-05,
+      "loss": 0.6597,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8479094394741646,
+      "grad_norm": 0.9316428899765015,
+      "learning_rate": 3.27529005282047e-05,
+      "loss": 0.6913,
+      "step": 2322
+    },
+    {
+      "epoch": 0.848274602884791,
+      "grad_norm": 1.0353072881698608,
+      "learning_rate": 3.274184672766673e-05,
+      "loss": 0.6635,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8486397662954172,
+      "grad_norm": 0.8999189138412476,
+      "learning_rate": 3.2730786371959906e-05,
+      "loss": 0.6754,
+      "step": 2324
+    },
+    {
+      "epoch": 0.8490049297060435,
+      "grad_norm": 1.006822943687439,
+      "learning_rate": 3.271971946677436e-05,
+      "loss": 0.6785,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8493700931166697,
+      "grad_norm": 0.8299421072006226,
+      "learning_rate": 3.270864601780355e-05,
+      "loss": 0.6778,
+      "step": 2326
+    },
+    {
+      "epoch": 0.849735256527296,
+      "grad_norm": 1.0007952451705933,
+      "learning_rate": 3.269756603074433e-05,
+      "loss": 0.6829,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8501004199379222,
+      "grad_norm": 0.8802875876426697,
+      "learning_rate": 3.268647951129692e-05,
+      "loss": 0.6588,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8504655833485485,
+      "grad_norm": 1.0025578737258911,
+      "learning_rate": 3.267538646516487e-05,
+      "loss": 0.65,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8508307467591747,
+      "grad_norm": 0.8563314080238342,
+      "learning_rate": 3.266428689805512e-05,
+      "loss": 0.681,
+      "step": 2330
+    },
+    {
+      "epoch": 0.851195910169801,
+      "grad_norm": 1.299754023551941,
+      "learning_rate": 3.265318081567794e-05,
+      "loss": 0.678,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8515610735804272,
+      "grad_norm": 0.7509905099868774,
+      "learning_rate": 3.2642068223746975e-05,
+      "loss": 0.6325,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8519262369910535,
+      "grad_norm": 0.9449114203453064,
+      "learning_rate": 3.2630949127979204e-05,
+      "loss": 0.6769,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8522914004016797,
+      "grad_norm": 0.8018974661827087,
+      "learning_rate": 3.2619823534094956e-05,
+      "loss": 0.6597,
+      "step": 2334
+    },
+    {
+      "epoch": 0.852656563812306,
+      "grad_norm": 1.187442421913147,
+      "learning_rate": 3.26086914478179e-05,
+      "loss": 0.6749,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8530217272229322,
+      "grad_norm": 1.0720336437225342,
+      "learning_rate": 3.259755287487505e-05,
+      "loss": 0.6584,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8533868906335585,
+      "grad_norm": 0.6564728617668152,
+      "learning_rate": 3.258640782099675e-05,
+      "loss": 0.6418,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8537520540441847,
+      "grad_norm": 0.9528270363807678,
+      "learning_rate": 3.257525629191669e-05,
+      "loss": 0.6414,
+      "step": 2338
+    },
+    {
+      "epoch": 0.8541172174548111,
+      "grad_norm": 1.947891116142273,
+      "learning_rate": 3.2564098293371884e-05,
+      "loss": 0.6577,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8544823808654373,
+      "grad_norm": 1.029895305633545,
+      "learning_rate": 3.255293383110267e-05,
+      "loss": 0.6412,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8548475442760636,
+      "grad_norm": 0.9232720136642456,
+      "learning_rate": 3.2541762910852716e-05,
+      "loss": 0.6858,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8552127076866898,
+      "grad_norm": 1.1750651597976685,
+      "learning_rate": 3.253058553836902e-05,
+      "loss": 0.6769,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8555778710973161,
+      "grad_norm": 1.049420952796936,
+      "learning_rate": 3.251940171940188e-05,
+      "loss": 0.6663,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8559430345079423,
+      "grad_norm": 0.8713171482086182,
+      "learning_rate": 3.250821145970493e-05,
+      "loss": 0.6758,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8563081979185686,
+      "grad_norm": 1.214188814163208,
+      "learning_rate": 3.2497014765035105e-05,
+      "loss": 0.6747,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8566733613291948,
+      "grad_norm": 1.0018802881240845,
+      "learning_rate": 3.2485811641152655e-05,
+      "loss": 0.6509,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8570385247398211,
+      "grad_norm": 0.8035647869110107,
+      "learning_rate": 3.2474602093821145e-05,
+      "loss": 0.6946,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8574036881504473,
+      "grad_norm": 1.55748450756073,
+      "learning_rate": 3.246338612880743e-05,
+      "loss": 0.6693,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8577688515610736,
+      "grad_norm": 1.0306609869003296,
+      "learning_rate": 3.245216375188168e-05,
+      "loss": 0.6867,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8581340149716998,
+      "grad_norm": 0.909618079662323,
+      "learning_rate": 3.2440934968817355e-05,
+      "loss": 0.6755,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8584991783823261,
+      "grad_norm": 1.1233172416687012,
+      "learning_rate": 3.2429699785391205e-05,
+      "loss": 0.6541,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8588643417929523,
+      "grad_norm": 0.6875085234642029,
+      "learning_rate": 3.241845820738329e-05,
+      "loss": 0.6486,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8592295052035785,
+      "grad_norm": 1.0278838872909546,
+      "learning_rate": 3.240721024057695e-05,
+      "loss": 0.6849,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8595946686142049,
+      "grad_norm": 0.888755202293396,
+      "learning_rate": 3.239595589075881e-05,
+      "loss": 0.6599,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8599598320248311,
+      "grad_norm": 1.1310205459594727,
+      "learning_rate": 3.238469516371879e-05,
+      "loss": 0.686,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8603249954354574,
+      "grad_norm": 1.0039758682250977,
+      "learning_rate": 3.237342806525007e-05,
+      "loss": 0.6724,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8606901588460836,
+      "grad_norm": 0.7611724138259888,
+      "learning_rate": 3.236215460114913e-05,
+      "loss": 0.6803,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8610553222567099,
+      "grad_norm": 1.0401297807693481,
+      "learning_rate": 3.23508747772157e-05,
+      "loss": 0.7059,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8614204856673361,
+      "grad_norm": 0.7942869663238525,
+      "learning_rate": 3.233958859925282e-05,
+      "loss": 0.675,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8617856490779624,
+      "grad_norm": 0.9569483399391174,
+      "learning_rate": 3.232829607306675e-05,
+      "loss": 0.6613,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8621508124885886,
+      "grad_norm": 1.113763689994812,
+      "learning_rate": 3.231699720446706e-05,
+      "loss": 0.6732,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8625159758992149,
+      "grad_norm": 0.8510305285453796,
+      "learning_rate": 3.230569199926656e-05,
+      "loss": 0.6957,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8628811393098411,
+      "grad_norm": 0.9915270805358887,
+      "learning_rate": 3.2294380463281315e-05,
+      "loss": 0.6472,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8632463027204674,
+      "grad_norm": 0.9330098032951355,
+      "learning_rate": 3.228306260233067e-05,
+      "loss": 0.6506,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8636114661310936,
+      "grad_norm": 0.9357939958572388,
+      "learning_rate": 3.227173842223721e-05,
+      "loss": 0.7097,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8639766295417199,
+      "grad_norm": 1.0856742858886719,
+      "learning_rate": 3.226040792882676e-05,
+      "loss": 0.6552,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8643417929523461,
+      "grad_norm": 1.2186148166656494,
+      "learning_rate": 3.224907112792841e-05,
+      "loss": 0.6762,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8647069563629725,
+      "grad_norm": 0.9334109425544739,
+      "learning_rate": 3.223772802537449e-05,
+      "loss": 0.7025,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8650721197735987,
+      "grad_norm": 0.8101790547370911,
+      "learning_rate": 3.2226378627000574e-05,
+      "loss": 0.6598,
+      "step": 2369
+    },
+    {
+      "epoch": 0.865437283184225,
+      "grad_norm": 1.0096237659454346,
+      "learning_rate": 3.2215022938645465e-05,
+      "loss": 0.6761,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8658024465948512,
+      "grad_norm": 0.7132946252822876,
+      "learning_rate": 3.2203660966151206e-05,
+      "loss": 0.6661,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8661676100054775,
+      "grad_norm": 0.8075200319290161,
+      "learning_rate": 3.219229271536309e-05,
+      "loss": 0.6654,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8665327734161037,
+      "grad_norm": 3.128431797027588,
+      "learning_rate": 3.218091819212962e-05,
+      "loss": 0.697,
+      "step": 2373
+    },
+    {
+      "epoch": 0.86689793682673,
+      "grad_norm": 1.0566328763961792,
+      "learning_rate": 3.2169537402302525e-05,
+      "loss": 0.6853,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8672631002373562,
+      "grad_norm": 1.0124942064285278,
+      "learning_rate": 3.215815035173678e-05,
+      "loss": 0.6753,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8676282636479825,
+      "grad_norm": 0.8646689057350159,
+      "learning_rate": 3.214675704629054e-05,
+      "loss": 0.6724,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8679934270586087,
+      "grad_norm": 1.017532467842102,
+      "learning_rate": 3.213535749182523e-05,
+      "loss": 0.6837,
+      "step": 2377
+    },
+    {
+      "epoch": 0.868358590469235,
+      "grad_norm": 0.8565889596939087,
+      "learning_rate": 3.212395169420544e-05,
+      "loss": 0.6915,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8687237538798612,
+      "grad_norm": 0.7826008200645447,
+      "learning_rate": 3.211253965929902e-05,
+      "loss": 0.6804,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8690889172904875,
+      "grad_norm": 1.0141047239303589,
+      "learning_rate": 3.2101121392976986e-05,
+      "loss": 0.6706,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8694540807011137,
+      "grad_norm": 0.8811666965484619,
+      "learning_rate": 3.2089696901113576e-05,
+      "loss": 0.6869,
+      "step": 2381
+    },
+    {
+      "epoch": 0.86981924411174,
+      "grad_norm": 1.180303931236267,
+      "learning_rate": 3.2078266189586256e-05,
+      "loss": 0.7188,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8701844075223663,
+      "grad_norm": 1.0213888883590698,
+      "learning_rate": 3.2066829264275644e-05,
+      "loss": 0.6647,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8705495709329926,
+      "grad_norm": 0.7763113975524902,
+      "learning_rate": 3.205538613106558e-05,
+      "loss": 0.6511,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8709147343436188,
+      "grad_norm": 1.2403777837753296,
+      "learning_rate": 3.204393679584311e-05,
+      "loss": 0.6908,
+      "step": 2385
+    },
+    {
+      "epoch": 0.871279897754245,
+      "grad_norm": 0.8211830854415894,
+      "learning_rate": 3.203248126449845e-05,
+      "loss": 0.6514,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8716450611648713,
+      "grad_norm": 0.8894174098968506,
+      "learning_rate": 3.2021019542925015e-05,
+      "loss": 0.6599,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8720102245754975,
+      "grad_norm": 0.8494507670402527,
+      "learning_rate": 3.20095516370194e-05,
+      "loss": 0.6791,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8723753879861238,
+      "grad_norm": 1.1291923522949219,
+      "learning_rate": 3.1998077552681387e-05,
+      "loss": 0.7286,
+      "step": 2389
+    },
+    {
+      "epoch": 0.87274055139675,
+      "grad_norm": 0.8444809913635254,
+      "learning_rate": 3.198659729581391e-05,
+      "loss": 0.6595,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8731057148073763,
+      "grad_norm": 0.9420198202133179,
+      "learning_rate": 3.197511087232313e-05,
+      "loss": 0.6908,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8734708782180025,
+      "grad_norm": 1.2120648622512817,
+      "learning_rate": 3.1963618288118334e-05,
+      "loss": 0.6942,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8738360416286288,
+      "grad_norm": 0.9157645106315613,
+      "learning_rate": 3.195211954911199e-05,
+      "loss": 0.7096,
+      "step": 2393
+    },
+    {
+      "epoch": 0.874201205039255,
+      "grad_norm": 0.8355812430381775,
+      "learning_rate": 3.194061466121976e-05,
+      "loss": 0.6897,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8745663684498813,
+      "grad_norm": 0.9789050221443176,
+      "learning_rate": 3.192910363036043e-05,
+      "loss": 0.6702,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8749315318605075,
+      "grad_norm": 1.2427998781204224,
+      "learning_rate": 3.191758646245596e-05,
+      "loss": 0.6486,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8752966952711339,
+      "grad_norm": 1.148383378982544,
+      "learning_rate": 3.1906063163431485e-05,
+      "loss": 0.6512,
+      "step": 2397
+    },
+    {
+      "epoch": 0.87566185868176,
+      "grad_norm": 0.9681849479675293,
+      "learning_rate": 3.189453373921527e-05,
+      "loss": 0.6417,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8760270220923864,
+      "grad_norm": 0.7402408123016357,
+      "learning_rate": 3.1882998195738744e-05,
+      "loss": 0.6536,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8763921855030126,
+      "grad_norm": 0.9436914920806885,
+      "learning_rate": 3.187145653893648e-05,
+      "loss": 0.6771,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8767573489136389,
+      "grad_norm": 1.0563963651657104,
+      "learning_rate": 3.1859908774746205e-05,
+      "loss": 0.6899,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8771225123242651,
+      "grad_norm": 0.9120274186134338,
+      "learning_rate": 3.184835490910877e-05,
+      "loss": 0.7038,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8774876757348914,
+      "grad_norm": 0.7715966701507568,
+      "learning_rate": 3.1836794947968175e-05,
+      "loss": 0.6648,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8778528391455176,
+      "grad_norm": 1.047858715057373,
+      "learning_rate": 3.182522889727157e-05,
+      "loss": 0.6703,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8782180025561439,
+      "grad_norm": 0.652629017829895,
+      "learning_rate": 3.1813656762969206e-05,
+      "loss": 0.666,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8785831659667701,
+      "grad_norm": 0.8451604843139648,
+      "learning_rate": 3.180207855101449e-05,
+      "loss": 0.7081,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8789483293773964,
+      "grad_norm": 0.7865225076675415,
+      "learning_rate": 3.1790494267363954e-05,
+      "loss": 0.6756,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8793134927880226,
+      "grad_norm": 1.1484609842300415,
+      "learning_rate": 3.177890391797724e-05,
+      "loss": 0.6816,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8796786561986489,
+      "grad_norm": 0.7304513454437256,
+      "learning_rate": 3.176730750881711e-05,
+      "loss": 0.69,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8800438196092751,
+      "grad_norm": 1.2418588399887085,
+      "learning_rate": 3.1755705045849465e-05,
+      "loss": 0.6796,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8804089830199014,
+      "grad_norm": 0.636386513710022,
+      "learning_rate": 3.17440965350433e-05,
+      "loss": 0.6747,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8807741464305276,
+      "grad_norm": 0.9517824649810791,
+      "learning_rate": 3.173248198237073e-05,
+      "loss": 0.673,
+      "step": 2412
+    },
+    {
+      "epoch": 0.881139309841154,
+      "grad_norm": 1.097337007522583,
+      "learning_rate": 3.172086139380698e-05,
+      "loss": 0.6783,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8815044732517802,
+      "grad_norm": 0.9089690446853638,
+      "learning_rate": 3.170923477533036e-05,
+      "loss": 0.6724,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8818696366624065,
+      "grad_norm": 1.1122137308120728,
+      "learning_rate": 3.169760213292232e-05,
+      "loss": 0.6299,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8822348000730327,
+      "grad_norm": 0.730311930179596,
+      "learning_rate": 3.168596347256737e-05,
+      "loss": 0.6903,
+      "step": 2416
+    },
+    {
+      "epoch": 0.882599963483659,
+      "grad_norm": 0.7182562351226807,
+      "learning_rate": 3.1674318800253146e-05,
+      "loss": 0.6815,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8829651268942852,
+      "grad_norm": 0.8872962594032288,
+      "learning_rate": 3.166266812197036e-05,
+      "loss": 0.658,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8833302903049115,
+      "grad_norm": 1.317947268486023,
+      "learning_rate": 3.1651011443712825e-05,
+      "loss": 0.6667,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8836954537155377,
+      "grad_norm": 0.9207883477210999,
+      "learning_rate": 3.1639348771477424e-05,
+      "loss": 0.6953,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8840606171261639,
+      "grad_norm": 1.3670681715011597,
+      "learning_rate": 3.1627680111264134e-05,
+      "loss": 0.6684,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8844257805367902,
+      "grad_norm": 1.701436161994934,
+      "learning_rate": 3.161600546907602e-05,
+      "loss": 0.6449,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8847909439474164,
+      "grad_norm": 0.8770363330841064,
+      "learning_rate": 3.160432485091922e-05,
+      "loss": 0.6526,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8851561073580427,
+      "grad_norm": 0.7543882131576538,
+      "learning_rate": 3.1592638262802926e-05,
+      "loss": 0.6595,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8855212707686689,
+      "grad_norm": 1.0555789470672607,
+      "learning_rate": 3.1580945710739435e-05,
+      "loss": 0.6508,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8858864341792952,
+      "grad_norm": 0.9207205772399902,
+      "learning_rate": 3.156924720074408e-05,
+      "loss": 0.649,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8862515975899214,
+      "grad_norm": 0.8387048244476318,
+      "learning_rate": 3.1557542738835295e-05,
+      "loss": 0.6737,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8866167610005478,
+      "grad_norm": 0.9562423825263977,
+      "learning_rate": 3.154583233103455e-05,
+      "loss": 0.6929,
+      "step": 2428
+    },
+    {
+      "epoch": 0.886981924411174,
+      "grad_norm": 1.06221342086792,
+      "learning_rate": 3.153411598336637e-05,
+      "loss": 0.6701,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8873470878218003,
+      "grad_norm": 1.1556217670440674,
+      "learning_rate": 3.1522393701858353e-05,
+      "loss": 0.6714,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8877122512324265,
+      "grad_norm": 1.1550284624099731,
+      "learning_rate": 3.151066549254115e-05,
+      "loss": 0.6792,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8880774146430528,
+      "grad_norm": 1.2821412086486816,
+      "learning_rate": 3.149893136144843e-05,
+      "loss": 0.6675,
+      "step": 2432
+    },
+    {
+      "epoch": 0.888442578053679,
+      "grad_norm": 1.109316110610962,
+      "learning_rate": 3.148719131461695e-05,
+      "loss": 0.6572,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8888077414643053,
+      "grad_norm": 1.0211321115493774,
+      "learning_rate": 3.14754453580865e-05,
+      "loss": 0.7051,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8891729048749315,
+      "grad_norm": 0.8004649877548218,
+      "learning_rate": 3.1463693497899894e-05,
+      "loss": 0.6631,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8895380682855578,
+      "grad_norm": 0.8751883506774902,
+      "learning_rate": 3.145193574010298e-05,
+      "loss": 0.6337,
+      "step": 2436
+    },
+    {
+      "epoch": 0.889903231696184,
+      "grad_norm": 1.1010115146636963,
+      "learning_rate": 3.1440172090744674e-05,
+      "loss": 0.6954,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8902683951068103,
+      "grad_norm": 0.675520122051239,
+      "learning_rate": 3.1428402555876896e-05,
+      "loss": 0.6704,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8906335585174365,
+      "grad_norm": 1.0251437425613403,
+      "learning_rate": 3.1416627141554595e-05,
+      "loss": 0.6634,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8909987219280628,
+      "grad_norm": 0.8706310987472534,
+      "learning_rate": 3.1404845853835744e-05,
+      "loss": 0.709,
+      "step": 2440
+    },
+    {
+      "epoch": 0.891363885338689,
+      "grad_norm": 0.9390824437141418,
+      "learning_rate": 3.139305869878135e-05,
+      "loss": 0.6915,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8917290487493154,
+      "grad_norm": 0.9122499227523804,
+      "learning_rate": 3.1381265682455436e-05,
+      "loss": 0.6764,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8920942121599416,
+      "grad_norm": 1.045343279838562,
+      "learning_rate": 3.136946681092503e-05,
+      "loss": 0.6748,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8924593755705679,
+      "grad_norm": 1.004455804824829,
+      "learning_rate": 3.135766209026017e-05,
+      "loss": 0.6948,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8928245389811941,
+      "grad_norm": 0.9815378785133362,
+      "learning_rate": 3.134585152653393e-05,
+      "loss": 0.6882,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8931897023918204,
+      "grad_norm": 0.958922803401947,
+      "learning_rate": 3.133403512582236e-05,
+      "loss": 0.6949,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8935548658024466,
+      "grad_norm": 1.0713224411010742,
+      "learning_rate": 3.132221289420451e-05,
+      "loss": 0.6835,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8939200292130729,
+      "grad_norm": 1.062995433807373,
+      "learning_rate": 3.131038483776247e-05,
+      "loss": 0.6827,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8942851926236991,
+      "grad_norm": 1.309880018234253,
+      "learning_rate": 3.129855096258129e-05,
+      "loss": 0.6812,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8946503560343254,
+      "grad_norm": 0.7982645034790039,
+      "learning_rate": 3.128671127474902e-05,
+      "loss": 0.7031,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8950155194449516,
+      "grad_norm": 0.8103627562522888,
+      "learning_rate": 3.127486578035671e-05,
+      "loss": 0.6996,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8953806828555779,
+      "grad_norm": 1.1655442714691162,
+      "learning_rate": 3.1263014485498374e-05,
+      "loss": 0.6641,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8957458462662041,
+      "grad_norm": 1.0135167837142944,
+      "learning_rate": 3.1251157396271055e-05,
+      "loss": 0.6292,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8961110096768303,
+      "grad_norm": 1.009577751159668,
+      "learning_rate": 3.123929451877473e-05,
+      "loss": 0.675,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8964761730874566,
+      "grad_norm": 1.0211693048477173,
+      "learning_rate": 3.122742585911238e-05,
+      "loss": 0.6735,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8968413364980828,
+      "grad_norm": 1.1773436069488525,
+      "learning_rate": 3.121555142338996e-05,
+      "loss": 0.6976,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8972064999087092,
+      "grad_norm": 0.7748075127601624,
+      "learning_rate": 3.120367121771638e-05,
+      "loss": 0.6993,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8975716633193354,
+      "grad_norm": 0.9061272144317627,
+      "learning_rate": 3.119178524820354e-05,
+      "loss": 0.6478,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8979368267299617,
+      "grad_norm": 0.8899768590927124,
+      "learning_rate": 3.1179893520966276e-05,
+      "loss": 0.6632,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8983019901405879,
+      "grad_norm": 0.9813979864120483,
+      "learning_rate": 3.1167996042122426e-05,
+      "loss": 0.668,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8986671535512142,
+      "grad_norm": 0.9683778285980225,
+      "learning_rate": 3.1156092817792756e-05,
+      "loss": 0.646,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8990323169618404,
+      "grad_norm": 1.1068257093429565,
+      "learning_rate": 3.1144183854100996e-05,
+      "loss": 0.6536,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8993974803724667,
+      "grad_norm": 0.9029096364974976,
+      "learning_rate": 3.113226915717383e-05,
+      "loss": 0.6418,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8997626437830929,
+      "grad_norm": 0.7347602844238281,
+      "learning_rate": 3.1120348733140896e-05,
+      "loss": 0.6893,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9001278071937192,
+      "grad_norm": 0.8413745164871216,
+      "learning_rate": 3.110842258813477e-05,
+      "loss": 0.7046,
+      "step": 2465
+    },
+    {
+      "epoch": 0.9004929706043454,
+      "grad_norm": 2.552323818206787,
+      "learning_rate": 3.109649072829097e-05,
+      "loss": 0.6794,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9008581340149717,
+      "grad_norm": 1.1075729131698608,
+      "learning_rate": 3.108455315974796e-05,
+      "loss": 0.6911,
+      "step": 2467
+    },
+    {
+      "epoch": 0.9012232974255979,
+      "grad_norm": 0.8394867181777954,
+      "learning_rate": 3.107260988864716e-05,
+      "loss": 0.6769,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9015884608362242,
+      "grad_norm": 1.4787451028823853,
+      "learning_rate": 3.106066092113288e-05,
+      "loss": 0.6637,
+      "step": 2469
+    },
+    {
+      "epoch": 0.9019536242468504,
+      "grad_norm": 0.9691876769065857,
+      "learning_rate": 3.10487062633524e-05,
+      "loss": 0.6755,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9023187876574768,
+      "grad_norm": 1.1069458723068237,
+      "learning_rate": 3.1036745921455895e-05,
+      "loss": 0.7007,
+      "step": 2471
+    },
+    {
+      "epoch": 0.902683951068103,
+      "grad_norm": 1.4953409433364868,
+      "learning_rate": 3.1024779901596496e-05,
+      "loss": 0.6846,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9030491144787293,
+      "grad_norm": 0.9708583354949951,
+      "learning_rate": 3.101280820993023e-05,
+      "loss": 0.6655,
+      "step": 2473
+    },
+    {
+      "epoch": 0.9034142778893555,
+      "grad_norm": 0.8601775765419006,
+      "learning_rate": 3.100083085261606e-05,
+      "loss": 0.673,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9037794412999818,
+      "grad_norm": 0.8028748631477356,
+      "learning_rate": 3.098884783581586e-05,
+      "loss": 0.6709,
+      "step": 2475
+    },
+    {
+      "epoch": 0.904144604710608,
+      "grad_norm": 0.9961649775505066,
+      "learning_rate": 3.097685916569439e-05,
+      "loss": 0.656,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9045097681212343,
+      "grad_norm": 0.8735599517822266,
+      "learning_rate": 3.096486484841935e-05,
+      "loss": 0.673,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9048749315318605,
+      "grad_norm": 2.3302524089813232,
+      "learning_rate": 3.095286489016135e-05,
+      "loss": 0.6274,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9052400949424868,
+      "grad_norm": 1.027161717414856,
+      "learning_rate": 3.0940859297093874e-05,
+      "loss": 0.6816,
+      "step": 2479
+    },
+    {
+      "epoch": 0.905605258353113,
+      "grad_norm": 1.2939852476119995,
+      "learning_rate": 3.092884807539331e-05,
+      "loss": 0.6826,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9059704217637393,
+      "grad_norm": 0.9535837173461914,
+      "learning_rate": 3.091683123123897e-05,
+      "loss": 0.6702,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9063355851743655,
+      "grad_norm": 0.6913743019104004,
+      "learning_rate": 3.0904808770813024e-05,
+      "loss": 0.6533,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9067007485849918,
+      "grad_norm": 0.9279155731201172,
+      "learning_rate": 3.0892780700300544e-05,
+      "loss": 0.6641,
+      "step": 2483
+    },
+    {
+      "epoch": 0.907065911995618,
+      "grad_norm": 0.9381618499755859,
+      "learning_rate": 3.08807470258895e-05,
+      "loss": 0.6606,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9074310754062443,
+      "grad_norm": 1.03154718875885,
+      "learning_rate": 3.086870775377072e-05,
+      "loss": 0.7056,
+      "step": 2485
+    },
+    {
+      "epoch": 0.9077962388168705,
+      "grad_norm": 0.9431748390197754,
+      "learning_rate": 3.085666289013794e-05,
+      "loss": 0.6774,
+      "step": 2486
+    },
+    {
+      "epoch": 0.9081614022274968,
+      "grad_norm": 0.8811989426612854,
+      "learning_rate": 3.0844612441187755e-05,
+      "loss": 0.6566,
+      "step": 2487
+    },
+    {
+      "epoch": 0.9085265656381231,
+      "grad_norm": 0.7740017771720886,
+      "learning_rate": 3.083255641311963e-05,
+      "loss": 0.6962,
+      "step": 2488
+    },
+    {
+      "epoch": 0.9088917290487493,
+      "grad_norm": 0.8390200734138489,
+      "learning_rate": 3.082049481213592e-05,
+      "loss": 0.6742,
+      "step": 2489
+    },
+    {
+      "epoch": 0.9092568924593756,
+      "grad_norm": 0.7879806756973267,
+      "learning_rate": 3.0808427644441825e-05,
+      "loss": 0.6726,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9096220558700018,
+      "grad_norm": 0.8950016498565674,
+      "learning_rate": 3.079635491624542e-05,
+      "loss": 0.6259,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9099872192806281,
+      "grad_norm": 0.940693736076355,
+      "learning_rate": 3.078427663375765e-05,
+      "loss": 0.6816,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9103523826912543,
+      "grad_norm": 0.7302183508872986,
+      "learning_rate": 3.077219280319229e-05,
+      "loss": 0.6627,
+      "step": 2493
+    },
+    {
+      "epoch": 0.9107175461018806,
+      "grad_norm": 1.1500719785690308,
+      "learning_rate": 3.0760103430766e-05,
+      "loss": 0.6769,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9110827095125068,
+      "grad_norm": 1.0735278129577637,
+      "learning_rate": 3.0748008522698265e-05,
+      "loss": 0.6611,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9114478729231331,
+      "grad_norm": 1.0470819473266602,
+      "learning_rate": 3.073590808521144e-05,
+      "loss": 0.696,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9118130363337593,
+      "grad_norm": 0.9414476752281189,
+      "learning_rate": 3.072380212453071e-05,
+      "loss": 0.6763,
+      "step": 2497
+    },
+    {
+      "epoch": 0.9121781997443856,
+      "grad_norm": 1.0331350564956665,
+      "learning_rate": 3.07116906468841e-05,
+      "loss": 0.6649,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9125433631550118,
+      "grad_norm": 0.8421180844306946,
+      "learning_rate": 3.069957365850249e-05,
+      "loss": 0.6743,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9129085265656381,
+      "grad_norm": 1.2963193655014038,
+      "learning_rate": 3.0687451165619586e-05,
+      "loss": 0.6843,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9132736899762643,
+      "grad_norm": 0.7593322992324829,
+      "learning_rate": 3.0675323174471905e-05,
+      "loss": 0.6536,
+      "step": 2501
+    },
+    {
+      "epoch": 0.9136388533868907,
+      "grad_norm": 0.9185073375701904,
+      "learning_rate": 3.0663189691298836e-05,
+      "loss": 0.684,
+      "step": 2502
+    },
+    {
+      "epoch": 0.9140040167975169,
+      "grad_norm": 0.7559370994567871,
+      "learning_rate": 3.0651050722342554e-05,
+      "loss": 0.6346,
+      "step": 2503
+    },
+    {
+      "epoch": 0.9143691802081432,
+      "grad_norm": 1.0058975219726562,
+      "learning_rate": 3.0638906273848075e-05,
+      "loss": 0.6587,
+      "step": 2504
+    },
+    {
+      "epoch": 0.9147343436187694,
+      "grad_norm": 0.6809512376785278,
+      "learning_rate": 3.062675635206323e-05,
+      "loss": 0.6429,
+      "step": 2505
+    },
+    {
+      "epoch": 0.9150995070293957,
+      "grad_norm": 0.8577700853347778,
+      "learning_rate": 3.061460096323867e-05,
+      "loss": 0.6777,
+      "step": 2506
+    },
+    {
+      "epoch": 0.9154646704400219,
+      "grad_norm": 0.8358782529830933,
+      "learning_rate": 3.060244011362785e-05,
+      "loss": 0.6736,
+      "step": 2507
+    },
+    {
+      "epoch": 0.9158298338506482,
+      "grad_norm": 0.864080548286438,
+      "learning_rate": 3.0590273809487037e-05,
+      "loss": 0.6594,
+      "step": 2508
+    },
+    {
+      "epoch": 0.9161949972612744,
+      "grad_norm": 0.9189779758453369,
+      "learning_rate": 3.057810205707532e-05,
+      "loss": 0.6852,
+      "step": 2509
+    },
+    {
+      "epoch": 0.9165601606719007,
+      "grad_norm": 1.0136405229568481,
+      "learning_rate": 3.0565924862654556e-05,
+      "loss": 0.6826,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9169253240825269,
+      "grad_norm": 0.7467730045318604,
+      "learning_rate": 3.055374223248944e-05,
+      "loss": 0.6686,
+      "step": 2511
+    },
+    {
+      "epoch": 0.9172904874931532,
+      "grad_norm": 0.9998980164527893,
+      "learning_rate": 3.054155417284745e-05,
+      "loss": 0.6429,
+      "step": 2512
+    },
+    {
+      "epoch": 0.9176556509037794,
+      "grad_norm": 0.713403582572937,
+      "learning_rate": 3.0529360689998836e-05,
+      "loss": 0.6859,
+      "step": 2513
+    },
+    {
+      "epoch": 0.9180208143144057,
+      "grad_norm": 1.016985535621643,
+      "learning_rate": 3.051716179021666e-05,
+      "loss": 0.6797,
+      "step": 2514
+    },
+    {
+      "epoch": 0.9183859777250319,
+      "grad_norm": 0.7532426118850708,
+      "learning_rate": 3.050495747977677e-05,
+      "loss": 0.6455,
+      "step": 2515
+    },
+    {
+      "epoch": 0.9187511411356583,
+      "grad_norm": 0.7963048219680786,
+      "learning_rate": 3.0492747764957798e-05,
+      "loss": 0.6453,
+      "step": 2516
+    },
+    {
+      "epoch": 0.9191163045462845,
+      "grad_norm": 1.199378490447998,
+      "learning_rate": 3.0480532652041153e-05,
+      "loss": 0.6631,
+      "step": 2517
+    },
+    {
+      "epoch": 0.9194814679569108,
+      "grad_norm": 1.2027003765106201,
+      "learning_rate": 3.0468312147311007e-05,
+      "loss": 0.683,
+      "step": 2518
+    },
+    {
+      "epoch": 0.919846631367537,
+      "grad_norm": 0.9782599806785583,
+      "learning_rate": 3.0456086257054336e-05,
+      "loss": 0.6481,
+      "step": 2519
+    },
+    {
+      "epoch": 0.9202117947781633,
+      "grad_norm": 1.3586641550064087,
+      "learning_rate": 3.0443854987560856e-05,
+      "loss": 0.6474,
+      "step": 2520
+    },
+    {
+      "epoch": 0.9205769581887895,
+      "grad_norm": 1.0380148887634277,
+      "learning_rate": 3.043161834512308e-05,
+      "loss": 0.673,
+      "step": 2521
+    },
+    {
+      "epoch": 0.9209421215994157,
+      "grad_norm": 0.9029962420463562,
+      "learning_rate": 3.0419376336036252e-05,
+      "loss": 0.6766,
+      "step": 2522
+    },
+    {
+      "epoch": 0.921307285010042,
+      "grad_norm": 0.9196898341178894,
+      "learning_rate": 3.04071289665984e-05,
+      "loss": 0.6318,
+      "step": 2523
+    },
+    {
+      "epoch": 0.9216724484206682,
+      "grad_norm": 0.8275750875473022,
+      "learning_rate": 3.0394876243110318e-05,
+      "loss": 0.6523,
+      "step": 2524
+    },
+    {
+      "epoch": 0.9220376118312945,
+      "grad_norm": 0.7632416486740112,
+      "learning_rate": 3.038261817187552e-05,
+      "loss": 0.6415,
+      "step": 2525
+    },
+    {
+      "epoch": 0.9224027752419207,
+      "grad_norm": 0.7995944023132324,
+      "learning_rate": 3.0370354759200307e-05,
+      "loss": 0.6556,
+      "step": 2526
+    },
+    {
+      "epoch": 0.922767938652547,
+      "grad_norm": 0.802648663520813,
+      "learning_rate": 3.03580860113937e-05,
+      "loss": 0.6705,
+      "step": 2527
+    },
+    {
+      "epoch": 0.9231331020631732,
+      "grad_norm": 0.8473702073097229,
+      "learning_rate": 3.034581193476749e-05,
+      "loss": 0.6671,
+      "step": 2528
+    },
+    {
+      "epoch": 0.9234982654737995,
+      "grad_norm": 1.3099536895751953,
+      "learning_rate": 3.0333532535636193e-05,
+      "loss": 0.6554,
+      "step": 2529
+    },
+    {
+      "epoch": 0.9238634288844257,
+      "grad_norm": 0.9661362767219543,
+      "learning_rate": 3.032124782031706e-05,
+      "loss": 0.6268,
+      "step": 2530
+    },
+    {
+      "epoch": 0.924228592295052,
+      "grad_norm": 0.8286318182945251,
+      "learning_rate": 3.0308957795130092e-05,
+      "loss": 0.657,
+      "step": 2531
+    },
+    {
+      "epoch": 0.9245937557056783,
+      "grad_norm": 0.8090802431106567,
+      "learning_rate": 3.0296662466398005e-05,
+      "loss": 0.6227,
+      "step": 2532
+    },
+    {
+      "epoch": 0.9249589191163046,
+      "grad_norm": 1.15580415725708,
+      "learning_rate": 3.028436184044626e-05,
+      "loss": 0.6479,
+      "step": 2533
+    },
+    {
+      "epoch": 0.9253240825269308,
+      "grad_norm": 1.3873413801193237,
+      "learning_rate": 3.0272055923603046e-05,
+      "loss": 0.6664,
+      "step": 2534
+    },
+    {
+      "epoch": 0.9256892459375571,
+      "grad_norm": 0.9026642441749573,
+      "learning_rate": 3.025974472219924e-05,
+      "loss": 0.7167,
+      "step": 2535
+    },
+    {
+      "epoch": 0.9260544093481833,
+      "grad_norm": 0.8149577379226685,
+      "learning_rate": 3.024742824256848e-05,
+      "loss": 0.703,
+      "step": 2536
+    },
+    {
+      "epoch": 0.9264195727588096,
+      "grad_norm": 0.8617291450500488,
+      "learning_rate": 3.0235106491047078e-05,
+      "loss": 0.6335,
+      "step": 2537
+    },
+    {
+      "epoch": 0.9267847361694358,
+      "grad_norm": 1.062455654144287,
+      "learning_rate": 3.022277947397411e-05,
+      "loss": 0.6685,
+      "step": 2538
+    },
+    {
+      "epoch": 0.9271498995800621,
+      "grad_norm": 0.8623011708259583,
+      "learning_rate": 3.0210447197691317e-05,
+      "loss": 0.6544,
+      "step": 2539
+    },
+    {
+      "epoch": 0.9275150629906883,
+      "grad_norm": 0.9414846897125244,
+      "learning_rate": 3.019810966854315e-05,
+      "loss": 0.676,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9278802264013146,
+      "grad_norm": 1.4701042175292969,
+      "learning_rate": 3.018576689287679e-05,
+      "loss": 0.6616,
+      "step": 2541
+    },
+    {
+      "epoch": 0.9282453898119408,
+      "grad_norm": 0.8895155787467957,
+      "learning_rate": 3.0173418877042092e-05,
+      "loss": 0.6573,
+      "step": 2542
+    },
+    {
+      "epoch": 0.9286105532225671,
+      "grad_norm": 0.735467791557312,
+      "learning_rate": 3.0161065627391618e-05,
+      "loss": 0.6837,
+      "step": 2543
+    },
+    {
+      "epoch": 0.9289757166331933,
+      "grad_norm": 0.7775281667709351,
+      "learning_rate": 3.0148707150280613e-05,
+      "loss": 0.6456,
+      "step": 2544
+    },
+    {
+      "epoch": 0.9293408800438197,
+      "grad_norm": 0.8725507259368896,
+      "learning_rate": 3.0136343452067023e-05,
+      "loss": 0.6616,
+      "step": 2545
+    },
+    {
+      "epoch": 0.9297060434544459,
+      "grad_norm": 0.7502171397209167,
+      "learning_rate": 3.012397453911147e-05,
+      "loss": 0.6427,
+      "step": 2546
+    },
+    {
+      "epoch": 0.9300712068650722,
+      "grad_norm": 0.8129810690879822,
+      "learning_rate": 3.011160041777727e-05,
+      "loss": 0.689,
+      "step": 2547
+    },
+    {
+      "epoch": 0.9304363702756984,
+      "grad_norm": 0.8770913481712341,
+      "learning_rate": 3.009922109443041e-05,
+      "loss": 0.6851,
+      "step": 2548
+    },
+    {
+      "epoch": 0.9308015336863247,
+      "grad_norm": 1.027323842048645,
+      "learning_rate": 3.0086836575439554e-05,
+      "loss": 0.6841,
+      "step": 2549
+    },
+    {
+      "epoch": 0.9311666970969509,
+      "grad_norm": 1.640295386314392,
+      "learning_rate": 3.0074446867176035e-05,
+      "loss": 0.6509,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9315318605075772,
+      "grad_norm": 0.9461489915847778,
+      "learning_rate": 3.006205197601387e-05,
+      "loss": 0.6566,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9318970239182034,
+      "grad_norm": 0.89789879322052,
+      "learning_rate": 3.0049651908329724e-05,
+      "loss": 0.6818,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9322621873288297,
+      "grad_norm": 1.0178797245025635,
+      "learning_rate": 3.0037246670502943e-05,
+      "loss": 0.6646,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9326273507394559,
+      "grad_norm": 1.3303459882736206,
+      "learning_rate": 3.002483626891551e-05,
+      "loss": 0.6968,
+      "step": 2554
+    },
+    {
+      "epoch": 0.9329925141500821,
+      "grad_norm": 0.9349764585494995,
+      "learning_rate": 3.00124207099521e-05,
+      "loss": 0.6819,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9333576775607084,
+      "grad_norm": 0.8530802726745605,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.634,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9337228409713346,
+      "grad_norm": 1.2609745264053345,
+      "learning_rate": 2.998757414544918e-05,
+      "loss": 0.6855,
+      "step": 2557
+    },
+    {
+      "epoch": 0.9340880043819609,
+      "grad_norm": 1.2250069379806519,
+      "learning_rate": 2.9975143152692242e-05,
+      "loss": 0.6812,
+      "step": 2558
+    },
+    {
+      "epoch": 0.9344531677925871,
+      "grad_norm": 1.0234918594360352,
+      "learning_rate": 2.9962707028124417e-05,
+      "loss": 0.6678,
+      "step": 2559
+    },
+    {
+      "epoch": 0.9348183312032134,
+      "grad_norm": 1.4435856342315674,
+      "learning_rate": 2.9950265778143616e-05,
+      "loss": 0.6772,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9351834946138396,
+      "grad_norm": 0.6534599661827087,
+      "learning_rate": 2.9937819409150343e-05,
+      "loss": 0.6635,
+      "step": 2561
+    },
+    {
+      "epoch": 0.935548658024466,
+      "grad_norm": 0.7986537218093872,
+      "learning_rate": 2.9925367927547772e-05,
+      "loss": 0.6802,
+      "step": 2562
+    },
+    {
+      "epoch": 0.9359138214350922,
+      "grad_norm": 1.0671440362930298,
+      "learning_rate": 2.991291133974168e-05,
+      "loss": 0.6917,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9362789848457185,
+      "grad_norm": 1.031617522239685,
+      "learning_rate": 2.990044965214048e-05,
+      "loss": 0.688,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9366441482563447,
+      "grad_norm": 0.7745130658149719,
+      "learning_rate": 2.988798287115522e-05,
+      "loss": 0.6746,
+      "step": 2565
+    },
+    {
+      "epoch": 0.937009311666971,
+      "grad_norm": 0.9795826077461243,
+      "learning_rate": 2.9875511003199547e-05,
+      "loss": 0.655,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9373744750775972,
+      "grad_norm": 0.8267850875854492,
+      "learning_rate": 2.9863034054689744e-05,
+      "loss": 0.6918,
+      "step": 2567
+    },
+    {
+      "epoch": 0.9377396384882235,
+      "grad_norm": 0.771300733089447,
+      "learning_rate": 2.98505520320447e-05,
+      "loss": 0.6711,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9381048018988497,
+      "grad_norm": 0.9123200178146362,
+      "learning_rate": 2.9838064941685914e-05,
+      "loss": 0.7088,
+      "step": 2569
+    },
+    {
+      "epoch": 0.938469965309476,
+      "grad_norm": 0.9156253933906555,
+      "learning_rate": 2.9825572790037497e-05,
+      "loss": 0.6643,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9388351287201022,
+      "grad_norm": 1.439208745956421,
+      "learning_rate": 2.9813075583526146e-05,
+      "loss": 0.6512,
+      "step": 2571
+    },
+    {
+      "epoch": 0.9392002921307285,
+      "grad_norm": 3.506324529647827,
+      "learning_rate": 2.9800573328581187e-05,
+      "loss": 0.6599,
+      "step": 2572
+    },
+    {
+      "epoch": 0.9395654555413547,
+      "grad_norm": 0.8508874177932739,
+      "learning_rate": 2.9788066031634523e-05,
+      "loss": 0.6328,
+      "step": 2573
+    },
+    {
+      "epoch": 0.939930618951981,
+      "grad_norm": 1.151938796043396,
+      "learning_rate": 2.9775553699120654e-05,
+      "loss": 0.6804,
+      "step": 2574
+    },
+    {
+      "epoch": 0.9402957823626072,
+      "grad_norm": 0.9399933815002441,
+      "learning_rate": 2.976303633747668e-05,
+      "loss": 0.6741,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9406609457732336,
+      "grad_norm": 0.8826687932014465,
+      "learning_rate": 2.975051395314227e-05,
+      "loss": 0.6538,
+      "step": 2576
+    },
+    {
+      "epoch": 0.9410261091838598,
+      "grad_norm": 1.036671757698059,
+      "learning_rate": 2.97379865525597e-05,
+      "loss": 0.6794,
+      "step": 2577
+    },
+    {
+      "epoch": 0.9413912725944861,
+      "grad_norm": 0.9944286942481995,
+      "learning_rate": 2.9725454142173805e-05,
+      "loss": 0.6844,
+      "step": 2578
+    },
+    {
+      "epoch": 0.9417564360051123,
+      "grad_norm": 0.885906994342804,
+      "learning_rate": 2.9712916728432016e-05,
+      "loss": 0.6588,
+      "step": 2579
+    },
+    {
+      "epoch": 0.9421215994157386,
+      "grad_norm": 0.9254575967788696,
+      "learning_rate": 2.9700374317784326e-05,
+      "loss": 0.6608,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9424867628263648,
+      "grad_norm": 0.7100337743759155,
+      "learning_rate": 2.9687826916683293e-05,
+      "loss": 0.6473,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9428519262369911,
+      "grad_norm": 1.1397892236709595,
+      "learning_rate": 2.967527453158407e-05,
+      "loss": 0.681,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9432170896476173,
+      "grad_norm": 0.9657423496246338,
+      "learning_rate": 2.9662717168944343e-05,
+      "loss": 0.6649,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9435822530582436,
+      "grad_norm": 0.9102057814598083,
+      "learning_rate": 2.9650154835224373e-05,
+      "loss": 0.6494,
+      "step": 2584
+    },
+    {
+      "epoch": 0.9439474164688698,
+      "grad_norm": 0.7469964027404785,
+      "learning_rate": 2.963758753688697e-05,
+      "loss": 0.6558,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9443125798794961,
+      "grad_norm": 0.9794651865959167,
+      "learning_rate": 2.962501528039752e-05,
+      "loss": 0.6475,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9446777432901223,
+      "grad_norm": 0.8790867328643799,
+      "learning_rate": 2.9612438072223926e-05,
+      "loss": 0.6753,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9450429067007485,
+      "grad_norm": 0.8802550435066223,
+      "learning_rate": 2.9599855918836677e-05,
+      "loss": 0.6663,
+      "step": 2588
+    },
+    {
+      "epoch": 0.9454080701113748,
+      "grad_norm": 0.7142858505249023,
+      "learning_rate": 2.9587268826708774e-05,
+      "loss": 0.6452,
+      "step": 2589
+    },
+    {
+      "epoch": 0.945773233522001,
+      "grad_norm": 0.9301719665527344,
+      "learning_rate": 2.9574676802315775e-05,
+      "loss": 0.6924,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9461383969326274,
+      "grad_norm": 0.8197965025901794,
+      "learning_rate": 2.9562079852135767e-05,
+      "loss": 0.6494,
+      "step": 2591
+    },
+    {
+      "epoch": 0.9465035603432536,
+      "grad_norm": 0.8453877568244934,
+      "learning_rate": 2.9549477982649372e-05,
+      "loss": 0.6896,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9468687237538799,
+      "grad_norm": 0.869964599609375,
+      "learning_rate": 2.9536871200339754e-05,
+      "loss": 0.5818,
+      "step": 2593
+    },
+    {
+      "epoch": 0.9472338871645061,
+      "grad_norm": 0.7052625417709351,
+      "learning_rate": 2.95242595116926e-05,
+      "loss": 0.6368,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9475990505751324,
+      "grad_norm": 1.1453596353530884,
+      "learning_rate": 2.951164292319611e-05,
+      "loss": 0.6415,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9479642139857586,
+      "grad_norm": 0.9452725648880005,
+      "learning_rate": 2.9499021441341012e-05,
+      "loss": 0.6585,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9483293773963849,
+      "grad_norm": 0.8761558532714844,
+      "learning_rate": 2.9486395072620552e-05,
+      "loss": 0.6634,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9486945408070111,
+      "grad_norm": 0.9212718605995178,
+      "learning_rate": 2.947376382353049e-05,
+      "loss": 0.6774,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9490597042176374,
+      "grad_norm": 0.8499828577041626,
+      "learning_rate": 2.946112770056911e-05,
+      "loss": 0.675,
+      "step": 2599
+    },
+    {
+      "epoch": 0.9494248676282636,
+      "grad_norm": 1.040263056755066,
+      "learning_rate": 2.9448486710237173e-05,
+      "loss": 0.6536,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9497900310388899,
+      "grad_norm": 1.0057485103607178,
+      "learning_rate": 2.943584085903797e-05,
+      "loss": 0.6959,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9501551944495161,
+      "grad_norm": 0.9337158799171448,
+      "learning_rate": 2.942319015347728e-05,
+      "loss": 0.6506,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9505203578601424,
+      "grad_norm": 0.7997397184371948,
+      "learning_rate": 2.9410534600063387e-05,
+      "loss": 0.6647,
+      "step": 2603
+    },
+    {
+      "epoch": 0.9508855212707686,
+      "grad_norm": 1.0067977905273438,
+      "learning_rate": 2.939787420530706e-05,
+      "loss": 0.6667,
+      "step": 2604
+    },
+    {
+      "epoch": 0.951250684681395,
+      "grad_norm": 0.8577103018760681,
+      "learning_rate": 2.9385208975721568e-05,
+      "loss": 0.6488,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9516158480920212,
+      "grad_norm": 0.7341446280479431,
+      "learning_rate": 2.9372538917822666e-05,
+      "loss": 0.6837,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9519810115026475,
+      "grad_norm": 0.7933617830276489,
+      "learning_rate": 2.9359864038128586e-05,
+      "loss": 0.6474,
+      "step": 2607
+    },
+    {
+      "epoch": 0.9523461749132737,
+      "grad_norm": 1.0580298900604248,
+      "learning_rate": 2.934718434316005e-05,
+      "loss": 0.653,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9527113383239,
+      "grad_norm": 0.959298849105835,
+      "learning_rate": 2.933449983944024e-05,
+      "loss": 0.6759,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9530765017345262,
+      "grad_norm": 0.6825525760650635,
+      "learning_rate": 2.932181053349484e-05,
+      "loss": 0.6476,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9534416651451525,
+      "grad_norm": 1.1617248058319092,
+      "learning_rate": 2.930911643185198e-05,
+      "loss": 0.6554,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9538068285557787,
+      "grad_norm": 0.9940661191940308,
+      "learning_rate": 2.9296417541042267e-05,
+      "loss": 0.6627,
+      "step": 2612
+    },
+    {
+      "epoch": 0.954171991966405,
+      "grad_norm": 1.0652787685394287,
+      "learning_rate": 2.928371386759877e-05,
+      "loss": 0.6692,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9545371553770312,
+      "grad_norm": 0.8765406608581543,
+      "learning_rate": 2.9271005418057023e-05,
+      "loss": 0.6378,
+      "step": 2614
+    },
+    {
+      "epoch": 0.9549023187876575,
+      "grad_norm": 0.9065739512443542,
+      "learning_rate": 2.9258292198955004e-05,
+      "loss": 0.6375,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9552674821982837,
+      "grad_norm": 1.0190774202346802,
+      "learning_rate": 2.924557421683317e-05,
+      "loss": 0.6453,
+      "step": 2616
+    },
+    {
+      "epoch": 0.95563264560891,
+      "grad_norm": 0.9807654023170471,
+      "learning_rate": 2.9232851478234397e-05,
+      "loss": 0.624,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9559978090195362,
+      "grad_norm": 0.9084011316299438,
+      "learning_rate": 2.9220123989704034e-05,
+      "loss": 0.6714,
+      "step": 2618
+    },
+    {
+      "epoch": 0.9563629724301626,
+      "grad_norm": 0.7725682258605957,
+      "learning_rate": 2.9207391757789857e-05,
+      "loss": 0.6478,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9567281358407888,
+      "grad_norm": 0.8699987530708313,
+      "learning_rate": 2.9194654789042088e-05,
+      "loss": 0.639,
+      "step": 2620
+    },
+    {
+      "epoch": 0.957093299251415,
+      "grad_norm": 1.0805641412734985,
+      "learning_rate": 2.9181913090013386e-05,
+      "loss": 0.6514,
+      "step": 2621
+    },
+    {
+      "epoch": 0.9574584626620413,
+      "grad_norm": 1.4972708225250244,
+      "learning_rate": 2.9169166667258856e-05,
+      "loss": 0.6708,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9578236260726675,
+      "grad_norm": 0.7883728742599487,
+      "learning_rate": 2.9156415527336003e-05,
+      "loss": 0.6362,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9581887894832938,
+      "grad_norm": 0.9713582992553711,
+      "learning_rate": 2.9143659676804788e-05,
+      "loss": 0.6578,
+      "step": 2624
+    },
+    {
+      "epoch": 0.95855395289392,
+      "grad_norm": 0.9497501254081726,
+      "learning_rate": 2.9130899122227583e-05,
+      "loss": 0.686,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9589191163045463,
+      "grad_norm": 0.8625170588493347,
+      "learning_rate": 2.911813387016918e-05,
+      "loss": 0.6597,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9592842797151725,
+      "grad_norm": 1.3011974096298218,
+      "learning_rate": 2.910536392719679e-05,
+      "loss": 0.6885,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9596494431257988,
+      "grad_norm": 0.7818294763565063,
+      "learning_rate": 2.9092589299880028e-05,
+      "loss": 0.6331,
+      "step": 2628
+    },
+    {
+      "epoch": 0.960014606536425,
+      "grad_norm": 0.9671834707260132,
+      "learning_rate": 2.9079809994790937e-05,
+      "loss": 0.666,
+      "step": 2629
+    },
+    {
+      "epoch": 0.9603797699470513,
+      "grad_norm": 0.8328306674957275,
+      "learning_rate": 2.9067026018503956e-05,
+      "loss": 0.6589,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9607449333576775,
+      "grad_norm": 1.8160985708236694,
+      "learning_rate": 2.9054237377595926e-05,
+      "loss": 0.6638,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9611100967683038,
+      "grad_norm": 0.9076361060142517,
+      "learning_rate": 2.904144407864609e-05,
+      "loss": 0.6498,
+      "step": 2632
+    },
+    {
+      "epoch": 0.96147526017893,
+      "grad_norm": 0.821333110332489,
+      "learning_rate": 2.9028646128236083e-05,
+      "loss": 0.6732,
+      "step": 2633
+    },
+    {
+      "epoch": 0.9618404235895563,
+      "grad_norm": 0.9395877718925476,
+      "learning_rate": 2.901584353294994e-05,
+      "loss": 0.6672,
+      "step": 2634
+    },
+    {
+      "epoch": 0.9622055870001825,
+      "grad_norm": 1.0115352869033813,
+      "learning_rate": 2.9003036299374083e-05,
+      "loss": 0.6884,
+      "step": 2635
+    },
+    {
+      "epoch": 0.9625707504108089,
+      "grad_norm": 1.0721027851104736,
+      "learning_rate": 2.899022443409732e-05,
+      "loss": 0.6583,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9629359138214351,
+      "grad_norm": 0.67411869764328,
+      "learning_rate": 2.8977407943710846e-05,
+      "loss": 0.6597,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9633010772320614,
+      "grad_norm": 0.8598989844322205,
+      "learning_rate": 2.8964586834808214e-05,
+      "loss": 0.6729,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9636662406426876,
+      "grad_norm": 0.9835753440856934,
+      "learning_rate": 2.8951761113985393e-05,
+      "loss": 0.6843,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9640314040533139,
+      "grad_norm": 0.967534601688385,
+      "learning_rate": 2.8938930787840683e-05,
+      "loss": 0.6575,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9643965674639401,
+      "grad_norm": 0.8860105276107788,
+      "learning_rate": 2.8926095862974782e-05,
+      "loss": 0.6855,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9647617308745664,
+      "grad_norm": 0.7182053923606873,
+      "learning_rate": 2.8913256345990746e-05,
+      "loss": 0.673,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9651268942851926,
+      "grad_norm": 0.788271427154541,
+      "learning_rate": 2.890041224349398e-05,
+      "loss": 0.6594,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9654920576958189,
+      "grad_norm": 0.8986093997955322,
+      "learning_rate": 2.888756356209227e-05,
+      "loss": 0.6663,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9658572211064451,
+      "grad_norm": 0.8097816109657288,
+      "learning_rate": 2.8874710308395745e-05,
+      "loss": 0.6356,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9662223845170714,
+      "grad_norm": 0.7233504056930542,
+      "learning_rate": 2.8861852489016882e-05,
+      "loss": 0.6732,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9665875479276976,
+      "grad_norm": 0.8124412298202515,
+      "learning_rate": 2.884899011057052e-05,
+      "loss": 0.6401,
+      "step": 2647
+    },
+    {
+      "epoch": 0.966952711338324,
+      "grad_norm": 1.0130997896194458,
+      "learning_rate": 2.8836123179673828e-05,
+      "loss": 0.6628,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9673178747489501,
+      "grad_norm": 0.9995191097259521,
+      "learning_rate": 2.882325170294634e-05,
+      "loss": 0.6696,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9676830381595765,
+      "grad_norm": 1.2716678380966187,
+      "learning_rate": 2.88103756870099e-05,
+      "loss": 0.669,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9680482015702027,
+      "grad_norm": 0.9267953038215637,
+      "learning_rate": 2.879749513848871e-05,
+      "loss": 0.6901,
+      "step": 2651
+    },
+    {
+      "epoch": 0.968413364980829,
+      "grad_norm": 0.9453235268592834,
+      "learning_rate": 2.8784610064009297e-05,
+      "loss": 0.6676,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9687785283914552,
+      "grad_norm": 0.8978050947189331,
+      "learning_rate": 2.87717204702005e-05,
+      "loss": 0.644,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9691436918020815,
+      "grad_norm": 0.9811480045318604,
+      "learning_rate": 2.8758826363693516e-05,
+      "loss": 0.688,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9695088552127077,
+      "grad_norm": 1.0782561302185059,
+      "learning_rate": 2.8745927751121834e-05,
+      "loss": 0.6663,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9698740186233339,
+      "grad_norm": 1.0224454402923584,
+      "learning_rate": 2.8733024639121283e-05,
+      "loss": 0.647,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9702391820339602,
+      "grad_norm": 1.114777684211731,
+      "learning_rate": 2.8720117034329984e-05,
+      "loss": 0.6792,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9706043454445864,
+      "grad_norm": 0.8056793212890625,
+      "learning_rate": 2.8707204943388386e-05,
+      "loss": 0.637,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9709695088552127,
+      "grad_norm": 0.9775443077087402,
+      "learning_rate": 2.8694288372939243e-05,
+      "loss": 0.6696,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9713346722658389,
+      "grad_norm": 2.4969546794891357,
+      "learning_rate": 2.8681367329627617e-05,
+      "loss": 0.6862,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9716998356764652,
+      "grad_norm": 0.8990214467048645,
+      "learning_rate": 2.8668441820100857e-05,
+      "loss": 0.6729,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9720649990870914,
+      "grad_norm": 1.258520483970642,
+      "learning_rate": 2.865551185100863e-05,
+      "loss": 0.6596,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9724301624977177,
+      "grad_norm": 0.851305365562439,
+      "learning_rate": 2.864257742900287e-05,
+      "loss": 0.6483,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9727953259083439,
+      "grad_norm": 1.215659499168396,
+      "learning_rate": 2.8629638560737832e-05,
+      "loss": 0.6443,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9731604893189703,
+      "grad_norm": 0.7999595403671265,
+      "learning_rate": 2.8616695252870044e-05,
+      "loss": 0.6779,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9735256527295965,
+      "grad_norm": 0.8783779144287109,
+      "learning_rate": 2.8603747512058312e-05,
+      "loss": 0.673,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9738908161402228,
+      "grad_norm": 0.8891261219978333,
+      "learning_rate": 2.859079534496373e-05,
+      "loss": 0.6394,
+      "step": 2667
+    },
+    {
+      "epoch": 0.974255979550849,
+      "grad_norm": 0.9345478415489197,
+      "learning_rate": 2.8577838758249674e-05,
+      "loss": 0.6508,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9746211429614753,
+      "grad_norm": 1.225130319595337,
+      "learning_rate": 2.8564877758581784e-05,
+      "loss": 0.6678,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9749863063721015,
+      "grad_norm": 1.0890440940856934,
+      "learning_rate": 2.855191235262797e-05,
+      "loss": 0.6763,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9753514697827278,
+      "grad_norm": 1.1266112327575684,
+      "learning_rate": 2.8538942547058425e-05,
+      "loss": 0.6585,
+      "step": 2671
+    },
+    {
+      "epoch": 0.975716633193354,
+      "grad_norm": 0.8562392592430115,
+      "learning_rate": 2.8525968348545574e-05,
+      "loss": 0.6486,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9760817966039803,
+      "grad_norm": 0.8453388810157776,
+      "learning_rate": 2.8512989763764146e-05,
+      "loss": 0.6464,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9764469600146065,
+      "grad_norm": 0.9347715377807617,
+      "learning_rate": 2.850000679939108e-05,
+      "loss": 0.6783,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9768121234252328,
+      "grad_norm": 0.811362624168396,
+      "learning_rate": 2.8487019462105606e-05,
+      "loss": 0.6508,
+      "step": 2675
+    },
+    {
+      "epoch": 0.977177286835859,
+      "grad_norm": 1.0582472085952759,
+      "learning_rate": 2.847402775858918e-05,
+      "loss": 0.6731,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9775424502464853,
+      "grad_norm": 1.1253806352615356,
+      "learning_rate": 2.846103169552551e-05,
+      "loss": 0.6492,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9779076136571115,
+      "grad_norm": 1.0529634952545166,
+      "learning_rate": 2.844803127960056e-05,
+      "loss": 0.6732,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9782727770677379,
+      "grad_norm": 0.7079199552536011,
+      "learning_rate": 2.843502651750251e-05,
+      "loss": 0.646,
+      "step": 2679
+    },
+    {
+      "epoch": 0.978637940478364,
+      "grad_norm": 0.8783506155014038,
+      "learning_rate": 2.8422017415921793e-05,
+      "loss": 0.6696,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9790031038889904,
+      "grad_norm": 0.8645327091217041,
+      "learning_rate": 2.840900398155107e-05,
+      "loss": 0.6573,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9793682672996166,
+      "grad_norm": 0.7897984981536865,
+      "learning_rate": 2.839598622108523e-05,
+      "loss": 0.6962,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9797334307102429,
+      "grad_norm": 0.7942821979522705,
+      "learning_rate": 2.8382964141221396e-05,
+      "loss": 0.6689,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9800985941208691,
+      "grad_norm": 1.1115641593933105,
+      "learning_rate": 2.8369937748658892e-05,
+      "loss": 0.645,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9804637575314954,
+      "grad_norm": 1.2980235815048218,
+      "learning_rate": 2.8356907050099284e-05,
+      "loss": 0.6995,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9808289209421216,
+      "grad_norm": 0.9709264039993286,
+      "learning_rate": 2.834387205224634e-05,
+      "loss": 0.6954,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9811940843527479,
+      "grad_norm": 0.951606810092926,
+      "learning_rate": 2.833083276180604e-05,
+      "loss": 0.6944,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9815592477633741,
+      "grad_norm": 0.9172406792640686,
+      "learning_rate": 2.8317789185486587e-05,
+      "loss": 0.6792,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9819244111740003,
+      "grad_norm": 0.8468958735466003,
+      "learning_rate": 2.8304741329998364e-05,
+      "loss": 0.6753,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9822895745846266,
+      "grad_norm": 0.8902730345726013,
+      "learning_rate": 2.829168920205398e-05,
+      "loss": 0.6565,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9826547379952528,
+      "grad_norm": 0.8718501329421997,
+      "learning_rate": 2.8278632808368222e-05,
+      "loss": 0.6646,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9830199014058791,
+      "grad_norm": 0.9342532753944397,
+      "learning_rate": 2.826557215565809e-05,
+      "loss": 0.6182,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9833850648165053,
+      "grad_norm": 1.0939021110534668,
+      "learning_rate": 2.825250725064275e-05,
+      "loss": 0.6551,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9837502282271317,
+      "grad_norm": 1.0584012269973755,
+      "learning_rate": 2.8239438100043597e-05,
+      "loss": 0.6512,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9841153916377579,
+      "grad_norm": 1.0934642553329468,
+      "learning_rate": 2.822636471058416e-05,
+      "loss": 0.6543,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9844805550483842,
+      "grad_norm": 1.096698522567749,
+      "learning_rate": 2.8213287088990184e-05,
+      "loss": 0.6534,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9848457184590104,
+      "grad_norm": 0.8641692399978638,
+      "learning_rate": 2.8200205241989583e-05,
+      "loss": 0.6652,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9852108818696367,
+      "grad_norm": 1.1894280910491943,
+      "learning_rate": 2.818711917631243e-05,
+      "loss": 0.6411,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9855760452802629,
+      "grad_norm": 1.0246202945709229,
+      "learning_rate": 2.8174028898690998e-05,
+      "loss": 0.6628,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9859412086908892,
+      "grad_norm": 0.8037968277931213,
+      "learning_rate": 2.81609344158597e-05,
+      "loss": 0.6401,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9863063721015154,
+      "grad_norm": 0.7906293272972107,
+      "learning_rate": 2.8147835734555114e-05,
+      "loss": 0.6542,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9866715355121417,
+      "grad_norm": 1.5510473251342773,
+      "learning_rate": 2.813473286151601e-05,
+      "loss": 0.6641,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9870366989227679,
+      "grad_norm": 0.9942931532859802,
+      "learning_rate": 2.8121625803483264e-05,
+      "loss": 0.6519,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9874018623333942,
+      "grad_norm": 0.8478285074234009,
+      "learning_rate": 2.810851456719995e-05,
+      "loss": 0.662,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9877670257440204,
+      "grad_norm": 0.8662696480751038,
+      "learning_rate": 2.8095399159411258e-05,
+      "loss": 0.6445,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9881321891546467,
+      "grad_norm": 0.8632956147193909,
+      "learning_rate": 2.8082279586864548e-05,
+      "loss": 0.6529,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9884973525652729,
+      "grad_norm": 0.6771783232688904,
+      "learning_rate": 2.8069155856309316e-05,
+      "loss": 0.6331,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9888625159758992,
+      "grad_norm": 0.9262771010398865,
+      "learning_rate": 2.805602797449719e-05,
+      "loss": 0.6822,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9892276793865254,
+      "grad_norm": 0.9815447926521301,
+      "learning_rate": 2.8042895948181944e-05,
+      "loss": 0.6843,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9895928427971518,
+      "grad_norm": 0.898939311504364,
+      "learning_rate": 2.8029759784119465e-05,
+      "loss": 0.6542,
+      "step": 2710
+    },
+    {
+      "epoch": 0.989958006207778,
+      "grad_norm": 1.0056416988372803,
+      "learning_rate": 2.80166194890678e-05,
+      "loss": 0.6937,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9903231696184043,
+      "grad_norm": 0.7354155778884888,
+      "learning_rate": 2.8003475069787084e-05,
+      "loss": 0.6714,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9906883330290305,
+      "grad_norm": 0.9542761445045471,
+      "learning_rate": 2.799032653303961e-05,
+      "loss": 0.6639,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9910534964396568,
+      "grad_norm": 1.159591794013977,
+      "learning_rate": 2.7977173885589768e-05,
+      "loss": 0.6654,
+      "step": 2714
+    },
+    {
+      "epoch": 0.991418659850283,
+      "grad_norm": 0.8092490434646606,
+      "learning_rate": 2.796401713420406e-05,
+      "loss": 0.6491,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9917838232609093,
+      "grad_norm": 1.3644760847091675,
+      "learning_rate": 2.7950856285651124e-05,
+      "loss": 0.6834,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9921489866715355,
+      "grad_norm": 0.8467960953712463,
+      "learning_rate": 2.793769134670167e-05,
+      "loss": 0.6613,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9925141500821618,
+      "grad_norm": 1.0781224966049194,
+      "learning_rate": 2.7924522324128535e-05,
+      "loss": 0.6527,
+      "step": 2718
+    },
+    {
+      "epoch": 0.992879313492788,
+      "grad_norm": 0.8228006958961487,
+      "learning_rate": 2.791134922470666e-05,
+      "loss": 0.6564,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9932444769034143,
+      "grad_norm": 0.9506269097328186,
+      "learning_rate": 2.7898172055213067e-05,
+      "loss": 0.6997,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9936096403140405,
+      "grad_norm": 0.931259036064148,
+      "learning_rate": 2.788499082242689e-05,
+      "loss": 0.6121,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9939748037246667,
+      "grad_norm": 0.9871553182601929,
+      "learning_rate": 2.7871805533129337e-05,
+      "loss": 0.6499,
+      "step": 2722
+    },
+    {
+      "epoch": 0.994339967135293,
+      "grad_norm": 1.1735976934432983,
+      "learning_rate": 2.7858616194103714e-05,
+      "loss": 0.6495,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9947051305459192,
+      "grad_norm": 0.8646172285079956,
+      "learning_rate": 2.7845422812135406e-05,
+      "loss": 0.6281,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9950702939565456,
+      "grad_norm": 1.1625986099243164,
+      "learning_rate": 2.783222539401188e-05,
+      "loss": 0.6686,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9954354573671718,
+      "grad_norm": 0.7719307541847229,
+      "learning_rate": 2.7819023946522678e-05,
+      "loss": 0.663,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9958006207777981,
+      "grad_norm": 1.3788267374038696,
+      "learning_rate": 2.7805818476459404e-05,
+      "loss": 0.7081,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9961657841884243,
+      "grad_norm": 0.9009970426559448,
+      "learning_rate": 2.7792608990615763e-05,
+      "loss": 0.6738,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9965309475990506,
+      "grad_norm": 1.0724678039550781,
+      "learning_rate": 2.777939549578749e-05,
+      "loss": 0.6417,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9968961110096768,
+      "grad_norm": 0.9405642747879028,
+      "learning_rate": 2.77661779987724e-05,
+      "loss": 0.666,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9972612744203031,
+      "grad_norm": 0.9860209226608276,
+      "learning_rate": 2.7752956506370366e-05,
+      "loss": 0.673,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9976264378309293,
+      "grad_norm": 0.9241143465042114,
+      "learning_rate": 2.773973102538332e-05,
+      "loss": 0.6704,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9979916012415556,
+      "grad_norm": 1.0373433828353882,
+      "learning_rate": 2.7726501562615237e-05,
+      "loss": 0.6631,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9983567646521818,
+      "grad_norm": 0.9698441028594971,
+      "learning_rate": 2.7713268124872145e-05,
+      "loss": 0.6735,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9987219280628081,
+      "grad_norm": 1.5003585815429688,
+      "learning_rate": 2.770003071896212e-05,
+      "loss": 0.6591,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9990870914734343,
+      "grad_norm": 1.487725853919983,
+      "learning_rate": 2.768678935169527e-05,
+      "loss": 0.6617,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9994522548840606,
+      "grad_norm": 0.8343055248260498,
+      "learning_rate": 2.7673544029883745e-05,
+      "loss": 0.6484,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9998174182946868,
+      "grad_norm": 1.0973925590515137,
+      "learning_rate": 2.7660294760341744e-05,
+      "loss": 0.6735,
+      "step": 2738
+    },
+    {
+      "epoch": 1.0001825817053132,
+      "grad_norm": 1.0450365543365479,
+      "learning_rate": 2.7647041549885472e-05,
+      "loss": 0.6652,
+      "step": 2739
+    },
+    {
+      "epoch": 1.0005477451159395,
+      "grad_norm": 1.9960644245147705,
+      "learning_rate": 2.7633784405333183e-05,
+      "loss": 0.6205,
+      "step": 2740
+    },
+    {
+      "epoch": 1.0009129085265656,
+      "grad_norm": 0.8917062878608704,
+      "learning_rate": 2.7620523333505142e-05,
+      "loss": 0.6269,
+      "step": 2741
+    },
+    {
+      "epoch": 1.0012780719371919,
+      "grad_norm": 0.7969324588775635,
+      "learning_rate": 2.7607258341223636e-05,
+      "loss": 0.6,
+      "step": 2742
+    },
+    {
+      "epoch": 1.0016432353478182,
+      "grad_norm": 0.9513354897499084,
+      "learning_rate": 2.7593989435312976e-05,
+      "loss": 0.611,
+      "step": 2743
+    },
+    {
+      "epoch": 1.0020083987584445,
+      "grad_norm": 0.8164132833480835,
+      "learning_rate": 2.7580716622599478e-05,
+      "loss": 0.6138,
+      "step": 2744
+    },
+    {
+      "epoch": 1.0023735621690706,
+      "grad_norm": 0.9981603026390076,
+      "learning_rate": 2.756743990991148e-05,
+      "loss": 0.6108,
+      "step": 2745
+    },
+    {
+      "epoch": 1.002738725579697,
+      "grad_norm": 0.9675151705741882,
+      "learning_rate": 2.7554159304079298e-05,
+      "loss": 0.5851,
+      "step": 2746
+    },
+    {
+      "epoch": 1.0031038889903232,
+      "grad_norm": 0.8304097056388855,
+      "learning_rate": 2.7540874811935295e-05,
+      "loss": 0.5764,
+      "step": 2747
+    },
+    {
+      "epoch": 1.0034690524009495,
+      "grad_norm": 0.8475216627120972,
+      "learning_rate": 2.7527586440313786e-05,
+      "loss": 0.5999,
+      "step": 2748
+    },
+    {
+      "epoch": 1.0038342158115756,
+      "grad_norm": 1.0157322883605957,
+      "learning_rate": 2.7514294196051124e-05,
+      "loss": 0.6093,
+      "step": 2749
+    },
+    {
+      "epoch": 1.004199379222202,
+      "grad_norm": 0.9846054911613464,
+      "learning_rate": 2.750099808598563e-05,
+      "loss": 0.5847,
+      "step": 2750
+    },
+    {
+      "epoch": 1.0045645426328282,
+      "grad_norm": 0.8925500512123108,
+      "learning_rate": 2.7487698116957617e-05,
+      "loss": 0.6193,
+      "step": 2751
+    },
+    {
+      "epoch": 1.0049297060434546,
+      "grad_norm": 0.8369265198707581,
+      "learning_rate": 2.747439429580938e-05,
+      "loss": 0.6191,
+      "step": 2752
+    },
+    {
+      "epoch": 1.0052948694540806,
+      "grad_norm": 0.9499483108520508,
+      "learning_rate": 2.746108662938521e-05,
+      "loss": 0.5982,
+      "step": 2753
+    },
+    {
+      "epoch": 1.005660032864707,
+      "grad_norm": 1.0188134908676147,
+      "learning_rate": 2.7447775124531367e-05,
+      "loss": 0.5984,
+      "step": 2754
+    },
+    {
+      "epoch": 1.0060251962753333,
+      "grad_norm": 0.9362654089927673,
+      "learning_rate": 2.7434459788096077e-05,
+      "loss": 0.6281,
+      "step": 2755
+    },
+    {
+      "epoch": 1.0063903596859594,
+      "grad_norm": 0.8612992763519287,
+      "learning_rate": 2.7421140626929556e-05,
+      "loss": 0.621,
+      "step": 2756
+    },
+    {
+      "epoch": 1.0067555230965857,
+      "grad_norm": 1.003761649131775,
+      "learning_rate": 2.7407817647883973e-05,
+      "loss": 0.6154,
+      "step": 2757
+    },
+    {
+      "epoch": 1.007120686507212,
+      "grad_norm": 0.8995448350906372,
+      "learning_rate": 2.7394490857813467e-05,
+      "loss": 0.6032,
+      "step": 2758
+    },
+    {
+      "epoch": 1.0074858499178383,
+      "grad_norm": 0.9468995928764343,
+      "learning_rate": 2.738116026357414e-05,
+      "loss": 0.604,
+      "step": 2759
+    },
+    {
+      "epoch": 1.0078510133284644,
+      "grad_norm": 0.8836452960968018,
+      "learning_rate": 2.7367825872024042e-05,
+      "loss": 0.6608,
+      "step": 2760
+    },
+    {
+      "epoch": 1.0082161767390907,
+      "grad_norm": 0.9900194406509399,
+      "learning_rate": 2.7354487690023178e-05,
+      "loss": 0.6043,
+      "step": 2761
+    },
+    {
+      "epoch": 1.008581340149717,
+      "grad_norm": 3.625828981399536,
+      "learning_rate": 2.7341145724433532e-05,
+      "loss": 0.6314,
+      "step": 2762
+    },
+    {
+      "epoch": 1.0089465035603433,
+      "grad_norm": 0.79560786485672,
+      "learning_rate": 2.732779998211898e-05,
+      "loss": 0.599,
+      "step": 2763
+    },
+    {
+      "epoch": 1.0093116669709694,
+      "grad_norm": 0.996148407459259,
+      "learning_rate": 2.731445046994539e-05,
+      "loss": 0.6152,
+      "step": 2764
+    },
+    {
+      "epoch": 1.0096768303815957,
+      "grad_norm": 1.097036600112915,
+      "learning_rate": 2.7301097194780536e-05,
+      "loss": 0.6236,
+      "step": 2765
+    },
+    {
+      "epoch": 1.010041993792222,
+      "grad_norm": 0.758590042591095,
+      "learning_rate": 2.7287740163494153e-05,
+      "loss": 0.6147,
+      "step": 2766
+    },
+    {
+      "epoch": 1.0104071572028483,
+      "grad_norm": 0.7391862273216248,
+      "learning_rate": 2.7274379382957897e-05,
+      "loss": 0.5834,
+      "step": 2767
+    },
+    {
+      "epoch": 1.0107723206134744,
+      "grad_norm": 0.9706049561500549,
+      "learning_rate": 2.726101486004535e-05,
+      "loss": 0.5939,
+      "step": 2768
+    },
+    {
+      "epoch": 1.0111374840241008,
+      "grad_norm": 0.890767514705658,
+      "learning_rate": 2.724764660163203e-05,
+      "loss": 0.5643,
+      "step": 2769
+    },
+    {
+      "epoch": 1.011502647434727,
+      "grad_norm": 1.1014769077301025,
+      "learning_rate": 2.7234274614595353e-05,
+      "loss": 0.5817,
+      "step": 2770
+    },
+    {
+      "epoch": 1.0118678108453534,
+      "grad_norm": 0.8103865385055542,
+      "learning_rate": 2.722089890581469e-05,
+      "loss": 0.5939,
+      "step": 2771
+    },
+    {
+      "epoch": 1.0122329742559795,
+      "grad_norm": 1.0938106775283813,
+      "learning_rate": 2.7207519482171285e-05,
+      "loss": 0.6061,
+      "step": 2772
+    },
+    {
+      "epoch": 1.0125981376666058,
+      "grad_norm": 1.1287803649902344,
+      "learning_rate": 2.7194136350548332e-05,
+      "loss": 0.6071,
+      "step": 2773
+    },
+    {
+      "epoch": 1.012963301077232,
+      "grad_norm": 0.9278669357299805,
+      "learning_rate": 2.7180749517830912e-05,
+      "loss": 0.5779,
+      "step": 2774
+    },
+    {
+      "epoch": 1.0133284644878584,
+      "grad_norm": 0.9545652866363525,
+      "learning_rate": 2.7167358990906e-05,
+      "loss": 0.6039,
+      "step": 2775
+    },
+    {
+      "epoch": 1.0136936278984845,
+      "grad_norm": 0.8851467967033386,
+      "learning_rate": 2.7153964776662517e-05,
+      "loss": 0.5648,
+      "step": 2776
+    },
+    {
+      "epoch": 1.0140587913091108,
+      "grad_norm": 0.9389585852622986,
+      "learning_rate": 2.7140566881991213e-05,
+      "loss": 0.6088,
+      "step": 2777
+    },
+    {
+      "epoch": 1.0144239547197371,
+      "grad_norm": 1.1182056665420532,
+      "learning_rate": 2.712716531378478e-05,
+      "loss": 0.5398,
+      "step": 2778
+    },
+    {
+      "epoch": 1.0147891181303634,
+      "grad_norm": 0.7980071902275085,
+      "learning_rate": 2.7113760078937787e-05,
+      "loss": 0.6202,
+      "step": 2779
+    },
+    {
+      "epoch": 1.0151542815409895,
+      "grad_norm": 1.2338283061981201,
+      "learning_rate": 2.7100351184346694e-05,
+      "loss": 0.5673,
+      "step": 2780
+    },
+    {
+      "epoch": 1.0155194449516158,
+      "grad_norm": 0.8782559633255005,
+      "learning_rate": 2.708693863690984e-05,
+      "loss": 0.5843,
+      "step": 2781
+    },
+    {
+      "epoch": 1.0158846083622421,
+      "grad_norm": 1.2286125421524048,
+      "learning_rate": 2.707352244352744e-05,
+      "loss": 0.5917,
+      "step": 2782
+    },
+    {
+      "epoch": 1.0162497717728685,
+      "grad_norm": 0.7889975905418396,
+      "learning_rate": 2.7060102611101577e-05,
+      "loss": 0.6085,
+      "step": 2783
+    },
+    {
+      "epoch": 1.0166149351834946,
+      "grad_norm": 1.1056513786315918,
+      "learning_rate": 2.704667914653623e-05,
+      "loss": 0.6028,
+      "step": 2784
+    },
+    {
+      "epoch": 1.0169800985941209,
+      "grad_norm": 1.011452317237854,
+      "learning_rate": 2.7033252056737224e-05,
+      "loss": 0.6351,
+      "step": 2785
+    },
+    {
+      "epoch": 1.0173452620047472,
+      "grad_norm": 1.3591222763061523,
+      "learning_rate": 2.7019821348612265e-05,
+      "loss": 0.6138,
+      "step": 2786
+    },
+    {
+      "epoch": 1.0177104254153735,
+      "grad_norm": 1.2160711288452148,
+      "learning_rate": 2.7006387029070907e-05,
+      "loss": 0.5875,
+      "step": 2787
+    },
+    {
+      "epoch": 1.0180755888259996,
+      "grad_norm": 0.9456118941307068,
+      "learning_rate": 2.699294910502457e-05,
+      "loss": 0.5653,
+      "step": 2788
+    },
+    {
+      "epoch": 1.018440752236626,
+      "grad_norm": 0.6976199150085449,
+      "learning_rate": 2.6979507583386537e-05,
+      "loss": 0.6295,
+      "step": 2789
+    },
+    {
+      "epoch": 1.0188059156472522,
+      "grad_norm": 0.8012118935585022,
+      "learning_rate": 2.6966062471071914e-05,
+      "loss": 0.5898,
+      "step": 2790
+    },
+    {
+      "epoch": 1.0191710790578783,
+      "grad_norm": 0.8957707285881042,
+      "learning_rate": 2.6952613774997683e-05,
+      "loss": 0.6247,
+      "step": 2791
+    },
+    {
+      "epoch": 1.0195362424685046,
+      "grad_norm": 1.2239645719528198,
+      "learning_rate": 2.6939161502082653e-05,
+      "loss": 0.6144,
+      "step": 2792
+    },
+    {
+      "epoch": 1.019901405879131,
+      "grad_norm": 0.9017224907875061,
+      "learning_rate": 2.692570565924749e-05,
+      "loss": 0.6038,
+      "step": 2793
+    },
+    {
+      "epoch": 1.0202665692897572,
+      "grad_norm": 0.8203404545783997,
+      "learning_rate": 2.691224625341467e-05,
+      "loss": 0.5793,
+      "step": 2794
+    },
+    {
+      "epoch": 1.0206317327003833,
+      "grad_norm": 0.7525933980941772,
+      "learning_rate": 2.6898783291508524e-05,
+      "loss": 0.6202,
+      "step": 2795
+    },
+    {
+      "epoch": 1.0209968961110096,
+      "grad_norm": 0.9386183619499207,
+      "learning_rate": 2.6885316780455208e-05,
+      "loss": 0.6441,
+      "step": 2796
+    },
+    {
+      "epoch": 1.021362059521636,
+      "grad_norm": 0.9447058439254761,
+      "learning_rate": 2.6871846727182696e-05,
+      "loss": 0.6296,
+      "step": 2797
+    },
+    {
+      "epoch": 1.0217272229322623,
+      "grad_norm": 1.1461701393127441,
+      "learning_rate": 2.6858373138620794e-05,
+      "loss": 0.613,
+      "step": 2798
+    },
+    {
+      "epoch": 1.0220923863428883,
+      "grad_norm": 1.1132742166519165,
+      "learning_rate": 2.684489602170112e-05,
+      "loss": 0.5981,
+      "step": 2799
+    },
+    {
+      "epoch": 1.0224575497535147,
+      "grad_norm": 0.8169803023338318,
+      "learning_rate": 2.6831415383357113e-05,
+      "loss": 0.5908,
+      "step": 2800
+    },
+    {
+      "epoch": 1.022822713164141,
+      "grad_norm": 1.1900393962860107,
+      "learning_rate": 2.6817931230524016e-05,
+      "loss": 0.5781,
+      "step": 2801
+    },
+    {
+      "epoch": 1.0231878765747673,
+      "grad_norm": 0.9564783573150635,
+      "learning_rate": 2.6804443570138895e-05,
+      "loss": 0.5901,
+      "step": 2802
+    },
+    {
+      "epoch": 1.0235530399853934,
+      "grad_norm": 0.8402716517448425,
+      "learning_rate": 2.6790952409140597e-05,
+      "loss": 0.6099,
+      "step": 2803
+    },
+    {
+      "epoch": 1.0239182033960197,
+      "grad_norm": 0.952369749546051,
+      "learning_rate": 2.6777457754469788e-05,
+      "loss": 0.6042,
+      "step": 2804
+    },
+    {
+      "epoch": 1.024283366806646,
+      "grad_norm": 1.0410730838775635,
+      "learning_rate": 2.6763959613068933e-05,
+      "loss": 0.5655,
+      "step": 2805
+    },
+    {
+      "epoch": 1.0246485302172723,
+      "grad_norm": 1.1353759765625,
+      "learning_rate": 2.675045799188227e-05,
+      "loss": 0.5638,
+      "step": 2806
+    },
+    {
+      "epoch": 1.0250136936278984,
+      "grad_norm": 0.8030809164047241,
+      "learning_rate": 2.6736952897855856e-05,
+      "loss": 0.5865,
+      "step": 2807
+    },
+    {
+      "epoch": 1.0253788570385247,
+      "grad_norm": 0.9601109027862549,
+      "learning_rate": 2.6723444337937502e-05,
+      "loss": 0.617,
+      "step": 2808
+    },
+    {
+      "epoch": 1.025744020449151,
+      "grad_norm": 0.7185949087142944,
+      "learning_rate": 2.670993231907684e-05,
+      "loss": 0.5891,
+      "step": 2809
+    },
+    {
+      "epoch": 1.0261091838597773,
+      "grad_norm": 0.6851087212562561,
+      "learning_rate": 2.6696416848225256e-05,
+      "loss": 0.5972,
+      "step": 2810
+    },
+    {
+      "epoch": 1.0264743472704034,
+      "grad_norm": 1.0103535652160645,
+      "learning_rate": 2.6682897932335907e-05,
+      "loss": 0.6285,
+      "step": 2811
+    },
+    {
+      "epoch": 1.0268395106810297,
+      "grad_norm": 0.9711493253707886,
+      "learning_rate": 2.666937557836374e-05,
+      "loss": 0.5834,
+      "step": 2812
+    },
+    {
+      "epoch": 1.027204674091656,
+      "grad_norm": 0.7026539444923401,
+      "learning_rate": 2.665584979326546e-05,
+      "loss": 0.5943,
+      "step": 2813
+    },
+    {
+      "epoch": 1.0275698375022824,
+      "grad_norm": 0.8979662656784058,
+      "learning_rate": 2.6642320583999556e-05,
+      "loss": 0.6189,
+      "step": 2814
+    },
+    {
+      "epoch": 1.0279350009129085,
+      "grad_norm": 0.987153172492981,
+      "learning_rate": 2.662878795752624e-05,
+      "loss": 0.5969,
+      "step": 2815
+    },
+    {
+      "epoch": 1.0283001643235348,
+      "grad_norm": 0.8173993825912476,
+      "learning_rate": 2.6615251920807527e-05,
+      "loss": 0.6077,
+      "step": 2816
+    },
+    {
+      "epoch": 1.028665327734161,
+      "grad_norm": 0.9545841217041016,
+      "learning_rate": 2.6601712480807145e-05,
+      "loss": 0.5931,
+      "step": 2817
+    },
+    {
+      "epoch": 1.0290304911447874,
+      "grad_norm": 1.0173547267913818,
+      "learning_rate": 2.6588169644490608e-05,
+      "loss": 0.5833,
+      "step": 2818
+    },
+    {
+      "epoch": 1.0293956545554135,
+      "grad_norm": 0.948368489742279,
+      "learning_rate": 2.6574623418825152e-05,
+      "loss": 0.6158,
+      "step": 2819
+    },
+    {
+      "epoch": 1.0297608179660398,
+      "grad_norm": 0.8072060346603394,
+      "learning_rate": 2.656107381077977e-05,
+      "loss": 0.5652,
+      "step": 2820
+    },
+    {
+      "epoch": 1.0301259813766661,
+      "grad_norm": 0.9504181146621704,
+      "learning_rate": 2.6547520827325192e-05,
+      "loss": 0.5698,
+      "step": 2821
+    },
+    {
+      "epoch": 1.0304911447872924,
+      "grad_norm": 0.8799892067909241,
+      "learning_rate": 2.6533964475433886e-05,
+      "loss": 0.5751,
+      "step": 2822
+    },
+    {
+      "epoch": 1.0308563081979185,
+      "grad_norm": 1.240471363067627,
+      "learning_rate": 2.6520404762080048e-05,
+      "loss": 0.5967,
+      "step": 2823
+    },
+    {
+      "epoch": 1.0312214716085448,
+      "grad_norm": 0.7833214402198792,
+      "learning_rate": 2.650684169423961e-05,
+      "loss": 0.6486,
+      "step": 2824
+    },
+    {
+      "epoch": 1.0315866350191711,
+      "grad_norm": 1.1240694522857666,
+      "learning_rate": 2.649327527889022e-05,
+      "loss": 0.5483,
+      "step": 2825
+    },
+    {
+      "epoch": 1.0319517984297972,
+      "grad_norm": 0.8658917546272278,
+      "learning_rate": 2.6479705523011254e-05,
+      "loss": 0.6028,
+      "step": 2826
+    },
+    {
+      "epoch": 1.0323169618404235,
+      "grad_norm": 1.2881187200546265,
+      "learning_rate": 2.646613243358382e-05,
+      "loss": 0.5525,
+      "step": 2827
+    },
+    {
+      "epoch": 1.0326821252510499,
+      "grad_norm": 0.8781007528305054,
+      "learning_rate": 2.645255601759071e-05,
+      "loss": 0.5805,
+      "step": 2828
+    },
+    {
+      "epoch": 1.0330472886616762,
+      "grad_norm": 1.0463286638259888,
+      "learning_rate": 2.6438976282016465e-05,
+      "loss": 0.6097,
+      "step": 2829
+    },
+    {
+      "epoch": 1.0334124520723023,
+      "grad_norm": 0.8506337404251099,
+      "learning_rate": 2.642539323384729e-05,
+      "loss": 0.6221,
+      "step": 2830
+    },
+    {
+      "epoch": 1.0337776154829286,
+      "grad_norm": 1.032930612564087,
+      "learning_rate": 2.641180688007114e-05,
+      "loss": 0.5947,
+      "step": 2831
+    },
+    {
+      "epoch": 1.0341427788935549,
+      "grad_norm": 0.9797373414039612,
+      "learning_rate": 2.6398217227677636e-05,
+      "loss": 0.5948,
+      "step": 2832
+    },
+    {
+      "epoch": 1.0345079423041812,
+      "grad_norm": 1.3347671031951904,
+      "learning_rate": 2.638462428365811e-05,
+      "loss": 0.5745,
+      "step": 2833
+    },
+    {
+      "epoch": 1.0348731057148073,
+      "grad_norm": 0.9851678013801575,
+      "learning_rate": 2.63710280550056e-05,
+      "loss": 0.5747,
+      "step": 2834
+    },
+    {
+      "epoch": 1.0352382691254336,
+      "grad_norm": 0.8773434162139893,
+      "learning_rate": 2.63574285487148e-05,
+      "loss": 0.613,
+      "step": 2835
+    },
+    {
+      "epoch": 1.03560343253606,
+      "grad_norm": 0.829378068447113,
+      "learning_rate": 2.6343825771782125e-05,
+      "loss": 0.6033,
+      "step": 2836
+    },
+    {
+      "epoch": 1.0359685959466862,
+      "grad_norm": 0.990553617477417,
+      "learning_rate": 2.633021973120565e-05,
+      "loss": 0.549,
+      "step": 2837
+    },
+    {
+      "epoch": 1.0363337593573123,
+      "grad_norm": 1.1182706356048584,
+      "learning_rate": 2.6316610433985136e-05,
+      "loss": 0.6159,
+      "step": 2838
+    },
+    {
+      "epoch": 1.0366989227679386,
+      "grad_norm": 0.887545108795166,
+      "learning_rate": 2.6302997887122024e-05,
+      "loss": 0.5685,
+      "step": 2839
+    },
+    {
+      "epoch": 1.037064086178565,
+      "grad_norm": 0.770965576171875,
+      "learning_rate": 2.6289382097619426e-05,
+      "loss": 0.6278,
+      "step": 2840
+    },
+    {
+      "epoch": 1.0374292495891912,
+      "grad_norm": 0.8624748587608337,
+      "learning_rate": 2.6275763072482116e-05,
+      "loss": 0.5809,
+      "step": 2841
+    },
+    {
+      "epoch": 1.0377944129998173,
+      "grad_norm": 0.9106437563896179,
+      "learning_rate": 2.6262140818716537e-05,
+      "loss": 0.5551,
+      "step": 2842
+    },
+    {
+      "epoch": 1.0381595764104437,
+      "grad_norm": 0.9461293816566467,
+      "learning_rate": 2.6248515343330784e-05,
+      "loss": 0.5948,
+      "step": 2843
+    },
+    {
+      "epoch": 1.03852473982107,
+      "grad_norm": 0.9623094201087952,
+      "learning_rate": 2.6234886653334632e-05,
+      "loss": 0.5931,
+      "step": 2844
+    },
+    {
+      "epoch": 1.0388899032316963,
+      "grad_norm": 1.062417984008789,
+      "learning_rate": 2.622125475573948e-05,
+      "loss": 0.5694,
+      "step": 2845
+    },
+    {
+      "epoch": 1.0392550666423224,
+      "grad_norm": 0.860431969165802,
+      "learning_rate": 2.6207619657558404e-05,
+      "loss": 0.5704,
+      "step": 2846
+    },
+    {
+      "epoch": 1.0396202300529487,
+      "grad_norm": 0.764029324054718,
+      "learning_rate": 2.6193981365806108e-05,
+      "loss": 0.5839,
+      "step": 2847
+    },
+    {
+      "epoch": 1.039985393463575,
+      "grad_norm": 0.9730731844902039,
+      "learning_rate": 2.618033988749895e-05,
+      "loss": 0.6255,
+      "step": 2848
+    },
+    {
+      "epoch": 1.0403505568742013,
+      "grad_norm": 0.9711384773254395,
+      "learning_rate": 2.6166695229654923e-05,
+      "loss": 0.5496,
+      "step": 2849
+    },
+    {
+      "epoch": 1.0407157202848274,
+      "grad_norm": 1.0858523845672607,
+      "learning_rate": 2.6153047399293653e-05,
+      "loss": 0.5913,
+      "step": 2850
+    },
+    {
+      "epoch": 1.0410808836954537,
+      "grad_norm": 0.7481245994567871,
+      "learning_rate": 2.6139396403436404e-05,
+      "loss": 0.6019,
+      "step": 2851
+    },
+    {
+      "epoch": 1.04144604710608,
+      "grad_norm": 0.7096604704856873,
+      "learning_rate": 2.612574224910606e-05,
+      "loss": 0.6191,
+      "step": 2852
+    },
+    {
+      "epoch": 1.0418112105167063,
+      "grad_norm": 0.9994904398918152,
+      "learning_rate": 2.6112084943327146e-05,
+      "loss": 0.5891,
+      "step": 2853
+    },
+    {
+      "epoch": 1.0421763739273324,
+      "grad_norm": 0.9799776673316956,
+      "learning_rate": 2.609842449312578e-05,
+      "loss": 0.6093,
+      "step": 2854
+    },
+    {
+      "epoch": 1.0425415373379587,
+      "grad_norm": 0.7989087104797363,
+      "learning_rate": 2.608476090552974e-05,
+      "loss": 0.6325,
+      "step": 2855
+    },
+    {
+      "epoch": 1.042906700748585,
+      "grad_norm": 0.8211569786071777,
+      "learning_rate": 2.6071094187568374e-05,
+      "loss": 0.5905,
+      "step": 2856
+    },
+    {
+      "epoch": 1.0432718641592111,
+      "grad_norm": 1.1618964672088623,
+      "learning_rate": 2.6057424346272667e-05,
+      "loss": 0.5789,
+      "step": 2857
+    },
+    {
+      "epoch": 1.0436370275698374,
+      "grad_norm": 0.812633752822876,
+      "learning_rate": 2.6043751388675205e-05,
+      "loss": 0.5911,
+      "step": 2858
+    },
+    {
+      "epoch": 1.0440021909804638,
+      "grad_norm": 1.2491474151611328,
+      "learning_rate": 2.6030075321810166e-05,
+      "loss": 0.5936,
+      "step": 2859
+    },
+    {
+      "epoch": 1.04436735439109,
+      "grad_norm": 1.066473126411438,
+      "learning_rate": 2.6016396152713355e-05,
+      "loss": 0.5812,
+      "step": 2860
+    },
+    {
+      "epoch": 1.0447325178017162,
+      "grad_norm": 0.8640859127044678,
+      "learning_rate": 2.600271388842214e-05,
+      "loss": 0.5997,
+      "step": 2861
+    },
+    {
+      "epoch": 1.0450976812123425,
+      "grad_norm": 1.0497597455978394,
+      "learning_rate": 2.5989028535975508e-05,
+      "loss": 0.6148,
+      "step": 2862
+    },
+    {
+      "epoch": 1.0454628446229688,
+      "grad_norm": 0.8977965116500854,
+      "learning_rate": 2.5975340102414023e-05,
+      "loss": 0.6049,
+      "step": 2863
+    },
+    {
+      "epoch": 1.045828008033595,
+      "grad_norm": 1.0095138549804688,
+      "learning_rate": 2.5961648594779824e-05,
+      "loss": 0.5774,
+      "step": 2864
+    },
+    {
+      "epoch": 1.0461931714442212,
+      "grad_norm": 0.7467365264892578,
+      "learning_rate": 2.594795402011665e-05,
+      "loss": 0.6265,
+      "step": 2865
+    },
+    {
+      "epoch": 1.0465583348548475,
+      "grad_norm": 0.9556071758270264,
+      "learning_rate": 2.5934256385469807e-05,
+      "loss": 0.5911,
+      "step": 2866
+    },
+    {
+      "epoch": 1.0469234982654738,
+      "grad_norm": 0.7638899683952332,
+      "learning_rate": 2.5920555697886184e-05,
+      "loss": 0.5836,
+      "step": 2867
+    },
+    {
+      "epoch": 1.0472886616761001,
+      "grad_norm": 0.7609127759933472,
+      "learning_rate": 2.590685196441423e-05,
+      "loss": 0.6416,
+      "step": 2868
+    },
+    {
+      "epoch": 1.0476538250867262,
+      "grad_norm": 1.2816520929336548,
+      "learning_rate": 2.589314519210397e-05,
+      "loss": 0.62,
+      "step": 2869
+    },
+    {
+      "epoch": 1.0480189884973525,
+      "grad_norm": 0.920642614364624,
+      "learning_rate": 2.5879435388006986e-05,
+      "loss": 0.5876,
+      "step": 2870
+    },
+    {
+      "epoch": 1.0483841519079788,
+      "grad_norm": 1.1667977571487427,
+      "learning_rate": 2.586572255917642e-05,
+      "loss": 0.5988,
+      "step": 2871
+    },
+    {
+      "epoch": 1.0487493153186052,
+      "grad_norm": 1.0526721477508545,
+      "learning_rate": 2.5852006712666975e-05,
+      "loss": 0.5389,
+      "step": 2872
+    },
+    {
+      "epoch": 1.0491144787292312,
+      "grad_norm": 0.8131223917007446,
+      "learning_rate": 2.5838287855534895e-05,
+      "loss": 0.5482,
+      "step": 2873
+    },
+    {
+      "epoch": 1.0494796421398576,
+      "grad_norm": 0.8092973232269287,
+      "learning_rate": 2.5824565994838e-05,
+      "loss": 0.5826,
+      "step": 2874
+    },
+    {
+      "epoch": 1.0498448055504839,
+      "grad_norm": 0.751356303691864,
+      "learning_rate": 2.581084113763562e-05,
+      "loss": 0.6341,
+      "step": 2875
+    },
+    {
+      "epoch": 1.0502099689611102,
+      "grad_norm": 1.6021597385406494,
+      "learning_rate": 2.5797113290988655e-05,
+      "loss": 0.6091,
+      "step": 2876
+    },
+    {
+      "epoch": 1.0505751323717363,
+      "grad_norm": 1.7192530632019043,
+      "learning_rate": 2.5783382461959523e-05,
+      "loss": 0.5354,
+      "step": 2877
+    },
+    {
+      "epoch": 1.0509402957823626,
+      "grad_norm": 1.4987651109695435,
+      "learning_rate": 2.576964865761218e-05,
+      "loss": 0.5722,
+      "step": 2878
+    },
+    {
+      "epoch": 1.051305459192989,
+      "grad_norm": 0.7340034246444702,
+      "learning_rate": 2.575591188501213e-05,
+      "loss": 0.5746,
+      "step": 2879
+    },
+    {
+      "epoch": 1.0516706226036152,
+      "grad_norm": 1.107073426246643,
+      "learning_rate": 2.574217215122639e-05,
+      "loss": 0.6007,
+      "step": 2880
+    },
+    {
+      "epoch": 1.0520357860142413,
+      "grad_norm": 0.7700313329696655,
+      "learning_rate": 2.5728429463323487e-05,
+      "loss": 0.5551,
+      "step": 2881
+    },
+    {
+      "epoch": 1.0524009494248676,
+      "grad_norm": 0.7817539572715759,
+      "learning_rate": 2.57146838283735e-05,
+      "loss": 0.6017,
+      "step": 2882
+    },
+    {
+      "epoch": 1.052766112835494,
+      "grad_norm": 0.7932517528533936,
+      "learning_rate": 2.5700935253447998e-05,
+      "loss": 0.5914,
+      "step": 2883
+    },
+    {
+      "epoch": 1.0531312762461202,
+      "grad_norm": 0.8778212666511536,
+      "learning_rate": 2.5687183745620078e-05,
+      "loss": 0.5986,
+      "step": 2884
+    },
+    {
+      "epoch": 1.0534964396567463,
+      "grad_norm": 0.8803035616874695,
+      "learning_rate": 2.567342931196432e-05,
+      "loss": 0.6064,
+      "step": 2885
+    },
+    {
+      "epoch": 1.0538616030673726,
+      "grad_norm": 0.8971658945083618,
+      "learning_rate": 2.5659671959556848e-05,
+      "loss": 0.6267,
+      "step": 2886
+    },
+    {
+      "epoch": 1.054226766477999,
+      "grad_norm": 0.8893007040023804,
+      "learning_rate": 2.5645911695475264e-05,
+      "loss": 0.6143,
+      "step": 2887
+    },
+    {
+      "epoch": 1.054591929888625,
+      "grad_norm": 0.6497613787651062,
+      "learning_rate": 2.563214852679867e-05,
+      "loss": 0.6102,
+      "step": 2888
+    },
+    {
+      "epoch": 1.0549570932992514,
+      "grad_norm": 0.8695589303970337,
+      "learning_rate": 2.5618382460607666e-05,
+      "loss": 0.5969,
+      "step": 2889
+    },
+    {
+      "epoch": 1.0553222567098777,
+      "grad_norm": 1.3337715864181519,
+      "learning_rate": 2.5604613503984327e-05,
+      "loss": 0.6025,
+      "step": 2890
+    },
+    {
+      "epoch": 1.055687420120504,
+      "grad_norm": 0.9800289273262024,
+      "learning_rate": 2.559084166401224e-05,
+      "loss": 0.6266,
+      "step": 2891
+    },
+    {
+      "epoch": 1.05605258353113,
+      "grad_norm": 0.8625394105911255,
+      "learning_rate": 2.557706694777647e-05,
+      "loss": 0.6205,
+      "step": 2892
+    },
+    {
+      "epoch": 1.0564177469417564,
+      "grad_norm": 1.1771821975708008,
+      "learning_rate": 2.5563289362363547e-05,
+      "loss": 0.5863,
+      "step": 2893
+    },
+    {
+      "epoch": 1.0567829103523827,
+      "grad_norm": 0.7348043918609619,
+      "learning_rate": 2.554950891486149e-05,
+      "loss": 0.5781,
+      "step": 2894
+    },
+    {
+      "epoch": 1.057148073763009,
+      "grad_norm": 0.978421151638031,
+      "learning_rate": 2.5535725612359778e-05,
+      "loss": 0.6074,
+      "step": 2895
+    },
+    {
+      "epoch": 1.057513237173635,
+      "grad_norm": 1.1634827852249146,
+      "learning_rate": 2.5521939461949384e-05,
+      "loss": 0.5952,
+      "step": 2896
+    },
+    {
+      "epoch": 1.0578784005842614,
+      "grad_norm": 0.711153507232666,
+      "learning_rate": 2.5508150470722708e-05,
+      "loss": 0.6351,
+      "step": 2897
+    },
+    {
+      "epoch": 1.0582435639948877,
+      "grad_norm": 0.962415874004364,
+      "learning_rate": 2.5494358645773646e-05,
+      "loss": 0.6055,
+      "step": 2898
+    },
+    {
+      "epoch": 1.058608727405514,
+      "grad_norm": 1.2400012016296387,
+      "learning_rate": 2.548056399419754e-05,
+      "loss": 0.6092,
+      "step": 2899
+    },
+    {
+      "epoch": 1.0589738908161401,
+      "grad_norm": 0.7230969071388245,
+      "learning_rate": 2.5466766523091172e-05,
+      "loss": 0.5801,
+      "step": 2900
+    },
+    {
+      "epoch": 1.0593390542267664,
+      "grad_norm": 0.9523801803588867,
+      "learning_rate": 2.5452966239552802e-05,
+      "loss": 0.5948,
+      "step": 2901
+    },
+    {
+      "epoch": 1.0597042176373928,
+      "grad_norm": 0.8902936577796936,
+      "learning_rate": 2.543916315068211e-05,
+      "loss": 0.5399,
+      "step": 2902
+    },
+    {
+      "epoch": 1.060069381048019,
+      "grad_norm": 1.0887128114700317,
+      "learning_rate": 2.5425357263580246e-05,
+      "loss": 0.6105,
+      "step": 2903
+    },
+    {
+      "epoch": 1.0604345444586452,
+      "grad_norm": 0.8181251287460327,
+      "learning_rate": 2.5411548585349772e-05,
+      "loss": 0.5854,
+      "step": 2904
+    },
+    {
+      "epoch": 1.0607997078692715,
+      "grad_norm": 0.8453181982040405,
+      "learning_rate": 2.5397737123094697e-05,
+      "loss": 0.5965,
+      "step": 2905
+    },
+    {
+      "epoch": 1.0611648712798978,
+      "grad_norm": 1.112425446510315,
+      "learning_rate": 2.5383922883920476e-05,
+      "loss": 0.5735,
+      "step": 2906
+    },
+    {
+      "epoch": 1.061530034690524,
+      "grad_norm": 0.9731928110122681,
+      "learning_rate": 2.5370105874933972e-05,
+      "loss": 0.5959,
+      "step": 2907
+    },
+    {
+      "epoch": 1.0618951981011502,
+      "grad_norm": 0.917127251625061,
+      "learning_rate": 2.5356286103243485e-05,
+      "loss": 0.5693,
+      "step": 2908
+    },
+    {
+      "epoch": 1.0622603615117765,
+      "grad_norm": 1.1035484075546265,
+      "learning_rate": 2.5342463575958737e-05,
+      "loss": 0.559,
+      "step": 2909
+    },
+    {
+      "epoch": 1.0626255249224028,
+      "grad_norm": 1.1218947172164917,
+      "learning_rate": 2.5328638300190856e-05,
+      "loss": 0.6578,
+      "step": 2910
+    },
+    {
+      "epoch": 1.0629906883330291,
+      "grad_norm": 0.6395195722579956,
+      "learning_rate": 2.531481028305239e-05,
+      "loss": 0.6437,
+      "step": 2911
+    },
+    {
+      "epoch": 1.0633558517436552,
+      "grad_norm": 0.935915470123291,
+      "learning_rate": 2.5300979531657305e-05,
+      "loss": 0.5709,
+      "step": 2912
+    },
+    {
+      "epoch": 1.0637210151542815,
+      "grad_norm": 0.835087239742279,
+      "learning_rate": 2.528714605312097e-05,
+      "loss": 0.5789,
+      "step": 2913
+    },
+    {
+      "epoch": 1.0640861785649078,
+      "grad_norm": 2.0276432037353516,
+      "learning_rate": 2.5273309854560148e-05,
+      "loss": 0.5714,
+      "step": 2914
+    },
+    {
+      "epoch": 1.0644513419755341,
+      "grad_norm": 0.7264708876609802,
+      "learning_rate": 2.525947094309301e-05,
+      "loss": 0.5891,
+      "step": 2915
+    },
+    {
+      "epoch": 1.0648165053861602,
+      "grad_norm": 0.9568043351173401,
+      "learning_rate": 2.5245629325839125e-05,
+      "loss": 0.569,
+      "step": 2916
+    },
+    {
+      "epoch": 1.0651816687967866,
+      "grad_norm": 1.0530427694320679,
+      "learning_rate": 2.5231785009919437e-05,
+      "loss": 0.6086,
+      "step": 2917
+    },
+    {
+      "epoch": 1.0655468322074129,
+      "grad_norm": 0.8962780833244324,
+      "learning_rate": 2.521793800245631e-05,
+      "loss": 0.5636,
+      "step": 2918
+    },
+    {
+      "epoch": 1.0659119956180392,
+      "grad_norm": 1.098936676979065,
+      "learning_rate": 2.5204088310573455e-05,
+      "loss": 0.5918,
+      "step": 2919
+    },
+    {
+      "epoch": 1.0662771590286653,
+      "grad_norm": 0.9142637252807617,
+      "learning_rate": 2.5190235941395996e-05,
+      "loss": 0.532,
+      "step": 2920
+    },
+    {
+      "epoch": 1.0666423224392916,
+      "grad_norm": 0.8808488845825195,
+      "learning_rate": 2.5176380902050418e-05,
+      "loss": 0.6227,
+      "step": 2921
+    },
+    {
+      "epoch": 1.067007485849918,
+      "grad_norm": 1.0145164728164673,
+      "learning_rate": 2.5162523199664583e-05,
+      "loss": 0.6177,
+      "step": 2922
+    },
+    {
+      "epoch": 1.0673726492605442,
+      "grad_norm": 0.912358820438385,
+      "learning_rate": 2.5148662841367718e-05,
+      "loss": 0.5798,
+      "step": 2923
+    },
+    {
+      "epoch": 1.0677378126711703,
+      "grad_norm": 0.8597058057785034,
+      "learning_rate": 2.5134799834290417e-05,
+      "loss": 0.6071,
+      "step": 2924
+    },
+    {
+      "epoch": 1.0681029760817966,
+      "grad_norm": 0.7203514575958252,
+      "learning_rate": 2.512093418556466e-05,
+      "loss": 0.5608,
+      "step": 2925
+    },
+    {
+      "epoch": 1.068468139492423,
+      "grad_norm": 1.068557858467102,
+      "learning_rate": 2.510706590232374e-05,
+      "loss": 0.5716,
+      "step": 2926
+    },
+    {
+      "epoch": 1.068833302903049,
+      "grad_norm": 1.191929578781128,
+      "learning_rate": 2.5093194991702362e-05,
+      "loss": 0.584,
+      "step": 2927
+    },
+    {
+      "epoch": 1.0691984663136753,
+      "grad_norm": 0.7341733574867249,
+      "learning_rate": 2.5079321460836528e-05,
+      "loss": 0.6083,
+      "step": 2928
+    },
+    {
+      "epoch": 1.0695636297243016,
+      "grad_norm": 0.8296530842781067,
+      "learning_rate": 2.5065445316863627e-05,
+      "loss": 0.5883,
+      "step": 2929
+    },
+    {
+      "epoch": 1.069928793134928,
+      "grad_norm": 0.8199982047080994,
+      "learning_rate": 2.5051566566922377e-05,
+      "loss": 0.6203,
+      "step": 2930
+    },
+    {
+      "epoch": 1.070293956545554,
+      "grad_norm": 1.316584587097168,
+      "learning_rate": 2.503768521815283e-05,
+      "loss": 0.5904,
+      "step": 2931
+    },
+    {
+      "epoch": 1.0706591199561803,
+      "grad_norm": 0.8966913819313049,
+      "learning_rate": 2.5023801277696393e-05,
+      "loss": 0.6328,
+      "step": 2932
+    },
+    {
+      "epoch": 1.0710242833668067,
+      "grad_norm": 0.8157796859741211,
+      "learning_rate": 2.5009914752695785e-05,
+      "loss": 0.6207,
+      "step": 2933
+    },
+    {
+      "epoch": 1.071389446777433,
+      "grad_norm": 0.7004765868186951,
+      "learning_rate": 2.4996025650295072e-05,
+      "loss": 0.6138,
+      "step": 2934
+    },
+    {
+      "epoch": 1.071754610188059,
+      "grad_norm": 0.9260857701301575,
+      "learning_rate": 2.4982133977639644e-05,
+      "loss": 0.6054,
+      "step": 2935
+    },
+    {
+      "epoch": 1.0721197735986854,
+      "grad_norm": 0.8620515465736389,
+      "learning_rate": 2.4968239741876205e-05,
+      "loss": 0.5885,
+      "step": 2936
+    },
+    {
+      "epoch": 1.0724849370093117,
+      "grad_norm": 0.7664886713027954,
+      "learning_rate": 2.4954342950152786e-05,
+      "loss": 0.5781,
+      "step": 2937
+    },
+    {
+      "epoch": 1.072850100419938,
+      "grad_norm": 1.2562214136123657,
+      "learning_rate": 2.4940443609618713e-05,
+      "loss": 0.6151,
+      "step": 2938
+    },
+    {
+      "epoch": 1.073215263830564,
+      "grad_norm": 1.0961320400238037,
+      "learning_rate": 2.4926541727424663e-05,
+      "loss": 0.559,
+      "step": 2939
+    },
+    {
+      "epoch": 1.0735804272411904,
+      "grad_norm": 0.9252917170524597,
+      "learning_rate": 2.4912637310722575e-05,
+      "loss": 0.6067,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0739455906518167,
+      "grad_norm": 0.8604303002357483,
+      "learning_rate": 2.4898730366665724e-05,
+      "loss": 0.5856,
+      "step": 2941
+    },
+    {
+      "epoch": 1.074310754062443,
+      "grad_norm": 0.792181670665741,
+      "learning_rate": 2.488482090240868e-05,
+      "loss": 0.6186,
+      "step": 2942
+    },
+    {
+      "epoch": 1.0746759174730691,
+      "grad_norm": 1.0725411176681519,
+      "learning_rate": 2.487090892510729e-05,
+      "loss": 0.5721,
+      "step": 2943
+    },
+    {
+      "epoch": 1.0750410808836954,
+      "grad_norm": 0.7707020044326782,
+      "learning_rate": 2.4856994441918718e-05,
+      "loss": 0.6124,
+      "step": 2944
+    },
+    {
+      "epoch": 1.0754062442943217,
+      "grad_norm": 0.9772723913192749,
+      "learning_rate": 2.4843077460001405e-05,
+      "loss": 0.5997,
+      "step": 2945
+    },
+    {
+      "epoch": 1.075771407704948,
+      "grad_norm": 2.125711441040039,
+      "learning_rate": 2.482915798651507e-05,
+      "loss": 0.6111,
+      "step": 2946
+    },
+    {
+      "epoch": 1.0761365711155741,
+      "grad_norm": 0.6215631365776062,
+      "learning_rate": 2.481523602862075e-05,
+      "loss": 0.6021,
+      "step": 2947
+    },
+    {
+      "epoch": 1.0765017345262005,
+      "grad_norm": 0.9591418504714966,
+      "learning_rate": 2.48013115934807e-05,
+      "loss": 0.5849,
+      "step": 2948
+    },
+    {
+      "epoch": 1.0768668979368268,
+      "grad_norm": 1.1631020307540894,
+      "learning_rate": 2.4787384688258514e-05,
+      "loss": 0.6129,
+      "step": 2949
+    },
+    {
+      "epoch": 1.077232061347453,
+      "grad_norm": 0.7211815714836121,
+      "learning_rate": 2.4773455320119005e-05,
+      "loss": 0.5873,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0777798064633923,
+      "grad_norm": 1.2566742897033691,
+      "learning_rate": 2.475952349622828e-05,
+      "loss": 0.6654,
+      "step": 2951
+    },
+    {
+      "epoch": 1.0781449698740186,
+      "grad_norm": 1.202701449394226,
+      "learning_rate": 2.474558922375371e-05,
+      "loss": 0.6309,
+      "step": 2952
+    },
+    {
+      "epoch": 1.078510133284645,
+      "grad_norm": 1.013235092163086,
+      "learning_rate": 2.4731652509863904e-05,
+      "loss": 0.6301,
+      "step": 2953
+    },
+    {
+      "epoch": 1.078875296695271,
+      "grad_norm": 1.1638104915618896,
+      "learning_rate": 2.471771336172876e-05,
+      "loss": 0.631,
+      "step": 2954
+    },
+    {
+      "epoch": 1.0792404601058974,
+      "grad_norm": 3.4340641498565674,
+      "learning_rate": 2.4703771786519392e-05,
+      "loss": 0.6566,
+      "step": 2955
+    },
+    {
+      "epoch": 1.0796056235165237,
+      "grad_norm": 0.9713271260261536,
+      "learning_rate": 2.4689827791408198e-05,
+      "loss": 0.65,
+      "step": 2956
+    },
+    {
+      "epoch": 1.07997078692715,
+      "grad_norm": 1.1270604133605957,
+      "learning_rate": 2.4675881383568797e-05,
+      "loss": 0.6987,
+      "step": 2957
+    },
+    {
+      "epoch": 1.080335950337776,
+      "grad_norm": 0.7504619359970093,
+      "learning_rate": 2.4661932570176047e-05,
+      "loss": 0.6166,
+      "step": 2958
+    },
+    {
+      "epoch": 1.0807011137484024,
+      "grad_norm": 0.7504952549934387,
+      "learning_rate": 2.464798135840607e-05,
+      "loss": 0.642,
+      "step": 2959
+    },
+    {
+      "epoch": 1.0810662771590287,
+      "grad_norm": 0.9410606622695923,
+      "learning_rate": 2.4634027755436192e-05,
+      "loss": 0.6299,
+      "step": 2960
+    },
+    {
+      "epoch": 1.081431440569655,
+      "grad_norm": 0.9131185412406921,
+      "learning_rate": 2.4620071768444985e-05,
+      "loss": 0.6688,
+      "step": 2961
+    },
+    {
+      "epoch": 1.081796603980281,
+      "grad_norm": 0.9587386846542358,
+      "learning_rate": 2.4606113404612244e-05,
+      "loss": 0.6307,
+      "step": 2962
+    },
+    {
+      "epoch": 1.0821617673909074,
+      "grad_norm": 0.9157978296279907,
+      "learning_rate": 2.4592152671118993e-05,
+      "loss": 0.6702,
+      "step": 2963
+    },
+    {
+      "epoch": 1.0825269308015337,
+      "grad_norm": 0.8683468699455261,
+      "learning_rate": 2.4578189575147465e-05,
+      "loss": 0.6246,
+      "step": 2964
+    },
+    {
+      "epoch": 1.08289209421216,
+      "grad_norm": 0.956314206123352,
+      "learning_rate": 2.4564224123881103e-05,
+      "loss": 0.6446,
+      "step": 2965
+    },
+    {
+      "epoch": 1.0832572576227861,
+      "grad_norm": 1.0378687381744385,
+      "learning_rate": 2.4550256324504594e-05,
+      "loss": 0.6288,
+      "step": 2966
+    },
+    {
+      "epoch": 1.0836224210334124,
+      "grad_norm": 0.9109642505645752,
+      "learning_rate": 2.4536286184203783e-05,
+      "loss": 0.6291,
+      "step": 2967
+    },
+    {
+      "epoch": 1.0839875844440388,
+      "grad_norm": 0.8259170055389404,
+      "learning_rate": 2.4522313710165765e-05,
+      "loss": 0.6749,
+      "step": 2968
+    },
+    {
+      "epoch": 1.084352747854665,
+      "grad_norm": 0.7047184705734253,
+      "learning_rate": 2.4508338909578817e-05,
+      "loss": 0.6255,
+      "step": 2969
+    },
+    {
+      "epoch": 1.0847179112652912,
+      "grad_norm": 0.8270784020423889,
+      "learning_rate": 2.4494361789632405e-05,
+      "loss": 0.6633,
+      "step": 2970
+    },
+    {
+      "epoch": 1.0850830746759175,
+      "grad_norm": 0.8580787181854248,
+      "learning_rate": 2.4480382357517195e-05,
+      "loss": 0.6518,
+      "step": 2971
+    },
+    {
+      "epoch": 1.0854482380865438,
+      "grad_norm": 0.7827367782592773,
+      "learning_rate": 2.4466400620425054e-05,
+      "loss": 0.6218,
+      "step": 2972
+    },
+    {
+      "epoch": 1.08581340149717,
+      "grad_norm": 0.8759443163871765,
+      "learning_rate": 2.4452416585549018e-05,
+      "loss": 0.627,
+      "step": 2973
+    },
+    {
+      "epoch": 1.0861785649077962,
+      "grad_norm": 0.9673501253128052,
+      "learning_rate": 2.443843026008331e-05,
+      "loss": 0.6233,
+      "step": 2974
+    },
+    {
+      "epoch": 1.0865437283184225,
+      "grad_norm": 0.932388186454773,
+      "learning_rate": 2.4424441651223343e-05,
+      "loss": 0.6277,
+      "step": 2975
+    },
+    {
+      "epoch": 1.0869088917290488,
+      "grad_norm": 0.7365838289260864,
+      "learning_rate": 2.4410450766165688e-05,
+      "loss": 0.5957,
+      "step": 2976
+    },
+    {
+      "epoch": 1.087274055139675,
+      "grad_norm": 1.4103069305419922,
+      "learning_rate": 2.439645761210809e-05,
+      "loss": 0.661,
+      "step": 2977
+    },
+    {
+      "epoch": 1.0876392185503012,
+      "grad_norm": 0.9331754446029663,
+      "learning_rate": 2.438246219624947e-05,
+      "loss": 0.652,
+      "step": 2978
+    },
+    {
+      "epoch": 1.0880043819609275,
+      "grad_norm": 0.980294406414032,
+      "learning_rate": 2.4368464525789905e-05,
+      "loss": 0.645,
+      "step": 2979
+    },
+    {
+      "epoch": 1.0883695453715538,
+      "grad_norm": 0.8242958784103394,
+      "learning_rate": 2.435446460793064e-05,
+      "loss": 0.6616,
+      "step": 2980
+    },
+    {
+      "epoch": 1.08873470878218,
+      "grad_norm": 0.8177999258041382,
+      "learning_rate": 2.4340462449874063e-05,
+      "loss": 0.5975,
+      "step": 2981
+    },
+    {
+      "epoch": 1.0890998721928062,
+      "grad_norm": 1.0282783508300781,
+      "learning_rate": 2.4326458058823735e-05,
+      "loss": 0.6388,
+      "step": 2982
+    },
+    {
+      "epoch": 1.0894650356034326,
+      "grad_norm": 1.0401636362075806,
+      "learning_rate": 2.4312451441984344e-05,
+      "loss": 0.6249,
+      "step": 2983
+    },
+    {
+      "epoch": 1.0898301990140589,
+      "grad_norm": 0.8768249154090881,
+      "learning_rate": 2.429844260656173e-05,
+      "loss": 0.6978,
+      "step": 2984
+    },
+    {
+      "epoch": 1.090195362424685,
+      "grad_norm": 1.2320053577423096,
+      "learning_rate": 2.4284431559762888e-05,
+      "loss": 0.6473,
+      "step": 2985
+    },
+    {
+      "epoch": 1.0905605258353113,
+      "grad_norm": 1.0827654600143433,
+      "learning_rate": 2.4270418308795923e-05,
+      "loss": 0.6978,
+      "step": 2986
+    },
+    {
+      "epoch": 1.0909256892459376,
+      "grad_norm": 0.8920133113861084,
+      "learning_rate": 2.4256402860870107e-05,
+      "loss": 0.6119,
+      "step": 2987
+    },
+    {
+      "epoch": 1.091290852656564,
+      "grad_norm": 0.7745959758758545,
+      "learning_rate": 2.424238522319581e-05,
+      "loss": 0.6356,
+      "step": 2988
+    },
+    {
+      "epoch": 1.09165601606719,
+      "grad_norm": 0.8089456558227539,
+      "learning_rate": 2.4228365402984562e-05,
+      "loss": 0.6567,
+      "step": 2989
+    },
+    {
+      "epoch": 1.0920211794778163,
+      "grad_norm": 1.0786186456680298,
+      "learning_rate": 2.4214343407448984e-05,
+      "loss": 0.6417,
+      "step": 2990
+    },
+    {
+      "epoch": 1.0923863428884426,
+      "grad_norm": 0.6472024321556091,
+      "learning_rate": 2.4200319243802826e-05,
+      "loss": 0.6252,
+      "step": 2991
+    },
+    {
+      "epoch": 1.092751506299069,
+      "grad_norm": 1.3685519695281982,
+      "learning_rate": 2.4186292919260975e-05,
+      "loss": 0.6237,
+      "step": 2992
+    },
+    {
+      "epoch": 1.093116669709695,
+      "grad_norm": 0.9059715867042542,
+      "learning_rate": 2.41722644410394e-05,
+      "loss": 0.5992,
+      "step": 2993
+    },
+    {
+      "epoch": 1.0934818331203213,
+      "grad_norm": 1.2271138429641724,
+      "learning_rate": 2.4158233816355185e-05,
+      "loss": 0.626,
+      "step": 2994
+    },
+    {
+      "epoch": 1.0938469965309476,
+      "grad_norm": 1.1302318572998047,
+      "learning_rate": 2.4144201052426543e-05,
+      "loss": 0.6304,
+      "step": 2995
+    },
+    {
+      "epoch": 1.094212159941574,
+      "grad_norm": 1.024825930595398,
+      "learning_rate": 2.413016615647275e-05,
+      "loss": 0.6524,
+      "step": 2996
+    },
+    {
+      "epoch": 1.0945773233522,
+      "grad_norm": 0.9111654758453369,
+      "learning_rate": 2.4116129135714208e-05,
+      "loss": 0.6465,
+      "step": 2997
+    },
+    {
+      "epoch": 1.0949424867628264,
+      "grad_norm": 0.8728148341178894,
+      "learning_rate": 2.410208999737239e-05,
+      "loss": 0.6064,
+      "step": 2998
+    },
+    {
+      "epoch": 1.0953076501734527,
+      "grad_norm": 1.0016210079193115,
+      "learning_rate": 2.408804874866987e-05,
+      "loss": 0.6364,
+      "step": 2999
+    },
+    {
+      "epoch": 1.095672813584079,
+      "grad_norm": 0.9198909401893616,
+      "learning_rate": 2.4074005396830322e-05,
+      "loss": 0.6321,
+      "step": 3000
+    },
+    {
+      "epoch": 1.096037976994705,
+      "grad_norm": 1.169684648513794,
+      "learning_rate": 2.4059959949078467e-05,
+      "loss": 0.6309,
+      "step": 3001
+    },
+    {
+      "epoch": 1.0964031404053314,
+      "grad_norm": 0.8687719702720642,
+      "learning_rate": 2.4045912412640146e-05,
+      "loss": 0.6444,
+      "step": 3002
+    },
+    {
+      "epoch": 1.0967683038159577,
+      "grad_norm": 0.9157119393348694,
+      "learning_rate": 2.4031862794742238e-05,
+      "loss": 0.6494,
+      "step": 3003
+    },
+    {
+      "epoch": 1.097133467226584,
+      "grad_norm": 0.973338782787323,
+      "learning_rate": 2.401781110261271e-05,
+      "loss": 0.634,
+      "step": 3004
+    },
+    {
+      "epoch": 1.09749863063721,
+      "grad_norm": 0.8346941471099854,
+      "learning_rate": 2.400375734348059e-05,
+      "loss": 0.614,
+      "step": 3005
+    },
+    {
+      "epoch": 1.0978637940478364,
+      "grad_norm": 0.9992729425430298,
+      "learning_rate": 2.3989701524575976e-05,
+      "loss": 0.6757,
+      "step": 3006
+    },
+    {
+      "epoch": 1.0982289574584627,
+      "grad_norm": 0.917357325553894,
+      "learning_rate": 2.3975643653130032e-05,
+      "loss": 0.6708,
+      "step": 3007
+    },
+    {
+      "epoch": 1.0985941208690888,
+      "grad_norm": 0.9452908635139465,
+      "learning_rate": 2.3961583736374955e-05,
+      "loss": 0.6424,
+      "step": 3008
+    },
+    {
+      "epoch": 1.0989592842797151,
+      "grad_norm": 0.9059060215950012,
+      "learning_rate": 2.394752178154402e-05,
+      "loss": 0.6346,
+      "step": 3009
+    },
+    {
+      "epoch": 1.0993244476903414,
+      "grad_norm": 2.1735620498657227,
+      "learning_rate": 2.393345779587153e-05,
+      "loss": 0.6439,
+      "step": 3010
+    },
+    {
+      "epoch": 1.0996896111009677,
+      "grad_norm": 0.8743827939033508,
+      "learning_rate": 2.3919391786592842e-05,
+      "loss": 0.6852,
+      "step": 3011
+    },
+    {
+      "epoch": 1.1000547745115938,
+      "grad_norm": 0.6446078419685364,
+      "learning_rate": 2.3905323760944356e-05,
+      "loss": 0.6317,
+      "step": 3012
+    },
+    {
+      "epoch": 1.1004199379222201,
+      "grad_norm": 1.1510143280029297,
+      "learning_rate": 2.3891253726163505e-05,
+      "loss": 0.6357,
+      "step": 3013
+    },
+    {
+      "epoch": 1.1007851013328465,
+      "grad_norm": 1.0216180086135864,
+      "learning_rate": 2.387718168948876e-05,
+      "loss": 0.6357,
+      "step": 3014
+    },
+    {
+      "epoch": 1.1011502647434728,
+      "grad_norm": 1.133445143699646,
+      "learning_rate": 2.3863107658159614e-05,
+      "loss": 0.7019,
+      "step": 3015
+    },
+    {
+      "epoch": 1.1015154281540989,
+      "grad_norm": 1.2700881958007812,
+      "learning_rate": 2.38490316394166e-05,
+      "loss": 0.6473,
+      "step": 3016
+    },
+    {
+      "epoch": 1.1018805915647252,
+      "grad_norm": 0.7656792402267456,
+      "learning_rate": 2.3834953640501262e-05,
+      "loss": 0.6409,
+      "step": 3017
+    },
+    {
+      "epoch": 1.1022457549753515,
+      "grad_norm": 0.8528900146484375,
+      "learning_rate": 2.3820873668656154e-05,
+      "loss": 0.6619,
+      "step": 3018
+    },
+    {
+      "epoch": 1.1026109183859778,
+      "grad_norm": 1.3215256929397583,
+      "learning_rate": 2.380679173112487e-05,
+      "loss": 0.6511,
+      "step": 3019
+    },
+    {
+      "epoch": 1.102976081796604,
+      "grad_norm": 2.560433864593506,
+      "learning_rate": 2.3792707835151995e-05,
+      "loss": 0.6196,
+      "step": 3020
+    },
+    {
+      "epoch": 1.1033412452072302,
+      "grad_norm": 0.8991859555244446,
+      "learning_rate": 2.3778621987983133e-05,
+      "loss": 0.6516,
+      "step": 3021
+    },
+    {
+      "epoch": 1.1037064086178565,
+      "grad_norm": 0.9166661500930786,
+      "learning_rate": 2.3764534196864886e-05,
+      "loss": 0.6185,
+      "step": 3022
+    },
+    {
+      "epoch": 1.1040715720284828,
+      "grad_norm": 1.0352705717086792,
+      "learning_rate": 2.3750444469044856e-05,
+      "loss": 0.6555,
+      "step": 3023
+    },
+    {
+      "epoch": 1.104436735439109,
+      "grad_norm": 0.8666485548019409,
+      "learning_rate": 2.3736352811771647e-05,
+      "loss": 0.6609,
+      "step": 3024
+    },
+    {
+      "epoch": 1.1048018988497352,
+      "grad_norm": 0.8922779560089111,
+      "learning_rate": 2.3722259232294835e-05,
+      "loss": 0.6678,
+      "step": 3025
+    },
+    {
+      "epoch": 1.1051670622603615,
+      "grad_norm": 3.4590539932250977,
+      "learning_rate": 2.370816373786502e-05,
+      "loss": 0.6838,
+      "step": 3026
+    },
+    {
+      "epoch": 1.1055322256709879,
+      "grad_norm": 0.7968533635139465,
+      "learning_rate": 2.3694066335733758e-05,
+      "loss": 0.6302,
+      "step": 3027
+    },
+    {
+      "epoch": 1.105897389081614,
+      "grad_norm": 0.7470107674598694,
+      "learning_rate": 2.3679967033153605e-05,
+      "loss": 0.6379,
+      "step": 3028
+    },
+    {
+      "epoch": 1.1062625524922403,
+      "grad_norm": 1.3071082830429077,
+      "learning_rate": 2.366586583737808e-05,
+      "loss": 0.6602,
+      "step": 3029
+    },
+    {
+      "epoch": 1.1066277159028666,
+      "grad_norm": 1.1564059257507324,
+      "learning_rate": 2.3651762755661683e-05,
+      "loss": 0.6407,
+      "step": 3030
+    },
+    {
+      "epoch": 1.1069928793134929,
+      "grad_norm": 0.8905368447303772,
+      "learning_rate": 2.3637657795259883e-05,
+      "loss": 0.6196,
+      "step": 3031
+    },
+    {
+      "epoch": 1.107358042724119,
+      "grad_norm": 0.910971462726593,
+      "learning_rate": 2.3623550963429117e-05,
+      "loss": 0.6431,
+      "step": 3032
+    },
+    {
+      "epoch": 1.1077232061347453,
+      "grad_norm": 1.8840655088424683,
+      "learning_rate": 2.3609442267426787e-05,
+      "loss": 0.6377,
+      "step": 3033
+    },
+    {
+      "epoch": 1.1080883695453716,
+      "grad_norm": 1.1728452444076538,
+      "learning_rate": 2.3595331714511243e-05,
+      "loss": 0.6516,
+      "step": 3034
+    },
+    {
+      "epoch": 1.108453532955998,
+      "grad_norm": 0.8091674447059631,
+      "learning_rate": 2.3581219311941813e-05,
+      "loss": 0.6046,
+      "step": 3035
+    },
+    {
+      "epoch": 1.108818696366624,
+      "grad_norm": 0.9540202617645264,
+      "learning_rate": 2.3567105066978744e-05,
+      "loss": 0.6034,
+      "step": 3036
+    },
+    {
+      "epoch": 1.1091838597772503,
+      "grad_norm": 0.9614107012748718,
+      "learning_rate": 2.355298898688326e-05,
+      "loss": 0.6391,
+      "step": 3037
+    },
+    {
+      "epoch": 1.1095490231878766,
+      "grad_norm": 0.8953191041946411,
+      "learning_rate": 2.3538871078917514e-05,
+      "loss": 0.6394,
+      "step": 3038
+    },
+    {
+      "epoch": 1.1099141865985027,
+      "grad_norm": 0.7751514911651611,
+      "learning_rate": 2.3524751350344602e-05,
+      "loss": 0.6622,
+      "step": 3039
+    },
+    {
+      "epoch": 1.110279350009129,
+      "grad_norm": 0.9415932297706604,
+      "learning_rate": 2.3510629808428568e-05,
+      "loss": 0.6224,
+      "step": 3040
+    },
+    {
+      "epoch": 1.1106445134197553,
+      "grad_norm": 0.9658535122871399,
+      "learning_rate": 2.349650646043436e-05,
+      "loss": 0.6577,
+      "step": 3041
+    },
+    {
+      "epoch": 1.1110096768303817,
+      "grad_norm": 0.9596617221832275,
+      "learning_rate": 2.3482381313627886e-05,
+      "loss": 0.6557,
+      "step": 3042
+    },
+    {
+      "epoch": 1.111374840241008,
+      "grad_norm": 0.7701271176338196,
+      "learning_rate": 2.3468254375275973e-05,
+      "loss": 0.6545,
+      "step": 3043
+    },
+    {
+      "epoch": 1.111740003651634,
+      "grad_norm": 0.6224210262298584,
+      "learning_rate": 2.3454125652646348e-05,
+      "loss": 0.6248,
+      "step": 3044
+    },
+    {
+      "epoch": 1.1121051670622604,
+      "grad_norm": 0.8339946866035461,
+      "learning_rate": 2.343999515300769e-05,
+      "loss": 0.6459,
+      "step": 3045
+    },
+    {
+      "epoch": 1.1124703304728867,
+      "grad_norm": 0.9032711386680603,
+      "learning_rate": 2.3425862883629553e-05,
+      "loss": 0.6484,
+      "step": 3046
+    },
+    {
+      "epoch": 1.1128354938835128,
+      "grad_norm": 0.9951335191726685,
+      "learning_rate": 2.3411728851782442e-05,
+      "loss": 0.6472,
+      "step": 3047
+    },
+    {
+      "epoch": 1.113200657294139,
+      "grad_norm": 0.9903227686882019,
+      "learning_rate": 2.3397593064737737e-05,
+      "loss": 0.6292,
+      "step": 3048
+    },
+    {
+      "epoch": 1.1135658207047654,
+      "grad_norm": 0.7369117140769958,
+      "learning_rate": 2.338345552976774e-05,
+      "loss": 0.6253,
+      "step": 3049
+    },
+    {
+      "epoch": 1.1139309841153917,
+      "grad_norm": 1.0023393630981445,
+      "learning_rate": 2.3369316254145636e-05,
+      "loss": 0.6415,
+      "step": 3050
+    },
+    {
+      "epoch": 1.1142961475260178,
+      "grad_norm": 0.9049243927001953,
+      "learning_rate": 2.3355175245145526e-05,
+      "loss": 0.6213,
+      "step": 3051
+    },
+    {
+      "epoch": 1.1146613109366441,
+      "grad_norm": 0.9037688970565796,
+      "learning_rate": 2.3341032510042387e-05,
+      "loss": 0.6517,
+      "step": 3052
+    },
+    {
+      "epoch": 1.1150264743472704,
+      "grad_norm": 0.8866589069366455,
+      "learning_rate": 2.3326888056112086e-05,
+      "loss": 0.6544,
+      "step": 3053
+    },
+    {
+      "epoch": 1.1153916377578967,
+      "grad_norm": 1.0946450233459473,
+      "learning_rate": 2.3312741890631383e-05,
+      "loss": 0.5887,
+      "step": 3054
+    },
+    {
+      "epoch": 1.1157568011685228,
+      "grad_norm": 0.8456511497497559,
+      "learning_rate": 2.3298594020877913e-05,
+      "loss": 0.6501,
+      "step": 3055
+    },
+    {
+      "epoch": 1.1161219645791491,
+      "grad_norm": 1.074009656906128,
+      "learning_rate": 2.328444445413018e-05,
+      "loss": 0.6182,
+      "step": 3056
+    },
+    {
+      "epoch": 1.1164871279897755,
+      "grad_norm": 0.6773717999458313,
+      "learning_rate": 2.3270293197667573e-05,
+      "loss": 0.6216,
+      "step": 3057
+    },
+    {
+      "epoch": 1.1168522914004018,
+      "grad_norm": 0.8690649271011353,
+      "learning_rate": 2.325614025877034e-05,
+      "loss": 0.665,
+      "step": 3058
+    },
+    {
+      "epoch": 1.1172174548110279,
+      "grad_norm": 0.969440221786499,
+      "learning_rate": 2.3241985644719603e-05,
+      "loss": 0.6223,
+      "step": 3059
+    },
+    {
+      "epoch": 1.1175826182216542,
+      "grad_norm": 0.9241883754730225,
+      "learning_rate": 2.3227829362797355e-05,
+      "loss": 0.6335,
+      "step": 3060
+    },
+    {
+      "epoch": 1.1179477816322805,
+      "grad_norm": 0.9070743322372437,
+      "learning_rate": 2.3213671420286413e-05,
+      "loss": 0.6459,
+      "step": 3061
+    },
+    {
+      "epoch": 1.1183129450429068,
+      "grad_norm": 0.7353122234344482,
+      "learning_rate": 2.3199511824470487e-05,
+      "loss": 0.6281,
+      "step": 3062
+    },
+    {
+      "epoch": 1.1186781084535329,
+      "grad_norm": 0.9907974004745483,
+      "learning_rate": 2.318535058263412e-05,
+      "loss": 0.6541,
+      "step": 3063
+    },
+    {
+      "epoch": 1.1190432718641592,
+      "grad_norm": 1.0479435920715332,
+      "learning_rate": 2.3171187702062693e-05,
+      "loss": 0.6409,
+      "step": 3064
+    },
+    {
+      "epoch": 1.1194084352747855,
+      "grad_norm": 1.9881740808486938,
+      "learning_rate": 2.3157023190042448e-05,
+      "loss": 0.6329,
+      "step": 3065
+    },
+    {
+      "epoch": 1.1197735986854118,
+      "grad_norm": 1.7955461740493774,
+      "learning_rate": 2.3142857053860454e-05,
+      "loss": 0.6378,
+      "step": 3066
+    },
+    {
+      "epoch": 1.120138762096038,
+      "grad_norm": 0.7752870917320251,
+      "learning_rate": 2.312868930080462e-05,
+      "loss": 0.6451,
+      "step": 3067
+    },
+    {
+      "epoch": 1.1205039255066642,
+      "grad_norm": 0.704662561416626,
+      "learning_rate": 2.3114519938163683e-05,
+      "loss": 0.6353,
+      "step": 3068
+    },
+    {
+      "epoch": 1.1208690889172905,
+      "grad_norm": 1.5579748153686523,
+      "learning_rate": 2.3100348973227224e-05,
+      "loss": 0.6571,
+      "step": 3069
+    },
+    {
+      "epoch": 1.1212342523279166,
+      "grad_norm": 1.206935167312622,
+      "learning_rate": 2.3086176413285627e-05,
+      "loss": 0.6174,
+      "step": 3070
+    },
+    {
+      "epoch": 1.121599415738543,
+      "grad_norm": 1.1550734043121338,
+      "learning_rate": 2.3072002265630102e-05,
+      "loss": 0.6234,
+      "step": 3071
+    },
+    {
+      "epoch": 1.1219645791491692,
+      "grad_norm": 0.8067594170570374,
+      "learning_rate": 2.3057826537552684e-05,
+      "loss": 0.6423,
+      "step": 3072
+    },
+    {
+      "epoch": 1.1223297425597956,
+      "grad_norm": 1.1658196449279785,
+      "learning_rate": 2.3043649236346216e-05,
+      "loss": 0.6543,
+      "step": 3073
+    },
+    {
+      "epoch": 1.1226949059704219,
+      "grad_norm": 0.8631314039230347,
+      "learning_rate": 2.302947036930435e-05,
+      "loss": 0.6254,
+      "step": 3074
+    },
+    {
+      "epoch": 1.123060069381048,
+      "grad_norm": 0.8141615986824036,
+      "learning_rate": 2.301528994372154e-05,
+      "loss": 0.646,
+      "step": 3075
+    },
+    {
+      "epoch": 1.1234252327916743,
+      "grad_norm": 1.31234610080719,
+      "learning_rate": 2.3001107966893054e-05,
+      "loss": 0.6113,
+      "step": 3076
+    },
+    {
+      "epoch": 1.1237903962023006,
+      "grad_norm": 0.7371488213539124,
+      "learning_rate": 2.2986924446114947e-05,
+      "loss": 0.6483,
+      "step": 3077
+    },
+    {
+      "epoch": 1.1241555596129267,
+      "grad_norm": 0.7674611806869507,
+      "learning_rate": 2.2972739388684068e-05,
+      "loss": 0.6209,
+      "step": 3078
+    },
+    {
+      "epoch": 1.124520723023553,
+      "grad_norm": 4.130733013153076,
+      "learning_rate": 2.2958552801898068e-05,
+      "loss": 0.6488,
+      "step": 3079
+    },
+    {
+      "epoch": 1.1248858864341793,
+      "grad_norm": 0.8405309915542603,
+      "learning_rate": 2.294436469305536e-05,
+      "loss": 0.6263,
+      "step": 3080
+    },
+    {
+      "epoch": 1.1252510498448056,
+      "grad_norm": 0.8591222167015076,
+      "learning_rate": 2.2930175069455175e-05,
+      "loss": 0.653,
+      "step": 3081
+    },
+    {
+      "epoch": 1.125616213255432,
+      "grad_norm": 0.7199135422706604,
+      "learning_rate": 2.2915983938397494e-05,
+      "loss": 0.6292,
+      "step": 3082
+    },
+    {
+      "epoch": 1.125981376666058,
+      "grad_norm": 1.0843881368637085,
+      "learning_rate": 2.290179130718309e-05,
+      "loss": 0.6216,
+      "step": 3083
+    },
+    {
+      "epoch": 1.1263465400766843,
+      "grad_norm": 0.9486679434776306,
+      "learning_rate": 2.2887597183113503e-05,
+      "loss": 0.6062,
+      "step": 3084
+    },
+    {
+      "epoch": 1.1267117034873106,
+      "grad_norm": 1.3532755374908447,
+      "learning_rate": 2.2873401573491033e-05,
+      "loss": 0.6511,
+      "step": 3085
+    },
+    {
+      "epoch": 1.1270768668979367,
+      "grad_norm": 0.8381106853485107,
+      "learning_rate": 2.2859204485618758e-05,
+      "loss": 0.6216,
+      "step": 3086
+    },
+    {
+      "epoch": 1.127442030308563,
+      "grad_norm": 0.9955036640167236,
+      "learning_rate": 2.2845005926800502e-05,
+      "loss": 0.6321,
+      "step": 3087
+    },
+    {
+      "epoch": 1.1278071937191894,
+      "grad_norm": 0.7981411218643188,
+      "learning_rate": 2.2830805904340867e-05,
+      "loss": 0.6302,
+      "step": 3088
+    },
+    {
+      "epoch": 1.1281723571298157,
+      "grad_norm": 0.6035706400871277,
+      "learning_rate": 2.281660442554518e-05,
+      "loss": 0.639,
+      "step": 3089
+    },
+    {
+      "epoch": 1.1285375205404418,
+      "grad_norm": 0.7041566371917725,
+      "learning_rate": 2.2802401497719545e-05,
+      "loss": 0.6393,
+      "step": 3090
+    },
+    {
+      "epoch": 1.128902683951068,
+      "grad_norm": 0.9806358218193054,
+      "learning_rate": 2.2788197128170798e-05,
+      "loss": 0.5987,
+      "step": 3091
+    },
+    {
+      "epoch": 1.1292678473616944,
+      "grad_norm": 1.0443167686462402,
+      "learning_rate": 2.27739913242065e-05,
+      "loss": 0.6513,
+      "step": 3092
+    },
+    {
+      "epoch": 1.1296330107723207,
+      "grad_norm": 0.8455133438110352,
+      "learning_rate": 2.2759784093134987e-05,
+      "loss": 0.6214,
+      "step": 3093
+    },
+    {
+      "epoch": 1.1299981741829468,
+      "grad_norm": 0.9792264699935913,
+      "learning_rate": 2.2745575442265297e-05,
+      "loss": 0.6476,
+      "step": 3094
+    },
+    {
+      "epoch": 1.130363337593573,
+      "grad_norm": 0.7479596138000488,
+      "learning_rate": 2.273136537890722e-05,
+      "loss": 0.6263,
+      "step": 3095
+    },
+    {
+      "epoch": 1.1307285010041994,
+      "grad_norm": 2.3983466625213623,
+      "learning_rate": 2.271715391037126e-05,
+      "loss": 0.6525,
+      "step": 3096
+    },
+    {
+      "epoch": 1.1310936644148257,
+      "grad_norm": 0.9311935901641846,
+      "learning_rate": 2.2702941043968635e-05,
+      "loss": 0.6346,
+      "step": 3097
+    },
+    {
+      "epoch": 1.1314588278254518,
+      "grad_norm": 0.9133360981941223,
+      "learning_rate": 2.2688726787011315e-05,
+      "loss": 0.6202,
+      "step": 3098
+    },
+    {
+      "epoch": 1.1318239912360781,
+      "grad_norm": 0.6708499193191528,
+      "learning_rate": 2.267451114681195e-05,
+      "loss": 0.647,
+      "step": 3099
+    },
+    {
+      "epoch": 1.1321891546467044,
+      "grad_norm": 0.8047714829444885,
+      "learning_rate": 2.2660294130683923e-05,
+      "loss": 0.6539,
+      "step": 3100
+    },
+    {
+      "epoch": 1.1325543180573305,
+      "grad_norm": 0.9618435502052307,
+      "learning_rate": 2.2646075745941315e-05,
+      "loss": 0.6393,
+      "step": 3101
+    },
+    {
+      "epoch": 1.1329194814679568,
+      "grad_norm": 0.824882984161377,
+      "learning_rate": 2.2631855999898914e-05,
+      "loss": 0.6126,
+      "step": 3102
+    },
+    {
+      "epoch": 1.1332846448785832,
+      "grad_norm": 0.9720734357833862,
+      "learning_rate": 2.261763489987222e-05,
+      "loss": 0.5818,
+      "step": 3103
+    },
+    {
+      "epoch": 1.1336498082892095,
+      "grad_norm": 1.0631283521652222,
+      "learning_rate": 2.26034124531774e-05,
+      "loss": 0.6443,
+      "step": 3104
+    },
+    {
+      "epoch": 1.1340149716998358,
+      "grad_norm": 1.071035623550415,
+      "learning_rate": 2.2589188667131346e-05,
+      "loss": 0.6133,
+      "step": 3105
+    },
+    {
+      "epoch": 1.1343801351104619,
+      "grad_norm": 1.0741324424743652,
+      "learning_rate": 2.257496354905162e-05,
+      "loss": 0.5823,
+      "step": 3106
+    },
+    {
+      "epoch": 1.1347452985210882,
+      "grad_norm": 0.8549277186393738,
+      "learning_rate": 2.2560737106256472e-05,
+      "loss": 0.6235,
+      "step": 3107
+    },
+    {
+      "epoch": 1.1351104619317145,
+      "grad_norm": 0.995807945728302,
+      "learning_rate": 2.254650934606484e-05,
+      "loss": 0.6291,
+      "step": 3108
+    },
+    {
+      "epoch": 1.1354756253423406,
+      "grad_norm": 0.6567954421043396,
+      "learning_rate": 2.2532280275796333e-05,
+      "loss": 0.6292,
+      "step": 3109
+    },
+    {
+      "epoch": 1.135840788752967,
+      "grad_norm": 0.9623636603355408,
+      "learning_rate": 2.251804990277125e-05,
+      "loss": 0.6548,
+      "step": 3110
+    },
+    {
+      "epoch": 1.1362059521635932,
+      "grad_norm": 0.9804652333259583,
+      "learning_rate": 2.250381823431052e-05,
+      "loss": 0.6023,
+      "step": 3111
+    },
+    {
+      "epoch": 1.1365711155742195,
+      "grad_norm": 0.9299272298812866,
+      "learning_rate": 2.248958527773579e-05,
+      "loss": 0.6508,
+      "step": 3112
+    },
+    {
+      "epoch": 1.1369362789848458,
+      "grad_norm": 1.0050239562988281,
+      "learning_rate": 2.2475351040369327e-05,
+      "loss": 0.614,
+      "step": 3113
+    },
+    {
+      "epoch": 1.137301442395472,
+      "grad_norm": 0.9981377720832825,
+      "learning_rate": 2.2461115529534084e-05,
+      "loss": 0.6364,
+      "step": 3114
+    },
+    {
+      "epoch": 1.1376666058060982,
+      "grad_norm": 0.8583781719207764,
+      "learning_rate": 2.244687875255367e-05,
+      "loss": 0.6411,
+      "step": 3115
+    },
+    {
+      "epoch": 1.1380317692167246,
+      "grad_norm": 1.0073418617248535,
+      "learning_rate": 2.2432640716752316e-05,
+      "loss": 0.6619,
+      "step": 3116
+    },
+    {
+      "epoch": 1.1383969326273506,
+      "grad_norm": 0.8421041965484619,
+      "learning_rate": 2.241840142945494e-05,
+      "loss": 0.6255,
+      "step": 3117
+    },
+    {
+      "epoch": 1.138762096037977,
+      "grad_norm": 0.8554766178131104,
+      "learning_rate": 2.2404160897987056e-05,
+      "loss": 0.5789,
+      "step": 3118
+    },
+    {
+      "epoch": 1.1391272594486033,
+      "grad_norm": 0.9527449607849121,
+      "learning_rate": 2.2389919129674872e-05,
+      "loss": 0.6251,
+      "step": 3119
+    },
+    {
+      "epoch": 1.1394924228592296,
+      "grad_norm": 1.126502275466919,
+      "learning_rate": 2.2375676131845196e-05,
+      "loss": 0.6461,
+      "step": 3120
+    },
+    {
+      "epoch": 1.1398575862698557,
+      "grad_norm": 0.7084481120109558,
+      "learning_rate": 2.236143191182548e-05,
+      "loss": 0.625,
+      "step": 3121
+    },
+    {
+      "epoch": 1.140222749680482,
+      "grad_norm": 0.8147100806236267,
+      "learning_rate": 2.2347186476943805e-05,
+      "loss": 0.6227,
+      "step": 3122
+    },
+    {
+      "epoch": 1.1405879130911083,
+      "grad_norm": 1.3247547149658203,
+      "learning_rate": 2.2332939834528875e-05,
+      "loss": 0.6486,
+      "step": 3123
+    },
+    {
+      "epoch": 1.1409530765017346,
+      "grad_norm": 0.6833971738815308,
+      "learning_rate": 2.2318691991910014e-05,
+      "loss": 0.6393,
+      "step": 3124
+    },
+    {
+      "epoch": 1.1413182399123607,
+      "grad_norm": 1.0287319421768188,
+      "learning_rate": 2.2304442956417164e-05,
+      "loss": 0.6579,
+      "step": 3125
+    },
+    {
+      "epoch": 1.141683403322987,
+      "grad_norm": 0.6814072728157043,
+      "learning_rate": 2.229019273538089e-05,
+      "loss": 0.6415,
+      "step": 3126
+    },
+    {
+      "epoch": 1.1420485667336133,
+      "grad_norm": 0.8899560570716858,
+      "learning_rate": 2.227594133613235e-05,
+      "loss": 0.6276,
+      "step": 3127
+    },
+    {
+      "epoch": 1.1424137301442396,
+      "grad_norm": 1.134264349937439,
+      "learning_rate": 2.2261688766003317e-05,
+      "loss": 0.6347,
+      "step": 3128
+    },
+    {
+      "epoch": 1.1427788935548657,
+      "grad_norm": 1.2518939971923828,
+      "learning_rate": 2.2247435032326178e-05,
+      "loss": 0.6269,
+      "step": 3129
+    },
+    {
+      "epoch": 1.143144056965492,
+      "grad_norm": 0.8139986991882324,
+      "learning_rate": 2.2233180142433894e-05,
+      "loss": 0.6105,
+      "step": 3130
+    },
+    {
+      "epoch": 1.1435092203761184,
+      "grad_norm": 0.8457422256469727,
+      "learning_rate": 2.2218924103660035e-05,
+      "loss": 0.6326,
+      "step": 3131
+    },
+    {
+      "epoch": 1.1438743837867447,
+      "grad_norm": 0.9971619844436646,
+      "learning_rate": 2.2204666923338772e-05,
+      "loss": 0.6281,
+      "step": 3132
+    },
+    {
+      "epoch": 1.1442395471973708,
+      "grad_norm": 1.9557191133499146,
+      "learning_rate": 2.219040860880484e-05,
+      "loss": 0.6534,
+      "step": 3133
+    },
+    {
+      "epoch": 1.144604710607997,
+      "grad_norm": 0.7603287100791931,
+      "learning_rate": 2.217614916739358e-05,
+      "loss": 0.6303,
+      "step": 3134
+    },
+    {
+      "epoch": 1.1449698740186234,
+      "grad_norm": 0.8402383327484131,
+      "learning_rate": 2.2161888606440885e-05,
+      "loss": 0.639,
+      "step": 3135
+    },
+    {
+      "epoch": 1.1453350374292497,
+      "grad_norm": 1.0194815397262573,
+      "learning_rate": 2.2147626933283265e-05,
+      "loss": 0.6159,
+      "step": 3136
+    },
+    {
+      "epoch": 1.1457002008398758,
+      "grad_norm": 0.9987945556640625,
+      "learning_rate": 2.213336415525776e-05,
+      "loss": 0.5863,
+      "step": 3137
+    },
+    {
+      "epoch": 1.146065364250502,
+      "grad_norm": 1.2469425201416016,
+      "learning_rate": 2.2119100279702005e-05,
+      "loss": 0.6081,
+      "step": 3138
+    },
+    {
+      "epoch": 1.1464305276611284,
+      "grad_norm": 0.7377803921699524,
+      "learning_rate": 2.2104835313954193e-05,
+      "loss": 0.5949,
+      "step": 3139
+    },
+    {
+      "epoch": 1.1467956910717545,
+      "grad_norm": 0.714694619178772,
+      "learning_rate": 2.209056926535307e-05,
+      "loss": 0.6082,
+      "step": 3140
+    },
+    {
+      "epoch": 1.1471608544823808,
+      "grad_norm": 0.9076855778694153,
+      "learning_rate": 2.2076302141237953e-05,
+      "loss": 0.6224,
+      "step": 3141
+    },
+    {
+      "epoch": 1.1475260178930071,
+      "grad_norm": 0.9985514283180237,
+      "learning_rate": 2.2062033948948697e-05,
+      "loss": 0.6256,
+      "step": 3142
+    },
+    {
+      "epoch": 1.1478911813036334,
+      "grad_norm": 0.9677404761314392,
+      "learning_rate": 2.2047764695825725e-05,
+      "loss": 0.6055,
+      "step": 3143
+    },
+    {
+      "epoch": 1.1482563447142597,
+      "grad_norm": 0.9319841265678406,
+      "learning_rate": 2.2033494389209988e-05,
+      "loss": 0.6483,
+      "step": 3144
+    },
+    {
+      "epoch": 1.1486215081248858,
+      "grad_norm": 0.7451246976852417,
+      "learning_rate": 2.201922303644298e-05,
+      "loss": 0.6267,
+      "step": 3145
+    },
+    {
+      "epoch": 1.1489866715355121,
+      "grad_norm": 0.7892596125602722,
+      "learning_rate": 2.200495064486675e-05,
+      "loss": 0.6456,
+      "step": 3146
+    },
+    {
+      "epoch": 1.1493518349461385,
+      "grad_norm": 0.8904467821121216,
+      "learning_rate": 2.1990677221823865e-05,
+      "loss": 0.6443,
+      "step": 3147
+    },
+    {
+      "epoch": 1.1497169983567646,
+      "grad_norm": 0.7229827046394348,
+      "learning_rate": 2.1976402774657432e-05,
+      "loss": 0.6479,
+      "step": 3148
+    },
+    {
+      "epoch": 1.1500821617673909,
+      "grad_norm": 0.9828261137008667,
+      "learning_rate": 2.196212731071108e-05,
+      "loss": 0.6593,
+      "step": 3149
+    },
+    {
+      "epoch": 1.1504473251780172,
+      "grad_norm": 0.8736459612846375,
+      "learning_rate": 2.194785083732896e-05,
+      "loss": 0.6517,
+      "step": 3150
+    },
+    {
+      "epoch": 1.1508124885886435,
+      "grad_norm": 0.7535048723220825,
+      "learning_rate": 2.193357336185575e-05,
+      "loss": 0.6409,
+      "step": 3151
+    },
+    {
+      "epoch": 1.1511776519992696,
+      "grad_norm": 0.9029123783111572,
+      "learning_rate": 2.191929489163663e-05,
+      "loss": 0.6156,
+      "step": 3152
+    },
+    {
+      "epoch": 1.151542815409896,
+      "grad_norm": 1.1329669952392578,
+      "learning_rate": 2.1905015434017313e-05,
+      "loss": 0.6435,
+      "step": 3153
+    },
+    {
+      "epoch": 1.1519079788205222,
+      "grad_norm": 0.9592320322990417,
+      "learning_rate": 2.1890734996343985e-05,
+      "loss": 0.6366,
+      "step": 3154
+    },
+    {
+      "epoch": 1.1522731422311485,
+      "grad_norm": 0.7683483958244324,
+      "learning_rate": 2.1876453585963384e-05,
+      "loss": 0.634,
+      "step": 3155
+    },
+    {
+      "epoch": 1.1526383056417746,
+      "grad_norm": 0.794584333896637,
+      "learning_rate": 2.1862171210222708e-05,
+      "loss": 0.6211,
+      "step": 3156
+    },
+    {
+      "epoch": 1.153003469052401,
+      "grad_norm": 0.8724021315574646,
+      "learning_rate": 2.1847887876469666e-05,
+      "loss": 0.6047,
+      "step": 3157
+    },
+    {
+      "epoch": 1.1533686324630272,
+      "grad_norm": 1.1390639543533325,
+      "learning_rate": 2.1833603592052464e-05,
+      "loss": 0.6597,
+      "step": 3158
+    },
+    {
+      "epoch": 1.1537337958736535,
+      "grad_norm": 0.7585846185684204,
+      "learning_rate": 2.181931836431979e-05,
+      "loss": 0.6213,
+      "step": 3159
+    },
+    {
+      "epoch": 1.1540989592842796,
+      "grad_norm": 1.4673725366592407,
+      "learning_rate": 2.1805032200620824e-05,
+      "loss": 0.6419,
+      "step": 3160
+    },
+    {
+      "epoch": 1.154464122694906,
+      "grad_norm": 1.0087941884994507,
+      "learning_rate": 2.1790745108305222e-05,
+      "loss": 0.6164,
+      "step": 3161
+    },
+    {
+      "epoch": 1.1548292861055323,
+      "grad_norm": 0.998311460018158,
+      "learning_rate": 2.1776457094723115e-05,
+      "loss": 0.6334,
+      "step": 3162
+    },
+    {
+      "epoch": 1.1551944495161586,
+      "grad_norm": 1.0328058004379272,
+      "learning_rate": 2.176216816722513e-05,
+      "loss": 0.6338,
+      "step": 3163
+    },
+    {
+      "epoch": 1.1555596129267847,
+      "grad_norm": 0.9341315627098083,
+      "learning_rate": 2.1747878333162326e-05,
+      "loss": 0.6263,
+      "step": 3164
+    },
+    {
+      "epoch": 1.155924776337411,
+      "grad_norm": 0.8238332271575928,
+      "learning_rate": 2.173358759988626e-05,
+      "loss": 0.6186,
+      "step": 3165
+    },
+    {
+      "epoch": 1.1562899397480373,
+      "grad_norm": 1.0291343927383423,
+      "learning_rate": 2.1719295974748934e-05,
+      "loss": 0.6154,
+      "step": 3166
+    },
+    {
+      "epoch": 1.1566551031586636,
+      "grad_norm": 0.6935338377952576,
+      "learning_rate": 2.1705003465102818e-05,
+      "loss": 0.6333,
+      "step": 3167
+    },
+    {
+      "epoch": 1.1570202665692897,
+      "grad_norm": 0.8157773613929749,
+      "learning_rate": 2.1690710078300847e-05,
+      "loss": 0.628,
+      "step": 3168
+    },
+    {
+      "epoch": 1.157385429979916,
+      "grad_norm": 0.8037132620811462,
+      "learning_rate": 2.167641582169637e-05,
+      "loss": 0.6138,
+      "step": 3169
+    },
+    {
+      "epoch": 1.1577505933905423,
+      "grad_norm": 0.714047372341156,
+      "learning_rate": 2.166212070264324e-05,
+      "loss": 0.6261,
+      "step": 3170
+    },
+    {
+      "epoch": 1.1581157568011684,
+      "grad_norm": 0.7886608242988586,
+      "learning_rate": 2.1647824728495696e-05,
+      "loss": 0.6274,
+      "step": 3171
+    },
+    {
+      "epoch": 1.1584809202117947,
+      "grad_norm": 0.853618860244751,
+      "learning_rate": 2.1633527906608457e-05,
+      "loss": 0.5916,
+      "step": 3172
+    },
+    {
+      "epoch": 1.158846083622421,
+      "grad_norm": 1.2856926918029785,
+      "learning_rate": 2.1619230244336652e-05,
+      "loss": 0.6022,
+      "step": 3173
+    },
+    {
+      "epoch": 1.1592112470330473,
+      "grad_norm": 0.9644615650177002,
+      "learning_rate": 2.1604931749035865e-05,
+      "loss": 0.6204,
+      "step": 3174
+    },
+    {
+      "epoch": 1.1595764104436737,
+      "grad_norm": 0.8236731886863708,
+      "learning_rate": 2.1590632428062097e-05,
+      "loss": 0.6224,
+      "step": 3175
+    },
+    {
+      "epoch": 1.1599415738542997,
+      "grad_norm": 1.1724156141281128,
+      "learning_rate": 2.1576332288771776e-05,
+      "loss": 0.6074,
+      "step": 3176
+    },
+    {
+      "epoch": 1.160306737264926,
+      "grad_norm": 0.8694778084754944,
+      "learning_rate": 2.1562031338521745e-05,
+      "loss": 0.6204,
+      "step": 3177
+    },
+    {
+      "epoch": 1.1606719006755524,
+      "grad_norm": 0.8083097338676453,
+      "learning_rate": 2.1547729584669262e-05,
+      "loss": 0.6483,
+      "step": 3178
+    },
+    {
+      "epoch": 1.1610370640861785,
+      "grad_norm": 0.6772050857543945,
+      "learning_rate": 2.1533427034572022e-05,
+      "loss": 0.6016,
+      "step": 3179
+    },
+    {
+      "epoch": 1.1614022274968048,
+      "grad_norm": 1.0343115329742432,
+      "learning_rate": 2.1519123695588106e-05,
+      "loss": 0.6033,
+      "step": 3180
+    },
+    {
+      "epoch": 1.161767390907431,
+      "grad_norm": 0.8711481690406799,
+      "learning_rate": 2.1504819575076e-05,
+      "loss": 0.6153,
+      "step": 3181
+    },
+    {
+      "epoch": 1.1621325543180574,
+      "grad_norm": 0.8656068444252014,
+      "learning_rate": 2.1490514680394616e-05,
+      "loss": 0.6346,
+      "step": 3182
+    },
+    {
+      "epoch": 1.1624977177286837,
+      "grad_norm": 1.0679489374160767,
+      "learning_rate": 2.147620901890324e-05,
+      "loss": 0.6439,
+      "step": 3183
+    },
+    {
+      "epoch": 1.1628628811393098,
+      "grad_norm": 0.8058302998542786,
+      "learning_rate": 2.146190259796155e-05,
+      "loss": 0.6195,
+      "step": 3184
+    },
+    {
+      "epoch": 1.1632280445499361,
+      "grad_norm": 0.8528069853782654,
+      "learning_rate": 2.1447595424929647e-05,
+      "loss": 0.6425,
+      "step": 3185
+    },
+    {
+      "epoch": 1.1635932079605624,
+      "grad_norm": 0.697010338306427,
+      "learning_rate": 2.143328750716798e-05,
+      "loss": 0.6411,
+      "step": 3186
+    },
+    {
+      "epoch": 1.1639583713711885,
+      "grad_norm": 0.905610203742981,
+      "learning_rate": 2.141897885203741e-05,
+      "loss": 0.6241,
+      "step": 3187
+    },
+    {
+      "epoch": 1.1643235347818148,
+      "grad_norm": 0.9987830519676208,
+      "learning_rate": 2.140466946689915e-05,
+      "loss": 0.5917,
+      "step": 3188
+    },
+    {
+      "epoch": 1.1646886981924411,
+      "grad_norm": 1.0862075090408325,
+      "learning_rate": 2.1390359359114826e-05,
+      "loss": 0.6369,
+      "step": 3189
+    },
+    {
+      "epoch": 1.1650538616030675,
+      "grad_norm": 1.4116789102554321,
+      "learning_rate": 2.13760485360464e-05,
+      "loss": 0.6541,
+      "step": 3190
+    },
+    {
+      "epoch": 1.1654190250136935,
+      "grad_norm": 1.0040624141693115,
+      "learning_rate": 2.136173700505622e-05,
+      "loss": 0.6105,
+      "step": 3191
+    },
+    {
+      "epoch": 1.1657841884243199,
+      "grad_norm": 0.7681383490562439,
+      "learning_rate": 2.134742477350699e-05,
+      "loss": 0.581,
+      "step": 3192
+    },
+    {
+      "epoch": 1.1661493518349462,
+      "grad_norm": 0.7578985691070557,
+      "learning_rate": 2.133311184876179e-05,
+      "loss": 0.5801,
+      "step": 3193
+    },
+    {
+      "epoch": 1.1665145152455725,
+      "grad_norm": 0.8705235123634338,
+      "learning_rate": 2.1318798238184036e-05,
+      "loss": 0.6402,
+      "step": 3194
+    },
+    {
+      "epoch": 1.1668796786561986,
+      "grad_norm": 0.9274020195007324,
+      "learning_rate": 2.1304483949137503e-05,
+      "loss": 0.6138,
+      "step": 3195
+    },
+    {
+      "epoch": 1.1672448420668249,
+      "grad_norm": 0.9331055879592896,
+      "learning_rate": 2.1290168988986332e-05,
+      "loss": 0.655,
+      "step": 3196
+    },
+    {
+      "epoch": 1.1676100054774512,
+      "grad_norm": 0.8188350796699524,
+      "learning_rate": 2.127585336509498e-05,
+      "loss": 0.6006,
+      "step": 3197
+    },
+    {
+      "epoch": 1.1679751688880775,
+      "grad_norm": 1.0621793270111084,
+      "learning_rate": 2.1261537084828274e-05,
+      "loss": 0.6332,
+      "step": 3198
+    },
+    {
+      "epoch": 1.1683403322987036,
+      "grad_norm": 0.8340405225753784,
+      "learning_rate": 2.1247220155551357e-05,
+      "loss": 0.6096,
+      "step": 3199
+    },
+    {
+      "epoch": 1.16870549570933,
+      "grad_norm": 0.7741566896438599,
+      "learning_rate": 2.1232902584629716e-05,
+      "loss": 0.6539,
+      "step": 3200
+    },
+    {
+      "epoch": 1.1690706591199562,
+      "grad_norm": 0.9330441355705261,
+      "learning_rate": 2.121858437942917e-05,
+      "loss": 0.6381,
+      "step": 3201
+    },
+    {
+      "epoch": 1.1694358225305823,
+      "grad_norm": 1.090240240097046,
+      "learning_rate": 2.1204265547315862e-05,
+      "loss": 0.6595,
+      "step": 3202
+    },
+    {
+      "epoch": 1.1698009859412086,
+      "grad_norm": 0.7946920394897461,
+      "learning_rate": 2.1189946095656255e-05,
+      "loss": 0.6296,
+      "step": 3203
+    },
+    {
+      "epoch": 1.170166149351835,
+      "grad_norm": 0.6620867848396301,
+      "learning_rate": 2.117562603181713e-05,
+      "loss": 0.6378,
+      "step": 3204
+    },
+    {
+      "epoch": 1.1705313127624613,
+      "grad_norm": 0.7557404637336731,
+      "learning_rate": 2.116130536316558e-05,
+      "loss": 0.6008,
+      "step": 3205
+    },
+    {
+      "epoch": 1.1708964761730876,
+      "grad_norm": 0.7486023306846619,
+      "learning_rate": 2.114698409706903e-05,
+      "loss": 0.6088,
+      "step": 3206
+    },
+    {
+      "epoch": 1.1712616395837137,
+      "grad_norm": 0.8459969162940979,
+      "learning_rate": 2.1132662240895182e-05,
+      "loss": 0.6185,
+      "step": 3207
+    },
+    {
+      "epoch": 1.17162680299434,
+      "grad_norm": 1.013345718383789,
+      "learning_rate": 2.111833980201207e-05,
+      "loss": 0.6066,
+      "step": 3208
+    },
+    {
+      "epoch": 1.1719919664049663,
+      "grad_norm": 1.3973004817962646,
+      "learning_rate": 2.1104016787787994e-05,
+      "loss": 0.6286,
+      "step": 3209
+    },
+    {
+      "epoch": 1.1723571298155924,
+      "grad_norm": 0.8168177008628845,
+      "learning_rate": 2.108969320559159e-05,
+      "loss": 0.6224,
+      "step": 3210
+    },
+    {
+      "epoch": 1.1727222932262187,
+      "grad_norm": 0.964625895023346,
+      "learning_rate": 2.107536906279176e-05,
+      "loss": 0.6149,
+      "step": 3211
+    },
+    {
+      "epoch": 1.173087456636845,
+      "grad_norm": 0.956104576587677,
+      "learning_rate": 2.106104436675769e-05,
+      "loss": 0.6476,
+      "step": 3212
+    },
+    {
+      "epoch": 1.1734526200474713,
+      "grad_norm": 1.1153203248977661,
+      "learning_rate": 2.1046719124858882e-05,
+      "loss": 0.6167,
+      "step": 3213
+    },
+    {
+      "epoch": 1.1738177834580976,
+      "grad_norm": 0.7539756298065186,
+      "learning_rate": 2.1032393344465077e-05,
+      "loss": 0.6234,
+      "step": 3214
+    },
+    {
+      "epoch": 1.1741829468687237,
+      "grad_norm": 1.6226251125335693,
+      "learning_rate": 2.1018067032946327e-05,
+      "loss": 0.6169,
+      "step": 3215
+    },
+    {
+      "epoch": 1.17454811027935,
+      "grad_norm": 0.7395288348197937,
+      "learning_rate": 2.1003740197672946e-05,
+      "loss": 0.6255,
+      "step": 3216
+    },
+    {
+      "epoch": 1.1749132736899763,
+      "grad_norm": 0.896364152431488,
+      "learning_rate": 2.0989412846015504e-05,
+      "loss": 0.6047,
+      "step": 3217
+    },
+    {
+      "epoch": 1.1752784371006024,
+      "grad_norm": 0.9521870017051697,
+      "learning_rate": 2.0975084985344857e-05,
+      "loss": 0.6346,
+      "step": 3218
+    },
+    {
+      "epoch": 1.1756436005112287,
+      "grad_norm": 1.2138594388961792,
+      "learning_rate": 2.0960756623032114e-05,
+      "loss": 0.6334,
+      "step": 3219
+    },
+    {
+      "epoch": 1.176008763921855,
+      "grad_norm": 1.0154399871826172,
+      "learning_rate": 2.0946427766448642e-05,
+      "loss": 0.6374,
+      "step": 3220
+    },
+    {
+      "epoch": 1.1763739273324814,
+      "grad_norm": 0.9649248719215393,
+      "learning_rate": 2.093209842296606e-05,
+      "loss": 0.5993,
+      "step": 3221
+    },
+    {
+      "epoch": 1.1767390907431075,
+      "grad_norm": 0.915959358215332,
+      "learning_rate": 2.0917768599956236e-05,
+      "loss": 0.6214,
+      "step": 3222
+    },
+    {
+      "epoch": 1.1771042541537338,
+      "grad_norm": 0.8261663317680359,
+      "learning_rate": 2.090343830479131e-05,
+      "loss": 0.6181,
+      "step": 3223
+    },
+    {
+      "epoch": 1.17746941756436,
+      "grad_norm": 0.8212270736694336,
+      "learning_rate": 2.0889107544843615e-05,
+      "loss": 0.6365,
+      "step": 3224
+    },
+    {
+      "epoch": 1.1778345809749864,
+      "grad_norm": 0.7964750528335571,
+      "learning_rate": 2.0874776327485777e-05,
+      "loss": 0.6329,
+      "step": 3225
+    },
+    {
+      "epoch": 1.1781997443856125,
+      "grad_norm": 0.8948934674263,
+      "learning_rate": 2.0860444660090612e-05,
+      "loss": 0.6213,
+      "step": 3226
+    },
+    {
+      "epoch": 1.1785649077962388,
+      "grad_norm": 0.9234115481376648,
+      "learning_rate": 2.0846112550031198e-05,
+      "loss": 0.6049,
+      "step": 3227
+    },
+    {
+      "epoch": 1.178930071206865,
+      "grad_norm": 1.0438138246536255,
+      "learning_rate": 2.0831780004680834e-05,
+      "loss": 0.6008,
+      "step": 3228
+    },
+    {
+      "epoch": 1.1792952346174914,
+      "grad_norm": 0.8235229253768921,
+      "learning_rate": 2.081744703141303e-05,
+      "loss": 0.6046,
+      "step": 3229
+    },
+    {
+      "epoch": 1.1796603980281175,
+      "grad_norm": 1.143739938735962,
+      "learning_rate": 2.0803113637601543e-05,
+      "loss": 0.6107,
+      "step": 3230
+    },
+    {
+      "epoch": 1.1800255614387438,
+      "grad_norm": 0.7664543390274048,
+      "learning_rate": 2.07887798306203e-05,
+      "loss": 0.6403,
+      "step": 3231
+    },
+    {
+      "epoch": 1.1803907248493701,
+      "grad_norm": 1.0188530683517456,
+      "learning_rate": 2.0774445617843493e-05,
+      "loss": 0.6088,
+      "step": 3232
+    },
+    {
+      "epoch": 1.1807558882599964,
+      "grad_norm": 0.8213171362876892,
+      "learning_rate": 2.076011100664549e-05,
+      "loss": 0.6318,
+      "step": 3233
+    },
+    {
+      "epoch": 1.1811210516706225,
+      "grad_norm": 0.9105982184410095,
+      "learning_rate": 2.0745776004400876e-05,
+      "loss": 0.5911,
+      "step": 3234
+    },
+    {
+      "epoch": 1.1814862150812488,
+      "grad_norm": 0.8999936580657959,
+      "learning_rate": 2.0731440618484436e-05,
+      "loss": 0.6133,
+      "step": 3235
+    },
+    {
+      "epoch": 1.1818513784918752,
+      "grad_norm": 0.6579003930091858,
+      "learning_rate": 2.0717104856271152e-05,
+      "loss": 0.6404,
+      "step": 3236
+    },
+    {
+      "epoch": 1.1822165419025015,
+      "grad_norm": 0.705596923828125,
+      "learning_rate": 2.0702768725136192e-05,
+      "loss": 0.6221,
+      "step": 3237
+    },
+    {
+      "epoch": 1.1825817053131276,
+      "grad_norm": 0.9739445447921753,
+      "learning_rate": 2.068843223245492e-05,
+      "loss": 0.5918,
+      "step": 3238
+    },
+    {
+      "epoch": 1.1829468687237539,
+      "grad_norm": 1.2284148931503296,
+      "learning_rate": 2.0674095385602885e-05,
+      "loss": 0.6329,
+      "step": 3239
+    },
+    {
+      "epoch": 1.1833120321343802,
+      "grad_norm": 1.154626488685608,
+      "learning_rate": 2.0659758191955833e-05,
+      "loss": 0.6033,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1836771955450063,
+      "grad_norm": 0.774357795715332,
+      "learning_rate": 2.0645420658889662e-05,
+      "loss": 0.626,
+      "step": 3241
+    },
+    {
+      "epoch": 1.1840423589556326,
+      "grad_norm": 0.7407965064048767,
+      "learning_rate": 2.0631082793780464e-05,
+      "loss": 0.6146,
+      "step": 3242
+    },
+    {
+      "epoch": 1.184407522366259,
+      "grad_norm": 1.034700870513916,
+      "learning_rate": 2.0616744604004496e-05,
+      "loss": 0.6119,
+      "step": 3243
+    },
+    {
+      "epoch": 1.1847726857768852,
+      "grad_norm": 0.7460023164749146,
+      "learning_rate": 2.0602406096938168e-05,
+      "loss": 0.6153,
+      "step": 3244
+    },
+    {
+      "epoch": 1.1851378491875115,
+      "grad_norm": 0.7647458910942078,
+      "learning_rate": 2.058806727995808e-05,
+      "loss": 0.6318,
+      "step": 3245
+    },
+    {
+      "epoch": 1.1855030125981376,
+      "grad_norm": 1.2149773836135864,
+      "learning_rate": 2.0573728160440972e-05,
+      "loss": 0.5822,
+      "step": 3246
+    },
+    {
+      "epoch": 1.185868176008764,
+      "grad_norm": 0.7748311161994934,
+      "learning_rate": 2.0559388745763754e-05,
+      "loss": 0.5909,
+      "step": 3247
+    },
+    {
+      "epoch": 1.1862333394193902,
+      "grad_norm": 0.9910469055175781,
+      "learning_rate": 2.0545049043303463e-05,
+      "loss": 0.5994,
+      "step": 3248
+    },
+    {
+      "epoch": 1.1865985028300163,
+      "grad_norm": 0.857268750667572,
+      "learning_rate": 2.0530709060437323e-05,
+      "loss": 0.6029,
+      "step": 3249
+    },
+    {
+      "epoch": 1.1869636662406426,
+      "grad_norm": 1.1047351360321045,
+      "learning_rate": 2.0516368804542662e-05,
+      "loss": 0.6266,
+      "step": 3250
+    },
+    {
+      "epoch": 1.187328829651269,
+      "grad_norm": 0.9308289885520935,
+      "learning_rate": 2.050202828299697e-05,
+      "loss": 0.6114,
+      "step": 3251
+    },
+    {
+      "epoch": 1.1876939930618953,
+      "grad_norm": 0.881679356098175,
+      "learning_rate": 2.0487687503177874e-05,
+      "loss": 0.5991,
+      "step": 3252
+    },
+    {
+      "epoch": 1.1880591564725214,
+      "grad_norm": 0.7082407474517822,
+      "learning_rate": 2.0473346472463125e-05,
+      "loss": 0.6062,
+      "step": 3253
+    },
+    {
+      "epoch": 1.1884243198831477,
+      "grad_norm": 1.4497092962265015,
+      "learning_rate": 2.045900519823061e-05,
+      "loss": 0.6298,
+      "step": 3254
+    },
+    {
+      "epoch": 1.188789483293774,
+      "grad_norm": 1.1088813543319702,
+      "learning_rate": 2.044466368785834e-05,
+      "loss": 0.6503,
+      "step": 3255
+    },
+    {
+      "epoch": 1.1891546467044003,
+      "grad_norm": 0.7719933986663818,
+      "learning_rate": 2.0430321948724447e-05,
+      "loss": 0.6332,
+      "step": 3256
+    },
+    {
+      "epoch": 1.1895198101150264,
+      "grad_norm": 0.6202806830406189,
+      "learning_rate": 2.041597998820718e-05,
+      "loss": 0.6081,
+      "step": 3257
+    },
+    {
+      "epoch": 1.1898849735256527,
+      "grad_norm": 0.73605877161026,
+      "learning_rate": 2.0401637813684897e-05,
+      "loss": 0.5995,
+      "step": 3258
+    },
+    {
+      "epoch": 1.190250136936279,
+      "grad_norm": 0.9471248984336853,
+      "learning_rate": 2.038729543253608e-05,
+      "loss": 0.6298,
+      "step": 3259
+    },
+    {
+      "epoch": 1.1906153003469053,
+      "grad_norm": 1.0479601621627808,
+      "learning_rate": 2.0372952852139297e-05,
+      "loss": 0.6345,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1909804637575314,
+      "grad_norm": 0.7535351514816284,
+      "learning_rate": 2.0358610079873248e-05,
+      "loss": 0.64,
+      "step": 3261
+    },
+    {
+      "epoch": 1.1913456271681577,
+      "grad_norm": 1.083878755569458,
+      "learning_rate": 2.0344267123116697e-05,
+      "loss": 0.6246,
+      "step": 3262
+    },
+    {
+      "epoch": 1.191710790578784,
+      "grad_norm": 1.0116360187530518,
+      "learning_rate": 2.0329923989248525e-05,
+      "loss": 0.6339,
+      "step": 3263
+    },
+    {
+      "epoch": 1.1920759539894104,
+      "grad_norm": 0.9353020787239075,
+      "learning_rate": 2.0315580685647703e-05,
+      "loss": 0.6057,
+      "step": 3264
+    },
+    {
+      "epoch": 1.1924411174000364,
+      "grad_norm": 0.7070524096488953,
+      "learning_rate": 2.0301237219693278e-05,
+      "loss": 0.6249,
+      "step": 3265
+    },
+    {
+      "epoch": 1.1928062808106628,
+      "grad_norm": 0.8464820384979248,
+      "learning_rate": 2.0286893598764393e-05,
+      "loss": 0.5947,
+      "step": 3266
+    },
+    {
+      "epoch": 1.193171444221289,
+      "grad_norm": 0.9669051170349121,
+      "learning_rate": 2.0272549830240265e-05,
+      "loss": 0.588,
+      "step": 3267
+    },
+    {
+      "epoch": 1.1935366076319154,
+      "grad_norm": 0.8157206773757935,
+      "learning_rate": 2.0258205921500183e-05,
+      "loss": 0.5887,
+      "step": 3268
+    },
+    {
+      "epoch": 1.1939017710425415,
+      "grad_norm": 1.0829534530639648,
+      "learning_rate": 2.024386187992352e-05,
+      "loss": 0.5842,
+      "step": 3269
+    },
+    {
+      "epoch": 1.1942669344531678,
+      "grad_norm": 0.7082162499427795,
+      "learning_rate": 2.02295177128897e-05,
+      "loss": 0.6026,
+      "step": 3270
+    },
+    {
+      "epoch": 1.194632097863794,
+      "grad_norm": 0.8928137421607971,
+      "learning_rate": 2.0215173427778234e-05,
+      "loss": 0.6482,
+      "step": 3271
+    },
+    {
+      "epoch": 1.1949972612744202,
+      "grad_norm": 0.8363444209098816,
+      "learning_rate": 2.0200829031968667e-05,
+      "loss": 0.5997,
+      "step": 3272
+    },
+    {
+      "epoch": 1.1953624246850465,
+      "grad_norm": 0.9447734355926514,
+      "learning_rate": 2.018648453284062e-05,
+      "loss": 0.6057,
+      "step": 3273
+    },
+    {
+      "epoch": 1.1957275880956728,
+      "grad_norm": 1.049188494682312,
+      "learning_rate": 2.017213993777377e-05,
+      "loss": 0.6091,
+      "step": 3274
+    },
+    {
+      "epoch": 1.1960927515062991,
+      "grad_norm": 1.544588565826416,
+      "learning_rate": 2.0157795254147826e-05,
+      "loss": 0.6389,
+      "step": 3275
+    },
+    {
+      "epoch": 1.1964579149169254,
+      "grad_norm": 0.8541837930679321,
+      "learning_rate": 2.0143450489342563e-05,
+      "loss": 0.622,
+      "step": 3276
+    },
+    {
+      "epoch": 1.1968230783275515,
+      "grad_norm": 0.9997867941856384,
+      "learning_rate": 2.012910565073777e-05,
+      "loss": 0.5931,
+      "step": 3277
+    },
+    {
+      "epoch": 1.1971882417381778,
+      "grad_norm": 0.8798093795776367,
+      "learning_rate": 2.0114760745713305e-05,
+      "loss": 0.6131,
+      "step": 3278
+    },
+    {
+      "epoch": 1.1975534051488042,
+      "grad_norm": 0.7250221371650696,
+      "learning_rate": 2.010041578164904e-05,
+      "loss": 0.6046,
+      "step": 3279
+    },
+    {
+      "epoch": 1.1979185685594302,
+      "grad_norm": 0.967989981174469,
+      "learning_rate": 2.0086070765924886e-05,
+      "loss": 0.6251,
+      "step": 3280
+    },
+    {
+      "epoch": 1.1982837319700566,
+      "grad_norm": 1.0515309572219849,
+      "learning_rate": 2.0071725705920776e-05,
+      "loss": 0.5974,
+      "step": 3281
+    },
+    {
+      "epoch": 1.1986488953806829,
+      "grad_norm": 2.368988037109375,
+      "learning_rate": 2.0057380609016666e-05,
+      "loss": 0.6284,
+      "step": 3282
+    },
+    {
+      "epoch": 1.1990140587913092,
+      "grad_norm": 1.0001410245895386,
+      "learning_rate": 2.0043035482592543e-05,
+      "loss": 0.582,
+      "step": 3283
+    },
+    {
+      "epoch": 1.1993792222019355,
+      "grad_norm": 1.3888506889343262,
+      "learning_rate": 2.0028690334028384e-05,
+      "loss": 0.636,
+      "step": 3284
+    },
+    {
+      "epoch": 1.1997443856125616,
+      "grad_norm": 0.8603945970535278,
+      "learning_rate": 2.0014345170704202e-05,
+      "loss": 0.6385,
+      "step": 3285
+    },
+    {
+      "epoch": 1.200109549023188,
+      "grad_norm": 2.1296420097351074,
+      "learning_rate": 2e-05,
+      "loss": 0.6085,
+      "step": 3286
+    },
+    {
+      "epoch": 1.2004747124338142,
+      "grad_norm": 1.1616238355636597,
+      "learning_rate": 1.99856548292958e-05,
+      "loss": 0.6325,
+      "step": 3287
+    },
+    {
+      "epoch": 1.2008398758444403,
+      "grad_norm": 0.7582462430000305,
+      "learning_rate": 1.9971309665971623e-05,
+      "loss": 0.6052,
+      "step": 3288
+    },
+    {
+      "epoch": 1.2012050392550666,
+      "grad_norm": 0.8241375088691711,
+      "learning_rate": 1.995696451740746e-05,
+      "loss": 0.6307,
+      "step": 3289
+    },
+    {
+      "epoch": 1.201570202665693,
+      "grad_norm": 1.116804599761963,
+      "learning_rate": 1.9942619390983334e-05,
+      "loss": 0.5859,
+      "step": 3290
+    },
+    {
+      "epoch": 1.2019353660763192,
+      "grad_norm": 0.7887852191925049,
+      "learning_rate": 1.9928274294079227e-05,
+      "loss": 0.612,
+      "step": 3291
+    },
+    {
+      "epoch": 1.2023005294869453,
+      "grad_norm": 0.7873060703277588,
+      "learning_rate": 1.9913929234075117e-05,
+      "loss": 0.598,
+      "step": 3292
+    },
+    {
+      "epoch": 1.2026656928975716,
+      "grad_norm": 0.9905545711517334,
+      "learning_rate": 1.989958421835097e-05,
+      "loss": 0.6238,
+      "step": 3293
+    },
+    {
+      "epoch": 1.203030856308198,
+      "grad_norm": 0.8429453372955322,
+      "learning_rate": 1.9885239254286705e-05,
+      "loss": 0.642,
+      "step": 3294
+    },
+    {
+      "epoch": 1.2033960197188243,
+      "grad_norm": 1.3780677318572998,
+      "learning_rate": 1.987089434926224e-05,
+      "loss": 0.5864,
+      "step": 3295
+    },
+    {
+      "epoch": 1.2037611831294504,
+      "grad_norm": 1.0858463048934937,
+      "learning_rate": 1.9856549510657447e-05,
+      "loss": 0.6095,
+      "step": 3296
+    },
+    {
+      "epoch": 1.2041263465400767,
+      "grad_norm": 1.0253275632858276,
+      "learning_rate": 1.984220474585218e-05,
+      "loss": 0.6207,
+      "step": 3297
+    },
+    {
+      "epoch": 1.204491509950703,
+      "grad_norm": 0.932194173336029,
+      "learning_rate": 1.9827860062226236e-05,
+      "loss": 0.5787,
+      "step": 3298
+    },
+    {
+      "epoch": 1.2048566733613293,
+      "grad_norm": 0.9060006737709045,
+      "learning_rate": 1.9813515467159382e-05,
+      "loss": 0.619,
+      "step": 3299
+    },
+    {
+      "epoch": 1.2052218367719554,
+      "grad_norm": 0.8561171889305115,
+      "learning_rate": 1.979917096803134e-05,
+      "loss": 0.61,
+      "step": 3300
+    },
+    {
+      "epoch": 1.2055870001825817,
+      "grad_norm": 0.7535567283630371,
+      "learning_rate": 1.9784826572221773e-05,
+      "loss": 0.6067,
+      "step": 3301
+    },
+    {
+      "epoch": 1.205952163593208,
+      "grad_norm": 1.2940592765808105,
+      "learning_rate": 1.9770482287110305e-05,
+      "loss": 0.6376,
+      "step": 3302
+    },
+    {
+      "epoch": 1.206317327003834,
+      "grad_norm": 0.8710494637489319,
+      "learning_rate": 1.9756138120076484e-05,
+      "loss": 0.5845,
+      "step": 3303
+    },
+    {
+      "epoch": 1.2066824904144604,
+      "grad_norm": 0.9782663583755493,
+      "learning_rate": 1.974179407849982e-05,
+      "loss": 0.5915,
+      "step": 3304
+    },
+    {
+      "epoch": 1.2070476538250867,
+      "grad_norm": 0.6249943375587463,
+      "learning_rate": 1.9727450169759738e-05,
+      "loss": 0.6279,
+      "step": 3305
+    },
+    {
+      "epoch": 1.207412817235713,
+      "grad_norm": 1.0076680183410645,
+      "learning_rate": 1.971310640123561e-05,
+      "loss": 0.6116,
+      "step": 3306
+    },
+    {
+      "epoch": 1.2077779806463393,
+      "grad_norm": 1.613810658454895,
+      "learning_rate": 1.9698762780306732e-05,
+      "loss": 0.6313,
+      "step": 3307
+    },
+    {
+      "epoch": 1.2081431440569654,
+      "grad_norm": 0.8273488879203796,
+      "learning_rate": 1.9684419314352307e-05,
+      "loss": 0.6171,
+      "step": 3308
+    },
+    {
+      "epoch": 1.2085083074675917,
+      "grad_norm": 0.8147680759429932,
+      "learning_rate": 1.9670076010751478e-05,
+      "loss": 0.6255,
+      "step": 3309
+    },
+    {
+      "epoch": 1.208873470878218,
+      "grad_norm": 1.2720317840576172,
+      "learning_rate": 1.965573287688331e-05,
+      "loss": 0.6363,
+      "step": 3310
+    },
+    {
+      "epoch": 1.2092386342888441,
+      "grad_norm": 0.7890989184379578,
+      "learning_rate": 1.964138992012676e-05,
+      "loss": 0.5869,
+      "step": 3311
+    },
+    {
+      "epoch": 1.2096037976994705,
+      "grad_norm": 0.931955873966217,
+      "learning_rate": 1.9627047147860706e-05,
+      "loss": 0.6323,
+      "step": 3312
+    },
+    {
+      "epoch": 1.2099689611100968,
+      "grad_norm": 0.6453506946563721,
+      "learning_rate": 1.9612704567463926e-05,
+      "loss": 0.639,
+      "step": 3313
+    },
+    {
+      "epoch": 1.210334124520723,
+      "grad_norm": 0.8512594103813171,
+      "learning_rate": 1.959836218631511e-05,
+      "loss": 0.6142,
+      "step": 3314
+    },
+    {
+      "epoch": 1.2106992879313494,
+      "grad_norm": 1.2556297779083252,
+      "learning_rate": 1.9584020011792825e-05,
+      "loss": 0.6037,
+      "step": 3315
+    },
+    {
+      "epoch": 1.2110644513419755,
+      "grad_norm": 0.8126077651977539,
+      "learning_rate": 1.956967805127556e-05,
+      "loss": 0.6229,
+      "step": 3316
+    },
+    {
+      "epoch": 1.2114296147526018,
+      "grad_norm": 0.7425557374954224,
+      "learning_rate": 1.955533631214166e-05,
+      "loss": 0.5986,
+      "step": 3317
+    },
+    {
+      "epoch": 1.2117947781632281,
+      "grad_norm": 0.9697204828262329,
+      "learning_rate": 1.9540994801769392e-05,
+      "loss": 0.5703,
+      "step": 3318
+    },
+    {
+      "epoch": 1.2121599415738542,
+      "grad_norm": 0.7134398221969604,
+      "learning_rate": 1.952665352753688e-05,
+      "loss": 0.6407,
+      "step": 3319
+    },
+    {
+      "epoch": 1.2125251049844805,
+      "grad_norm": 1.0264909267425537,
+      "learning_rate": 1.9512312496822136e-05,
+      "loss": 0.6099,
+      "step": 3320
+    },
+    {
+      "epoch": 1.2128902683951068,
+      "grad_norm": 0.814012348651886,
+      "learning_rate": 1.949797171700304e-05,
+      "loss": 0.6287,
+      "step": 3321
+    },
+    {
+      "epoch": 1.2132554318057331,
+      "grad_norm": 0.7063661813735962,
+      "learning_rate": 1.9483631195457348e-05,
+      "loss": 0.5926,
+      "step": 3322
+    },
+    {
+      "epoch": 1.2136205952163592,
+      "grad_norm": 0.959875762462616,
+      "learning_rate": 1.9469290939562684e-05,
+      "loss": 0.5814,
+      "step": 3323
+    },
+    {
+      "epoch": 1.2139857586269855,
+      "grad_norm": 1.2501739263534546,
+      "learning_rate": 1.945495095669654e-05,
+      "loss": 0.6395,
+      "step": 3324
+    },
+    {
+      "epoch": 1.2143509220376119,
+      "grad_norm": 1.0051590204238892,
+      "learning_rate": 1.9440611254236253e-05,
+      "loss": 0.6177,
+      "step": 3325
+    },
+    {
+      "epoch": 1.2147160854482382,
+      "grad_norm": 0.9452838897705078,
+      "learning_rate": 1.942627183955903e-05,
+      "loss": 0.5999,
+      "step": 3326
+    },
+    {
+      "epoch": 1.2150812488588643,
+      "grad_norm": 0.8062846660614014,
+      "learning_rate": 1.9411932720041926e-05,
+      "loss": 0.5997,
+      "step": 3327
+    },
+    {
+      "epoch": 1.2154464122694906,
+      "grad_norm": 1.2123805284500122,
+      "learning_rate": 1.939759390306184e-05,
+      "loss": 0.593,
+      "step": 3328
+    },
+    {
+      "epoch": 1.2158115756801169,
+      "grad_norm": 0.7379060983657837,
+      "learning_rate": 1.9383255395995514e-05,
+      "loss": 0.63,
+      "step": 3329
+    },
+    {
+      "epoch": 1.2161767390907432,
+      "grad_norm": 0.9123709797859192,
+      "learning_rate": 1.9368917206219536e-05,
+      "loss": 0.6281,
+      "step": 3330
+    },
+    {
+      "epoch": 1.2165419025013693,
+      "grad_norm": 0.8074901103973389,
+      "learning_rate": 1.935457934111034e-05,
+      "loss": 0.6359,
+      "step": 3331
+    },
+    {
+      "epoch": 1.2169070659119956,
+      "grad_norm": 0.7655048370361328,
+      "learning_rate": 1.9340241808044167e-05,
+      "loss": 0.6021,
+      "step": 3332
+    },
+    {
+      "epoch": 1.217272229322622,
+      "grad_norm": 0.9054430723190308,
+      "learning_rate": 1.932590461439712e-05,
+      "loss": 0.5834,
+      "step": 3333
+    },
+    {
+      "epoch": 1.2176373927332482,
+      "grad_norm": 0.7461997270584106,
+      "learning_rate": 1.931156776754509e-05,
+      "loss": 0.6249,
+      "step": 3334
+    },
+    {
+      "epoch": 1.2180025561438743,
+      "grad_norm": 0.7512338161468506,
+      "learning_rate": 1.9297231274863818e-05,
+      "loss": 0.6102,
+      "step": 3335
+    },
+    {
+      "epoch": 1.2183677195545006,
+      "grad_norm": 0.8803744912147522,
+      "learning_rate": 1.9282895143728858e-05,
+      "loss": 0.6065,
+      "step": 3336
+    },
+    {
+      "epoch": 1.218732882965127,
+      "grad_norm": 0.9500965476036072,
+      "learning_rate": 1.9268559381515567e-05,
+      "loss": 0.5828,
+      "step": 3337
+    },
+    {
+      "epoch": 1.2190980463757533,
+      "grad_norm": 1.378042221069336,
+      "learning_rate": 1.925422399559913e-05,
+      "loss": 0.6112,
+      "step": 3338
+    },
+    {
+      "epoch": 1.2194632097863793,
+      "grad_norm": 1.0525225400924683,
+      "learning_rate": 1.9239888993354513e-05,
+      "loss": 0.5964,
+      "step": 3339
+    },
+    {
+      "epoch": 1.2198283731970057,
+      "grad_norm": 1.1789579391479492,
+      "learning_rate": 1.9225554382156514e-05,
+      "loss": 0.6137,
+      "step": 3340
+    },
+    {
+      "epoch": 1.220193536607632,
+      "grad_norm": 1.065134048461914,
+      "learning_rate": 1.9211220169379706e-05,
+      "loss": 0.6132,
+      "step": 3341
+    },
+    {
+      "epoch": 1.220558700018258,
+      "grad_norm": 1.0772281885147095,
+      "learning_rate": 1.9196886362398467e-05,
+      "loss": 0.6047,
+      "step": 3342
+    },
+    {
+      "epoch": 1.2209238634288844,
+      "grad_norm": 0.8940218091011047,
+      "learning_rate": 1.9182552968586973e-05,
+      "loss": 0.5607,
+      "step": 3343
+    },
+    {
+      "epoch": 1.2212890268395107,
+      "grad_norm": 1.1024268865585327,
+      "learning_rate": 1.9168219995319166e-05,
+      "loss": 0.6625,
+      "step": 3344
+    },
+    {
+      "epoch": 1.221654190250137,
+      "grad_norm": 0.9297084808349609,
+      "learning_rate": 1.9153887449968802e-05,
+      "loss": 0.5984,
+      "step": 3345
+    },
+    {
+      "epoch": 1.2220193536607633,
+      "grad_norm": 0.9231723546981812,
+      "learning_rate": 1.9139555339909388e-05,
+      "loss": 0.6176,
+      "step": 3346
+    },
+    {
+      "epoch": 1.2223845170713894,
+      "grad_norm": 0.9609585404396057,
+      "learning_rate": 1.9125223672514233e-05,
+      "loss": 0.6263,
+      "step": 3347
+    },
+    {
+      "epoch": 1.2227496804820157,
+      "grad_norm": 0.7409241199493408,
+      "learning_rate": 1.9110892455156395e-05,
+      "loss": 0.6207,
+      "step": 3348
+    },
+    {
+      "epoch": 1.223114843892642,
+      "grad_norm": 0.7970425486564636,
+      "learning_rate": 1.90965616952087e-05,
+      "loss": 0.6109,
+      "step": 3349
+    },
+    {
+      "epoch": 1.2234800073032681,
+      "grad_norm": 0.944929301738739,
+      "learning_rate": 1.908223140004377e-05,
+      "loss": 0.6056,
+      "step": 3350
+    },
+    {
+      "epoch": 1.2238451707138944,
+      "grad_norm": 1.079415202140808,
+      "learning_rate": 1.906790157703395e-05,
+      "loss": 0.5833,
+      "step": 3351
+    },
+    {
+      "epoch": 1.2242103341245207,
+      "grad_norm": 0.7950957417488098,
+      "learning_rate": 1.9053572233551365e-05,
+      "loss": 0.6001,
+      "step": 3352
+    },
+    {
+      "epoch": 1.224575497535147,
+      "grad_norm": 1.1039421558380127,
+      "learning_rate": 1.9039243376967893e-05,
+      "loss": 0.6146,
+      "step": 3353
+    },
+    {
+      "epoch": 1.2249406609457731,
+      "grad_norm": 0.9724546670913696,
+      "learning_rate": 1.9024915014655146e-05,
+      "loss": 0.6157,
+      "step": 3354
+    },
+    {
+      "epoch": 1.2253058243563995,
+      "grad_norm": 0.9508349895477295,
+      "learning_rate": 1.9010587153984503e-05,
+      "loss": 0.6254,
+      "step": 3355
+    },
+    {
+      "epoch": 1.2256709877670258,
+      "grad_norm": 0.9141985774040222,
+      "learning_rate": 1.899625980232706e-05,
+      "loss": 0.5978,
+      "step": 3356
+    },
+    {
+      "epoch": 1.226036151177652,
+      "grad_norm": 1.1930100917816162,
+      "learning_rate": 1.8981932967053677e-05,
+      "loss": 0.6092,
+      "step": 3357
+    },
+    {
+      "epoch": 1.2264013145882782,
+      "grad_norm": 0.7728860974311829,
+      "learning_rate": 1.8967606655534926e-05,
+      "loss": 0.6281,
+      "step": 3358
+    },
+    {
+      "epoch": 1.2267664779989045,
+      "grad_norm": 0.7239397168159485,
+      "learning_rate": 1.8953280875141125e-05,
+      "loss": 0.6024,
+      "step": 3359
+    },
+    {
+      "epoch": 1.2271316414095308,
+      "grad_norm": 0.8932539224624634,
+      "learning_rate": 1.893895563324232e-05,
+      "loss": 0.591,
+      "step": 3360
+    },
+    {
+      "epoch": 1.227496804820157,
+      "grad_norm": 0.9221148490905762,
+      "learning_rate": 1.892463093720825e-05,
+      "loss": 0.6439,
+      "step": 3361
+    },
+    {
+      "epoch": 1.2278619682307832,
+      "grad_norm": 0.7435846328735352,
+      "learning_rate": 1.891030679440842e-05,
+      "loss": 0.6223,
+      "step": 3362
+    },
+    {
+      "epoch": 1.2282271316414095,
+      "grad_norm": 1.098710298538208,
+      "learning_rate": 1.889598321221201e-05,
+      "loss": 0.6279,
+      "step": 3363
+    },
+    {
+      "epoch": 1.2285922950520358,
+      "grad_norm": 1.0755566358566284,
+      "learning_rate": 1.8881660197987937e-05,
+      "loss": 0.5842,
+      "step": 3364
+    },
+    {
+      "epoch": 1.2289574584626621,
+      "grad_norm": 1.1215485334396362,
+      "learning_rate": 1.886733775910482e-05,
+      "loss": 0.6005,
+      "step": 3365
+    },
+    {
+      "epoch": 1.2293226218732882,
+      "grad_norm": 0.8552205562591553,
+      "learning_rate": 1.8853015902930974e-05,
+      "loss": 0.6146,
+      "step": 3366
+    },
+    {
+      "epoch": 1.2296877852839145,
+      "grad_norm": 0.6829099655151367,
+      "learning_rate": 1.8838694636834423e-05,
+      "loss": 0.6265,
+      "step": 3367
+    },
+    {
+      "epoch": 1.2300529486945408,
+      "grad_norm": 1.0725380182266235,
+      "learning_rate": 1.8824373968182875e-05,
+      "loss": 0.5992,
+      "step": 3368
+    },
+    {
+      "epoch": 1.2304181121051672,
+      "grad_norm": 1.746687650680542,
+      "learning_rate": 1.881005390434375e-05,
+      "loss": 0.6139,
+      "step": 3369
+    },
+    {
+      "epoch": 1.2307832755157933,
+      "grad_norm": 1.0292305946350098,
+      "learning_rate": 1.879573445268414e-05,
+      "loss": 0.6255,
+      "step": 3370
+    },
+    {
+      "epoch": 1.2311484389264196,
+      "grad_norm": 1.2852039337158203,
+      "learning_rate": 1.8781415620570832e-05,
+      "loss": 0.6349,
+      "step": 3371
+    },
+    {
+      "epoch": 1.2315136023370459,
+      "grad_norm": 1.2481268644332886,
+      "learning_rate": 1.8767097415370287e-05,
+      "loss": 0.605,
+      "step": 3372
+    },
+    {
+      "epoch": 1.231878765747672,
+      "grad_norm": 0.859007716178894,
+      "learning_rate": 1.8752779844448653e-05,
+      "loss": 0.6245,
+      "step": 3373
+    },
+    {
+      "epoch": 1.2322439291582983,
+      "grad_norm": 1.0601139068603516,
+      "learning_rate": 1.8738462915171736e-05,
+      "loss": 0.6321,
+      "step": 3374
+    },
+    {
+      "epoch": 1.2326090925689246,
+      "grad_norm": 1.0997142791748047,
+      "learning_rate": 1.8724146634905026e-05,
+      "loss": 0.5712,
+      "step": 3375
+    },
+    {
+      "epoch": 1.232974255979551,
+      "grad_norm": 0.720138430595398,
+      "learning_rate": 1.8709831011013678e-05,
+      "loss": 0.5951,
+      "step": 3376
+    },
+    {
+      "epoch": 1.2333394193901772,
+      "grad_norm": 1.0345478057861328,
+      "learning_rate": 1.8695516050862504e-05,
+      "loss": 0.5701,
+      "step": 3377
+    },
+    {
+      "epoch": 1.2337045828008033,
+      "grad_norm": 1.1205919981002808,
+      "learning_rate": 1.8681201761815974e-05,
+      "loss": 0.6008,
+      "step": 3378
+    },
+    {
+      "epoch": 1.2340697462114296,
+      "grad_norm": 1.0405442714691162,
+      "learning_rate": 1.8666888151238217e-05,
+      "loss": 0.6039,
+      "step": 3379
+    },
+    {
+      "epoch": 1.234434909622056,
+      "grad_norm": 0.8944239020347595,
+      "learning_rate": 1.8652575226493012e-05,
+      "loss": 0.6328,
+      "step": 3380
+    },
+    {
+      "epoch": 1.234800073032682,
+      "grad_norm": 0.9236727952957153,
+      "learning_rate": 1.863826299494379e-05,
+      "loss": 0.5893,
+      "step": 3381
+    },
+    {
+      "epoch": 1.2351652364433083,
+      "grad_norm": 0.7447712421417236,
+      "learning_rate": 1.8623951463953605e-05,
+      "loss": 0.6182,
+      "step": 3382
+    },
+    {
+      "epoch": 1.2355303998539346,
+      "grad_norm": 0.8919333219528198,
+      "learning_rate": 1.8609640640885177e-05,
+      "loss": 0.6091,
+      "step": 3383
+    },
+    {
+      "epoch": 1.235895563264561,
+      "grad_norm": 0.9320427179336548,
+      "learning_rate": 1.859533053310085e-05,
+      "loss": 0.6072,
+      "step": 3384
+    },
+    {
+      "epoch": 1.236260726675187,
+      "grad_norm": 0.8821183443069458,
+      "learning_rate": 1.8581021147962593e-05,
+      "loss": 0.619,
+      "step": 3385
+    },
+    {
+      "epoch": 1.2366258900858134,
+      "grad_norm": 1.4499365091323853,
+      "learning_rate": 1.856671249283202e-05,
+      "loss": 0.6027,
+      "step": 3386
+    },
+    {
+      "epoch": 1.2369910534964397,
+      "grad_norm": 1.0097719430923462,
+      "learning_rate": 1.8552404575070363e-05,
+      "loss": 0.6203,
+      "step": 3387
+    },
+    {
+      "epoch": 1.237356216907066,
+      "grad_norm": 0.7847924828529358,
+      "learning_rate": 1.8538097402038452e-05,
+      "loss": 0.6049,
+      "step": 3388
+    },
+    {
+      "epoch": 1.237721380317692,
+      "grad_norm": 3.8774895668029785,
+      "learning_rate": 1.852379098109677e-05,
+      "loss": 0.5725,
+      "step": 3389
+    },
+    {
+      "epoch": 1.2380865437283184,
+      "grad_norm": 0.8568375110626221,
+      "learning_rate": 1.850948531960539e-05,
+      "loss": 0.6049,
+      "step": 3390
+    },
+    {
+      "epoch": 1.2384517071389447,
+      "grad_norm": 0.9716466665267944,
+      "learning_rate": 1.8495180424924003e-05,
+      "loss": 0.594,
+      "step": 3391
+    },
+    {
+      "epoch": 1.238816870549571,
+      "grad_norm": 1.1959747076034546,
+      "learning_rate": 1.84808763044119e-05,
+      "loss": 0.5873,
+      "step": 3392
+    },
+    {
+      "epoch": 1.239182033960197,
+      "grad_norm": 0.9050604104995728,
+      "learning_rate": 1.8466572965427984e-05,
+      "loss": 0.5921,
+      "step": 3393
+    },
+    {
+      "epoch": 1.2395471973708234,
+      "grad_norm": 0.9194764494895935,
+      "learning_rate": 1.845227041533074e-05,
+      "loss": 0.587,
+      "step": 3394
+    },
+    {
+      "epoch": 1.2399123607814497,
+      "grad_norm": 0.8811522126197815,
+      "learning_rate": 1.8437968661478262e-05,
+      "loss": 0.6271,
+      "step": 3395
+    },
+    {
+      "epoch": 1.240277524192076,
+      "grad_norm": 1.0084917545318604,
+      "learning_rate": 1.842366771122823e-05,
+      "loss": 0.5811,
+      "step": 3396
+    },
+    {
+      "epoch": 1.2406426876027021,
+      "grad_norm": 0.7907148003578186,
+      "learning_rate": 1.8409367571937903e-05,
+      "loss": 0.6132,
+      "step": 3397
+    },
+    {
+      "epoch": 1.2410078510133284,
+      "grad_norm": 0.6632513403892517,
+      "learning_rate": 1.8395068250964138e-05,
+      "loss": 0.6138,
+      "step": 3398
+    },
+    {
+      "epoch": 1.2413730144239548,
+      "grad_norm": 0.6575708985328674,
+      "learning_rate": 1.8380769755663348e-05,
+      "loss": 0.6143,
+      "step": 3399
+    },
+    {
+      "epoch": 1.241738177834581,
+      "grad_norm": 0.6493009328842163,
+      "learning_rate": 1.8366472093391553e-05,
+      "loss": 0.6385,
+      "step": 3400
+    },
+    {
+      "epoch": 1.2421033412452072,
+      "grad_norm": 0.8245896697044373,
+      "learning_rate": 1.8352175271504314e-05,
+      "loss": 0.5871,
+      "step": 3401
+    },
+    {
+      "epoch": 1.2424685046558335,
+      "grad_norm": 1.0025814771652222,
+      "learning_rate": 1.833787929735677e-05,
+      "loss": 0.597,
+      "step": 3402
+    },
+    {
+      "epoch": 1.2428336680664598,
+      "grad_norm": 0.953640341758728,
+      "learning_rate": 1.8323584178303632e-05,
+      "loss": 0.5999,
+      "step": 3403
+    },
+    {
+      "epoch": 1.2431988314770859,
+      "grad_norm": 0.8461502194404602,
+      "learning_rate": 1.8309289921699163e-05,
+      "loss": 0.5781,
+      "step": 3404
+    },
+    {
+      "epoch": 1.2435639948877122,
+      "grad_norm": 0.5869039297103882,
+      "learning_rate": 1.8294996534897185e-05,
+      "loss": 0.6208,
+      "step": 3405
+    },
+    {
+      "epoch": 1.2439291582983385,
+      "grad_norm": 1.093299388885498,
+      "learning_rate": 1.8280704025251076e-05,
+      "loss": 0.6124,
+      "step": 3406
+    },
+    {
+      "epoch": 1.2442943217089648,
+      "grad_norm": 1.0694578886032104,
+      "learning_rate": 1.8266412400113747e-05,
+      "loss": 0.5899,
+      "step": 3407
+    },
+    {
+      "epoch": 1.2446594851195911,
+      "grad_norm": 1.018556833267212,
+      "learning_rate": 1.825212166683768e-05,
+      "loss": 0.5649,
+      "step": 3408
+    },
+    {
+      "epoch": 1.2450246485302172,
+      "grad_norm": 1.0864132642745972,
+      "learning_rate": 1.8237831832774877e-05,
+      "loss": 0.5956,
+      "step": 3409
+    },
+    {
+      "epoch": 1.2453898119408435,
+      "grad_norm": 1.1059077978134155,
+      "learning_rate": 1.8223542905276885e-05,
+      "loss": 0.5708,
+      "step": 3410
+    },
+    {
+      "epoch": 1.2457549753514698,
+      "grad_norm": 1.5342904329299927,
+      "learning_rate": 1.820925489169478e-05,
+      "loss": 0.5978,
+      "step": 3411
+    },
+    {
+      "epoch": 1.246120138762096,
+      "grad_norm": 0.9663258790969849,
+      "learning_rate": 1.819496779937918e-05,
+      "loss": 0.5566,
+      "step": 3412
+    },
+    {
+      "epoch": 1.2464853021727222,
+      "grad_norm": 0.8954864740371704,
+      "learning_rate": 1.818068163568022e-05,
+      "loss": 0.6285,
+      "step": 3413
+    },
+    {
+      "epoch": 1.2468504655833486,
+      "grad_norm": 1.2142242193222046,
+      "learning_rate": 1.8166396407947546e-05,
+      "loss": 0.5819,
+      "step": 3414
+    },
+    {
+      "epoch": 1.2472156289939749,
+      "grad_norm": 0.9968467354774475,
+      "learning_rate": 1.8152112123530345e-05,
+      "loss": 0.6378,
+      "step": 3415
+    },
+    {
+      "epoch": 1.2475807924046012,
+      "grad_norm": 0.8761385679244995,
+      "learning_rate": 1.8137828789777302e-05,
+      "loss": 0.616,
+      "step": 3416
+    },
+    {
+      "epoch": 1.2479459558152273,
+      "grad_norm": 0.755793035030365,
+      "learning_rate": 1.8123546414036623e-05,
+      "loss": 0.5758,
+      "step": 3417
+    },
+    {
+      "epoch": 1.2483111192258536,
+      "grad_norm": 1.3821804523468018,
+      "learning_rate": 1.810926500365602e-05,
+      "loss": 0.5734,
+      "step": 3418
+    },
+    {
+      "epoch": 1.24867628263648,
+      "grad_norm": 0.9680818915367126,
+      "learning_rate": 1.8094984565982697e-05,
+      "loss": 0.6304,
+      "step": 3419
+    },
+    {
+      "epoch": 1.249041446047106,
+      "grad_norm": 0.9872563481330872,
+      "learning_rate": 1.8080705108363376e-05,
+      "loss": 0.595,
+      "step": 3420
+    },
+    {
+      "epoch": 1.2494066094577323,
+      "grad_norm": 0.8364111185073853,
+      "learning_rate": 1.8066426638144253e-05,
+      "loss": 0.5807,
+      "step": 3421
+    },
+    {
+      "epoch": 1.2497717728683586,
+      "grad_norm": 1.2990816831588745,
+      "learning_rate": 1.8052149162671045e-05,
+      "loss": 0.6285,
+      "step": 3422
+    },
+    {
+      "epoch": 1.250136936278985,
+      "grad_norm": 0.6941619515419006,
+      "learning_rate": 1.8037872689288923e-05,
+      "loss": 0.6104,
+      "step": 3423
+    },
+    {
+      "epoch": 1.2505020996896112,
+      "grad_norm": 1.0219366550445557,
+      "learning_rate": 1.802359722534257e-05,
+      "loss": 0.629,
+      "step": 3424
+    },
+    {
+      "epoch": 1.2508672631002373,
+      "grad_norm": 0.88230961561203,
+      "learning_rate": 1.800932277817614e-05,
+      "loss": 0.6431,
+      "step": 3425
+    },
+    {
+      "epoch": 1.2512324265108636,
+      "grad_norm": 0.7840175032615662,
+      "learning_rate": 1.7995049355133254e-05,
+      "loss": 0.6221,
+      "step": 3426
+    },
+    {
+      "epoch": 1.2515975899214897,
+      "grad_norm": 1.0608835220336914,
+      "learning_rate": 1.798077696355703e-05,
+      "loss": 0.6035,
+      "step": 3427
+    },
+    {
+      "epoch": 1.251962753332116,
+      "grad_norm": 0.9382961988449097,
+      "learning_rate": 1.7966505610790022e-05,
+      "loss": 0.5972,
+      "step": 3428
+    },
+    {
+      "epoch": 1.2523279167427424,
+      "grad_norm": 0.9159266352653503,
+      "learning_rate": 1.795223530417428e-05,
+      "loss": 0.5949,
+      "step": 3429
+    },
+    {
+      "epoch": 1.2526930801533687,
+      "grad_norm": 0.7968629002571106,
+      "learning_rate": 1.7937966051051306e-05,
+      "loss": 0.6394,
+      "step": 3430
+    },
+    {
+      "epoch": 1.253058243563995,
+      "grad_norm": 0.9000434875488281,
+      "learning_rate": 1.7923697858762054e-05,
+      "loss": 0.6027,
+      "step": 3431
+    },
+    {
+      "epoch": 1.253423406974621,
+      "grad_norm": 1.0707937479019165,
+      "learning_rate": 1.7909430734646936e-05,
+      "loss": 0.5795,
+      "step": 3432
+    },
+    {
+      "epoch": 1.2537885703852474,
+      "grad_norm": 0.9548970460891724,
+      "learning_rate": 1.789516468604581e-05,
+      "loss": 0.5969,
+      "step": 3433
+    },
+    {
+      "epoch": 1.2541537337958737,
+      "grad_norm": 0.6696003079414368,
+      "learning_rate": 1.7880899720297998e-05,
+      "loss": 0.6089,
+      "step": 3434
+    },
+    {
+      "epoch": 1.2545188972064998,
+      "grad_norm": 0.9154466390609741,
+      "learning_rate": 1.7866635844742243e-05,
+      "loss": 0.6233,
+      "step": 3435
+    },
+    {
+      "epoch": 1.254884060617126,
+      "grad_norm": 0.8010258674621582,
+      "learning_rate": 1.785237306671674e-05,
+      "loss": 0.6129,
+      "step": 3436
+    },
+    {
+      "epoch": 1.2552492240277524,
+      "grad_norm": 0.9692965149879456,
+      "learning_rate": 1.7838111393559115e-05,
+      "loss": 0.577,
+      "step": 3437
+    },
+    {
+      "epoch": 1.2556143874383787,
+      "grad_norm": 0.7379612326622009,
+      "learning_rate": 1.7823850832606425e-05,
+      "loss": 0.6067,
+      "step": 3438
+    },
+    {
+      "epoch": 1.255979550849005,
+      "grad_norm": 0.8523130416870117,
+      "learning_rate": 1.7809591391195162e-05,
+      "loss": 0.5649,
+      "step": 3439
+    },
+    {
+      "epoch": 1.2563447142596311,
+      "grad_norm": 0.9611847996711731,
+      "learning_rate": 1.7795333076661238e-05,
+      "loss": 0.5861,
+      "step": 3440
+    },
+    {
+      "epoch": 1.2567098776702574,
+      "grad_norm": 0.828985869884491,
+      "learning_rate": 1.7781075896339968e-05,
+      "loss": 0.616,
+      "step": 3441
+    },
+    {
+      "epoch": 1.2570750410808837,
+      "grad_norm": 1.1675282716751099,
+      "learning_rate": 1.7766819857566116e-05,
+      "loss": 0.5979,
+      "step": 3442
+    },
+    {
+      "epoch": 1.2574402044915098,
+      "grad_norm": 1.1937335729599,
+      "learning_rate": 1.7752564967673832e-05,
+      "loss": 0.6245,
+      "step": 3443
+    },
+    {
+      "epoch": 1.2578053679021362,
+      "grad_norm": 0.7989965081214905,
+      "learning_rate": 1.7738311233996686e-05,
+      "loss": 0.6193,
+      "step": 3444
+    },
+    {
+      "epoch": 1.2581705313127625,
+      "grad_norm": 2.223703384399414,
+      "learning_rate": 1.7724058663867656e-05,
+      "loss": 0.6288,
+      "step": 3445
+    },
+    {
+      "epoch": 1.2585356947233888,
+      "grad_norm": 0.8401123881340027,
+      "learning_rate": 1.770980726461912e-05,
+      "loss": 0.618,
+      "step": 3446
+    },
+    {
+      "epoch": 1.258900858134015,
+      "grad_norm": 1.1787495613098145,
+      "learning_rate": 1.769555704358284e-05,
+      "loss": 0.5753,
+      "step": 3447
+    },
+    {
+      "epoch": 1.2592660215446412,
+      "grad_norm": 1.0285699367523193,
+      "learning_rate": 1.7681308008089993e-05,
+      "loss": 0.5934,
+      "step": 3448
+    },
+    {
+      "epoch": 1.2596311849552675,
+      "grad_norm": 0.884979248046875,
+      "learning_rate": 1.766706016547113e-05,
+      "loss": 0.6229,
+      "step": 3449
+    },
+    {
+      "epoch": 1.2599963483658938,
+      "grad_norm": 0.7082614898681641,
+      "learning_rate": 1.7652813523056195e-05,
+      "loss": 0.6307,
+      "step": 3450
+    },
+    {
+      "epoch": 1.26036151177652,
+      "grad_norm": 1.0871187448501587,
+      "learning_rate": 1.763856808817452e-05,
+      "loss": 0.5687,
+      "step": 3451
+    },
+    {
+      "epoch": 1.2607266751871462,
+      "grad_norm": 1.1111515760421753,
+      "learning_rate": 1.7624323868154804e-05,
+      "loss": 0.5856,
+      "step": 3452
+    },
+    {
+      "epoch": 1.2610918385977725,
+      "grad_norm": 1.0516471862792969,
+      "learning_rate": 1.7610080870325135e-05,
+      "loss": 0.5662,
+      "step": 3453
+    },
+    {
+      "epoch": 1.2614570020083988,
+      "grad_norm": 0.9718304872512817,
+      "learning_rate": 1.7595839102012954e-05,
+      "loss": 0.6116,
+      "step": 3454
+    },
+    {
+      "epoch": 1.2618221654190251,
+      "grad_norm": 0.9986475706100464,
+      "learning_rate": 1.7581598570545075e-05,
+      "loss": 0.5891,
+      "step": 3455
+    },
+    {
+      "epoch": 1.2621873288296512,
+      "grad_norm": 0.9629244804382324,
+      "learning_rate": 1.756735928324769e-05,
+      "loss": 0.5981,
+      "step": 3456
+    },
+    {
+      "epoch": 1.2625524922402775,
+      "grad_norm": 1.0716477632522583,
+      "learning_rate": 1.7553121247446337e-05,
+      "loss": 0.5955,
+      "step": 3457
+    },
+    {
+      "epoch": 1.2629176556509039,
+      "grad_norm": 0.8771053552627563,
+      "learning_rate": 1.753888447046592e-05,
+      "loss": 0.579,
+      "step": 3458
+    },
+    {
+      "epoch": 1.26328281906153,
+      "grad_norm": 1.0961087942123413,
+      "learning_rate": 1.7524648959630676e-05,
+      "loss": 0.5978,
+      "step": 3459
+    },
+    {
+      "epoch": 1.2636479824721563,
+      "grad_norm": 2.062915325164795,
+      "learning_rate": 1.7510414722264217e-05,
+      "loss": 0.6278,
+      "step": 3460
+    },
+    {
+      "epoch": 1.2640131458827826,
+      "grad_norm": 0.8420235514640808,
+      "learning_rate": 1.7496181765689485e-05,
+      "loss": 0.5904,
+      "step": 3461
+    },
+    {
+      "epoch": 1.2643783092934089,
+      "grad_norm": 0.8386824727058411,
+      "learning_rate": 1.7481950097228757e-05,
+      "loss": 0.607,
+      "step": 3462
+    },
+    {
+      "epoch": 1.264743472704035,
+      "grad_norm": 0.8218091726303101,
+      "learning_rate": 1.7467719724203667e-05,
+      "loss": 0.5942,
+      "step": 3463
+    },
+    {
+      "epoch": 1.2651086361146613,
+      "grad_norm": 0.8152195811271667,
+      "learning_rate": 1.7453490653935162e-05,
+      "loss": 0.5744,
+      "step": 3464
+    },
+    {
+      "epoch": 1.2654737995252876,
+      "grad_norm": 0.853538990020752,
+      "learning_rate": 1.743926289374353e-05,
+      "loss": 0.6172,
+      "step": 3465
+    },
+    {
+      "epoch": 1.2658389629359137,
+      "grad_norm": 0.6543163657188416,
+      "learning_rate": 1.7425036450948383e-05,
+      "loss": 0.6091,
+      "step": 3466
+    },
+    {
+      "epoch": 1.26620412634654,
+      "grad_norm": 1.062475323677063,
+      "learning_rate": 1.7410811332868664e-05,
+      "loss": 0.586,
+      "step": 3467
+    },
+    {
+      "epoch": 1.2665692897571663,
+      "grad_norm": 0.7827355861663818,
+      "learning_rate": 1.739658754682261e-05,
+      "loss": 0.6071,
+      "step": 3468
+    },
+    {
+      "epoch": 1.2669344531677926,
+      "grad_norm": 0.6963713765144348,
+      "learning_rate": 1.738236510012779e-05,
+      "loss": 0.5623,
+      "step": 3469
+    },
+    {
+      "epoch": 1.267299616578419,
+      "grad_norm": 0.8926814794540405,
+      "learning_rate": 1.7368144000101093e-05,
+      "loss": 0.6138,
+      "step": 3470
+    },
+    {
+      "epoch": 1.267664779989045,
+      "grad_norm": 0.79139244556427,
+      "learning_rate": 1.7353924254058695e-05,
+      "loss": 0.6263,
+      "step": 3471
+    },
+    {
+      "epoch": 1.2680299433996713,
+      "grad_norm": 0.8635314702987671,
+      "learning_rate": 1.7339705869316083e-05,
+      "loss": 0.6343,
+      "step": 3472
+    },
+    {
+      "epoch": 1.2683951068102977,
+      "grad_norm": 0.8019347786903381,
+      "learning_rate": 1.732548885318806e-05,
+      "loss": 0.6206,
+      "step": 3473
+    },
+    {
+      "epoch": 1.2687602702209237,
+      "grad_norm": 0.8886422514915466,
+      "learning_rate": 1.7311273212988692e-05,
+      "loss": 0.6003,
+      "step": 3474
+    },
+    {
+      "epoch": 1.26912543363155,
+      "grad_norm": 1.609634280204773,
+      "learning_rate": 1.729705895603137e-05,
+      "loss": 0.5965,
+      "step": 3475
+    },
+    {
+      "epoch": 1.2694905970421764,
+      "grad_norm": 1.058031439781189,
+      "learning_rate": 1.728284608962875e-05,
+      "loss": 0.5999,
+      "step": 3476
+    },
+    {
+      "epoch": 1.2698557604528027,
+      "grad_norm": 0.8010134696960449,
+      "learning_rate": 1.7268634621092786e-05,
+      "loss": 0.5939,
+      "step": 3477
+    },
+    {
+      "epoch": 1.270220923863429,
+      "grad_norm": 0.746239185333252,
+      "learning_rate": 1.7254424557734703e-05,
+      "loss": 0.5873,
+      "step": 3478
+    },
+    {
+      "epoch": 1.270586087274055,
+      "grad_norm": 0.766264796257019,
+      "learning_rate": 1.7240215906865016e-05,
+      "loss": 0.6213,
+      "step": 3479
+    },
+    {
+      "epoch": 1.2709512506846814,
+      "grad_norm": 0.900159478187561,
+      "learning_rate": 1.722600867579351e-05,
+      "loss": 0.603,
+      "step": 3480
+    },
+    {
+      "epoch": 1.2713164140953077,
+      "grad_norm": 3.7984745502471924,
+      "learning_rate": 1.7211802871829216e-05,
+      "loss": 0.6069,
+      "step": 3481
+    },
+    {
+      "epoch": 1.2716815775059338,
+      "grad_norm": 0.9797561168670654,
+      "learning_rate": 1.719759850228046e-05,
+      "loss": 0.5898,
+      "step": 3482
+    },
+    {
+      "epoch": 1.2720467409165601,
+      "grad_norm": 0.7241623997688293,
+      "learning_rate": 1.7183395574454823e-05,
+      "loss": 0.5902,
+      "step": 3483
+    },
+    {
+      "epoch": 1.2724119043271864,
+      "grad_norm": 1.2979917526245117,
+      "learning_rate": 1.716919409565914e-05,
+      "loss": 0.5622,
+      "step": 3484
+    },
+    {
+      "epoch": 1.2727770677378127,
+      "grad_norm": 0.8148595690727234,
+      "learning_rate": 1.71549940731995e-05,
+      "loss": 0.592,
+      "step": 3485
+    },
+    {
+      "epoch": 1.273142231148439,
+      "grad_norm": 0.7827801704406738,
+      "learning_rate": 1.714079551438125e-05,
+      "loss": 0.6179,
+      "step": 3486
+    },
+    {
+      "epoch": 1.2735073945590651,
+      "grad_norm": 0.8408051133155823,
+      "learning_rate": 1.7126598426508974e-05,
+      "loss": 0.5924,
+      "step": 3487
+    },
+    {
+      "epoch": 1.2738725579696915,
+      "grad_norm": 0.909966766834259,
+      "learning_rate": 1.7112402816886504e-05,
+      "loss": 0.573,
+      "step": 3488
+    },
+    {
+      "epoch": 1.2742377213803178,
+      "grad_norm": 0.9594547748565674,
+      "learning_rate": 1.7098208692816915e-05,
+      "loss": 0.5775,
+      "step": 3489
+    },
+    {
+      "epoch": 1.2746028847909439,
+      "grad_norm": 0.6666066646575928,
+      "learning_rate": 1.708401606160251e-05,
+      "loss": 0.6097,
+      "step": 3490
+    },
+    {
+      "epoch": 1.2749680482015702,
+      "grad_norm": 0.9995865821838379,
+      "learning_rate": 1.706982493054483e-05,
+      "loss": 0.6027,
+      "step": 3491
+    },
+    {
+      "epoch": 1.2753332116121965,
+      "grad_norm": 1.255068063735962,
+      "learning_rate": 1.705563530694464e-05,
+      "loss": 0.5425,
+      "step": 3492
+    },
+    {
+      "epoch": 1.2756983750228228,
+      "grad_norm": 0.9269176721572876,
+      "learning_rate": 1.7041447198101946e-05,
+      "loss": 0.618,
+      "step": 3493
+    },
+    {
+      "epoch": 1.276063538433449,
+      "grad_norm": 0.8610042929649353,
+      "learning_rate": 1.7027260611315936e-05,
+      "loss": 0.6,
+      "step": 3494
+    },
+    {
+      "epoch": 1.2764287018440752,
+      "grad_norm": 0.9045096039772034,
+      "learning_rate": 1.7013075553885063e-05,
+      "loss": 0.5956,
+      "step": 3495
+    },
+    {
+      "epoch": 1.2767938652547015,
+      "grad_norm": 0.6672929525375366,
+      "learning_rate": 1.699889203310695e-05,
+      "loss": 0.5851,
+      "step": 3496
+    },
+    {
+      "epoch": 1.2771590286653276,
+      "grad_norm": 1.0661526918411255,
+      "learning_rate": 1.6984710056278462e-05,
+      "loss": 0.5876,
+      "step": 3497
+    },
+    {
+      "epoch": 1.277524192075954,
+      "grad_norm": 0.7957583665847778,
+      "learning_rate": 1.6970529630695656e-05,
+      "loss": 0.5843,
+      "step": 3498
+    },
+    {
+      "epoch": 1.2778893554865802,
+      "grad_norm": 0.9293861985206604,
+      "learning_rate": 1.695635076365379e-05,
+      "loss": 0.5713,
+      "step": 3499
+    },
+    {
+      "epoch": 1.2782545188972065,
+      "grad_norm": 0.9029990434646606,
+      "learning_rate": 1.694217346244732e-05,
+      "loss": 0.6146,
+      "step": 3500
+    },
+    {
+      "epoch": 1.2786196823078328,
+      "grad_norm": 1.06729257106781,
+      "learning_rate": 1.6927997734369904e-05,
+      "loss": 0.5658,
+      "step": 3501
+    },
+    {
+      "epoch": 1.278984845718459,
+      "grad_norm": 0.5850042104721069,
+      "learning_rate": 1.691382358671438e-05,
+      "loss": 0.6087,
+      "step": 3502
+    },
+    {
+      "epoch": 1.2793500091290853,
+      "grad_norm": 1.2588943243026733,
+      "learning_rate": 1.6899651026772776e-05,
+      "loss": 0.6163,
+      "step": 3503
+    },
+    {
+      "epoch": 1.2797151725397116,
+      "grad_norm": 1.1200400590896606,
+      "learning_rate": 1.6885480061836314e-05,
+      "loss": 0.5824,
+      "step": 3504
+    },
+    {
+      "epoch": 1.2800803359503377,
+      "grad_norm": 0.8963230848312378,
+      "learning_rate": 1.687131069919538e-05,
+      "loss": 0.5888,
+      "step": 3505
+    },
+    {
+      "epoch": 1.280445499360964,
+      "grad_norm": 0.977247416973114,
+      "learning_rate": 1.685714294613955e-05,
+      "loss": 0.5614,
+      "step": 3506
+    },
+    {
+      "epoch": 1.2808106627715903,
+      "grad_norm": 0.7008729577064514,
+      "learning_rate": 1.6842976809957562e-05,
+      "loss": 0.5833,
+      "step": 3507
+    },
+    {
+      "epoch": 1.2811758261822166,
+      "grad_norm": 1.0665124654769897,
+      "learning_rate": 1.6828812297937314e-05,
+      "loss": 0.563,
+      "step": 3508
+    },
+    {
+      "epoch": 1.281540989592843,
+      "grad_norm": 0.7580817937850952,
+      "learning_rate": 1.681464941736589e-05,
+      "loss": 0.5966,
+      "step": 3509
+    },
+    {
+      "epoch": 1.281906153003469,
+      "grad_norm": 2.03531813621521,
+      "learning_rate": 1.6800488175529516e-05,
+      "loss": 0.5923,
+      "step": 3510
+    },
+    {
+      "epoch": 1.2822713164140953,
+      "grad_norm": 0.7629796266555786,
+      "learning_rate": 1.6786328579713593e-05,
+      "loss": 0.5746,
+      "step": 3511
+    },
+    {
+      "epoch": 1.2826364798247216,
+      "grad_norm": 0.9172356724739075,
+      "learning_rate": 1.6772170637202655e-05,
+      "loss": 0.581,
+      "step": 3512
+    },
+    {
+      "epoch": 1.2830016432353477,
+      "grad_norm": 1.3666951656341553,
+      "learning_rate": 1.67580143552804e-05,
+      "loss": 0.5851,
+      "step": 3513
+    },
+    {
+      "epoch": 1.283366806645974,
+      "grad_norm": 0.8977357149124146,
+      "learning_rate": 1.6743859741229667e-05,
+      "loss": 0.615,
+      "step": 3514
+    },
+    {
+      "epoch": 1.2837319700566003,
+      "grad_norm": 0.7211498022079468,
+      "learning_rate": 1.6729706802332433e-05,
+      "loss": 0.5903,
+      "step": 3515
+    },
+    {
+      "epoch": 1.2840971334672266,
+      "grad_norm": 0.8175953030586243,
+      "learning_rate": 1.6715555545869827e-05,
+      "loss": 0.5858,
+      "step": 3516
+    },
+    {
+      "epoch": 1.284462296877853,
+      "grad_norm": 1.1230109930038452,
+      "learning_rate": 1.670140597912209e-05,
+      "loss": 0.5659,
+      "step": 3517
+    },
+    {
+      "epoch": 1.284827460288479,
+      "grad_norm": 1.0331265926361084,
+      "learning_rate": 1.6687258109368617e-05,
+      "loss": 0.5712,
+      "step": 3518
+    },
+    {
+      "epoch": 1.2851926236991054,
+      "grad_norm": 0.9673627614974976,
+      "learning_rate": 1.667311194388791e-05,
+      "loss": 0.597,
+      "step": 3519
+    },
+    {
+      "epoch": 1.2855577871097317,
+      "grad_norm": 0.77398282289505,
+      "learning_rate": 1.665896748995762e-05,
+      "loss": 0.6122,
+      "step": 3520
+    },
+    {
+      "epoch": 1.2859229505203578,
+      "grad_norm": 0.8281552791595459,
+      "learning_rate": 1.6644824754854484e-05,
+      "loss": 0.6006,
+      "step": 3521
+    },
+    {
+      "epoch": 1.286288113930984,
+      "grad_norm": 0.8036080002784729,
+      "learning_rate": 1.663068374585437e-05,
+      "loss": 0.593,
+      "step": 3522
+    },
+    {
+      "epoch": 1.2866532773416104,
+      "grad_norm": 0.6916384100914001,
+      "learning_rate": 1.661654447023227e-05,
+      "loss": 0.5752,
+      "step": 3523
+    },
+    {
+      "epoch": 1.2870184407522367,
+      "grad_norm": 0.7672572731971741,
+      "learning_rate": 1.6602406935262273e-05,
+      "loss": 0.5778,
+      "step": 3524
+    },
+    {
+      "epoch": 1.287383604162863,
+      "grad_norm": 0.9159708023071289,
+      "learning_rate": 1.658827114821756e-05,
+      "loss": 0.5782,
+      "step": 3525
+    },
+    {
+      "epoch": 1.287748767573489,
+      "grad_norm": 0.6897777915000916,
+      "learning_rate": 1.657413711637045e-05,
+      "loss": 0.6111,
+      "step": 3526
+    },
+    {
+      "epoch": 1.2881139309841154,
+      "grad_norm": 0.8971490859985352,
+      "learning_rate": 1.656000484699232e-05,
+      "loss": 0.5769,
+      "step": 3527
+    },
+    {
+      "epoch": 1.2884790943947415,
+      "grad_norm": 0.7451416850090027,
+      "learning_rate": 1.6545874347353655e-05,
+      "loss": 0.6108,
+      "step": 3528
+    },
+    {
+      "epoch": 1.2888442578053678,
+      "grad_norm": 0.9213032722473145,
+      "learning_rate": 1.653174562472403e-05,
+      "loss": 0.6112,
+      "step": 3529
+    },
+    {
+      "epoch": 1.2892094212159941,
+      "grad_norm": 0.91603684425354,
+      "learning_rate": 1.6517618686372114e-05,
+      "loss": 0.5648,
+      "step": 3530
+    },
+    {
+      "epoch": 1.2895745846266204,
+      "grad_norm": 0.8848807215690613,
+      "learning_rate": 1.6503493539565642e-05,
+      "loss": 0.6,
+      "step": 3531
+    },
+    {
+      "epoch": 1.2899397480372468,
+      "grad_norm": 0.9526199698448181,
+      "learning_rate": 1.648937019157144e-05,
+      "loss": 0.5811,
+      "step": 3532
+    },
+    {
+      "epoch": 1.2903049114478728,
+      "grad_norm": 0.939439058303833,
+      "learning_rate": 1.6475248649655398e-05,
+      "loss": 0.5511,
+      "step": 3533
+    },
+    {
+      "epoch": 1.2906700748584992,
+      "grad_norm": 1.0261359214782715,
+      "learning_rate": 1.6461128921082496e-05,
+      "loss": 0.5912,
+      "step": 3534
+    },
+    {
+      "epoch": 1.2910352382691255,
+      "grad_norm": 0.8147271275520325,
+      "learning_rate": 1.6447011013116753e-05,
+      "loss": 0.5732,
+      "step": 3535
+    },
+    {
+      "epoch": 1.2914004016797516,
+      "grad_norm": 0.9904578924179077,
+      "learning_rate": 1.6432894933021266e-05,
+      "loss": 0.5649,
+      "step": 3536
+    },
+    {
+      "epoch": 1.2917655650903779,
+      "grad_norm": 0.7666850090026855,
+      "learning_rate": 1.6418780688058197e-05,
+      "loss": 0.608,
+      "step": 3537
+    },
+    {
+      "epoch": 1.2921307285010042,
+      "grad_norm": 1.113914966583252,
+      "learning_rate": 1.6404668285488763e-05,
+      "loss": 0.5859,
+      "step": 3538
+    },
+    {
+      "epoch": 1.2924958919116305,
+      "grad_norm": 0.7998248934745789,
+      "learning_rate": 1.6390557732573217e-05,
+      "loss": 0.5944,
+      "step": 3539
+    },
+    {
+      "epoch": 1.2928610553222568,
+      "grad_norm": 0.9326393008232117,
+      "learning_rate": 1.637644903657089e-05,
+      "loss": 0.5851,
+      "step": 3540
+    },
+    {
+      "epoch": 1.293226218732883,
+      "grad_norm": 0.6946517825126648,
+      "learning_rate": 1.6362342204740124e-05,
+      "loss": 0.5579,
+      "step": 3541
+    },
+    {
+      "epoch": 1.2935913821435092,
+      "grad_norm": 1.0338560342788696,
+      "learning_rate": 1.6348237244338324e-05,
+      "loss": 0.5815,
+      "step": 3542
+    },
+    {
+      "epoch": 1.2939565455541355,
+      "grad_norm": 0.8650919198989868,
+      "learning_rate": 1.6334134162621923e-05,
+      "loss": 0.5754,
+      "step": 3543
+    },
+    {
+      "epoch": 1.2943217089647616,
+      "grad_norm": 0.7965771555900574,
+      "learning_rate": 1.63200329668464e-05,
+      "loss": 0.6223,
+      "step": 3544
+    },
+    {
+      "epoch": 1.294686872375388,
+      "grad_norm": 0.8801450133323669,
+      "learning_rate": 1.6305933664266242e-05,
+      "loss": 0.5831,
+      "step": 3545
+    },
+    {
+      "epoch": 1.2950520357860142,
+      "grad_norm": 0.7547784447669983,
+      "learning_rate": 1.629183626213498e-05,
+      "loss": 0.6145,
+      "step": 3546
+    },
+    {
+      "epoch": 1.2954171991966406,
+      "grad_norm": 1.00898277759552,
+      "learning_rate": 1.627774076770517e-05,
+      "loss": 0.586,
+      "step": 3547
+    },
+    {
+      "epoch": 1.2957823626072669,
+      "grad_norm": 0.8119264245033264,
+      "learning_rate": 1.6263647188228366e-05,
+      "loss": 0.5754,
+      "step": 3548
+    },
+    {
+      "epoch": 1.296147526017893,
+      "grad_norm": 0.8790151476860046,
+      "learning_rate": 1.624955553095515e-05,
+      "loss": 0.5873,
+      "step": 3549
+    },
+    {
+      "epoch": 1.2965126894285193,
+      "grad_norm": 0.7947791218757629,
+      "learning_rate": 1.623546580313512e-05,
+      "loss": 0.5911,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2968778528391456,
+      "grad_norm": 0.9465892910957336,
+      "learning_rate": 1.622137801201687e-05,
+      "loss": 0.6387,
+      "step": 3551
+    },
+    {
+      "epoch": 1.2972430162497717,
+      "grad_norm": 0.9719758629798889,
+      "learning_rate": 1.620729216484801e-05,
+      "loss": 0.5508,
+      "step": 3552
+    },
+    {
+      "epoch": 1.297608179660398,
+      "grad_norm": 0.9086902141571045,
+      "learning_rate": 1.6193208268875133e-05,
+      "loss": 0.5989,
+      "step": 3553
+    },
+    {
+      "epoch": 1.2979733430710243,
+      "grad_norm": 0.8419565558433533,
+      "learning_rate": 1.617912633134385e-05,
+      "loss": 0.568,
+      "step": 3554
+    },
+    {
+      "epoch": 1.2983385064816506,
+      "grad_norm": 0.9197947978973389,
+      "learning_rate": 1.6165046359498748e-05,
+      "loss": 0.562,
+      "step": 3555
+    },
+    {
+      "epoch": 1.298703669892277,
+      "grad_norm": 0.9575653672218323,
+      "learning_rate": 1.6150968360583404e-05,
+      "loss": 0.5974,
+      "step": 3556
+    },
+    {
+      "epoch": 1.299068833302903,
+      "grad_norm": 1.4620823860168457,
+      "learning_rate": 1.6136892341840386e-05,
+      "loss": 0.6057,
+      "step": 3557
+    },
+    {
+      "epoch": 1.2994339967135293,
+      "grad_norm": 0.8264151811599731,
+      "learning_rate": 1.612281831051124e-05,
+      "loss": 0.5728,
+      "step": 3558
+    },
+    {
+      "epoch": 1.2997991601241556,
+      "grad_norm": 1.0922088623046875,
+      "learning_rate": 1.6108746273836495e-05,
+      "loss": 0.5734,
+      "step": 3559
+    },
+    {
+      "epoch": 1.3001643235347817,
+      "grad_norm": 0.9147183895111084,
+      "learning_rate": 1.6094676239055654e-05,
+      "loss": 0.5759,
+      "step": 3560
+    },
+    {
+      "epoch": 1.300529486945408,
+      "grad_norm": 0.938529372215271,
+      "learning_rate": 1.6080608213407164e-05,
+      "loss": 0.5618,
+      "step": 3561
+    },
+    {
+      "epoch": 1.3008946503560344,
+      "grad_norm": 0.7035357356071472,
+      "learning_rate": 1.606654220412848e-05,
+      "loss": 0.5869,
+      "step": 3562
+    },
+    {
+      "epoch": 1.3012598137666607,
+      "grad_norm": 0.8934658169746399,
+      "learning_rate": 1.6052478218455986e-05,
+      "loss": 0.5558,
+      "step": 3563
+    },
+    {
+      "epoch": 1.3016249771772868,
+      "grad_norm": 0.8569304943084717,
+      "learning_rate": 1.603841626362505e-05,
+      "loss": 0.5476,
+      "step": 3564
+    },
+    {
+      "epoch": 1.301990140587913,
+      "grad_norm": 0.8856249451637268,
+      "learning_rate": 1.6024356346869975e-05,
+      "loss": 0.554,
+      "step": 3565
+    },
+    {
+      "epoch": 1.3023553039985394,
+      "grad_norm": 0.8537352085113525,
+      "learning_rate": 1.6010298475424028e-05,
+      "loss": 0.5645,
+      "step": 3566
+    },
+    {
+      "epoch": 1.3027204674091655,
+      "grad_norm": 0.9424476027488708,
+      "learning_rate": 1.5996242656519418e-05,
+      "loss": 0.6075,
+      "step": 3567
+    },
+    {
+      "epoch": 1.3030856308197918,
+      "grad_norm": 0.8585494756698608,
+      "learning_rate": 1.5982188897387296e-05,
+      "loss": 0.5877,
+      "step": 3568
+    },
+    {
+      "epoch": 1.303450794230418,
+      "grad_norm": 0.8759137988090515,
+      "learning_rate": 1.596813720525777e-05,
+      "loss": 0.5927,
+      "step": 3569
+    },
+    {
+      "epoch": 1.3038159576410444,
+      "grad_norm": 1.1474668979644775,
+      "learning_rate": 1.5954087587359857e-05,
+      "loss": 0.5372,
+      "step": 3570
+    },
+    {
+      "epoch": 1.3041811210516707,
+      "grad_norm": 0.9488094449043274,
+      "learning_rate": 1.594004005092153e-05,
+      "loss": 0.5874,
+      "step": 3571
+    },
+    {
+      "epoch": 1.3045462844622968,
+      "grad_norm": 0.6511358022689819,
+      "learning_rate": 1.5925994603169678e-05,
+      "loss": 0.5779,
+      "step": 3572
+    },
+    {
+      "epoch": 1.3049114478729231,
+      "grad_norm": 0.8309807777404785,
+      "learning_rate": 1.5911951251330127e-05,
+      "loss": 0.5756,
+      "step": 3573
+    },
+    {
+      "epoch": 1.3052766112835494,
+      "grad_norm": 0.8680068850517273,
+      "learning_rate": 1.5897910002627625e-05,
+      "loss": 0.593,
+      "step": 3574
+    },
+    {
+      "epoch": 1.3056417746941755,
+      "grad_norm": 0.7424835562705994,
+      "learning_rate": 1.5883870864285806e-05,
+      "loss": 0.5756,
+      "step": 3575
+    },
+    {
+      "epoch": 1.3060069381048018,
+      "grad_norm": 0.8280590772628784,
+      "learning_rate": 1.586983384352726e-05,
+      "loss": 0.5773,
+      "step": 3576
+    },
+    {
+      "epoch": 1.3063721015154282,
+      "grad_norm": 1.1244232654571533,
+      "learning_rate": 1.5855798947573464e-05,
+      "loss": 0.6116,
+      "step": 3577
+    },
+    {
+      "epoch": 1.3067372649260545,
+      "grad_norm": 0.8125243186950684,
+      "learning_rate": 1.584176618364482e-05,
+      "loss": 0.559,
+      "step": 3578
+    },
+    {
+      "epoch": 1.3071024283366808,
+      "grad_norm": 0.8949918150901794,
+      "learning_rate": 1.582773555896061e-05,
+      "loss": 0.5954,
+      "step": 3579
+    },
+    {
+      "epoch": 1.3074675917473069,
+      "grad_norm": 1.0998412370681763,
+      "learning_rate": 1.5813707080739028e-05,
+      "loss": 0.5421,
+      "step": 3580
+    },
+    {
+      "epoch": 1.3078327551579332,
+      "grad_norm": 0.9270798563957214,
+      "learning_rate": 1.5799680756197177e-05,
+      "loss": 0.6094,
+      "step": 3581
+    },
+    {
+      "epoch": 1.3081979185685595,
+      "grad_norm": 0.7078328132629395,
+      "learning_rate": 1.5785656592551022e-05,
+      "loss": 0.5704,
+      "step": 3582
+    },
+    {
+      "epoch": 1.3085630819791856,
+      "grad_norm": 0.7362704277038574,
+      "learning_rate": 1.5771634597015445e-05,
+      "loss": 0.5872,
+      "step": 3583
+    },
+    {
+      "epoch": 1.308928245389812,
+      "grad_norm": 1.009394645690918,
+      "learning_rate": 1.575761477680419e-05,
+      "loss": 0.5833,
+      "step": 3584
+    },
+    {
+      "epoch": 1.3092934088004382,
+      "grad_norm": 1.5687860250473022,
+      "learning_rate": 1.57435971391299e-05,
+      "loss": 0.5931,
+      "step": 3585
+    },
+    {
+      "epoch": 1.3096585722110645,
+      "grad_norm": 0.6487178206443787,
+      "learning_rate": 1.572958169120408e-05,
+      "loss": 0.5676,
+      "step": 3586
+    },
+    {
+      "epoch": 1.3100237356216908,
+      "grad_norm": 0.9973515272140503,
+      "learning_rate": 1.5715568440237122e-05,
+      "loss": 0.5599,
+      "step": 3587
+    },
+    {
+      "epoch": 1.310388899032317,
+      "grad_norm": 0.9869074821472168,
+      "learning_rate": 1.5701557393438277e-05,
+      "loss": 0.6007,
+      "step": 3588
+    },
+    {
+      "epoch": 1.3107540624429432,
+      "grad_norm": 0.6144683361053467,
+      "learning_rate": 1.5687548558015663e-05,
+      "loss": 0.5912,
+      "step": 3589
+    },
+    {
+      "epoch": 1.3111192258535695,
+      "grad_norm": 0.8956754803657532,
+      "learning_rate": 1.567354194117627e-05,
+      "loss": 0.5551,
+      "step": 3590
+    },
+    {
+      "epoch": 1.3114843892641956,
+      "grad_norm": 0.9467038512229919,
+      "learning_rate": 1.565953755012594e-05,
+      "loss": 0.5912,
+      "step": 3591
+    },
+    {
+      "epoch": 1.311849552674822,
+      "grad_norm": 0.9374017715454102,
+      "learning_rate": 1.5645535392069366e-05,
+      "loss": 0.5633,
+      "step": 3592
+    },
+    {
+      "epoch": 1.3122147160854483,
+      "grad_norm": 1.8570634126663208,
+      "learning_rate": 1.56315354742101e-05,
+      "loss": 0.6488,
+      "step": 3593
+    },
+    {
+      "epoch": 1.3125798794960746,
+      "grad_norm": 0.7921823263168335,
+      "learning_rate": 1.5617537803750538e-05,
+      "loss": 0.5729,
+      "step": 3594
+    },
+    {
+      "epoch": 1.3129450429067007,
+      "grad_norm": 0.7369295358657837,
+      "learning_rate": 1.560354238789192e-05,
+      "loss": 0.5887,
+      "step": 3595
+    },
+    {
+      "epoch": 1.313310206317327,
+      "grad_norm": 0.7375071048736572,
+      "learning_rate": 1.558954923383432e-05,
+      "loss": 0.5477,
+      "step": 3596
+    },
+    {
+      "epoch": 1.3136753697279533,
+      "grad_norm": 0.6595288515090942,
+      "learning_rate": 1.5575558348776664e-05,
+      "loss": 0.5816,
+      "step": 3597
+    },
+    {
+      "epoch": 1.3140405331385794,
+      "grad_norm": 0.7850028276443481,
+      "learning_rate": 1.556156973991669e-05,
+      "loss": 0.5391,
+      "step": 3598
+    },
+    {
+      "epoch": 1.3144056965492057,
+      "grad_norm": 0.8331600427627563,
+      "learning_rate": 1.5547583414450985e-05,
+      "loss": 0.6069,
+      "step": 3599
+    },
+    {
+      "epoch": 1.314770859959832,
+      "grad_norm": 0.8196750283241272,
+      "learning_rate": 1.5533599379574956e-05,
+      "loss": 0.5251,
+      "step": 3600
+    },
+    {
+      "epoch": 1.3151360233704583,
+      "grad_norm": 0.9406454563140869,
+      "learning_rate": 1.551961764248281e-05,
+      "loss": 0.5274,
+      "step": 3601
+    },
+    {
+      "epoch": 1.3155011867810846,
+      "grad_norm": 0.644262969493866,
+      "learning_rate": 1.5505638210367605e-05,
+      "loss": 0.5952,
+      "step": 3602
+    },
+    {
+      "epoch": 1.3158663501917107,
+      "grad_norm": 0.9494520425796509,
+      "learning_rate": 1.5491661090421193e-05,
+      "loss": 0.59,
+      "step": 3603
+    },
+    {
+      "epoch": 1.316231513602337,
+      "grad_norm": 0.9427124261856079,
+      "learning_rate": 1.5477686289834238e-05,
+      "loss": 0.5742,
+      "step": 3604
+    },
+    {
+      "epoch": 1.3165966770129633,
+      "grad_norm": 0.8440623879432678,
+      "learning_rate": 1.5463713815796223e-05,
+      "loss": 0.5854,
+      "step": 3605
+    },
+    {
+      "epoch": 1.3169618404235894,
+      "grad_norm": 0.7992889881134033,
+      "learning_rate": 1.5449743675495416e-05,
+      "loss": 0.5585,
+      "step": 3606
+    },
+    {
+      "epoch": 1.3173270038342157,
+      "grad_norm": 0.8490431308746338,
+      "learning_rate": 1.54357758761189e-05,
+      "loss": 0.6118,
+      "step": 3607
+    },
+    {
+      "epoch": 1.317692167244842,
+      "grad_norm": 0.8494915962219238,
+      "learning_rate": 1.5421810424852542e-05,
+      "loss": 0.567,
+      "step": 3608
+    },
+    {
+      "epoch": 1.3180573306554684,
+      "grad_norm": 1.1305323839187622,
+      "learning_rate": 1.5407847328881013e-05,
+      "loss": 0.5708,
+      "step": 3609
+    },
+    {
+      "epoch": 1.3184224940660947,
+      "grad_norm": 1.355056643486023,
+      "learning_rate": 1.5393886595387756e-05,
+      "loss": 0.5883,
+      "step": 3610
+    },
+    {
+      "epoch": 1.3187876574767208,
+      "grad_norm": 1.0465811491012573,
+      "learning_rate": 1.5379928231555014e-05,
+      "loss": 0.5835,
+      "step": 3611
+    },
+    {
+      "epoch": 1.319152820887347,
+      "grad_norm": 1.0105823278427124,
+      "learning_rate": 1.536597224456381e-05,
+      "loss": 0.564,
+      "step": 3612
+    },
+    {
+      "epoch": 1.3195179842979734,
+      "grad_norm": 1.0827776193618774,
+      "learning_rate": 1.5352018641593933e-05,
+      "loss": 0.5918,
+      "step": 3613
+    },
+    {
+      "epoch": 1.3198831477085995,
+      "grad_norm": 0.5904019474983215,
+      "learning_rate": 1.5338067429823956e-05,
+      "loss": 0.6219,
+      "step": 3614
+    },
+    {
+      "epoch": 1.3202483111192258,
+      "grad_norm": 0.9047408103942871,
+      "learning_rate": 1.5324118616431216e-05,
+      "loss": 0.5801,
+      "step": 3615
+    },
+    {
+      "epoch": 1.3206134745298521,
+      "grad_norm": 0.8691691756248474,
+      "learning_rate": 1.531017220859181e-05,
+      "loss": 0.5646,
+      "step": 3616
+    },
+    {
+      "epoch": 1.3209786379404784,
+      "grad_norm": 0.9058032631874084,
+      "learning_rate": 1.5296228213480615e-05,
+      "loss": 0.5858,
+      "step": 3617
+    },
+    {
+      "epoch": 1.3213438013511047,
+      "grad_norm": 0.896232008934021,
+      "learning_rate": 1.5282286638271248e-05,
+      "loss": 0.5699,
+      "step": 3618
+    },
+    {
+      "epoch": 1.3217089647617308,
+      "grad_norm": 0.97877436876297,
+      "learning_rate": 1.5268347490136102e-05,
+      "loss": 0.574,
+      "step": 3619
+    },
+    {
+      "epoch": 1.3220741281723571,
+      "grad_norm": 0.7933487296104431,
+      "learning_rate": 1.5254410776246299e-05,
+      "loss": 0.5931,
+      "step": 3620
+    },
+    {
+      "epoch": 1.3224392915829835,
+      "grad_norm": 0.6697726249694824,
+      "learning_rate": 1.5240476503771726e-05,
+      "loss": 0.5953,
+      "step": 3621
+    },
+    {
+      "epoch": 1.3228044549936095,
+      "grad_norm": 0.7056508660316467,
+      "learning_rate": 1.5226544679881e-05,
+      "loss": 0.5887,
+      "step": 3622
+    },
+    {
+      "epoch": 1.3231696184042359,
+      "grad_norm": 0.9026390314102173,
+      "learning_rate": 1.5212615311741488e-05,
+      "loss": 0.5659,
+      "step": 3623
+    },
+    {
+      "epoch": 1.3235347818148622,
+      "grad_norm": 0.9017171859741211,
+      "learning_rate": 1.5198688406519297e-05,
+      "loss": 0.5895,
+      "step": 3624
+    },
+    {
+      "epoch": 1.3238999452254885,
+      "grad_norm": 0.7538767457008362,
+      "learning_rate": 1.5184763971379255e-05,
+      "loss": 0.5979,
+      "step": 3625
+    },
+    {
+      "epoch": 1.3242651086361148,
+      "grad_norm": 1.9563366174697876,
+      "learning_rate": 1.5170842013484928e-05,
+      "loss": 0.5507,
+      "step": 3626
+    },
+    {
+      "epoch": 1.3246302720467409,
+      "grad_norm": 0.705376386642456,
+      "learning_rate": 1.5156922539998609e-05,
+      "loss": 0.5605,
+      "step": 3627
+    },
+    {
+      "epoch": 1.3249954354573672,
+      "grad_norm": 0.6387503743171692,
+      "learning_rate": 1.5143005558081292e-05,
+      "loss": 0.5996,
+      "step": 3628
+    },
+    {
+      "epoch": 1.3253605988679933,
+      "grad_norm": 0.7741349935531616,
+      "learning_rate": 1.5129091074892721e-05,
+      "loss": 0.5544,
+      "step": 3629
+    },
+    {
+      "epoch": 1.3257257622786196,
+      "grad_norm": 0.865826427936554,
+      "learning_rate": 1.5115179097591331e-05,
+      "loss": 0.5783,
+      "step": 3630
+    },
+    {
+      "epoch": 1.326090925689246,
+      "grad_norm": 0.9152030348777771,
+      "learning_rate": 1.5101269633334284e-05,
+      "loss": 0.579,
+      "step": 3631
+    },
+    {
+      "epoch": 1.3264560890998722,
+      "grad_norm": 0.7293338775634766,
+      "learning_rate": 1.5087362689277431e-05,
+      "loss": 0.5459,
+      "step": 3632
+    },
+    {
+      "epoch": 1.3268212525104985,
+      "grad_norm": 0.7804079651832581,
+      "learning_rate": 1.5073458272575345e-05,
+      "loss": 0.6124,
+      "step": 3633
+    },
+    {
+      "epoch": 1.3271864159211246,
+      "grad_norm": 0.7898918986320496,
+      "learning_rate": 1.5059556390381289e-05,
+      "loss": 0.569,
+      "step": 3634
+    },
+    {
+      "epoch": 1.327551579331751,
+      "grad_norm": 0.6254060864448547,
+      "learning_rate": 1.5045657049847223e-05,
+      "loss": 0.6038,
+      "step": 3635
+    },
+    {
+      "epoch": 1.3279167427423773,
+      "grad_norm": 0.7602380514144897,
+      "learning_rate": 1.50317602581238e-05,
+      "loss": 0.5654,
+      "step": 3636
+    },
+    {
+      "epoch": 1.3282819061530033,
+      "grad_norm": 0.8868459463119507,
+      "learning_rate": 1.5017866022360356e-05,
+      "loss": 0.5845,
+      "step": 3637
+    },
+    {
+      "epoch": 1.3286470695636297,
+      "grad_norm": 0.9397568702697754,
+      "learning_rate": 1.5003974349704931e-05,
+      "loss": 0.5626,
+      "step": 3638
+    },
+    {
+      "epoch": 1.329012232974256,
+      "grad_norm": 0.755822479724884,
+      "learning_rate": 1.4990085247304218e-05,
+      "loss": 0.5749,
+      "step": 3639
+    },
+    {
+      "epoch": 1.3293773963848823,
+      "grad_norm": 0.8150222897529602,
+      "learning_rate": 1.4976198722303619e-05,
+      "loss": 0.553,
+      "step": 3640
+    },
+    {
+      "epoch": 1.3297425597955086,
+      "grad_norm": 1.1800682544708252,
+      "learning_rate": 1.496231478184718e-05,
+      "loss": 0.585,
+      "step": 3641
+    },
+    {
+      "epoch": 1.3301077232061347,
+      "grad_norm": 0.8337178826332092,
+      "learning_rate": 1.4948433433077632e-05,
+      "loss": 0.5671,
+      "step": 3642
+    },
+    {
+      "epoch": 1.330472886616761,
+      "grad_norm": 0.7551597952842712,
+      "learning_rate": 1.4934554683136382e-05,
+      "loss": 0.5992,
+      "step": 3643
+    },
+    {
+      "epoch": 1.3308380500273873,
+      "grad_norm": 0.6858837604522705,
+      "learning_rate": 1.4920678539163479e-05,
+      "loss": 0.5988,
+      "step": 3644
+    },
+    {
+      "epoch": 1.3312032134380134,
+      "grad_norm": 1.4082273244857788,
+      "learning_rate": 1.4906805008297645e-05,
+      "loss": 0.5713,
+      "step": 3645
+    },
+    {
+      "epoch": 1.3315683768486397,
+      "grad_norm": 0.8915179967880249,
+      "learning_rate": 1.4892934097676262e-05,
+      "loss": 0.5796,
+      "step": 3646
+    },
+    {
+      "epoch": 1.331933540259266,
+      "grad_norm": 0.8021594285964966,
+      "learning_rate": 1.4879065814435349e-05,
+      "loss": 0.5968,
+      "step": 3647
+    },
+    {
+      "epoch": 1.3322987036698923,
+      "grad_norm": 0.8213039636611938,
+      "learning_rate": 1.4865200165709588e-05,
+      "loss": 0.5697,
+      "step": 3648
+    },
+    {
+      "epoch": 1.3326638670805186,
+      "grad_norm": 0.6211522817611694,
+      "learning_rate": 1.485133715863229e-05,
+      "loss": 0.5959,
+      "step": 3649
+    },
+    {
+      "epoch": 1.3330290304911447,
+      "grad_norm": 0.7799485325813293,
+      "learning_rate": 1.4837476800335427e-05,
+      "loss": 0.5653,
+      "step": 3650
+    },
+    {
+      "epoch": 1.333394193901771,
+      "grad_norm": 0.9671348333358765,
+      "learning_rate": 1.4823619097949584e-05,
+      "loss": 0.5603,
+      "step": 3651
+    },
+    {
+      "epoch": 1.3337593573123974,
+      "grad_norm": 0.9694781303405762,
+      "learning_rate": 1.4809764058604006e-05,
+      "loss": 0.5805,
+      "step": 3652
+    },
+    {
+      "epoch": 1.3341245207230235,
+      "grad_norm": 1.7140861749649048,
+      "learning_rate": 1.4795911689426543e-05,
+      "loss": 0.5508,
+      "step": 3653
+    },
+    {
+      "epoch": 1.3344896841336498,
+      "grad_norm": 0.8197154998779297,
+      "learning_rate": 1.4782061997543699e-05,
+      "loss": 0.5751,
+      "step": 3654
+    },
+    {
+      "epoch": 1.334854847544276,
+      "grad_norm": 0.7604065537452698,
+      "learning_rate": 1.476821499008057e-05,
+      "loss": 0.5731,
+      "step": 3655
+    },
+    {
+      "epoch": 1.3352200109549024,
+      "grad_norm": 0.8634833097457886,
+      "learning_rate": 1.4754370674160885e-05,
+      "loss": 0.5941,
+      "step": 3656
+    },
+    {
+      "epoch": 1.3355851743655287,
+      "grad_norm": 0.946653425693512,
+      "learning_rate": 1.4740529056906994e-05,
+      "loss": 0.5927,
+      "step": 3657
+    },
+    {
+      "epoch": 1.3359503377761548,
+      "grad_norm": 0.8148925304412842,
+      "learning_rate": 1.4726690145439858e-05,
+      "loss": 0.5824,
+      "step": 3658
+    },
+    {
+      "epoch": 1.336315501186781,
+      "grad_norm": 1.1249194145202637,
+      "learning_rate": 1.4712853946879035e-05,
+      "loss": 0.5413,
+      "step": 3659
+    },
+    {
+      "epoch": 1.3366806645974074,
+      "grad_norm": 0.791527271270752,
+      "learning_rate": 1.46990204683427e-05,
+      "loss": 0.5885,
+      "step": 3660
+    },
+    {
+      "epoch": 1.3370458280080335,
+      "grad_norm": 0.8487535119056702,
+      "learning_rate": 1.4685189716947614e-05,
+      "loss": 0.5484,
+      "step": 3661
+    },
+    {
+      "epoch": 1.3374109914186598,
+      "grad_norm": 0.6759431958198547,
+      "learning_rate": 1.4671361699809153e-05,
+      "loss": 0.6078,
+      "step": 3662
+    },
+    {
+      "epoch": 1.3377761548292861,
+      "grad_norm": 1.246487021446228,
+      "learning_rate": 1.4657536424041268e-05,
+      "loss": 0.5653,
+      "step": 3663
+    },
+    {
+      "epoch": 1.3381413182399124,
+      "grad_norm": 0.8097664713859558,
+      "learning_rate": 1.4643713896756518e-05,
+      "loss": 0.5686,
+      "step": 3664
+    },
+    {
+      "epoch": 1.3385064816505385,
+      "grad_norm": 0.6811612844467163,
+      "learning_rate": 1.4629894125066028e-05,
+      "loss": 0.5887,
+      "step": 3665
+    },
+    {
+      "epoch": 1.3388716450611649,
+      "grad_norm": 0.9761230945587158,
+      "learning_rate": 1.4616077116079524e-05,
+      "loss": 0.5722,
+      "step": 3666
+    },
+    {
+      "epoch": 1.3392368084717912,
+      "grad_norm": 0.793403148651123,
+      "learning_rate": 1.4602262876905306e-05,
+      "loss": 0.5549,
+      "step": 3667
+    },
+    {
+      "epoch": 1.3396019718824173,
+      "grad_norm": 0.8240900039672852,
+      "learning_rate": 1.458845141465024e-05,
+      "loss": 0.608,
+      "step": 3668
+    },
+    {
+      "epoch": 1.3399671352930436,
+      "grad_norm": 1.1468547582626343,
+      "learning_rate": 1.4574642736419763e-05,
+      "loss": 0.5709,
+      "step": 3669
+    },
+    {
+      "epoch": 1.3403322987036699,
+      "grad_norm": 0.8393019437789917,
+      "learning_rate": 1.4560836849317895e-05,
+      "loss": 0.5826,
+      "step": 3670
+    },
+    {
+      "epoch": 1.3406974621142962,
+      "grad_norm": 1.20576012134552,
+      "learning_rate": 1.4547033760447202e-05,
+      "loss": 0.5608,
+      "step": 3671
+    },
+    {
+      "epoch": 1.3410626255249225,
+      "grad_norm": 0.8008895516395569,
+      "learning_rate": 1.4533233476908835e-05,
+      "loss": 0.5823,
+      "step": 3672
+    },
+    {
+      "epoch": 1.3414277889355486,
+      "grad_norm": 0.9509861469268799,
+      "learning_rate": 1.451943600580247e-05,
+      "loss": 0.5393,
+      "step": 3673
+    },
+    {
+      "epoch": 1.341792952346175,
+      "grad_norm": 0.9399440288543701,
+      "learning_rate": 1.450564135422636e-05,
+      "loss": 0.5958,
+      "step": 3674
+    },
+    {
+      "epoch": 1.3421581157568012,
+      "grad_norm": 1.1058588027954102,
+      "learning_rate": 1.4491849529277295e-05,
+      "loss": 0.556,
+      "step": 3675
+    },
+    {
+      "epoch": 1.3425232791674273,
+      "grad_norm": 0.7471702098846436,
+      "learning_rate": 1.4478060538050622e-05,
+      "loss": 0.6152,
+      "step": 3676
+    },
+    {
+      "epoch": 1.3428884425780536,
+      "grad_norm": 0.9087870717048645,
+      "learning_rate": 1.4464274387640224e-05,
+      "loss": 0.5765,
+      "step": 3677
+    },
+    {
+      "epoch": 1.34325360598868,
+      "grad_norm": 1.1029573678970337,
+      "learning_rate": 1.4450491085138514e-05,
+      "loss": 0.5664,
+      "step": 3678
+    },
+    {
+      "epoch": 1.3436187693993062,
+      "grad_norm": 0.8007150292396545,
+      "learning_rate": 1.4436710637636456e-05,
+      "loss": 0.5482,
+      "step": 3679
+    },
+    {
+      "epoch": 1.3439839328099326,
+      "grad_norm": 1.6680142879486084,
+      "learning_rate": 1.442293305222354e-05,
+      "loss": 0.6091,
+      "step": 3680
+    },
+    {
+      "epoch": 1.3443490962205586,
+      "grad_norm": 1.0721821784973145,
+      "learning_rate": 1.4409158335987763e-05,
+      "loss": 0.6155,
+      "step": 3681
+    },
+    {
+      "epoch": 1.344714259631185,
+      "grad_norm": 0.790995180606842,
+      "learning_rate": 1.4395386496015685e-05,
+      "loss": 0.5885,
+      "step": 3682
+    },
+    {
+      "epoch": 1.3450794230418113,
+      "grad_norm": 0.9400860667228699,
+      "learning_rate": 1.4381617539392347e-05,
+      "loss": 0.5948,
+      "step": 3683
+    },
+    {
+      "epoch": 1.3454445864524374,
+      "grad_norm": 0.8138934373855591,
+      "learning_rate": 1.436785147320134e-05,
+      "loss": 0.5776,
+      "step": 3684
+    },
+    {
+      "epoch": 1.3458097498630637,
+      "grad_norm": 1.228329062461853,
+      "learning_rate": 1.4354088304524739e-05,
+      "loss": 0.5466,
+      "step": 3685
+    },
+    {
+      "epoch": 1.34617491327369,
+      "grad_norm": 0.8579606413841248,
+      "learning_rate": 1.4340328040443154e-05,
+      "loss": 0.5759,
+      "step": 3686
+    },
+    {
+      "epoch": 1.3465400766843163,
+      "grad_norm": 1.073363184928894,
+      "learning_rate": 1.4326570688035682e-05,
+      "loss": 0.5757,
+      "step": 3687
+    },
+    {
+      "epoch": 1.3469052400949426,
+      "grad_norm": 0.8040978908538818,
+      "learning_rate": 1.4312816254379928e-05,
+      "loss": 0.5711,
+      "step": 3688
+    },
+    {
+      "epoch": 1.3472704035055687,
+      "grad_norm": 0.5264380574226379,
+      "learning_rate": 1.4299064746552005e-05,
+      "loss": 0.6295,
+      "step": 3689
+    },
+    {
+      "epoch": 1.347635566916195,
+      "grad_norm": 1.2140576839447021,
+      "learning_rate": 1.42853161716265e-05,
+      "loss": 0.5482,
+      "step": 3690
+    },
+    {
+      "epoch": 1.3480007303268213,
+      "grad_norm": 0.8451759219169617,
+      "learning_rate": 1.4271570536676513e-05,
+      "loss": 0.6004,
+      "step": 3691
+    },
+    {
+      "epoch": 1.3483658937374474,
+      "grad_norm": 0.6922603845596313,
+      "learning_rate": 1.4257827848773613e-05,
+      "loss": 0.5847,
+      "step": 3692
+    },
+    {
+      "epoch": 1.3487310571480737,
+      "grad_norm": 0.8059636354446411,
+      "learning_rate": 1.424408811498787e-05,
+      "loss": 0.5731,
+      "step": 3693
+    },
+    {
+      "epoch": 1.3490962205587,
+      "grad_norm": 1.1259942054748535,
+      "learning_rate": 1.4230351342387827e-05,
+      "loss": 0.5465,
+      "step": 3694
+    },
+    {
+      "epoch": 1.3494613839693264,
+      "grad_norm": 0.9364913105964661,
+      "learning_rate": 1.4216617538040488e-05,
+      "loss": 0.5436,
+      "step": 3695
+    },
+    {
+      "epoch": 1.3498265473799524,
+      "grad_norm": 0.8418157696723938,
+      "learning_rate": 1.4202886709011357e-05,
+      "loss": 0.5953,
+      "step": 3696
+    },
+    {
+      "epoch": 1.3501917107905788,
+      "grad_norm": 0.9355802536010742,
+      "learning_rate": 1.4189158862364386e-05,
+      "loss": 0.6121,
+      "step": 3697
+    },
+    {
+      "epoch": 1.350556874201205,
+      "grad_norm": 0.7521985769271851,
+      "learning_rate": 1.417543400516201e-05,
+      "loss": 0.5795,
+      "step": 3698
+    },
+    {
+      "epoch": 1.3509220376118312,
+      "grad_norm": 0.953596830368042,
+      "learning_rate": 1.4161712144465108e-05,
+      "loss": 0.5792,
+      "step": 3699
+    },
+    {
+      "epoch": 1.3512872010224575,
+      "grad_norm": 0.640637218952179,
+      "learning_rate": 1.4147993287333032e-05,
+      "loss": 0.5792,
+      "step": 3700
+    },
+    {
+      "epoch": 1.3516523644330838,
+      "grad_norm": 0.8487382531166077,
+      "learning_rate": 1.413427744082359e-05,
+      "loss": 0.5794,
+      "step": 3701
+    },
+    {
+      "epoch": 1.35201752784371,
+      "grad_norm": 0.7688397765159607,
+      "learning_rate": 1.412056461199302e-05,
+      "loss": 0.6128,
+      "step": 3702
+    },
+    {
+      "epoch": 1.3523826912543364,
+      "grad_norm": 0.9199714660644531,
+      "learning_rate": 1.4106854807896035e-05,
+      "loss": 0.5534,
+      "step": 3703
+    },
+    {
+      "epoch": 1.3527478546649625,
+      "grad_norm": 0.8518757820129395,
+      "learning_rate": 1.4093148035585774e-05,
+      "loss": 0.5898,
+      "step": 3704
+    },
+    {
+      "epoch": 1.3531130180755888,
+      "grad_norm": 1.321502923965454,
+      "learning_rate": 1.4079444302113821e-05,
+      "loss": 0.5872,
+      "step": 3705
+    },
+    {
+      "epoch": 1.3534781814862151,
+      "grad_norm": 0.7342141270637512,
+      "learning_rate": 1.4065743614530193e-05,
+      "loss": 0.5936,
+      "step": 3706
+    },
+    {
+      "epoch": 1.3538433448968412,
+      "grad_norm": 0.9883751273155212,
+      "learning_rate": 1.405204597988336e-05,
+      "loss": 0.5825,
+      "step": 3707
+    },
+    {
+      "epoch": 1.3542085083074675,
+      "grad_norm": 0.7839553356170654,
+      "learning_rate": 1.403835140522019e-05,
+      "loss": 0.5811,
+      "step": 3708
+    },
+    {
+      "epoch": 1.3545736717180938,
+      "grad_norm": 1.0226047039031982,
+      "learning_rate": 1.4024659897585989e-05,
+      "loss": 0.5851,
+      "step": 3709
+    },
+    {
+      "epoch": 1.3549388351287202,
+      "grad_norm": 0.95647794008255,
+      "learning_rate": 1.4010971464024494e-05,
+      "loss": 0.5844,
+      "step": 3710
+    },
+    {
+      "epoch": 1.3553039985393465,
+      "grad_norm": 0.9128265976905823,
+      "learning_rate": 1.3997286111577864e-05,
+      "loss": 0.6129,
+      "step": 3711
+    },
+    {
+      "epoch": 1.3556691619499726,
+      "grad_norm": 0.8987192511558533,
+      "learning_rate": 1.3983603847286648e-05,
+      "loss": 0.5786,
+      "step": 3712
+    },
+    {
+      "epoch": 1.3560343253605989,
+      "grad_norm": 0.8189154863357544,
+      "learning_rate": 1.3969924678189837e-05,
+      "loss": 0.5656,
+      "step": 3713
+    },
+    {
+      "epoch": 1.3563994887712252,
+      "grad_norm": 0.8834811449050903,
+      "learning_rate": 1.3956248611324803e-05,
+      "loss": 0.5743,
+      "step": 3714
+    },
+    {
+      "epoch": 1.3567646521818513,
+      "grad_norm": 0.9477565288543701,
+      "learning_rate": 1.3942575653727341e-05,
+      "loss": 0.5716,
+      "step": 3715
+    },
+    {
+      "epoch": 1.3571298155924776,
+      "grad_norm": 0.9446448087692261,
+      "learning_rate": 1.392890581243163e-05,
+      "loss": 0.567,
+      "step": 3716
+    },
+    {
+      "epoch": 1.357494979003104,
+      "grad_norm": 0.7631604075431824,
+      "learning_rate": 1.3915239094470268e-05,
+      "loss": 0.5714,
+      "step": 3717
+    },
+    {
+      "epoch": 1.3578601424137302,
+      "grad_norm": 0.6586843132972717,
+      "learning_rate": 1.3901575506874218e-05,
+      "loss": 0.5876,
+      "step": 3718
+    },
+    {
+      "epoch": 1.3582253058243565,
+      "grad_norm": 0.8762580752372742,
+      "learning_rate": 1.3887915056672863e-05,
+      "loss": 0.6044,
+      "step": 3719
+    },
+    {
+      "epoch": 1.3585904692349826,
+      "grad_norm": 0.9814974069595337,
+      "learning_rate": 1.387425775089395e-05,
+      "loss": 0.5732,
+      "step": 3720
+    },
+    {
+      "epoch": 1.358955632645609,
+      "grad_norm": 0.6957576870918274,
+      "learning_rate": 1.3860603596563606e-05,
+      "loss": 0.5784,
+      "step": 3721
+    },
+    {
+      "epoch": 1.3593207960562352,
+      "grad_norm": 0.841157078742981,
+      "learning_rate": 1.3846952600706354e-05,
+      "loss": 0.5605,
+      "step": 3722
+    },
+    {
+      "epoch": 1.3596859594668613,
+      "grad_norm": 0.8217060565948486,
+      "learning_rate": 1.3833304770345084e-05,
+      "loss": 0.5889,
+      "step": 3723
+    },
+    {
+      "epoch": 1.3600511228774876,
+      "grad_norm": 1.0585644245147705,
+      "learning_rate": 1.3819660112501054e-05,
+      "loss": 0.5721,
+      "step": 3724
+    },
+    {
+      "epoch": 1.360416286288114,
+      "grad_norm": 0.8639219403266907,
+      "learning_rate": 1.3806018634193899e-05,
+      "loss": 0.5401,
+      "step": 3725
+    },
+    {
+      "epoch": 1.3607814496987403,
+      "grad_norm": 0.7209730744361877,
+      "learning_rate": 1.3792380342441601e-05,
+      "loss": 0.5725,
+      "step": 3726
+    },
+    {
+      "epoch": 1.3611466131093666,
+      "grad_norm": 0.782489538192749,
+      "learning_rate": 1.3778745244260528e-05,
+      "loss": 0.5829,
+      "step": 3727
+    },
+    {
+      "epoch": 1.3615117765199927,
+      "grad_norm": 0.9441948533058167,
+      "learning_rate": 1.3765113346665375e-05,
+      "loss": 0.5483,
+      "step": 3728
+    },
+    {
+      "epoch": 1.361876939930619,
+      "grad_norm": 0.7579976320266724,
+      "learning_rate": 1.3751484656669223e-05,
+      "loss": 0.5814,
+      "step": 3729
+    },
+    {
+      "epoch": 1.362242103341245,
+      "grad_norm": 0.8741255402565002,
+      "learning_rate": 1.3737859181283471e-05,
+      "loss": 0.5351,
+      "step": 3730
+    },
+    {
+      "epoch": 1.3626072667518714,
+      "grad_norm": 1.0973438024520874,
+      "learning_rate": 1.3724236927517887e-05,
+      "loss": 0.5543,
+      "step": 3731
+    },
+    {
+      "epoch": 1.3629724301624977,
+      "grad_norm": 0.7737992405891418,
+      "learning_rate": 1.3710617902380579e-05,
+      "loss": 0.593,
+      "step": 3732
+    },
+    {
+      "epoch": 1.363337593573124,
+      "grad_norm": 1.0185577869415283,
+      "learning_rate": 1.3697002112877975e-05,
+      "loss": 0.5996,
+      "step": 3733
+    },
+    {
+      "epoch": 1.3637027569837503,
+      "grad_norm": 1.013541579246521,
+      "learning_rate": 1.3683389566014871e-05,
+      "loss": 0.5299,
+      "step": 3734
+    },
+    {
+      "epoch": 1.3640679203943764,
+      "grad_norm": 0.8510308861732483,
+      "learning_rate": 1.3669780268794362e-05,
+      "loss": 0.5641,
+      "step": 3735
+    },
+    {
+      "epoch": 1.3644330838050027,
+      "grad_norm": 1.2188884019851685,
+      "learning_rate": 1.3656174228217883e-05,
+      "loss": 0.5375,
+      "step": 3736
+    },
+    {
+      "epoch": 1.364798247215629,
+      "grad_norm": 1.3204916715621948,
+      "learning_rate": 1.3642571451285207e-05,
+      "loss": 0.5532,
+      "step": 3737
+    },
+    {
+      "epoch": 1.3651634106262551,
+      "grad_norm": 0.905667781829834,
+      "learning_rate": 1.3628971944994407e-05,
+      "loss": 0.5627,
+      "step": 3738
+    },
+    {
+      "epoch": 1.3655285740368814,
+      "grad_norm": 0.7779754400253296,
+      "learning_rate": 1.3615375716341893e-05,
+      "loss": 0.5475,
+      "step": 3739
+    },
+    {
+      "epoch": 1.3658937374475077,
+      "grad_norm": 1.3370355367660522,
+      "learning_rate": 1.3601782772322368e-05,
+      "loss": 0.5469,
+      "step": 3740
+    },
+    {
+      "epoch": 1.366258900858134,
+      "grad_norm": 0.7945521473884583,
+      "learning_rate": 1.3588193119928868e-05,
+      "loss": 0.5854,
+      "step": 3741
+    },
+    {
+      "epoch": 1.3666240642687604,
+      "grad_norm": 1.4507278203964233,
+      "learning_rate": 1.3574606766152712e-05,
+      "loss": 0.5408,
+      "step": 3742
+    },
+    {
+      "epoch": 1.3669892276793865,
+      "grad_norm": 0.8589543104171753,
+      "learning_rate": 1.3561023717983541e-05,
+      "loss": 0.5685,
+      "step": 3743
+    },
+    {
+      "epoch": 1.3673543910900128,
+      "grad_norm": 0.8070398569107056,
+      "learning_rate": 1.3547443982409291e-05,
+      "loss": 0.5871,
+      "step": 3744
+    },
+    {
+      "epoch": 1.367719554500639,
+      "grad_norm": 0.8978539109230042,
+      "learning_rate": 1.3533867566416184e-05,
+      "loss": 0.5471,
+      "step": 3745
+    },
+    {
+      "epoch": 1.3680847179112652,
+      "grad_norm": 0.7683486342430115,
+      "learning_rate": 1.3520294476988747e-05,
+      "loss": 0.5642,
+      "step": 3746
+    },
+    {
+      "epoch": 1.3684498813218915,
+      "grad_norm": 0.7854225635528564,
+      "learning_rate": 1.3506724721109792e-05,
+      "loss": 0.5832,
+      "step": 3747
+    },
+    {
+      "epoch": 1.3688150447325178,
+      "grad_norm": 0.8262268304824829,
+      "learning_rate": 1.3493158305760401e-05,
+      "loss": 0.5329,
+      "step": 3748
+    },
+    {
+      "epoch": 1.3691802081431441,
+      "grad_norm": 0.8037892580032349,
+      "learning_rate": 1.3479595237919963e-05,
+      "loss": 0.5434,
+      "step": 3749
+    },
+    {
+      "epoch": 1.3695453715537704,
+      "grad_norm": 0.862305223941803,
+      "learning_rate": 1.346603552456612e-05,
+      "loss": 0.5833,
+      "step": 3750
+    },
+    {
+      "epoch": 1.3699105349643965,
+      "grad_norm": 1.2398102283477783,
+      "learning_rate": 1.3452479172674817e-05,
+      "loss": 0.5561,
+      "step": 3751
+    },
+    {
+      "epoch": 1.3702756983750228,
+      "grad_norm": 0.859796404838562,
+      "learning_rate": 1.3438926189220239e-05,
+      "loss": 0.5598,
+      "step": 3752
+    },
+    {
+      "epoch": 1.3706408617856491,
+      "grad_norm": 0.9049490690231323,
+      "learning_rate": 1.3425376581174855e-05,
+      "loss": 0.5749,
+      "step": 3753
+    },
+    {
+      "epoch": 1.3710060251962752,
+      "grad_norm": 0.8754092454910278,
+      "learning_rate": 1.34118303555094e-05,
+      "loss": 0.5948,
+      "step": 3754
+    },
+    {
+      "epoch": 1.3713711886069015,
+      "grad_norm": 1.0824122428894043,
+      "learning_rate": 1.3398287519192858e-05,
+      "loss": 0.5371,
+      "step": 3755
+    },
+    {
+      "epoch": 1.3717363520175279,
+      "grad_norm": 0.76957768201828,
+      "learning_rate": 1.3384748079192482e-05,
+      "loss": 0.5688,
+      "step": 3756
+    },
+    {
+      "epoch": 1.3721015154281542,
+      "grad_norm": 0.7734008431434631,
+      "learning_rate": 1.337121204247376e-05,
+      "loss": 0.6054,
+      "step": 3757
+    },
+    {
+      "epoch": 1.3724666788387805,
+      "grad_norm": 0.9477915167808533,
+      "learning_rate": 1.335767941600045e-05,
+      "loss": 0.5388,
+      "step": 3758
+    },
+    {
+      "epoch": 1.3728318422494066,
+      "grad_norm": 0.7826148867607117,
+      "learning_rate": 1.3344150206734537e-05,
+      "loss": 0.5506,
+      "step": 3759
+    },
+    {
+      "epoch": 1.3731970056600329,
+      "grad_norm": 0.9231326580047607,
+      "learning_rate": 1.3330624421636265e-05,
+      "loss": 0.5547,
+      "step": 3760
+    },
+    {
+      "epoch": 1.3735621690706592,
+      "grad_norm": 0.8588625192642212,
+      "learning_rate": 1.3317102067664104e-05,
+      "loss": 0.5501,
+      "step": 3761
+    },
+    {
+      "epoch": 1.3739273324812853,
+      "grad_norm": 0.9153193235397339,
+      "learning_rate": 1.3303583151774758e-05,
+      "loss": 0.5639,
+      "step": 3762
+    },
+    {
+      "epoch": 1.3742924958919116,
+      "grad_norm": 0.8094452023506165,
+      "learning_rate": 1.3290067680923169e-05,
+      "loss": 0.5513,
+      "step": 3763
+    },
+    {
+      "epoch": 1.374657659302538,
+      "grad_norm": 0.8183193802833557,
+      "learning_rate": 1.3276555662062503e-05,
+      "loss": 0.5526,
+      "step": 3764
+    },
+    {
+      "epoch": 1.3750228227131642,
+      "grad_norm": 0.8727033734321594,
+      "learning_rate": 1.3263047102144154e-05,
+      "loss": 0.5302,
+      "step": 3765
+    },
+    {
+      "epoch": 1.3753879861237903,
+      "grad_norm": 0.8611675500869751,
+      "learning_rate": 1.3249542008117737e-05,
+      "loss": 0.5409,
+      "step": 3766
+    },
+    {
+      "epoch": 1.3757531495344166,
+      "grad_norm": 1.183205485343933,
+      "learning_rate": 1.3236040386931075e-05,
+      "loss": 0.5649,
+      "step": 3767
+    },
+    {
+      "epoch": 1.376118312945043,
+      "grad_norm": 0.7820980548858643,
+      "learning_rate": 1.3222542245530217e-05,
+      "loss": 0.5738,
+      "step": 3768
+    },
+    {
+      "epoch": 1.376483476355669,
+      "grad_norm": 1.0162458419799805,
+      "learning_rate": 1.3209047590859407e-05,
+      "loss": 0.5634,
+      "step": 3769
+    },
+    {
+      "epoch": 1.3768486397662953,
+      "grad_norm": 0.9946308732032776,
+      "learning_rate": 1.3195556429861112e-05,
+      "loss": 0.5799,
+      "step": 3770
+    },
+    {
+      "epoch": 1.3772138031769217,
+      "grad_norm": 0.9185833930969238,
+      "learning_rate": 1.3182068769475984e-05,
+      "loss": 0.5247,
+      "step": 3771
+    },
+    {
+      "epoch": 1.377578966587548,
+      "grad_norm": 0.7531657814979553,
+      "learning_rate": 1.316858461664289e-05,
+      "loss": 0.5931,
+      "step": 3772
+    },
+    {
+      "epoch": 1.3779441299981743,
+      "grad_norm": 0.9958174228668213,
+      "learning_rate": 1.3155103978298882e-05,
+      "loss": 0.5481,
+      "step": 3773
+    },
+    {
+      "epoch": 1.3783092934088004,
+      "grad_norm": 0.9429916739463806,
+      "learning_rate": 1.3141626861379215e-05,
+      "loss": 0.557,
+      "step": 3774
+    },
+    {
+      "epoch": 1.3786744568194267,
+      "grad_norm": 0.6601399779319763,
+      "learning_rate": 1.3128153272817312e-05,
+      "loss": 0.6037,
+      "step": 3775
+    },
+    {
+      "epoch": 1.379039620230053,
+      "grad_norm": 0.8233461380004883,
+      "learning_rate": 1.31146832195448e-05,
+      "loss": 0.5806,
+      "step": 3776
+    },
+    {
+      "epoch": 1.379404783640679,
+      "grad_norm": 0.8602706789970398,
+      "learning_rate": 1.3101216708491482e-05,
+      "loss": 0.5287,
+      "step": 3777
+    },
+    {
+      "epoch": 1.3797699470513054,
+      "grad_norm": 0.8407243490219116,
+      "learning_rate": 1.308775374658534e-05,
+      "loss": 0.5904,
+      "step": 3778
+    },
+    {
+      "epoch": 1.3801351104619317,
+      "grad_norm": 0.896361231803894,
+      "learning_rate": 1.3074294340752518e-05,
+      "loss": 0.5925,
+      "step": 3779
+    },
+    {
+      "epoch": 1.380500273872558,
+      "grad_norm": 0.9036267995834351,
+      "learning_rate": 1.306083849791735e-05,
+      "loss": 0.5231,
+      "step": 3780
+    },
+    {
+      "epoch": 1.3808654372831843,
+      "grad_norm": 0.8295239210128784,
+      "learning_rate": 1.304738622500232e-05,
+      "loss": 0.5306,
+      "step": 3781
+    },
+    {
+      "epoch": 1.3812306006938104,
+      "grad_norm": 1.0325911045074463,
+      "learning_rate": 1.3033937528928093e-05,
+      "loss": 0.5435,
+      "step": 3782
+    },
+    {
+      "epoch": 1.3815957641044367,
+      "grad_norm": 0.9571465849876404,
+      "learning_rate": 1.3020492416613468e-05,
+      "loss": 0.5611,
+      "step": 3783
+    },
+    {
+      "epoch": 1.381960927515063,
+      "grad_norm": 0.8533524870872498,
+      "learning_rate": 1.3007050894975433e-05,
+      "loss": 0.5857,
+      "step": 3784
+    },
+    {
+      "epoch": 1.3823260909256891,
+      "grad_norm": 0.7776716351509094,
+      "learning_rate": 1.2993612970929097e-05,
+      "loss": 0.5905,
+      "step": 3785
+    },
+    {
+      "epoch": 1.3826912543363155,
+      "grad_norm": 0.806546688079834,
+      "learning_rate": 1.2980178651387738e-05,
+      "loss": 0.5858,
+      "step": 3786
+    },
+    {
+      "epoch": 1.3830564177469418,
+      "grad_norm": 0.9101070165634155,
+      "learning_rate": 1.2966747943262786e-05,
+      "loss": 0.567,
+      "step": 3787
+    },
+    {
+      "epoch": 1.383421581157568,
+      "grad_norm": 1.0257490873336792,
+      "learning_rate": 1.2953320853463782e-05,
+      "loss": 0.5518,
+      "step": 3788
+    },
+    {
+      "epoch": 1.3837867445681944,
+      "grad_norm": 0.9479292035102844,
+      "learning_rate": 1.293989738889843e-05,
+      "loss": 0.5491,
+      "step": 3789
+    },
+    {
+      "epoch": 1.3841519079788205,
+      "grad_norm": 0.9954557418823242,
+      "learning_rate": 1.2926477556472573e-05,
+      "loss": 0.5419,
+      "step": 3790
+    },
+    {
+      "epoch": 1.3845170713894468,
+      "grad_norm": 0.9924768209457397,
+      "learning_rate": 1.2913061363090166e-05,
+      "loss": 0.5404,
+      "step": 3791
+    },
+    {
+      "epoch": 1.384882234800073,
+      "grad_norm": 0.8244521021842957,
+      "learning_rate": 1.289964881565331e-05,
+      "loss": 0.5723,
+      "step": 3792
+    },
+    {
+      "epoch": 1.3852473982106992,
+      "grad_norm": 0.9758168458938599,
+      "learning_rate": 1.2886239921062216e-05,
+      "loss": 0.5497,
+      "step": 3793
+    },
+    {
+      "epoch": 1.3856125616213255,
+      "grad_norm": 0.7866721749305725,
+      "learning_rate": 1.2872834686215227e-05,
+      "loss": 0.5958,
+      "step": 3794
+    },
+    {
+      "epoch": 1.3859777250319518,
+      "grad_norm": 0.6681350469589233,
+      "learning_rate": 1.2859433118008796e-05,
+      "loss": 0.5775,
+      "step": 3795
+    },
+    {
+      "epoch": 1.3863428884425781,
+      "grad_norm": 0.7156676054000854,
+      "learning_rate": 1.284603522333749e-05,
+      "loss": 0.6006,
+      "step": 3796
+    },
+    {
+      "epoch": 1.3867080518532042,
+      "grad_norm": 0.9035090208053589,
+      "learning_rate": 1.2832641009093995e-05,
+      "loss": 0.5595,
+      "step": 3797
+    },
+    {
+      "epoch": 1.3870732152638305,
+      "grad_norm": 0.8540573120117188,
+      "learning_rate": 1.281925048216909e-05,
+      "loss": 0.5612,
+      "step": 3798
+    },
+    {
+      "epoch": 1.3874383786744569,
+      "grad_norm": 0.8145067095756531,
+      "learning_rate": 1.2805863649451671e-05,
+      "loss": 0.5763,
+      "step": 3799
+    },
+    {
+      "epoch": 1.387803542085083,
+      "grad_norm": 0.8422971367835999,
+      "learning_rate": 1.2792480517828714e-05,
+      "loss": 0.5636,
+      "step": 3800
+    },
+    {
+      "epoch": 1.3881687054957093,
+      "grad_norm": 1.2129441499710083,
+      "learning_rate": 1.2779101094185322e-05,
+      "loss": 0.5776,
+      "step": 3801
+    },
+    {
+      "epoch": 1.3885338689063356,
+      "grad_norm": 0.9555438756942749,
+      "learning_rate": 1.2765725385404655e-05,
+      "loss": 0.5233,
+      "step": 3802
+    },
+    {
+      "epoch": 1.3888990323169619,
+      "grad_norm": 1.0055880546569824,
+      "learning_rate": 1.2752353398367982e-05,
+      "loss": 0.5875,
+      "step": 3803
+    },
+    {
+      "epoch": 1.3892641957275882,
+      "grad_norm": 0.8483778238296509,
+      "learning_rate": 1.2738985139954658e-05,
+      "loss": 0.6154,
+      "step": 3804
+    },
+    {
+      "epoch": 1.3896293591382143,
+      "grad_norm": 0.9318357110023499,
+      "learning_rate": 1.272562061704211e-05,
+      "loss": 0.5445,
+      "step": 3805
+    },
+    {
+      "epoch": 1.3899945225488406,
+      "grad_norm": 0.9205259084701538,
+      "learning_rate": 1.2712259836505854e-05,
+      "loss": 0.4951,
+      "step": 3806
+    },
+    {
+      "epoch": 1.390359685959467,
+      "grad_norm": 0.8118085265159607,
+      "learning_rate": 1.269890280521947e-05,
+      "loss": 0.5425,
+      "step": 3807
+    },
+    {
+      "epoch": 1.390724849370093,
+      "grad_norm": 0.737595796585083,
+      "learning_rate": 1.2685549530054617e-05,
+      "loss": 0.5488,
+      "step": 3808
+    },
+    {
+      "epoch": 1.3910900127807193,
+      "grad_norm": 1.112452507019043,
+      "learning_rate": 1.2672200017881027e-05,
+      "loss": 0.5408,
+      "step": 3809
+    },
+    {
+      "epoch": 1.3914551761913456,
+      "grad_norm": 0.7649756669998169,
+      "learning_rate": 1.2658854275566475e-05,
+      "loss": 0.585,
+      "step": 3810
+    },
+    {
+      "epoch": 1.391820339601972,
+      "grad_norm": 0.9654197692871094,
+      "learning_rate": 1.264551230997682e-05,
+      "loss": 0.5782,
+      "step": 3811
+    },
+    {
+      "epoch": 1.3921855030125982,
+      "grad_norm": 0.8528624176979065,
+      "learning_rate": 1.2632174127975963e-05,
+      "loss": 0.5843,
+      "step": 3812
+    },
+    {
+      "epoch": 1.3925506664232243,
+      "grad_norm": 0.8950546979904175,
+      "learning_rate": 1.2618839736425867e-05,
+      "loss": 0.5574,
+      "step": 3813
+    },
+    {
+      "epoch": 1.3929158298338506,
+      "grad_norm": 0.9986966252326965,
+      "learning_rate": 1.2605509142186543e-05,
+      "loss": 0.5403,
+      "step": 3814
+    },
+    {
+      "epoch": 1.393280993244477,
+      "grad_norm": 0.8956006765365601,
+      "learning_rate": 1.2592182352116037e-05,
+      "loss": 0.598,
+      "step": 3815
+    },
+    {
+      "epoch": 1.393646156655103,
+      "grad_norm": 1.0453033447265625,
+      "learning_rate": 1.2578859373070453e-05,
+      "loss": 0.5408,
+      "step": 3816
+    },
+    {
+      "epoch": 1.3940113200657294,
+      "grad_norm": 0.9151045680046082,
+      "learning_rate": 1.2565540211903931e-05,
+      "loss": 0.578,
+      "step": 3817
+    },
+    {
+      "epoch": 1.3943764834763557,
+      "grad_norm": 0.9833351969718933,
+      "learning_rate": 1.2552224875468642e-05,
+      "loss": 0.5793,
+      "step": 3818
+    },
+    {
+      "epoch": 1.394741646886982,
+      "grad_norm": 0.8560035228729248,
+      "learning_rate": 1.2538913370614795e-05,
+      "loss": 0.558,
+      "step": 3819
+    },
+    {
+      "epoch": 1.3951068102976083,
+      "grad_norm": 1.030977725982666,
+      "learning_rate": 1.2525605704190622e-05,
+      "loss": 0.5409,
+      "step": 3820
+    },
+    {
+      "epoch": 1.3954719737082344,
+      "grad_norm": 0.9003497958183289,
+      "learning_rate": 1.251230188304239e-05,
+      "loss": 0.5715,
+      "step": 3821
+    },
+    {
+      "epoch": 1.3958371371188607,
+      "grad_norm": 1.1778000593185425,
+      "learning_rate": 1.2499001914014373e-05,
+      "loss": 0.5281,
+      "step": 3822
+    },
+    {
+      "epoch": 1.396202300529487,
+      "grad_norm": 0.8389256000518799,
+      "learning_rate": 1.2485705803948877e-05,
+      "loss": 0.5568,
+      "step": 3823
+    },
+    {
+      "epoch": 1.396567463940113,
+      "grad_norm": 0.9334737658500671,
+      "learning_rate": 1.2472413559686212e-05,
+      "loss": 0.5723,
+      "step": 3824
+    },
+    {
+      "epoch": 1.3969326273507394,
+      "grad_norm": 0.7888615131378174,
+      "learning_rate": 1.2459125188064713e-05,
+      "loss": 0.5525,
+      "step": 3825
+    },
+    {
+      "epoch": 1.3972977907613657,
+      "grad_norm": 0.8109673857688904,
+      "learning_rate": 1.24458406959207e-05,
+      "loss": 0.5836,
+      "step": 3826
+    },
+    {
+      "epoch": 1.397662954171992,
+      "grad_norm": 0.8338742256164551,
+      "learning_rate": 1.2432560090088533e-05,
+      "loss": 0.5601,
+      "step": 3827
+    },
+    {
+      "epoch": 1.3980281175826184,
+      "grad_norm": 0.8548614382743835,
+      "learning_rate": 1.241928337740053e-05,
+      "loss": 0.5526,
+      "step": 3828
+    },
+    {
+      "epoch": 1.3983932809932444,
+      "grad_norm": 1.429435133934021,
+      "learning_rate": 1.240601056468703e-05,
+      "loss": 0.563,
+      "step": 3829
+    },
+    {
+      "epoch": 1.3987584444038708,
+      "grad_norm": 1.2365398406982422,
+      "learning_rate": 1.2392741658776368e-05,
+      "loss": 0.5715,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3991236078144969,
+      "grad_norm": 0.976597249507904,
+      "learning_rate": 1.2379476666494866e-05,
+      "loss": 0.5486,
+      "step": 3831
+    },
+    {
+      "epoch": 1.3994887712251232,
+      "grad_norm": 0.9714939594268799,
+      "learning_rate": 1.2366215594666822e-05,
+      "loss": 0.5401,
+      "step": 3832
+    },
+    {
+      "epoch": 1.3998539346357495,
+      "grad_norm": 0.7221394777297974,
+      "learning_rate": 1.2352958450114532e-05,
+      "loss": 0.5636,
+      "step": 3833
+    },
+    {
+      "epoch": 1.4002190980463758,
+      "grad_norm": 0.7672884464263916,
+      "learning_rate": 1.2339705239658262e-05,
+      "loss": 0.5564,
+      "step": 3834
+    },
+    {
+      "epoch": 1.400584261457002,
+      "grad_norm": 0.6529276967048645,
+      "learning_rate": 1.232645597011626e-05,
+      "loss": 0.5907,
+      "step": 3835
+    },
+    {
+      "epoch": 1.4009494248676282,
+      "grad_norm": 0.9142425656318665,
+      "learning_rate": 1.2313210648304739e-05,
+      "loss": 0.5649,
+      "step": 3836
+    },
+    {
+      "epoch": 1.4013145882782545,
+      "grad_norm": 0.6880323886871338,
+      "learning_rate": 1.229996928103789e-05,
+      "loss": 0.5952,
+      "step": 3837
+    },
+    {
+      "epoch": 1.4016797516888808,
+      "grad_norm": 0.587834894657135,
+      "learning_rate": 1.2286731875127858e-05,
+      "loss": 0.5665,
+      "step": 3838
+    },
+    {
+      "epoch": 1.402044915099507,
+      "grad_norm": 1.0331255197525024,
+      "learning_rate": 1.2273498437384763e-05,
+      "loss": 0.5191,
+      "step": 3839
+    },
+    {
+      "epoch": 1.4024100785101332,
+      "grad_norm": 0.9560843110084534,
+      "learning_rate": 1.2260268974616683e-05,
+      "loss": 0.5837,
+      "step": 3840
+    },
+    {
+      "epoch": 1.4027752419207595,
+      "grad_norm": 0.7214618921279907,
+      "learning_rate": 1.224704349362964e-05,
+      "loss": 0.6163,
+      "step": 3841
+    },
+    {
+      "epoch": 1.4031404053313858,
+      "grad_norm": 0.9066854119300842,
+      "learning_rate": 1.2233822001227606e-05,
+      "loss": 0.5465,
+      "step": 3842
+    },
+    {
+      "epoch": 1.4035055687420122,
+      "grad_norm": 0.9757594466209412,
+      "learning_rate": 1.2220604504212519e-05,
+      "loss": 0.5285,
+      "step": 3843
+    },
+    {
+      "epoch": 1.4038707321526382,
+      "grad_norm": 0.8267287015914917,
+      "learning_rate": 1.2207391009384244e-05,
+      "loss": 0.5702,
+      "step": 3844
+    },
+    {
+      "epoch": 1.4042358955632646,
+      "grad_norm": 0.6605767011642456,
+      "learning_rate": 1.2194181523540601e-05,
+      "loss": 0.5706,
+      "step": 3845
+    },
+    {
+      "epoch": 1.4046010589738909,
+      "grad_norm": 0.9179114103317261,
+      "learning_rate": 1.2180976053477332e-05,
+      "loss": 0.5471,
+      "step": 3846
+    },
+    {
+      "epoch": 1.404966222384517,
+      "grad_norm": 0.9867071509361267,
+      "learning_rate": 1.2167774605988126e-05,
+      "loss": 0.525,
+      "step": 3847
+    },
+    {
+      "epoch": 1.4053313857951433,
+      "grad_norm": 0.8408907651901245,
+      "learning_rate": 1.2154577187864595e-05,
+      "loss": 0.5522,
+      "step": 3848
+    },
+    {
+      "epoch": 1.4056965492057696,
+      "grad_norm": 1.0048277378082275,
+      "learning_rate": 1.214138380589629e-05,
+      "loss": 0.555,
+      "step": 3849
+    },
+    {
+      "epoch": 1.406061712616396,
+      "grad_norm": 1.020862102508545,
+      "learning_rate": 1.2128194466870666e-05,
+      "loss": 0.5525,
+      "step": 3850
+    },
+    {
+      "epoch": 1.4064268760270222,
+      "grad_norm": 0.812157928943634,
+      "learning_rate": 1.2115009177573112e-05,
+      "loss": 0.5604,
+      "step": 3851
+    },
+    {
+      "epoch": 1.4067920394376483,
+      "grad_norm": 0.7309326529502869,
+      "learning_rate": 1.2101827944786936e-05,
+      "loss": 0.5843,
+      "step": 3852
+    },
+    {
+      "epoch": 1.4071572028482746,
+      "grad_norm": 0.7321484088897705,
+      "learning_rate": 1.2088650775293344e-05,
+      "loss": 0.5268,
+      "step": 3853
+    },
+    {
+      "epoch": 1.407522366258901,
+      "grad_norm": 1.0029045343399048,
+      "learning_rate": 1.2075477675871472e-05,
+      "loss": 0.5575,
+      "step": 3854
+    },
+    {
+      "epoch": 1.407887529669527,
+      "grad_norm": 0.7887302041053772,
+      "learning_rate": 1.2062308653298343e-05,
+      "loss": 0.5568,
+      "step": 3855
+    },
+    {
+      "epoch": 1.4082526930801533,
+      "grad_norm": 0.85622638463974,
+      "learning_rate": 1.2049143714348886e-05,
+      "loss": 0.5516,
+      "step": 3856
+    },
+    {
+      "epoch": 1.4086178564907796,
+      "grad_norm": 0.9460580348968506,
+      "learning_rate": 1.2035982865795944e-05,
+      "loss": 0.5249,
+      "step": 3857
+    },
+    {
+      "epoch": 1.408983019901406,
+      "grad_norm": 0.7221261858940125,
+      "learning_rate": 1.202282611441024e-05,
+      "loss": 0.5618,
+      "step": 3858
+    },
+    {
+      "epoch": 1.4093481833120323,
+      "grad_norm": 0.6986485719680786,
+      "learning_rate": 1.20096734669604e-05,
+      "loss": 0.5591,
+      "step": 3859
+    },
+    {
+      "epoch": 1.4097133467226584,
+      "grad_norm": 0.9845605492591858,
+      "learning_rate": 1.1996524930212921e-05,
+      "loss": 0.5518,
+      "step": 3860
+    },
+    {
+      "epoch": 1.4100785101332847,
+      "grad_norm": 1.0616382360458374,
+      "learning_rate": 1.198338051093221e-05,
+      "loss": 0.5635,
+      "step": 3861
+    },
+    {
+      "epoch": 1.410443673543911,
+      "grad_norm": 0.7955450415611267,
+      "learning_rate": 1.197024021588054e-05,
+      "loss": 0.5477,
+      "step": 3862
+    },
+    {
+      "epoch": 1.410808836954537,
+      "grad_norm": 1.042151689529419,
+      "learning_rate": 1.1957104051818063e-05,
+      "loss": 0.5869,
+      "step": 3863
+    },
+    {
+      "epoch": 1.4111740003651634,
+      "grad_norm": 0.8910982608795166,
+      "learning_rate": 1.1943972025502815e-05,
+      "loss": 0.5279,
+      "step": 3864
+    },
+    {
+      "epoch": 1.4115391637757897,
+      "grad_norm": 0.716566264629364,
+      "learning_rate": 1.1930844143690686e-05,
+      "loss": 0.5704,
+      "step": 3865
+    },
+    {
+      "epoch": 1.411904327186416,
+      "grad_norm": 1.04740571975708,
+      "learning_rate": 1.1917720413135454e-05,
+      "loss": 0.5475,
+      "step": 3866
+    },
+    {
+      "epoch": 1.412269490597042,
+      "grad_norm": 0.5876196622848511,
+      "learning_rate": 1.1904600840588752e-05,
+      "loss": 0.5816,
+      "step": 3867
+    },
+    {
+      "epoch": 1.4126346540076684,
+      "grad_norm": 0.8768644332885742,
+      "learning_rate": 1.189148543280006e-05,
+      "loss": 0.5486,
+      "step": 3868
+    },
+    {
+      "epoch": 1.4129998174182947,
+      "grad_norm": 0.6389560103416443,
+      "learning_rate": 1.1878374196516745e-05,
+      "loss": 0.5966,
+      "step": 3869
+    },
+    {
+      "epoch": 1.4133649808289208,
+      "grad_norm": 0.7460922598838806,
+      "learning_rate": 1.1865267138484e-05,
+      "loss": 0.5564,
+      "step": 3870
+    },
+    {
+      "epoch": 1.4137301442395471,
+      "grad_norm": 0.7749945521354675,
+      "learning_rate": 1.185216426544489e-05,
+      "loss": 0.5663,
+      "step": 3871
+    },
+    {
+      "epoch": 1.4140953076501734,
+      "grad_norm": 0.9464659094810486,
+      "learning_rate": 1.1839065584140308e-05,
+      "loss": 0.5497,
+      "step": 3872
+    },
+    {
+      "epoch": 1.4144604710607998,
+      "grad_norm": 0.9089056253433228,
+      "learning_rate": 1.1825971101309007e-05,
+      "loss": 0.5338,
+      "step": 3873
+    },
+    {
+      "epoch": 1.414825634471426,
+      "grad_norm": 0.7740195989608765,
+      "learning_rate": 1.1812880823687574e-05,
+      "loss": 0.5854,
+      "step": 3874
+    },
+    {
+      "epoch": 1.4151907978820522,
+      "grad_norm": 1.0285331010818481,
+      "learning_rate": 1.1799794758010425e-05,
+      "loss": 0.5397,
+      "step": 3875
+    },
+    {
+      "epoch": 1.4155559612926785,
+      "grad_norm": 0.7956613898277283,
+      "learning_rate": 1.1786712911009821e-05,
+      "loss": 0.5657,
+      "step": 3876
+    },
+    {
+      "epoch": 1.4159211247033048,
+      "grad_norm": 1.1669707298278809,
+      "learning_rate": 1.1773635289415846e-05,
+      "loss": 0.5629,
+      "step": 3877
+    },
+    {
+      "epoch": 1.4162862881139309,
+      "grad_norm": 1.1565542221069336,
+      "learning_rate": 1.1760561899956412e-05,
+      "loss": 0.6029,
+      "step": 3878
+    },
+    {
+      "epoch": 1.4166514515245572,
+      "grad_norm": 0.7887046933174133,
+      "learning_rate": 1.1747492749357248e-05,
+      "loss": 0.5464,
+      "step": 3879
+    },
+    {
+      "epoch": 1.4170166149351835,
+      "grad_norm": 0.8616452813148499,
+      "learning_rate": 1.1734427844341916e-05,
+      "loss": 0.5793,
+      "step": 3880
+    },
+    {
+      "epoch": 1.4173817783458098,
+      "grad_norm": 0.9074066877365112,
+      "learning_rate": 1.1721367191631788e-05,
+      "loss": 0.549,
+      "step": 3881
+    },
+    {
+      "epoch": 1.4177469417564361,
+      "grad_norm": 0.9903778433799744,
+      "learning_rate": 1.1708310797946028e-05,
+      "loss": 0.5272,
+      "step": 3882
+    },
+    {
+      "epoch": 1.4181121051670622,
+      "grad_norm": 0.9381337761878967,
+      "learning_rate": 1.169525867000164e-05,
+      "loss": 0.561,
+      "step": 3883
+    },
+    {
+      "epoch": 1.4184772685776885,
+      "grad_norm": 0.9176742434501648,
+      "learning_rate": 1.1682210814513422e-05,
+      "loss": 0.5579,
+      "step": 3884
+    },
+    {
+      "epoch": 1.4188424319883148,
+      "grad_norm": 2.271436929702759,
+      "learning_rate": 1.1669167238193965e-05,
+      "loss": 0.5552,
+      "step": 3885
+    },
+    {
+      "epoch": 1.419207595398941,
+      "grad_norm": 0.9941447973251343,
+      "learning_rate": 1.1656127947753668e-05,
+      "loss": 0.5275,
+      "step": 3886
+    },
+    {
+      "epoch": 1.4195727588095672,
+      "grad_norm": 1.0468906164169312,
+      "learning_rate": 1.1643092949900721e-05,
+      "loss": 0.5659,
+      "step": 3887
+    },
+    {
+      "epoch": 1.4199379222201935,
+      "grad_norm": 0.768421471118927,
+      "learning_rate": 1.1630062251341108e-05,
+      "loss": 0.5684,
+      "step": 3888
+    },
+    {
+      "epoch": 1.4203030856308199,
+      "grad_norm": 0.7507640719413757,
+      "learning_rate": 1.161703585877861e-05,
+      "loss": 0.5627,
+      "step": 3889
+    },
+    {
+      "epoch": 1.4206682490414462,
+      "grad_norm": 0.7245061993598938,
+      "learning_rate": 1.1604013778914771e-05,
+      "loss": 0.5872,
+      "step": 3890
+    },
+    {
+      "epoch": 1.4210334124520723,
+      "grad_norm": 1.1061756610870361,
+      "learning_rate": 1.159099601844893e-05,
+      "loss": 0.5411,
+      "step": 3891
+    },
+    {
+      "epoch": 1.4213985758626986,
+      "grad_norm": 0.8531544804573059,
+      "learning_rate": 1.1577982584078207e-05,
+      "loss": 0.5502,
+      "step": 3892
+    },
+    {
+      "epoch": 1.421763739273325,
+      "grad_norm": 0.7668573260307312,
+      "learning_rate": 1.156497348249749e-05,
+      "loss": 0.569,
+      "step": 3893
+    },
+    {
+      "epoch": 1.422128902683951,
+      "grad_norm": 0.6851369142532349,
+      "learning_rate": 1.1551968720399444e-05,
+      "loss": 0.5666,
+      "step": 3894
+    },
+    {
+      "epoch": 1.4224940660945773,
+      "grad_norm": 0.9241596460342407,
+      "learning_rate": 1.1538968304474499e-05,
+      "loss": 0.5724,
+      "step": 3895
+    },
+    {
+      "epoch": 1.4228592295052036,
+      "grad_norm": 0.6880790591239929,
+      "learning_rate": 1.1525972241410827e-05,
+      "loss": 0.5906,
+      "step": 3896
+    },
+    {
+      "epoch": 1.42322439291583,
+      "grad_norm": 0.8247973918914795,
+      "learning_rate": 1.15129805378944e-05,
+      "loss": 0.5439,
+      "step": 3897
+    },
+    {
+      "epoch": 1.423589556326456,
+      "grad_norm": 0.9598363041877747,
+      "learning_rate": 1.1499993200608921e-05,
+      "loss": 0.5766,
+      "step": 3898
+    },
+    {
+      "epoch": 1.4239547197370823,
+      "grad_norm": 1.1017124652862549,
+      "learning_rate": 1.1487010236235865e-05,
+      "loss": 0.5811,
+      "step": 3899
+    },
+    {
+      "epoch": 1.4243198831477086,
+      "grad_norm": 0.5694159269332886,
+      "learning_rate": 1.147403165145443e-05,
+      "loss": 0.6052,
+      "step": 3900
+    },
+    {
+      "epoch": 1.4246850465583347,
+      "grad_norm": 1.05720055103302,
+      "learning_rate": 1.1461057452941584e-05,
+      "loss": 0.5368,
+      "step": 3901
+    },
+    {
+      "epoch": 1.425050209968961,
+      "grad_norm": 1.0480480194091797,
+      "learning_rate": 1.1448087647372032e-05,
+      "loss": 0.514,
+      "step": 3902
+    },
+    {
+      "epoch": 1.4254153733795873,
+      "grad_norm": 1.006308913230896,
+      "learning_rate": 1.1435122241418224e-05,
+      "loss": 0.5287,
+      "step": 3903
+    },
+    {
+      "epoch": 1.4257805367902137,
+      "grad_norm": 0.7137683033943176,
+      "learning_rate": 1.142216124175033e-05,
+      "loss": 0.588,
+      "step": 3904
+    },
+    {
+      "epoch": 1.42614570020084,
+      "grad_norm": 0.9662609696388245,
+      "learning_rate": 1.1409204655036272e-05,
+      "loss": 0.5659,
+      "step": 3905
+    },
+    {
+      "epoch": 1.426510863611466,
+      "grad_norm": 0.7355804443359375,
+      "learning_rate": 1.139625248794169e-05,
+      "loss": 0.5449,
+      "step": 3906
+    },
+    {
+      "epoch": 1.4268760270220924,
+      "grad_norm": 0.9621326327323914,
+      "learning_rate": 1.1383304747129964e-05,
+      "loss": 0.5663,
+      "step": 3907
+    },
+    {
+      "epoch": 1.4272411904327187,
+      "grad_norm": 0.9017486572265625,
+      "learning_rate": 1.137036143926217e-05,
+      "loss": 0.5773,
+      "step": 3908
+    },
+    {
+      "epoch": 1.4276063538433448,
+      "grad_norm": 0.7580010890960693,
+      "learning_rate": 1.1357422570997138e-05,
+      "loss": 0.5631,
+      "step": 3909
+    },
+    {
+      "epoch": 1.427971517253971,
+      "grad_norm": 0.9669618010520935,
+      "learning_rate": 1.134448814899138e-05,
+      "loss": 0.5405,
+      "step": 3910
+    },
+    {
+      "epoch": 1.4283366806645974,
+      "grad_norm": 0.8277207016944885,
+      "learning_rate": 1.1331558179899148e-05,
+      "loss": 0.5593,
+      "step": 3911
+    },
+    {
+      "epoch": 1.4287018440752237,
+      "grad_norm": 0.7959915399551392,
+      "learning_rate": 1.1318632670372388e-05,
+      "loss": 0.5966,
+      "step": 3912
+    },
+    {
+      "epoch": 1.42906700748585,
+      "grad_norm": 0.9532966017723083,
+      "learning_rate": 1.1305711627060765e-05,
+      "loss": 0.5316,
+      "step": 3913
+    },
+    {
+      "epoch": 1.4294321708964761,
+      "grad_norm": 0.7921164631843567,
+      "learning_rate": 1.1292795056611621e-05,
+      "loss": 0.5753,
+      "step": 3914
+    },
+    {
+      "epoch": 1.4297973343071024,
+      "grad_norm": 0.8244426250457764,
+      "learning_rate": 1.1279882965670024e-05,
+      "loss": 0.5699,
+      "step": 3915
+    },
+    {
+      "epoch": 1.4301624977177287,
+      "grad_norm": 0.9989232420921326,
+      "learning_rate": 1.1266975360878723e-05,
+      "loss": 0.5644,
+      "step": 3916
+    },
+    {
+      "epoch": 1.4305276611283548,
+      "grad_norm": 0.886845588684082,
+      "learning_rate": 1.1254072248878164e-05,
+      "loss": 0.5635,
+      "step": 3917
+    },
+    {
+      "epoch": 1.4308928245389811,
+      "grad_norm": 0.6461564302444458,
+      "learning_rate": 1.1241173636306488e-05,
+      "loss": 0.5755,
+      "step": 3918
+    },
+    {
+      "epoch": 1.4312579879496075,
+      "grad_norm": 0.7689158320426941,
+      "learning_rate": 1.1228279529799501e-05,
+      "loss": 0.5815,
+      "step": 3919
+    },
+    {
+      "epoch": 1.4316231513602338,
+      "grad_norm": 0.874688446521759,
+      "learning_rate": 1.1215389935990708e-05,
+      "loss": 0.5505,
+      "step": 3920
+    },
+    {
+      "epoch": 1.43198831477086,
+      "grad_norm": 1.1357669830322266,
+      "learning_rate": 1.1202504861511296e-05,
+      "loss": 0.4918,
+      "step": 3921
+    },
+    {
+      "epoch": 1.4323534781814862,
+      "grad_norm": 0.6703139543533325,
+      "learning_rate": 1.1189624312990103e-05,
+      "loss": 0.585,
+      "step": 3922
+    },
+    {
+      "epoch": 1.4327186415921125,
+      "grad_norm": 1.2244127988815308,
+      "learning_rate": 1.1176748297053672e-05,
+      "loss": 0.5466,
+      "step": 3923
+    },
+    {
+      "epoch": 1.4330838050027388,
+      "grad_norm": 0.8089661002159119,
+      "learning_rate": 1.1163876820326179e-05,
+      "loss": 0.5803,
+      "step": 3924
+    },
+    {
+      "epoch": 1.4334489684133649,
+      "grad_norm": 0.7841872572898865,
+      "learning_rate": 1.1151009889429489e-05,
+      "loss": 0.5667,
+      "step": 3925
+    },
+    {
+      "epoch": 1.4338141318239912,
+      "grad_norm": 0.8316547870635986,
+      "learning_rate": 1.1138147510983121e-05,
+      "loss": 0.5403,
+      "step": 3926
+    },
+    {
+      "epoch": 1.4341792952346175,
+      "grad_norm": 0.8686932921409607,
+      "learning_rate": 1.112528969160426e-05,
+      "loss": 0.581,
+      "step": 3927
+    },
+    {
+      "epoch": 1.4345444586452438,
+      "grad_norm": 0.9217099547386169,
+      "learning_rate": 1.1112436437907737e-05,
+      "loss": 0.5358,
+      "step": 3928
+    },
+    {
+      "epoch": 1.4349096220558701,
+      "grad_norm": 0.8455994725227356,
+      "learning_rate": 1.1099587756506022e-05,
+      "loss": 0.5579,
+      "step": 3929
+    },
+    {
+      "epoch": 1.4352747854664962,
+      "grad_norm": 0.7332953214645386,
+      "learning_rate": 1.1086743654009257e-05,
+      "loss": 0.5773,
+      "step": 3930
+    },
+    {
+      "epoch": 1.4356399488771225,
+      "grad_norm": 0.8323222994804382,
+      "learning_rate": 1.1073904137025218e-05,
+      "loss": 0.5199,
+      "step": 3931
+    },
+    {
+      "epoch": 1.4360051122877486,
+      "grad_norm": 0.8535659313201904,
+      "learning_rate": 1.106106921215932e-05,
+      "loss": 0.5389,
+      "step": 3932
+    },
+    {
+      "epoch": 1.436370275698375,
+      "grad_norm": 0.6618138551712036,
+      "learning_rate": 1.1048238886014616e-05,
+      "loss": 0.5463,
+      "step": 3933
+    },
+    {
+      "epoch": 1.4367354391090013,
+      "grad_norm": 0.8994616866111755,
+      "learning_rate": 1.1035413165191792e-05,
+      "loss": 0.568,
+      "step": 3934
+    },
+    {
+      "epoch": 1.4371006025196276,
+      "grad_norm": 0.833524763584137,
+      "learning_rate": 1.1022592056289168e-05,
+      "loss": 0.5856,
+      "step": 3935
+    },
+    {
+      "epoch": 1.4374657659302539,
+      "grad_norm": 0.6980143785476685,
+      "learning_rate": 1.1009775565902686e-05,
+      "loss": 0.5832,
+      "step": 3936
+    },
+    {
+      "epoch": 1.43783092934088,
+      "grad_norm": 0.8310022354125977,
+      "learning_rate": 1.099696370062592e-05,
+      "loss": 0.5659,
+      "step": 3937
+    },
+    {
+      "epoch": 1.4381960927515063,
+      "grad_norm": 1.0758943557739258,
+      "learning_rate": 1.098415646705007e-05,
+      "loss": 0.5527,
+      "step": 3938
+    },
+    {
+      "epoch": 1.4385612561621326,
+      "grad_norm": 0.6438793540000916,
+      "learning_rate": 1.0971353871763925e-05,
+      "loss": 0.587,
+      "step": 3939
+    },
+    {
+      "epoch": 1.4389264195727587,
+      "grad_norm": 0.9396200180053711,
+      "learning_rate": 1.0958555921353918e-05,
+      "loss": 0.5582,
+      "step": 3940
+    },
+    {
+      "epoch": 1.439291582983385,
+      "grad_norm": 0.7176880836486816,
+      "learning_rate": 1.0945762622404078e-05,
+      "loss": 0.5728,
+      "step": 3941
+    },
+    {
+      "epoch": 1.4396567463940113,
+      "grad_norm": 0.6318895816802979,
+      "learning_rate": 1.0932973981496051e-05,
+      "loss": 0.592,
+      "step": 3942
+    },
+    {
+      "epoch": 1.4400219098046376,
+      "grad_norm": 0.7599021196365356,
+      "learning_rate": 1.0920190005209066e-05,
+      "loss": 0.5544,
+      "step": 3943
+    },
+    {
+      "epoch": 1.440387073215264,
+      "grad_norm": 1.7072486877441406,
+      "learning_rate": 1.0907410700119976e-05,
+      "loss": 0.5623,
+      "step": 3944
+    },
+    {
+      "epoch": 1.44075223662589,
+      "grad_norm": 0.8058611154556274,
+      "learning_rate": 1.0894636072803214e-05,
+      "loss": 0.5729,
+      "step": 3945
+    },
+    {
+      "epoch": 1.4411174000365163,
+      "grad_norm": 0.8422569632530212,
+      "learning_rate": 1.0881866129830829e-05,
+      "loss": 0.5716,
+      "step": 3946
+    },
+    {
+      "epoch": 1.4414825634471427,
+      "grad_norm": 0.8342012166976929,
+      "learning_rate": 1.086910087777242e-05,
+      "loss": 0.5366,
+      "step": 3947
+    },
+    {
+      "epoch": 1.4418477268577687,
+      "grad_norm": 0.7607089877128601,
+      "learning_rate": 1.085634032319522e-05,
+      "loss": 0.5294,
+      "step": 3948
+    },
+    {
+      "epoch": 1.442212890268395,
+      "grad_norm": 0.8281744718551636,
+      "learning_rate": 1.0843584472664004e-05,
+      "loss": 0.5604,
+      "step": 3949
+    },
+    {
+      "epoch": 1.4425780536790214,
+      "grad_norm": 0.8964477777481079,
+      "learning_rate": 1.0830833332741154e-05,
+      "loss": 0.5499,
+      "step": 3950
+    },
+    {
+      "epoch": 1.4429432170896477,
+      "grad_norm": 0.8936501145362854,
+      "learning_rate": 1.0818086909986613e-05,
+      "loss": 0.5402,
+      "step": 3951
+    },
+    {
+      "epoch": 1.443308380500274,
+      "grad_norm": 0.7300297617912292,
+      "learning_rate": 1.080534521095792e-05,
+      "loss": 0.5587,
+      "step": 3952
+    },
+    {
+      "epoch": 1.4436735439109,
+      "grad_norm": 0.8601420521736145,
+      "learning_rate": 1.0792608242210151e-05,
+      "loss": 0.568,
+      "step": 3953
+    },
+    {
+      "epoch": 1.4440387073215264,
+      "grad_norm": 0.6742613315582275,
+      "learning_rate": 1.0779876010295971e-05,
+      "loss": 0.5932,
+      "step": 3954
+    },
+    {
+      "epoch": 1.4444038707321527,
+      "grad_norm": 1.2348051071166992,
+      "learning_rate": 1.0767148521765604e-05,
+      "loss": 0.5711,
+      "step": 3955
+    },
+    {
+      "epoch": 1.4447690341427788,
+      "grad_norm": 0.7522004842758179,
+      "learning_rate": 1.0754425783166837e-05,
+      "loss": 0.5435,
+      "step": 3956
+    },
+    {
+      "epoch": 1.445134197553405,
+      "grad_norm": 0.8764027953147888,
+      "learning_rate": 1.0741707801044998e-05,
+      "loss": 0.5298,
+      "step": 3957
+    },
+    {
+      "epoch": 1.4454993609640314,
+      "grad_norm": 0.9089523553848267,
+      "learning_rate": 1.0728994581942982e-05,
+      "loss": 0.5717,
+      "step": 3958
+    },
+    {
+      "epoch": 1.4458645243746577,
+      "grad_norm": 0.7274268269538879,
+      "learning_rate": 1.0716286132401232e-05,
+      "loss": 0.5364,
+      "step": 3959
+    },
+    {
+      "epoch": 1.446229687785284,
+      "grad_norm": 0.6975461840629578,
+      "learning_rate": 1.0703582458957733e-05,
+      "loss": 0.5533,
+      "step": 3960
+    },
+    {
+      "epoch": 1.4465948511959101,
+      "grad_norm": 0.8617317080497742,
+      "learning_rate": 1.0690883568148025e-05,
+      "loss": 0.5683,
+      "step": 3961
+    },
+    {
+      "epoch": 1.4469600146065364,
+      "grad_norm": 1.0526492595672607,
+      "learning_rate": 1.0678189466505172e-05,
+      "loss": 0.5812,
+      "step": 3962
+    },
+    {
+      "epoch": 1.4473251780171628,
+      "grad_norm": 0.8932110667228699,
+      "learning_rate": 1.0665500160559765e-05,
+      "loss": 0.553,
+      "step": 3963
+    },
+    {
+      "epoch": 1.4476903414277889,
+      "grad_norm": 0.7149062156677246,
+      "learning_rate": 1.065281565683996e-05,
+      "loss": 0.5448,
+      "step": 3964
+    },
+    {
+      "epoch": 1.4480555048384152,
+      "grad_norm": 0.8641566634178162,
+      "learning_rate": 1.0640135961871417e-05,
+      "loss": 0.5274,
+      "step": 3965
+    },
+    {
+      "epoch": 1.4484206682490415,
+      "grad_norm": 1.0292166471481323,
+      "learning_rate": 1.0627461082177342e-05,
+      "loss": 0.5384,
+      "step": 3966
+    },
+    {
+      "epoch": 1.4487858316596678,
+      "grad_norm": 0.7775523066520691,
+      "learning_rate": 1.0614791024278437e-05,
+      "loss": 0.5234,
+      "step": 3967
+    },
+    {
+      "epoch": 1.4491509950702939,
+      "grad_norm": 0.7780976295471191,
+      "learning_rate": 1.0602125794692943e-05,
+      "loss": 0.5346,
+      "step": 3968
+    },
+    {
+      "epoch": 1.4495161584809202,
+      "grad_norm": 0.8889197707176208,
+      "learning_rate": 1.0589465399936616e-05,
+      "loss": 0.5238,
+      "step": 3969
+    },
+    {
+      "epoch": 1.4498813218915465,
+      "grad_norm": 0.848050594329834,
+      "learning_rate": 1.0576809846522721e-05,
+      "loss": 0.5987,
+      "step": 3970
+    },
+    {
+      "epoch": 1.4502464853021726,
+      "grad_norm": 0.8156909346580505,
+      "learning_rate": 1.0564159140962036e-05,
+      "loss": 0.5767,
+      "step": 3971
+    },
+    {
+      "epoch": 1.450611648712799,
+      "grad_norm": 1.0465294122695923,
+      "learning_rate": 1.0551513289762832e-05,
+      "loss": 0.5284,
+      "step": 3972
+    },
+    {
+      "epoch": 1.4509768121234252,
+      "grad_norm": 0.7149264216423035,
+      "learning_rate": 1.0538872299430892e-05,
+      "loss": 0.5391,
+      "step": 3973
+    },
+    {
+      "epoch": 1.4513419755340515,
+      "grad_norm": 0.5393648743629456,
+      "learning_rate": 1.052623617646951e-05,
+      "loss": 0.582,
+      "step": 3974
+    },
+    {
+      "epoch": 1.4517071389446778,
+      "grad_norm": 0.8497831225395203,
+      "learning_rate": 1.0513604927379455e-05,
+      "loss": 0.5229,
+      "step": 3975
+    },
+    {
+      "epoch": 1.452072302355304,
+      "grad_norm": 0.666197657585144,
+      "learning_rate": 1.0500978558659001e-05,
+      "loss": 0.5528,
+      "step": 3976
+    },
+    {
+      "epoch": 1.4524374657659302,
+      "grad_norm": 0.864449679851532,
+      "learning_rate": 1.0488357076803903e-05,
+      "loss": 0.538,
+      "step": 3977
+    },
+    {
+      "epoch": 1.4528026291765566,
+      "grad_norm": 0.6567715406417847,
+      "learning_rate": 1.047574048830741e-05,
+      "loss": 0.5606,
+      "step": 3978
+    },
+    {
+      "epoch": 1.4531677925871826,
+      "grad_norm": 0.6790978908538818,
+      "learning_rate": 1.046312879966025e-05,
+      "loss": 0.5479,
+      "step": 3979
+    },
+    {
+      "epoch": 1.453532955997809,
+      "grad_norm": 0.7013542652130127,
+      "learning_rate": 1.045052201735063e-05,
+      "loss": 0.5931,
+      "step": 3980
+    },
+    {
+      "epoch": 1.4538981194084353,
+      "grad_norm": 0.908169686794281,
+      "learning_rate": 1.0437920147864245e-05,
+      "loss": 0.5725,
+      "step": 3981
+    },
+    {
+      "epoch": 1.4542632828190616,
+      "grad_norm": 0.7810689210891724,
+      "learning_rate": 1.0425323197684233e-05,
+      "loss": 0.5818,
+      "step": 3982
+    },
+    {
+      "epoch": 1.454628446229688,
+      "grad_norm": 0.9803763031959534,
+      "learning_rate": 1.0412731173291229e-05,
+      "loss": 0.5341,
+      "step": 3983
+    },
+    {
+      "epoch": 1.454993609640314,
+      "grad_norm": 0.7520824074745178,
+      "learning_rate": 1.0400144081163321e-05,
+      "loss": 0.5522,
+      "step": 3984
+    },
+    {
+      "epoch": 1.4553587730509403,
+      "grad_norm": 1.0242899656295776,
+      "learning_rate": 1.0387561927776075e-05,
+      "loss": 0.5183,
+      "step": 3985
+    },
+    {
+      "epoch": 1.4557239364615666,
+      "grad_norm": 0.9487613439559937,
+      "learning_rate": 1.0374984719602486e-05,
+      "loss": 0.564,
+      "step": 3986
+    },
+    {
+      "epoch": 1.4560890998721927,
+      "grad_norm": 0.9238029718399048,
+      "learning_rate": 1.036241246311303e-05,
+      "loss": 0.5687,
+      "step": 3987
+    },
+    {
+      "epoch": 1.456454263282819,
+      "grad_norm": 0.9257331490516663,
+      "learning_rate": 1.0349845164775639e-05,
+      "loss": 0.5799,
+      "step": 3988
+    },
+    {
+      "epoch": 1.4568194266934453,
+      "grad_norm": 1.1217631101608276,
+      "learning_rate": 1.0337282831055664e-05,
+      "loss": 0.551,
+      "step": 3989
+    },
+    {
+      "epoch": 1.4571845901040716,
+      "grad_norm": 0.8550851345062256,
+      "learning_rate": 1.0324725468415942e-05,
+      "loss": 0.5538,
+      "step": 3990
+    },
+    {
+      "epoch": 1.457549753514698,
+      "grad_norm": 0.7631929516792297,
+      "learning_rate": 1.0312173083316712e-05,
+      "loss": 0.5674,
+      "step": 3991
+    },
+    {
+      "epoch": 1.457914916925324,
+      "grad_norm": 0.8212781548500061,
+      "learning_rate": 1.0299625682215684e-05,
+      "loss": 0.5521,
+      "step": 3992
+    },
+    {
+      "epoch": 1.4582800803359504,
+      "grad_norm": 0.8766182065010071,
+      "learning_rate": 1.028708327156799e-05,
+      "loss": 0.5665,
+      "step": 3993
+    },
+    {
+      "epoch": 1.4586452437465767,
+      "grad_norm": 1.0717082023620605,
+      "learning_rate": 1.0274545857826195e-05,
+      "loss": 0.5619,
+      "step": 3994
+    },
+    {
+      "epoch": 1.4590104071572028,
+      "grad_norm": 1.0999945402145386,
+      "learning_rate": 1.0262013447440311e-05,
+      "loss": 0.5492,
+      "step": 3995
+    },
+    {
+      "epoch": 1.459375570567829,
+      "grad_norm": 0.9581376314163208,
+      "learning_rate": 1.0249486046857735e-05,
+      "loss": 0.5228,
+      "step": 3996
+    },
+    {
+      "epoch": 1.4597407339784554,
+      "grad_norm": 0.8297455310821533,
+      "learning_rate": 1.0236963662523328e-05,
+      "loss": 0.5572,
+      "step": 3997
+    },
+    {
+      "epoch": 1.4601058973890817,
+      "grad_norm": 0.9138409495353699,
+      "learning_rate": 1.0224446300879344e-05,
+      "loss": 0.5201,
+      "step": 3998
+    },
+    {
+      "epoch": 1.4604710607997078,
+      "grad_norm": 0.7559996247291565,
+      "learning_rate": 1.0211933968365484e-05,
+      "loss": 0.5313,
+      "step": 3999
+    },
+    {
+      "epoch": 1.460836224210334,
+      "grad_norm": 0.8069090843200684,
+      "learning_rate": 1.0199426671418818e-05,
+      "loss": 0.5323,
+      "step": 4000
+    },
+    {
+      "epoch": 1.4612013876209604,
+      "grad_norm": 0.8363916277885437,
+      "learning_rate": 1.0186924416473862e-05,
+      "loss": 0.5317,
+      "step": 4001
+    },
+    {
+      "epoch": 1.4615665510315865,
+      "grad_norm": 0.877045750617981,
+      "learning_rate": 1.0174427209962513e-05,
+      "loss": 0.5178,
+      "step": 4002
+    },
+    {
+      "epoch": 1.4619317144422128,
+      "grad_norm": 0.961313784122467,
+      "learning_rate": 1.0161935058314087e-05,
+      "loss": 0.5294,
+      "step": 4003
+    },
+    {
+      "epoch": 1.4622968778528391,
+      "grad_norm": 1.438890814781189,
+      "learning_rate": 1.01494479679553e-05,
+      "loss": 0.602,
+      "step": 4004
+    },
+    {
+      "epoch": 1.4626620412634654,
+      "grad_norm": 1.0085265636444092,
+      "learning_rate": 1.0136965945310262e-05,
+      "loss": 0.5579,
+      "step": 4005
+    },
+    {
+      "epoch": 1.4630272046740918,
+      "grad_norm": 0.838114321231842,
+      "learning_rate": 1.0124488996800456e-05,
+      "loss": 0.5563,
+      "step": 4006
+    },
+    {
+      "epoch": 1.4633923680847178,
+      "grad_norm": 0.6769111156463623,
+      "learning_rate": 1.0112017128844784e-05,
+      "loss": 0.5522,
+      "step": 4007
+    },
+    {
+      "epoch": 1.4637575314953442,
+      "grad_norm": 0.7011337280273438,
+      "learning_rate": 1.0099550347859522e-05,
+      "loss": 0.572,
+      "step": 4008
+    },
+    {
+      "epoch": 1.4641226949059705,
+      "grad_norm": 0.8495879173278809,
+      "learning_rate": 1.008708866025833e-05,
+      "loss": 0.5612,
+      "step": 4009
+    },
+    {
+      "epoch": 1.4644878583165966,
+      "grad_norm": 0.8406143188476562,
+      "learning_rate": 1.0074632072452233e-05,
+      "loss": 0.5535,
+      "step": 4010
+    },
+    {
+      "epoch": 1.4648530217272229,
+      "grad_norm": 0.9218531847000122,
+      "learning_rate": 1.0062180590849655e-05,
+      "loss": 0.5561,
+      "step": 4011
+    },
+    {
+      "epoch": 1.4652181851378492,
+      "grad_norm": 0.8530716896057129,
+      "learning_rate": 1.0049734221856387e-05,
+      "loss": 0.553,
+      "step": 4012
+    },
+    {
+      "epoch": 1.4655833485484755,
+      "grad_norm": 0.6426557302474976,
+      "learning_rate": 1.003729297187558e-05,
+      "loss": 0.5647,
+      "step": 4013
+    },
+    {
+      "epoch": 1.4659485119591018,
+      "grad_norm": 2.151454448699951,
+      "learning_rate": 1.0024856847307766e-05,
+      "loss": 0.5303,
+      "step": 4014
+    },
+    {
+      "epoch": 1.466313675369728,
+      "grad_norm": 1.0635218620300293,
+      "learning_rate": 1.001242585455083e-05,
+      "loss": 0.4997,
+      "step": 4015
+    },
+    {
+      "epoch": 1.4666788387803542,
+      "grad_norm": 0.7495030760765076,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.5418,
+      "step": 4016
+    },
+    {
+      "epoch": 1.4670440021909805,
+      "grad_norm": 0.7568851113319397,
+      "learning_rate": 9.987579290047906e-06,
+      "loss": 0.597,
+      "step": 4017
+    },
+    {
+      "epoch": 1.4674091656016066,
+      "grad_norm": 0.9760790467262268,
+      "learning_rate": 9.97516373108449e-06,
+      "loss": 0.5502,
+      "step": 4018
+    },
+    {
+      "epoch": 1.467774329012233,
+      "grad_norm": 0.9156248569488525,
+      "learning_rate": 9.962753329497069e-06,
+      "loss": 0.5353,
+      "step": 4019
+    },
+    {
+      "epoch": 1.4681394924228592,
+      "grad_norm": 0.5740343332290649,
+      "learning_rate": 9.950348091670281e-06,
+      "loss": 0.5579,
+      "step": 4020
+    },
+    {
+      "epoch": 1.4685046558334856,
+      "grad_norm": 0.9513039588928223,
+      "learning_rate": 9.937948023986135e-06,
+      "loss": 0.5347,
+      "step": 4021
+    },
+    {
+      "epoch": 1.4688698192441119,
+      "grad_norm": 0.9432803988456726,
+      "learning_rate": 9.925553132823967e-06,
+      "loss": 0.5591,
+      "step": 4022
+    },
+    {
+      "epoch": 1.469234982654738,
+      "grad_norm": 0.9032891392707825,
+      "learning_rate": 9.913163424560446e-06,
+      "loss": 0.5587,
+      "step": 4023
+    },
+    {
+      "epoch": 1.4696001460653643,
+      "grad_norm": 0.9759991765022278,
+      "learning_rate": 9.900778905569592e-06,
+      "loss": 0.5101,
+      "step": 4024
+    },
+    {
+      "epoch": 1.4699653094759906,
+      "grad_norm": 0.5861659646034241,
+      "learning_rate": 9.88839958222273e-06,
+      "loss": 0.575,
+      "step": 4025
+    },
+    {
+      "epoch": 1.4703304728866167,
+      "grad_norm": 1.0142534971237183,
+      "learning_rate": 9.876025460888528e-06,
+      "loss": 0.5567,
+      "step": 4026
+    },
+    {
+      "epoch": 1.470695636297243,
+      "grad_norm": 1.1566455364227295,
+      "learning_rate": 9.863656547932976e-06,
+      "loss": 0.5341,
+      "step": 4027
+    },
+    {
+      "epoch": 1.4710607997078693,
+      "grad_norm": 0.8980649709701538,
+      "learning_rate": 9.851292849719392e-06,
+      "loss": 0.5689,
+      "step": 4028
+    },
+    {
+      "epoch": 1.4714259631184956,
+      "grad_norm": 0.6560776829719543,
+      "learning_rate": 9.838934372608394e-06,
+      "loss": 0.5469,
+      "step": 4029
+    },
+    {
+      "epoch": 1.471791126529122,
+      "grad_norm": 1.4901273250579834,
+      "learning_rate": 9.826581122957915e-06,
+      "loss": 0.513,
+      "step": 4030
+    },
+    {
+      "epoch": 1.472156289939748,
+      "grad_norm": 1.2591936588287354,
+      "learning_rate": 9.814233107123215e-06,
+      "loss": 0.5422,
+      "step": 4031
+    },
+    {
+      "epoch": 1.4725214533503743,
+      "grad_norm": 0.7727863788604736,
+      "learning_rate": 9.801890331456851e-06,
+      "loss": 0.5754,
+      "step": 4032
+    },
+    {
+      "epoch": 1.4728866167610004,
+      "grad_norm": 1.1301239728927612,
+      "learning_rate": 9.789552802308697e-06,
+      "loss": 0.5648,
+      "step": 4033
+    },
+    {
+      "epoch": 1.4732517801716267,
+      "grad_norm": 0.7674900889396667,
+      "learning_rate": 9.777220526025897e-06,
+      "loss": 0.5556,
+      "step": 4034
+    },
+    {
+      "epoch": 1.473616943582253,
+      "grad_norm": 1.0524770021438599,
+      "learning_rate": 9.764893508952924e-06,
+      "loss": 0.5739,
+      "step": 4035
+    },
+    {
+      "epoch": 1.4739821069928793,
+      "grad_norm": 1.1383737325668335,
+      "learning_rate": 9.752571757431528e-06,
+      "loss": 0.5726,
+      "step": 4036
+    },
+    {
+      "epoch": 1.4743472704035057,
+      "grad_norm": 1.0157060623168945,
+      "learning_rate": 9.740255277800761e-06,
+      "loss": 0.4792,
+      "step": 4037
+    },
+    {
+      "epoch": 1.4747124338141318,
+      "grad_norm": 0.8764933347702026,
+      "learning_rate": 9.727944076396962e-06,
+      "loss": 0.523,
+      "step": 4038
+    },
+    {
+      "epoch": 1.475077597224758,
+      "grad_norm": 1.1367887258529663,
+      "learning_rate": 9.715638159553737e-06,
+      "loss": 0.5216,
+      "step": 4039
+    },
+    {
+      "epoch": 1.4754427606353844,
+      "grad_norm": 0.9799183011054993,
+      "learning_rate": 9.703337533601995e-06,
+      "loss": 0.5226,
+      "step": 4040
+    },
+    {
+      "epoch": 1.4758079240460105,
+      "grad_norm": 0.8010982275009155,
+      "learning_rate": 9.691042204869918e-06,
+      "loss": 0.5846,
+      "step": 4041
+    },
+    {
+      "epoch": 1.4761730874566368,
+      "grad_norm": 0.9556754231452942,
+      "learning_rate": 9.678752179682947e-06,
+      "loss": 0.5359,
+      "step": 4042
+    },
+    {
+      "epoch": 1.476538250867263,
+      "grad_norm": 0.8769105672836304,
+      "learning_rate": 9.666467464363822e-06,
+      "loss": 0.5394,
+      "step": 4043
+    },
+    {
+      "epoch": 1.4769034142778894,
+      "grad_norm": 1.2128407955169678,
+      "learning_rate": 9.65418806523252e-06,
+      "loss": 0.5633,
+      "step": 4044
+    },
+    {
+      "epoch": 1.4772685776885157,
+      "grad_norm": 1.1116158962249756,
+      "learning_rate": 9.641913988606304e-06,
+      "loss": 0.5294,
+      "step": 4045
+    },
+    {
+      "epoch": 1.4776337410991418,
+      "grad_norm": 2.208587169647217,
+      "learning_rate": 9.629645240799698e-06,
+      "loss": 0.5724,
+      "step": 4046
+    },
+    {
+      "epoch": 1.4779989045097681,
+      "grad_norm": 1.1062947511672974,
+      "learning_rate": 9.617381828124482e-06,
+      "loss": 0.5328,
+      "step": 4047
+    },
+    {
+      "epoch": 1.4783640679203944,
+      "grad_norm": 1.1361697912216187,
+      "learning_rate": 9.605123756889692e-06,
+      "loss": 0.541,
+      "step": 4048
+    },
+    {
+      "epoch": 1.4787292313310205,
+      "grad_norm": 0.9591314792633057,
+      "learning_rate": 9.5928710334016e-06,
+      "loss": 0.5646,
+      "step": 4049
+    },
+    {
+      "epoch": 1.4790943947416468,
+      "grad_norm": 2.141479015350342,
+      "learning_rate": 9.580623663963753e-06,
+      "loss": 0.5472,
+      "step": 4050
+    },
+    {
+      "epoch": 1.4794595581522731,
+      "grad_norm": 0.926795482635498,
+      "learning_rate": 9.568381654876924e-06,
+      "loss": 0.4863,
+      "step": 4051
+    },
+    {
+      "epoch": 1.4798247215628995,
+      "grad_norm": 0.8757331967353821,
+      "learning_rate": 9.55614501243915e-06,
+      "loss": 0.5847,
+      "step": 4052
+    },
+    {
+      "epoch": 1.4801898849735258,
+      "grad_norm": 1.0324450731277466,
+      "learning_rate": 9.54391374294567e-06,
+      "loss": 0.5141,
+      "step": 4053
+    },
+    {
+      "epoch": 1.4805550483841519,
+      "grad_norm": 0.8291592597961426,
+      "learning_rate": 9.531687852689003e-06,
+      "loss": 0.5546,
+      "step": 4054
+    },
+    {
+      "epoch": 1.4809202117947782,
+      "grad_norm": 1.4921578168869019,
+      "learning_rate": 9.519467347958857e-06,
+      "loss": 0.5553,
+      "step": 4055
+    },
+    {
+      "epoch": 1.4812853752054045,
+      "grad_norm": 0.980554461479187,
+      "learning_rate": 9.507252235042205e-06,
+      "loss": 0.5258,
+      "step": 4056
+    },
+    {
+      "epoch": 1.4816505386160306,
+      "grad_norm": 0.7031359672546387,
+      "learning_rate": 9.495042520223233e-06,
+      "loss": 0.5606,
+      "step": 4057
+    },
+    {
+      "epoch": 1.482015702026657,
+      "grad_norm": 0.9737385511398315,
+      "learning_rate": 9.482838209783351e-06,
+      "loss": 0.5645,
+      "step": 4058
+    },
+    {
+      "epoch": 1.4823808654372832,
+      "grad_norm": 0.7892079949378967,
+      "learning_rate": 9.470639310001176e-06,
+      "loss": 0.5356,
+      "step": 4059
+    },
+    {
+      "epoch": 1.4827460288479095,
+      "grad_norm": 0.8065754771232605,
+      "learning_rate": 9.458445827152558e-06,
+      "loss": 0.5406,
+      "step": 4060
+    },
+    {
+      "epoch": 1.4831111922585358,
+      "grad_norm": 1.0150619745254517,
+      "learning_rate": 9.446257767510559e-06,
+      "loss": 0.4894,
+      "step": 4061
+    },
+    {
+      "epoch": 1.483476355669162,
+      "grad_norm": 0.8664150238037109,
+      "learning_rate": 9.434075137345447e-06,
+      "loss": 0.5523,
+      "step": 4062
+    },
+    {
+      "epoch": 1.4838415190797882,
+      "grad_norm": 0.9920085668563843,
+      "learning_rate": 9.421897942924687e-06,
+      "loss": 0.5421,
+      "step": 4063
+    },
+    {
+      "epoch": 1.4842066824904145,
+      "grad_norm": 0.914808988571167,
+      "learning_rate": 9.409726190512962e-06,
+      "loss": 0.5664,
+      "step": 4064
+    },
+    {
+      "epoch": 1.4845718459010406,
+      "grad_norm": 0.8107731342315674,
+      "learning_rate": 9.397559886372152e-06,
+      "loss": 0.5396,
+      "step": 4065
+    },
+    {
+      "epoch": 1.484937009311667,
+      "grad_norm": 0.9913045167922974,
+      "learning_rate": 9.385399036761329e-06,
+      "loss": 0.5233,
+      "step": 4066
+    },
+    {
+      "epoch": 1.4853021727222933,
+      "grad_norm": 0.9688202142715454,
+      "learning_rate": 9.373243647936773e-06,
+      "loss": 0.5298,
+      "step": 4067
+    },
+    {
+      "epoch": 1.4856673361329196,
+      "grad_norm": 1.4763659238815308,
+      "learning_rate": 9.361093726151935e-06,
+      "loss": 0.576,
+      "step": 4068
+    },
+    {
+      "epoch": 1.4860324995435457,
+      "grad_norm": 0.9279058575630188,
+      "learning_rate": 9.348949277657455e-06,
+      "loss": 0.5326,
+      "step": 4069
+    },
+    {
+      "epoch": 1.486397662954172,
+      "grad_norm": 0.8254098296165466,
+      "learning_rate": 9.33681030870117e-06,
+      "loss": 0.5876,
+      "step": 4070
+    },
+    {
+      "epoch": 1.4867628263647983,
+      "grad_norm": 1.0469989776611328,
+      "learning_rate": 9.324676825528095e-06,
+      "loss": 0.5327,
+      "step": 4071
+    },
+    {
+      "epoch": 1.4871279897754244,
+      "grad_norm": 0.7550424337387085,
+      "learning_rate": 9.312548834380429e-06,
+      "loss": 0.5599,
+      "step": 4072
+    },
+    {
+      "epoch": 1.4874931531860507,
+      "grad_norm": 0.8626777529716492,
+      "learning_rate": 9.300426341497515e-06,
+      "loss": 0.4987,
+      "step": 4073
+    },
+    {
+      "epoch": 1.487858316596677,
+      "grad_norm": 0.7991651892662048,
+      "learning_rate": 9.288309353115903e-06,
+      "loss": 0.5625,
+      "step": 4074
+    },
+    {
+      "epoch": 1.4882234800073033,
+      "grad_norm": 1.0298484563827515,
+      "learning_rate": 9.276197875469298e-06,
+      "loss": 0.5132,
+      "step": 4075
+    },
+    {
+      "epoch": 1.4885886434179296,
+      "grad_norm": 0.7250978350639343,
+      "learning_rate": 9.264091914788572e-06,
+      "loss": 0.589,
+      "step": 4076
+    },
+    {
+      "epoch": 1.4889538068285557,
+      "grad_norm": 0.6632958650588989,
+      "learning_rate": 9.251991477301742e-06,
+      "loss": 0.5673,
+      "step": 4077
+    },
+    {
+      "epoch": 1.489318970239182,
+      "grad_norm": 0.8273259997367859,
+      "learning_rate": 9.239896569234008e-06,
+      "loss": 0.5549,
+      "step": 4078
+    },
+    {
+      "epoch": 1.4896841336498083,
+      "grad_norm": 0.7699050307273865,
+      "learning_rate": 9.227807196807711e-06,
+      "loss": 0.5649,
+      "step": 4079
+    },
+    {
+      "epoch": 1.4900492970604344,
+      "grad_norm": 0.9873746037483215,
+      "learning_rate": 9.215723366242352e-06,
+      "loss": 0.5673,
+      "step": 4080
+    },
+    {
+      "epoch": 1.4904144604710607,
+      "grad_norm": 0.8614767789840698,
+      "learning_rate": 9.203645083754581e-06,
+      "loss": 0.5473,
+      "step": 4081
+    },
+    {
+      "epoch": 1.490779623881687,
+      "grad_norm": 0.7521244287490845,
+      "learning_rate": 9.191572355558187e-06,
+      "loss": 0.5263,
+      "step": 4082
+    },
+    {
+      "epoch": 1.4911447872923134,
+      "grad_norm": 1.188138723373413,
+      "learning_rate": 9.17950518786409e-06,
+      "loss": 0.5385,
+      "step": 4083
+    },
+    {
+      "epoch": 1.4915099507029397,
+      "grad_norm": 0.7209674715995789,
+      "learning_rate": 9.167443586880376e-06,
+      "loss": 0.5397,
+      "step": 4084
+    },
+    {
+      "epoch": 1.4918751141135658,
+      "grad_norm": 0.6747141480445862,
+      "learning_rate": 9.155387558812252e-06,
+      "loss": 0.5808,
+      "step": 4085
+    },
+    {
+      "epoch": 1.492240277524192,
+      "grad_norm": 1.232947826385498,
+      "learning_rate": 9.14333710986207e-06,
+      "loss": 0.5044,
+      "step": 4086
+    },
+    {
+      "epoch": 1.4926054409348184,
+      "grad_norm": 0.6864703893661499,
+      "learning_rate": 9.131292246229286e-06,
+      "loss": 0.5612,
+      "step": 4087
+    },
+    {
+      "epoch": 1.4929706043454445,
+      "grad_norm": 0.889289915561676,
+      "learning_rate": 9.119252974110508e-06,
+      "loss": 0.5536,
+      "step": 4088
+    },
+    {
+      "epoch": 1.4933357677560708,
+      "grad_norm": 0.8222779631614685,
+      "learning_rate": 9.107219299699459e-06,
+      "loss": 0.5777,
+      "step": 4089
+    },
+    {
+      "epoch": 1.493700931166697,
+      "grad_norm": 0.8773689270019531,
+      "learning_rate": 9.095191229186977e-06,
+      "loss": 0.5237,
+      "step": 4090
+    },
+    {
+      "epoch": 1.4940660945773234,
+      "grad_norm": 1.2188704013824463,
+      "learning_rate": 9.083168768761035e-06,
+      "loss": 0.5388,
+      "step": 4091
+    },
+    {
+      "epoch": 1.4944312579879497,
+      "grad_norm": 0.9776803255081177,
+      "learning_rate": 9.071151924606688e-06,
+      "loss": 0.5234,
+      "step": 4092
+    },
+    {
+      "epoch": 1.4947964213985758,
+      "grad_norm": 1.0659689903259277,
+      "learning_rate": 9.059140702906128e-06,
+      "loss": 0.5246,
+      "step": 4093
+    },
+    {
+      "epoch": 1.4951615848092021,
+      "grad_norm": 0.952974259853363,
+      "learning_rate": 9.047135109838654e-06,
+      "loss": 0.5211,
+      "step": 4094
+    },
+    {
+      "epoch": 1.4955267482198285,
+      "grad_norm": 1.0041263103485107,
+      "learning_rate": 9.035135151580649e-06,
+      "loss": 0.4886,
+      "step": 4095
+    },
+    {
+      "epoch": 1.4958919116304545,
+      "grad_norm": 0.918709397315979,
+      "learning_rate": 9.023140834305621e-06,
+      "loss": 0.5502,
+      "step": 4096
+    },
+    {
+      "epoch": 1.4962570750410809,
+      "grad_norm": 1.0465898513793945,
+      "learning_rate": 9.011152164184157e-06,
+      "loss": 0.5168,
+      "step": 4097
+    },
+    {
+      "epoch": 1.4966222384517072,
+      "grad_norm": 0.7389519810676575,
+      "learning_rate": 8.999169147383943e-06,
+      "loss": 0.5679,
+      "step": 4098
+    },
+    {
+      "epoch": 1.4969874018623335,
+      "grad_norm": 0.8351882696151733,
+      "learning_rate": 8.987191790069771e-06,
+      "loss": 0.5687,
+      "step": 4099
+    },
+    {
+      "epoch": 1.4973525652729596,
+      "grad_norm": 1.0429890155792236,
+      "learning_rate": 8.975220098403507e-06,
+      "loss": 0.5707,
+      "step": 4100
+    },
+    {
+      "epoch": 1.4977177286835859,
+      "grad_norm": 0.9522386193275452,
+      "learning_rate": 8.963254078544112e-06,
+      "loss": 0.5316,
+      "step": 4101
+    },
+    {
+      "epoch": 1.4980828920942122,
+      "grad_norm": 0.617000162601471,
+      "learning_rate": 8.951293736647608e-06,
+      "loss": 0.5643,
+      "step": 4102
+    },
+    {
+      "epoch": 1.4984480555048383,
+      "grad_norm": 0.8510279059410095,
+      "learning_rate": 8.93933907886712e-06,
+      "loss": 0.5594,
+      "step": 4103
+    },
+    {
+      "epoch": 1.4988132189154646,
+      "grad_norm": 0.9541760683059692,
+      "learning_rate": 8.92739011135284e-06,
+      "loss": 0.5601,
+      "step": 4104
+    },
+    {
+      "epoch": 1.499178382326091,
+      "grad_norm": 0.7965231537818909,
+      "learning_rate": 8.91544684025204e-06,
+      "loss": 0.5201,
+      "step": 4105
+    },
+    {
+      "epoch": 1.4995435457367172,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 8.903509271709035e-06,
+      "loss": 0.5137,
+      "step": 4106
+    },
+    {
+      "epoch": 1.4999087091473435,
+      "grad_norm": 0.8249757289886475,
+      "learning_rate": 8.891577411865237e-06,
+      "loss": 0.5493,
+      "step": 4107
+    },
+    {
+      "epoch": 1.5002738725579698,
+      "grad_norm": 1.1597923040390015,
+      "learning_rate": 8.879651266859116e-06,
+      "loss": 0.5199,
+      "step": 4108
+    },
+    {
+      "epoch": 1.500639035968596,
+      "grad_norm": 1.1648917198181152,
+      "learning_rate": 8.867730842826177e-06,
+      "loss": 0.5038,
+      "step": 4109
+    },
+    {
+      "epoch": 1.5010041993792222,
+      "grad_norm": 0.7594660520553589,
+      "learning_rate": 8.855816145899016e-06,
+      "loss": 0.5369,
+      "step": 4110
+    },
+    {
+      "epoch": 1.5013693627898483,
+      "grad_norm": 1.0559147596359253,
+      "learning_rate": 8.843907182207254e-06,
+      "loss": 0.5389,
+      "step": 4111
+    },
+    {
+      "epoch": 1.5017345262004747,
+      "grad_norm": 0.7653175592422485,
+      "learning_rate": 8.832003957877579e-06,
+      "loss": 0.5305,
+      "step": 4112
+    },
+    {
+      "epoch": 1.502099689611101,
+      "grad_norm": 0.9201658368110657,
+      "learning_rate": 8.820106479033725e-06,
+      "loss": 0.5267,
+      "step": 4113
+    },
+    {
+      "epoch": 1.5024648530217273,
+      "grad_norm": 0.8921972513198853,
+      "learning_rate": 8.808214751796467e-06,
+      "loss": 0.5702,
+      "step": 4114
+    },
+    {
+      "epoch": 1.5028300164323536,
+      "grad_norm": 0.8630145788192749,
+      "learning_rate": 8.796328782283627e-06,
+      "loss": 0.4911,
+      "step": 4115
+    },
+    {
+      "epoch": 1.5031951798429797,
+      "grad_norm": 0.8433764576911926,
+      "learning_rate": 8.784448576610045e-06,
+      "loss": 0.5311,
+      "step": 4116
+    },
+    {
+      "epoch": 1.503560343253606,
+      "grad_norm": 0.9963394999504089,
+      "learning_rate": 8.77257414088762e-06,
+      "loss": 0.5367,
+      "step": 4117
+    },
+    {
+      "epoch": 1.503925506664232,
+      "grad_norm": 0.8315544128417969,
+      "learning_rate": 8.760705481225271e-06,
+      "loss": 0.5399,
+      "step": 4118
+    },
+    {
+      "epoch": 1.5042906700748584,
+      "grad_norm": 0.6828075051307678,
+      "learning_rate": 8.748842603728954e-06,
+      "loss": 0.5844,
+      "step": 4119
+    },
+    {
+      "epoch": 1.5046558334854847,
+      "grad_norm": 1.0928466320037842,
+      "learning_rate": 8.736985514501627e-06,
+      "loss": 0.5097,
+      "step": 4120
+    },
+    {
+      "epoch": 1.505020996896111,
+      "grad_norm": 0.8490285277366638,
+      "learning_rate": 8.725134219643307e-06,
+      "loss": 0.5106,
+      "step": 4121
+    },
+    {
+      "epoch": 1.5053861603067373,
+      "grad_norm": 0.9667754173278809,
+      "learning_rate": 8.71328872525099e-06,
+      "loss": 0.5697,
+      "step": 4122
+    },
+    {
+      "epoch": 1.5057513237173636,
+      "grad_norm": 0.8549676537513733,
+      "learning_rate": 8.701449037418717e-06,
+      "loss": 0.5666,
+      "step": 4123
+    },
+    {
+      "epoch": 1.5061164871279897,
+      "grad_norm": 0.7192597389221191,
+      "learning_rate": 8.68961516223753e-06,
+      "loss": 0.515,
+      "step": 4124
+    },
+    {
+      "epoch": 1.506481650538616,
+      "grad_norm": 0.8196496367454529,
+      "learning_rate": 8.677787105795494e-06,
+      "loss": 0.564,
+      "step": 4125
+    },
+    {
+      "epoch": 1.5068468139492421,
+      "grad_norm": 0.8854207396507263,
+      "learning_rate": 8.66596487417765e-06,
+      "loss": 0.5716,
+      "step": 4126
+    },
+    {
+      "epoch": 1.5072119773598684,
+      "grad_norm": 0.9815844297409058,
+      "learning_rate": 8.654148473466075e-06,
+      "loss": 0.5479,
+      "step": 4127
+    },
+    {
+      "epoch": 1.5075771407704948,
+      "grad_norm": 0.9503182172775269,
+      "learning_rate": 8.642337909739826e-06,
+      "loss": 0.5236,
+      "step": 4128
+    },
+    {
+      "epoch": 1.507942304181121,
+      "grad_norm": 0.8710759878158569,
+      "learning_rate": 8.630533189074979e-06,
+      "loss": 0.5655,
+      "step": 4129
+    },
+    {
+      "epoch": 1.5083074675917474,
+      "grad_norm": 0.9771408438682556,
+      "learning_rate": 8.618734317544569e-06,
+      "loss": 0.5138,
+      "step": 4130
+    },
+    {
+      "epoch": 1.5086726310023737,
+      "grad_norm": 1.1011977195739746,
+      "learning_rate": 8.60694130121865e-06,
+      "loss": 0.5382,
+      "step": 4131
+    },
+    {
+      "epoch": 1.5090377944129998,
+      "grad_norm": 0.9117840528488159,
+      "learning_rate": 8.595154146164257e-06,
+      "loss": 0.5525,
+      "step": 4132
+    },
+    {
+      "epoch": 1.509402957823626,
+      "grad_norm": 0.9619202017784119,
+      "learning_rate": 8.583372858445408e-06,
+      "loss": 0.4971,
+      "step": 4133
+    },
+    {
+      "epoch": 1.5097681212342522,
+      "grad_norm": 0.99234539270401,
+      "learning_rate": 8.57159744412311e-06,
+      "loss": 0.5446,
+      "step": 4134
+    },
+    {
+      "epoch": 1.5101332846448785,
+      "grad_norm": 0.9445732831954956,
+      "learning_rate": 8.559827909255333e-06,
+      "loss": 0.5789,
+      "step": 4135
+    },
+    {
+      "epoch": 1.5104984480555048,
+      "grad_norm": 0.8710625767707825,
+      "learning_rate": 8.548064259897024e-06,
+      "loss": 0.5395,
+      "step": 4136
+    },
+    {
+      "epoch": 1.5108636114661311,
+      "grad_norm": 0.867486834526062,
+      "learning_rate": 8.536306502100118e-06,
+      "loss": 0.5536,
+      "step": 4137
+    },
+    {
+      "epoch": 1.5112287748767574,
+      "grad_norm": 1.0533684492111206,
+      "learning_rate": 8.524554641913504e-06,
+      "loss": 0.5558,
+      "step": 4138
+    },
+    {
+      "epoch": 1.5115939382873838,
+      "grad_norm": 0.8343387246131897,
+      "learning_rate": 8.512808685383056e-06,
+      "loss": 0.5719,
+      "step": 4139
+    },
+    {
+      "epoch": 1.5119591016980098,
+      "grad_norm": 0.9419680833816528,
+      "learning_rate": 8.501068638551577e-06,
+      "loss": 0.5453,
+      "step": 4140
+    },
+    {
+      "epoch": 1.5123242651086362,
+      "grad_norm": 0.8943369388580322,
+      "learning_rate": 8.489334507458862e-06,
+      "loss": 0.5189,
+      "step": 4141
+    },
+    {
+      "epoch": 1.5126894285192622,
+      "grad_norm": 0.9612909555435181,
+      "learning_rate": 8.47760629814165e-06,
+      "loss": 0.5728,
+      "step": 4142
+    },
+    {
+      "epoch": 1.5130545919298886,
+      "grad_norm": 1.2583495378494263,
+      "learning_rate": 8.465884016633629e-06,
+      "loss": 0.5344,
+      "step": 4143
+    },
+    {
+      "epoch": 1.5134197553405149,
+      "grad_norm": 1.2594666481018066,
+      "learning_rate": 8.454167668965457e-06,
+      "loss": 0.5131,
+      "step": 4144
+    },
+    {
+      "epoch": 1.5137849187511412,
+      "grad_norm": 0.8043733835220337,
+      "learning_rate": 8.442457261164705e-06,
+      "loss": 0.5537,
+      "step": 4145
+    },
+    {
+      "epoch": 1.5141500821617675,
+      "grad_norm": 0.8278462886810303,
+      "learning_rate": 8.430752799255918e-06,
+      "loss": 0.5605,
+      "step": 4146
+    },
+    {
+      "epoch": 1.5145152455723938,
+      "grad_norm": 0.7966932058334351,
+      "learning_rate": 8.419054289260569e-06,
+      "loss": 0.5403,
+      "step": 4147
+    },
+    {
+      "epoch": 1.51488040898302,
+      "grad_norm": 0.7923164367675781,
+      "learning_rate": 8.407361737197079e-06,
+      "loss": 0.5519,
+      "step": 4148
+    },
+    {
+      "epoch": 1.5152455723936462,
+      "grad_norm": 0.9273017048835754,
+      "learning_rate": 8.395675149080795e-06,
+      "loss": 0.5249,
+      "step": 4149
+    },
+    {
+      "epoch": 1.5156107358042723,
+      "grad_norm": 0.9245014190673828,
+      "learning_rate": 8.383994530923987e-06,
+      "loss": 0.5533,
+      "step": 4150
+    },
+    {
+      "epoch": 1.5159758992148986,
+      "grad_norm": 1.4228578805923462,
+      "learning_rate": 8.372319888735872e-06,
+      "loss": 0.52,
+      "step": 4151
+    },
+    {
+      "epoch": 1.516341062625525,
+      "grad_norm": 0.9283208847045898,
+      "learning_rate": 8.360651228522583e-06,
+      "loss": 0.5454,
+      "step": 4152
+    },
+    {
+      "epoch": 1.5167062260361512,
+      "grad_norm": 0.8381152749061584,
+      "learning_rate": 8.348988556287185e-06,
+      "loss": 0.5594,
+      "step": 4153
+    },
+    {
+      "epoch": 1.5170713894467776,
+      "grad_norm": 1.025354027748108,
+      "learning_rate": 8.337331878029644e-06,
+      "loss": 0.4973,
+      "step": 4154
+    },
+    {
+      "epoch": 1.5174365528574036,
+      "grad_norm": 0.7261088490486145,
+      "learning_rate": 8.325681199746856e-06,
+      "loss": 0.5406,
+      "step": 4155
+    },
+    {
+      "epoch": 1.51780171626803,
+      "grad_norm": 0.9314051270484924,
+      "learning_rate": 8.314036527432631e-06,
+      "loss": 0.5681,
+      "step": 4156
+    },
+    {
+      "epoch": 1.518166879678656,
+      "grad_norm": 1.000509262084961,
+      "learning_rate": 8.302397867077683e-06,
+      "loss": 0.5456,
+      "step": 4157
+    },
+    {
+      "epoch": 1.5185320430892824,
+      "grad_norm": 1.0051593780517578,
+      "learning_rate": 8.290765224669646e-06,
+      "loss": 0.5569,
+      "step": 4158
+    },
+    {
+      "epoch": 1.5188972064999087,
+      "grad_norm": 0.9077114462852478,
+      "learning_rate": 8.27913860619303e-06,
+      "loss": 0.5591,
+      "step": 4159
+    },
+    {
+      "epoch": 1.519262369910535,
+      "grad_norm": 0.7231484055519104,
+      "learning_rate": 8.267518017629272e-06,
+      "loss": 0.5533,
+      "step": 4160
+    },
+    {
+      "epoch": 1.5196275333211613,
+      "grad_norm": 0.8238896131515503,
+      "learning_rate": 8.255903464956707e-06,
+      "loss": 0.5546,
+      "step": 4161
+    },
+    {
+      "epoch": 1.5199926967317876,
+      "grad_norm": 0.8406558632850647,
+      "learning_rate": 8.24429495415054e-06,
+      "loss": 0.5359,
+      "step": 4162
+    },
+    {
+      "epoch": 1.5203578601424137,
+      "grad_norm": 0.7404271364212036,
+      "learning_rate": 8.232692491182898e-06,
+      "loss": 0.5751,
+      "step": 4163
+    },
+    {
+      "epoch": 1.52072302355304,
+      "grad_norm": 0.8278073668479919,
+      "learning_rate": 8.221096082022773e-06,
+      "loss": 0.5676,
+      "step": 4164
+    },
+    {
+      "epoch": 1.521088186963666,
+      "grad_norm": 0.9355740547180176,
+      "learning_rate": 8.209505732636052e-06,
+      "loss": 0.5716,
+      "step": 4165
+    },
+    {
+      "epoch": 1.5214533503742924,
+      "grad_norm": 0.9325588345527649,
+      "learning_rate": 8.197921448985512e-06,
+      "loss": 0.5132,
+      "step": 4166
+    },
+    {
+      "epoch": 1.5218185137849187,
+      "grad_norm": 0.8123670220375061,
+      "learning_rate": 8.186343237030795e-06,
+      "loss": 0.5347,
+      "step": 4167
+    },
+    {
+      "epoch": 1.522183677195545,
+      "grad_norm": 0.8848276734352112,
+      "learning_rate": 8.174771102728438e-06,
+      "loss": 0.5711,
+      "step": 4168
+    },
+    {
+      "epoch": 1.5225488406061713,
+      "grad_norm": 0.9114759564399719,
+      "learning_rate": 8.163205052031827e-06,
+      "loss": 0.5122,
+      "step": 4169
+    },
+    {
+      "epoch": 1.5229140040167977,
+      "grad_norm": 0.8145031929016113,
+      "learning_rate": 8.151645090891234e-06,
+      "loss": 0.5796,
+      "step": 4170
+    },
+    {
+      "epoch": 1.5232791674274238,
+      "grad_norm": 0.678084135055542,
+      "learning_rate": 8.140091225253797e-06,
+      "loss": 0.5652,
+      "step": 4171
+    },
+    {
+      "epoch": 1.52364433083805,
+      "grad_norm": 0.8058270812034607,
+      "learning_rate": 8.128543461063523e-06,
+      "loss": 0.563,
+      "step": 4172
+    },
+    {
+      "epoch": 1.5240094942486762,
+      "grad_norm": 0.8949602842330933,
+      "learning_rate": 8.11700180426126e-06,
+      "loss": 0.5804,
+      "step": 4173
+    },
+    {
+      "epoch": 1.5243746576593025,
+      "grad_norm": 0.9329439401626587,
+      "learning_rate": 8.105466260784733e-06,
+      "loss": 0.5499,
+      "step": 4174
+    },
+    {
+      "epoch": 1.5247398210699288,
+      "grad_norm": 1.1913632154464722,
+      "learning_rate": 8.093936836568523e-06,
+      "loss": 0.5648,
+      "step": 4175
+    },
+    {
+      "epoch": 1.525104984480555,
+      "grad_norm": 0.7555683255195618,
+      "learning_rate": 8.082413537544045e-06,
+      "loss": 0.5427,
+      "step": 4176
+    },
+    {
+      "epoch": 1.5254701478911814,
+      "grad_norm": 1.0422959327697754,
+      "learning_rate": 8.070896369639578e-06,
+      "loss": 0.5305,
+      "step": 4177
+    },
+    {
+      "epoch": 1.5258353113018077,
+      "grad_norm": 0.8744900822639465,
+      "learning_rate": 8.05938533878025e-06,
+      "loss": 0.5235,
+      "step": 4178
+    },
+    {
+      "epoch": 1.5262004747124338,
+      "grad_norm": 0.6947089433670044,
+      "learning_rate": 8.047880450888013e-06,
+      "loss": 0.5809,
+      "step": 4179
+    },
+    {
+      "epoch": 1.5265656381230601,
+      "grad_norm": 1.1186779737472534,
+      "learning_rate": 8.036381711881674e-06,
+      "loss": 0.5539,
+      "step": 4180
+    },
+    {
+      "epoch": 1.5269308015336862,
+      "grad_norm": 0.7364650368690491,
+      "learning_rate": 8.024889127676874e-06,
+      "loss": 0.5533,
+      "step": 4181
+    },
+    {
+      "epoch": 1.5272959649443125,
+      "grad_norm": 1.0077247619628906,
+      "learning_rate": 8.013402704186095e-06,
+      "loss": 0.5028,
+      "step": 4182
+    },
+    {
+      "epoch": 1.5276611283549388,
+      "grad_norm": 0.8376016616821289,
+      "learning_rate": 8.001922447318624e-06,
+      "loss": 0.5817,
+      "step": 4183
+    },
+    {
+      "epoch": 1.5280262917655651,
+      "grad_norm": 0.9573248028755188,
+      "learning_rate": 7.990448362980601e-06,
+      "loss": 0.5298,
+      "step": 4184
+    },
+    {
+      "epoch": 1.5283914551761915,
+      "grad_norm": 0.9927940368652344,
+      "learning_rate": 7.978980457074983e-06,
+      "loss": 0.5438,
+      "step": 4185
+    },
+    {
+      "epoch": 1.5287566185868176,
+      "grad_norm": 1.1189831495285034,
+      "learning_rate": 7.967518735501545e-06,
+      "loss": 0.4682,
+      "step": 4186
+    },
+    {
+      "epoch": 1.5291217819974439,
+      "grad_norm": 1.0482594966888428,
+      "learning_rate": 7.956063204156892e-06,
+      "loss": 0.527,
+      "step": 4187
+    },
+    {
+      "epoch": 1.52948694540807,
+      "grad_norm": 1.1917227506637573,
+      "learning_rate": 7.944613868934428e-06,
+      "loss": 0.528,
+      "step": 4188
+    },
+    {
+      "epoch": 1.5298521088186963,
+      "grad_norm": 0.9152950048446655,
+      "learning_rate": 7.93317073572437e-06,
+      "loss": 0.5598,
+      "step": 4189
+    },
+    {
+      "epoch": 1.5302172722293226,
+      "grad_norm": 0.9191230535507202,
+      "learning_rate": 7.921733810413754e-06,
+      "loss": 0.4855,
+      "step": 4190
+    },
+    {
+      "epoch": 1.530582435639949,
+      "grad_norm": 0.8195654153823853,
+      "learning_rate": 7.910303098886422e-06,
+      "loss": 0.5565,
+      "step": 4191
+    },
+    {
+      "epoch": 1.5309475990505752,
+      "grad_norm": 0.7268591523170471,
+      "learning_rate": 7.898878607023024e-06,
+      "loss": 0.5603,
+      "step": 4192
+    },
+    {
+      "epoch": 1.5313127624612015,
+      "grad_norm": 1.0640649795532227,
+      "learning_rate": 7.887460340700988e-06,
+      "loss": 0.561,
+      "step": 4193
+    },
+    {
+      "epoch": 1.5316779258718276,
+      "grad_norm": 0.7788658738136292,
+      "learning_rate": 7.87604830579456e-06,
+      "loss": 0.6047,
+      "step": 4194
+    },
+    {
+      "epoch": 1.532043089282454,
+      "grad_norm": 0.5427097082138062,
+      "learning_rate": 7.864642508174778e-06,
+      "loss": 0.586,
+      "step": 4195
+    },
+    {
+      "epoch": 1.53240825269308,
+      "grad_norm": 0.9323387742042542,
+      "learning_rate": 7.853242953709467e-06,
+      "loss": 0.5583,
+      "step": 4196
+    },
+    {
+      "epoch": 1.5327734161037063,
+      "grad_norm": 0.647345781326294,
+      "learning_rate": 7.841849648263233e-06,
+      "loss": 0.5873,
+      "step": 4197
+    },
+    {
+      "epoch": 1.5331385795143326,
+      "grad_norm": 1.0061805248260498,
+      "learning_rate": 7.830462597697476e-06,
+      "loss": 0.5473,
+      "step": 4198
+    },
+    {
+      "epoch": 1.533503742924959,
+      "grad_norm": 0.8690460324287415,
+      "learning_rate": 7.819081807870383e-06,
+      "loss": 0.5322,
+      "step": 4199
+    },
+    {
+      "epoch": 1.5338689063355853,
+      "grad_norm": 0.814120352268219,
+      "learning_rate": 7.807707284636906e-06,
+      "loss": 0.5649,
+      "step": 4200
+    },
+    {
+      "epoch": 1.5342340697462116,
+      "grad_norm": 0.8226633667945862,
+      "learning_rate": 7.796339033848797e-06,
+      "loss": 0.5178,
+      "step": 4201
+    },
+    {
+      "epoch": 1.5345992331568377,
+      "grad_norm": 0.8266099691390991,
+      "learning_rate": 7.784977061354548e-06,
+      "loss": 0.5394,
+      "step": 4202
+    },
+    {
+      "epoch": 1.534964396567464,
+      "grad_norm": 0.9530923962593079,
+      "learning_rate": 7.773621372999437e-06,
+      "loss": 0.5309,
+      "step": 4203
+    },
+    {
+      "epoch": 1.53532955997809,
+      "grad_norm": 1.0963695049285889,
+      "learning_rate": 7.762271974625516e-06,
+      "loss": 0.4514,
+      "step": 4204
+    },
+    {
+      "epoch": 1.5356947233887164,
+      "grad_norm": 0.8349227905273438,
+      "learning_rate": 7.750928872071594e-06,
+      "loss": 0.5683,
+      "step": 4205
+    },
+    {
+      "epoch": 1.5360598867993427,
+      "grad_norm": 1.0259308815002441,
+      "learning_rate": 7.73959207117325e-06,
+      "loss": 0.5078,
+      "step": 4206
+    },
+    {
+      "epoch": 1.536425050209969,
+      "grad_norm": 0.7931904196739197,
+      "learning_rate": 7.728261577762798e-06,
+      "loss": 0.5291,
+      "step": 4207
+    },
+    {
+      "epoch": 1.5367902136205953,
+      "grad_norm": 0.8693494200706482,
+      "learning_rate": 7.716937397669333e-06,
+      "loss": 0.5345,
+      "step": 4208
+    },
+    {
+      "epoch": 1.5371553770312216,
+      "grad_norm": 0.7820866703987122,
+      "learning_rate": 7.705619536718685e-06,
+      "loss": 0.585,
+      "step": 4209
+    },
+    {
+      "epoch": 1.5375205404418477,
+      "grad_norm": 0.6628980040550232,
+      "learning_rate": 7.694308000733443e-06,
+      "loss": 0.5522,
+      "step": 4210
+    },
+    {
+      "epoch": 1.537885703852474,
+      "grad_norm": 0.8858047127723694,
+      "learning_rate": 7.683002795532947e-06,
+      "loss": 0.5928,
+      "step": 4211
+    },
+    {
+      "epoch": 1.5382508672631001,
+      "grad_norm": 0.9911432266235352,
+      "learning_rate": 7.671703926933253e-06,
+      "loss": 0.4933,
+      "step": 4212
+    },
+    {
+      "epoch": 1.5386160306737264,
+      "grad_norm": 1.3404275178909302,
+      "learning_rate": 7.660411400747188e-06,
+      "loss": 0.4949,
+      "step": 4213
+    },
+    {
+      "epoch": 1.5389811940843527,
+      "grad_norm": 0.7487965226173401,
+      "learning_rate": 7.649125222784298e-06,
+      "loss": 0.549,
+      "step": 4214
+    },
+    {
+      "epoch": 1.539346357494979,
+      "grad_norm": 0.7452327609062195,
+      "learning_rate": 7.637845398850879e-06,
+      "loss": 0.548,
+      "step": 4215
+    },
+    {
+      "epoch": 1.5397115209056054,
+      "grad_norm": 1.9849321842193604,
+      "learning_rate": 7.6265719347499376e-06,
+      "loss": 0.5586,
+      "step": 4216
+    },
+    {
+      "epoch": 1.5400766843162315,
+      "grad_norm": 1.342613935470581,
+      "learning_rate": 7.6153048362812166e-06,
+      "loss": 0.5374,
+      "step": 4217
+    },
+    {
+      "epoch": 1.5404418477268578,
+      "grad_norm": 0.7283867597579956,
+      "learning_rate": 7.604044109241191e-06,
+      "loss": 0.5539,
+      "step": 4218
+    },
+    {
+      "epoch": 1.5408070111374839,
+      "grad_norm": 0.8096727728843689,
+      "learning_rate": 7.592789759423049e-06,
+      "loss": 0.579,
+      "step": 4219
+    },
+    {
+      "epoch": 1.5411721745481102,
+      "grad_norm": 0.7584173679351807,
+      "learning_rate": 7.581541792616709e-06,
+      "loss": 0.5505,
+      "step": 4220
+    },
+    {
+      "epoch": 1.5415373379587365,
+      "grad_norm": 0.7899060249328613,
+      "learning_rate": 7.570300214608801e-06,
+      "loss": 0.5756,
+      "step": 4221
+    },
+    {
+      "epoch": 1.5419025013693628,
+      "grad_norm": 0.9617719054222107,
+      "learning_rate": 7.559065031182653e-06,
+      "loss": 0.5525,
+      "step": 4222
+    },
+    {
+      "epoch": 1.5422676647799891,
+      "grad_norm": 0.7457582354545593,
+      "learning_rate": 7.547836248118321e-06,
+      "loss": 0.5548,
+      "step": 4223
+    },
+    {
+      "epoch": 1.5426328281906154,
+      "grad_norm": 0.7833566665649414,
+      "learning_rate": 7.536613871192566e-06,
+      "loss": 0.5673,
+      "step": 4224
+    },
+    {
+      "epoch": 1.5429979916012415,
+      "grad_norm": 0.9961585402488708,
+      "learning_rate": 7.525397906178858e-06,
+      "loss": 0.4791,
+      "step": 4225
+    },
+    {
+      "epoch": 1.5433631550118678,
+      "grad_norm": 0.8389790058135986,
+      "learning_rate": 7.514188358847345e-06,
+      "loss": 0.5434,
+      "step": 4226
+    },
+    {
+      "epoch": 1.543728318422494,
+      "grad_norm": 0.7960191965103149,
+      "learning_rate": 7.502985234964897e-06,
+      "loss": 0.5604,
+      "step": 4227
+    },
+    {
+      "epoch": 1.5440934818331202,
+      "grad_norm": 0.7586323618888855,
+      "learning_rate": 7.491788540295077e-06,
+      "loss": 0.5443,
+      "step": 4228
+    },
+    {
+      "epoch": 1.5444586452437465,
+      "grad_norm": 1.1452782154083252,
+      "learning_rate": 7.480598280598126e-06,
+      "loss": 0.5582,
+      "step": 4229
+    },
+    {
+      "epoch": 1.5448238086543729,
+      "grad_norm": 0.8075771927833557,
+      "learning_rate": 7.4694144616309835e-06,
+      "loss": 0.5435,
+      "step": 4230
+    },
+    {
+      "epoch": 1.5451889720649992,
+      "grad_norm": 0.8981938362121582,
+      "learning_rate": 7.458237089147289e-06,
+      "loss": 0.5468,
+      "step": 4231
+    },
+    {
+      "epoch": 1.5455541354756255,
+      "grad_norm": 0.9306362271308899,
+      "learning_rate": 7.447066168897334e-06,
+      "loss": 0.4987,
+      "step": 4232
+    },
+    {
+      "epoch": 1.5459192988862516,
+      "grad_norm": 1.1683992147445679,
+      "learning_rate": 7.435901706628119e-06,
+      "loss": 0.5735,
+      "step": 4233
+    },
+    {
+      "epoch": 1.5462844622968779,
+      "grad_norm": 0.7260207533836365,
+      "learning_rate": 7.424743708083308e-06,
+      "loss": 0.5517,
+      "step": 4234
+    },
+    {
+      "epoch": 1.546649625707504,
+      "grad_norm": 2.088953971862793,
+      "learning_rate": 7.413592179003255e-06,
+      "loss": 0.5201,
+      "step": 4235
+    },
+    {
+      "epoch": 1.5470147891181303,
+      "grad_norm": 0.9240289330482483,
+      "learning_rate": 7.402447125124956e-06,
+      "loss": 0.5305,
+      "step": 4236
+    },
+    {
+      "epoch": 1.5473799525287566,
+      "grad_norm": 1.0445443391799927,
+      "learning_rate": 7.391308552182104e-06,
+      "loss": 0.4925,
+      "step": 4237
+    },
+    {
+      "epoch": 1.547745115939383,
+      "grad_norm": 0.9031312465667725,
+      "learning_rate": 7.380176465905047e-06,
+      "loss": 0.5323,
+      "step": 4238
+    },
+    {
+      "epoch": 1.5481102793500092,
+      "grad_norm": 0.9161219000816345,
+      "learning_rate": 7.369050872020802e-06,
+      "loss": 0.5013,
+      "step": 4239
+    },
+    {
+      "epoch": 1.5484754427606355,
+      "grad_norm": 0.9719271659851074,
+      "learning_rate": 7.357931776253027e-06,
+      "loss": 0.5147,
+      "step": 4240
+    },
+    {
+      "epoch": 1.5488406061712616,
+      "grad_norm": 1.3940577507019043,
+      "learning_rate": 7.346819184322067e-06,
+      "loss": 0.5789,
+      "step": 4241
+    },
+    {
+      "epoch": 1.549205769581888,
+      "grad_norm": 0.7420766353607178,
+      "learning_rate": 7.3357131019448906e-06,
+      "loss": 0.5715,
+      "step": 4242
+    },
+    {
+      "epoch": 1.549570932992514,
+      "grad_norm": 0.8596094846725464,
+      "learning_rate": 7.324613534835134e-06,
+      "loss": 0.5281,
+      "step": 4243
+    },
+    {
+      "epoch": 1.5499360964031403,
+      "grad_norm": 0.8126491904258728,
+      "learning_rate": 7.313520488703083e-06,
+      "loss": 0.5525,
+      "step": 4244
+    },
+    {
+      "epoch": 1.5503012598137667,
+      "grad_norm": 0.6757723093032837,
+      "learning_rate": 7.3024339692556714e-06,
+      "loss": 0.5642,
+      "step": 4245
+    },
+    {
+      "epoch": 1.550666423224393,
+      "grad_norm": 0.7929092645645142,
+      "learning_rate": 7.291353982196454e-06,
+      "loss": 0.5198,
+      "step": 4246
+    },
+    {
+      "epoch": 1.5510315866350193,
+      "grad_norm": 0.9600295424461365,
+      "learning_rate": 7.280280533225648e-06,
+      "loss": 0.5576,
+      "step": 4247
+    },
+    {
+      "epoch": 1.5513967500456456,
+      "grad_norm": 0.8763399720191956,
+      "learning_rate": 7.269213628040095e-06,
+      "loss": 0.5662,
+      "step": 4248
+    },
+    {
+      "epoch": 1.5517619134562717,
+      "grad_norm": 1.014392375946045,
+      "learning_rate": 7.258153272333281e-06,
+      "loss": 0.5445,
+      "step": 4249
+    },
+    {
+      "epoch": 1.552127076866898,
+      "grad_norm": 0.8908765912055969,
+      "learning_rate": 7.247099471795307e-06,
+      "loss": 0.548,
+      "step": 4250
+    },
+    {
+      "epoch": 1.552492240277524,
+      "grad_norm": 0.7699535489082336,
+      "learning_rate": 7.236052232112912e-06,
+      "loss": 0.5413,
+      "step": 4251
+    },
+    {
+      "epoch": 1.5528574036881504,
+      "grad_norm": 0.9604656100273132,
+      "learning_rate": 7.225011558969457e-06,
+      "loss": 0.5506,
+      "step": 4252
+    },
+    {
+      "epoch": 1.5532225670987767,
+      "grad_norm": 0.8450736999511719,
+      "learning_rate": 7.213977458044925e-06,
+      "loss": 0.5228,
+      "step": 4253
+    },
+    {
+      "epoch": 1.553587730509403,
+      "grad_norm": 1.0771404504776,
+      "learning_rate": 7.202949935015928e-06,
+      "loss": 0.4957,
+      "step": 4254
+    },
+    {
+      "epoch": 1.5539528939200293,
+      "grad_norm": 0.9505804777145386,
+      "learning_rate": 7.191928995555677e-06,
+      "loss": 0.5399,
+      "step": 4255
+    },
+    {
+      "epoch": 1.5543180573306554,
+      "grad_norm": 0.8374097347259521,
+      "learning_rate": 7.180914645333994e-06,
+      "loss": 0.5684,
+      "step": 4256
+    },
+    {
+      "epoch": 1.5546832207412817,
+      "grad_norm": 0.6433123350143433,
+      "learning_rate": 7.1699068900173286e-06,
+      "loss": 0.5648,
+      "step": 4257
+    },
+    {
+      "epoch": 1.5550483841519078,
+      "grad_norm": 0.89088374376297,
+      "learning_rate": 7.158905735268728e-06,
+      "loss": 0.4925,
+      "step": 4258
+    },
+    {
+      "epoch": 1.5554135475625341,
+      "grad_norm": 0.7068061232566833,
+      "learning_rate": 7.147911186747853e-06,
+      "loss": 0.5532,
+      "step": 4259
+    },
+    {
+      "epoch": 1.5557787109731605,
+      "grad_norm": 0.7148109674453735,
+      "learning_rate": 7.136923250110943e-06,
+      "loss": 0.5472,
+      "step": 4260
+    },
+    {
+      "epoch": 1.5561438743837868,
+      "grad_norm": 1.200268268585205,
+      "learning_rate": 7.125941931010858e-06,
+      "loss": 0.5366,
+      "step": 4261
+    },
+    {
+      "epoch": 1.556509037794413,
+      "grad_norm": 1.0052423477172852,
+      "learning_rate": 7.114967235097046e-06,
+      "loss": 0.4856,
+      "step": 4262
+    },
+    {
+      "epoch": 1.5568742012050394,
+      "grad_norm": 0.6853835582733154,
+      "learning_rate": 7.103999168015548e-06,
+      "loss": 0.6011,
+      "step": 4263
+    },
+    {
+      "epoch": 1.5572393646156655,
+      "grad_norm": 0.8341567516326904,
+      "learning_rate": 7.093037735408998e-06,
+      "loss": 0.5544,
+      "step": 4264
+    },
+    {
+      "epoch": 1.5576045280262918,
+      "grad_norm": 0.8696557283401489,
+      "learning_rate": 7.082082942916604e-06,
+      "loss": 0.4904,
+      "step": 4265
+    },
+    {
+      "epoch": 1.5579696914369179,
+      "grad_norm": 0.7784153819084167,
+      "learning_rate": 7.071134796174171e-06,
+      "loss": 0.522,
+      "step": 4266
+    },
+    {
+      "epoch": 1.5583348548475442,
+      "grad_norm": 0.8596316576004028,
+      "learning_rate": 7.060193300814085e-06,
+      "loss": 0.5118,
+      "step": 4267
+    },
+    {
+      "epoch": 1.5587000182581705,
+      "grad_norm": 0.9530545473098755,
+      "learning_rate": 7.049258462465307e-06,
+      "loss": 0.5403,
+      "step": 4268
+    },
+    {
+      "epoch": 1.5590651816687968,
+      "grad_norm": 1.1124368906021118,
+      "learning_rate": 7.03833028675337e-06,
+      "loss": 0.4995,
+      "step": 4269
+    },
+    {
+      "epoch": 1.5594303450794231,
+      "grad_norm": 0.8794035911560059,
+      "learning_rate": 7.027408779300375e-06,
+      "loss": 0.514,
+      "step": 4270
+    },
+    {
+      "epoch": 1.5597955084900494,
+      "grad_norm": 0.7167550325393677,
+      "learning_rate": 7.016493945725007e-06,
+      "loss": 0.5587,
+      "step": 4271
+    },
+    {
+      "epoch": 1.5601606719006755,
+      "grad_norm": 0.8408735990524292,
+      "learning_rate": 7.005585791642506e-06,
+      "loss": 0.524,
+      "step": 4272
+    },
+    {
+      "epoch": 1.5605258353113018,
+      "grad_norm": 0.8484911918640137,
+      "learning_rate": 6.994684322664682e-06,
+      "loss": 0.5269,
+      "step": 4273
+    },
+    {
+      "epoch": 1.560890998721928,
+      "grad_norm": 0.7512218356132507,
+      "learning_rate": 6.983789544399911e-06,
+      "loss": 0.5798,
+      "step": 4274
+    },
+    {
+      "epoch": 1.5612561621325542,
+      "grad_norm": 1.0278340578079224,
+      "learning_rate": 6.9729014624531035e-06,
+      "loss": 0.5275,
+      "step": 4275
+    },
+    {
+      "epoch": 1.5616213255431806,
+      "grad_norm": 0.7132531404495239,
+      "learning_rate": 6.962020082425749e-06,
+      "loss": 0.5517,
+      "step": 4276
+    },
+    {
+      "epoch": 1.5619864889538069,
+      "grad_norm": 1.00215744972229,
+      "learning_rate": 6.951145409915881e-06,
+      "loss": 0.5051,
+      "step": 4277
+    },
+    {
+      "epoch": 1.5623516523644332,
+      "grad_norm": 0.9334973692893982,
+      "learning_rate": 6.940277450518089e-06,
+      "loss": 0.5254,
+      "step": 4278
+    },
+    {
+      "epoch": 1.5627168157750595,
+      "grad_norm": 0.837821364402771,
+      "learning_rate": 6.929416209823485e-06,
+      "loss": 0.5042,
+      "step": 4279
+    },
+    {
+      "epoch": 1.5630819791856856,
+      "grad_norm": 0.9468808770179749,
+      "learning_rate": 6.918561693419754e-06,
+      "loss": 0.5677,
+      "step": 4280
+    },
+    {
+      "epoch": 1.563447142596312,
+      "grad_norm": 0.9204182028770447,
+      "learning_rate": 6.907713906891114e-06,
+      "loss": 0.5432,
+      "step": 4281
+    },
+    {
+      "epoch": 1.563812306006938,
+      "grad_norm": 0.8597743511199951,
+      "learning_rate": 6.896872855818298e-06,
+      "loss": 0.5634,
+      "step": 4282
+    },
+    {
+      "epoch": 1.5641774694175643,
+      "grad_norm": 0.696017861366272,
+      "learning_rate": 6.886038545778611e-06,
+      "loss": 0.5752,
+      "step": 4283
+    },
+    {
+      "epoch": 1.5645426328281906,
+      "grad_norm": 0.833747386932373,
+      "learning_rate": 6.875210982345855e-06,
+      "loss": 0.5622,
+      "step": 4284
+    },
+    {
+      "epoch": 1.564907796238817,
+      "grad_norm": 0.7575204372406006,
+      "learning_rate": 6.8643901710903825e-06,
+      "loss": 0.5383,
+      "step": 4285
+    },
+    {
+      "epoch": 1.5652729596494432,
+      "grad_norm": 0.9199098944664001,
+      "learning_rate": 6.8535761175790686e-06,
+      "loss": 0.5327,
+      "step": 4286
+    },
+    {
+      "epoch": 1.5656381230600693,
+      "grad_norm": 0.7201038599014282,
+      "learning_rate": 6.842768827375308e-06,
+      "loss": 0.5562,
+      "step": 4287
+    },
+    {
+      "epoch": 1.5660032864706956,
+      "grad_norm": 1.0112850666046143,
+      "learning_rate": 6.831968306039025e-06,
+      "loss": 0.5761,
+      "step": 4288
+    },
+    {
+      "epoch": 1.5663684498813217,
+      "grad_norm": 0.9436196684837341,
+      "learning_rate": 6.821174559126644e-06,
+      "loss": 0.5137,
+      "step": 4289
+    },
+    {
+      "epoch": 1.566733613291948,
+      "grad_norm": 0.8342386484146118,
+      "learning_rate": 6.8103875921911185e-06,
+      "loss": 0.5444,
+      "step": 4290
+    },
+    {
+      "epoch": 1.5670987767025744,
+      "grad_norm": 1.1027247905731201,
+      "learning_rate": 6.79960741078191e-06,
+      "loss": 0.5704,
+      "step": 4291
+    },
+    {
+      "epoch": 1.5674639401132007,
+      "grad_norm": 0.9045964479446411,
+      "learning_rate": 6.788834020444997e-06,
+      "loss": 0.5527,
+      "step": 4292
+    },
+    {
+      "epoch": 1.567829103523827,
+      "grad_norm": 0.7952874898910522,
+      "learning_rate": 6.778067426722841e-06,
+      "loss": 0.5579,
+      "step": 4293
+    },
+    {
+      "epoch": 1.5681942669344533,
+      "grad_norm": 1.68389892578125,
+      "learning_rate": 6.767307635154432e-06,
+      "loss": 0.5771,
+      "step": 4294
+    },
+    {
+      "epoch": 1.5685594303450794,
+      "grad_norm": 0.82811039686203,
+      "learning_rate": 6.7565546512752575e-06,
+      "loss": 0.5667,
+      "step": 4295
+    },
+    {
+      "epoch": 1.5689245937557057,
+      "grad_norm": 0.8454976081848145,
+      "learning_rate": 6.74580848061728e-06,
+      "loss": 0.5093,
+      "step": 4296
+    },
+    {
+      "epoch": 1.5692897571663318,
+      "grad_norm": 1.0146894454956055,
+      "learning_rate": 6.735069128708984e-06,
+      "loss": 0.518,
+      "step": 4297
+    },
+    {
+      "epoch": 1.569654920576958,
+      "grad_norm": 0.9876936674118042,
+      "learning_rate": 6.724336601075339e-06,
+      "loss": 0.566,
+      "step": 4298
+    },
+    {
+      "epoch": 1.5700200839875844,
+      "grad_norm": 0.9768635630607605,
+      "learning_rate": 6.713610903237784e-06,
+      "loss": 0.5191,
+      "step": 4299
+    },
+    {
+      "epoch": 1.5703852473982107,
+      "grad_norm": 0.8869251012802124,
+      "learning_rate": 6.702892040714273e-06,
+      "loss": 0.5574,
+      "step": 4300
+    },
+    {
+      "epoch": 1.570750410808837,
+      "grad_norm": 0.7400929927825928,
+      "learning_rate": 6.692180019019226e-06,
+      "loss": 0.5781,
+      "step": 4301
+    },
+    {
+      "epoch": 1.5711155742194634,
+      "grad_norm": 0.9121541380882263,
+      "learning_rate": 6.681474843663556e-06,
+      "loss": 0.5522,
+      "step": 4302
+    },
+    {
+      "epoch": 1.5714807376300894,
+      "grad_norm": 0.8536775708198547,
+      "learning_rate": 6.670776520154634e-06,
+      "loss": 0.554,
+      "step": 4303
+    },
+    {
+      "epoch": 1.5718459010407158,
+      "grad_norm": 1.0310137271881104,
+      "learning_rate": 6.6600850539963215e-06,
+      "loss": 0.5031,
+      "step": 4304
+    },
+    {
+      "epoch": 1.5722110644513418,
+      "grad_norm": 0.9023722410202026,
+      "learning_rate": 6.6494004506889545e-06,
+      "loss": 0.5627,
+      "step": 4305
+    },
+    {
+      "epoch": 1.5725762278619682,
+      "grad_norm": 0.7008075714111328,
+      "learning_rate": 6.638722715729327e-06,
+      "loss": 0.5958,
+      "step": 4306
+    },
+    {
+      "epoch": 1.5729413912725945,
+      "grad_norm": 0.7555179595947266,
+      "learning_rate": 6.628051854610715e-06,
+      "loss": 0.5398,
+      "step": 4307
+    },
+    {
+      "epoch": 1.5733065546832208,
+      "grad_norm": 0.8288542032241821,
+      "learning_rate": 6.617387872822842e-06,
+      "loss": 0.5487,
+      "step": 4308
+    },
+    {
+      "epoch": 1.573671718093847,
+      "grad_norm": 0.8955180644989014,
+      "learning_rate": 6.606730775851891e-06,
+      "loss": 0.5454,
+      "step": 4309
+    },
+    {
+      "epoch": 1.5740368815044734,
+      "grad_norm": 0.9104514718055725,
+      "learning_rate": 6.596080569180517e-06,
+      "loss": 0.5197,
+      "step": 4310
+    },
+    {
+      "epoch": 1.5744020449150995,
+      "grad_norm": 0.9378561973571777,
+      "learning_rate": 6.585437258287823e-06,
+      "loss": 0.5162,
+      "step": 4311
+    },
+    {
+      "epoch": 1.5747672083257258,
+      "grad_norm": 0.8490752577781677,
+      "learning_rate": 6.574800848649374e-06,
+      "loss": 0.5299,
+      "step": 4312
+    },
+    {
+      "epoch": 1.575132371736352,
+      "grad_norm": 0.866038978099823,
+      "learning_rate": 6.564171345737163e-06,
+      "loss": 0.4706,
+      "step": 4313
+    },
+    {
+      "epoch": 1.5754975351469782,
+      "grad_norm": 0.9095643758773804,
+      "learning_rate": 6.553548755019648e-06,
+      "loss": 0.5261,
+      "step": 4314
+    },
+    {
+      "epoch": 1.5758626985576045,
+      "grad_norm": 1.3830903768539429,
+      "learning_rate": 6.542933081961724e-06,
+      "loss": 0.5293,
+      "step": 4315
+    },
+    {
+      "epoch": 1.5762278619682308,
+      "grad_norm": 1.1060370206832886,
+      "learning_rate": 6.532324332024733e-06,
+      "loss": 0.4509,
+      "step": 4316
+    },
+    {
+      "epoch": 1.5765930253788571,
+      "grad_norm": 1.4499001502990723,
+      "learning_rate": 6.521722510666457e-06,
+      "loss": 0.5313,
+      "step": 4317
+    },
+    {
+      "epoch": 1.5769581887894832,
+      "grad_norm": 0.9172452092170715,
+      "learning_rate": 6.511127623341091e-06,
+      "loss": 0.5488,
+      "step": 4318
+    },
+    {
+      "epoch": 1.5773233522001096,
+      "grad_norm": 0.8071887493133545,
+      "learning_rate": 6.5005396754992885e-06,
+      "loss": 0.5442,
+      "step": 4319
+    },
+    {
+      "epoch": 1.5776885156107356,
+      "grad_norm": 1.0549798011779785,
+      "learning_rate": 6.4899586725881235e-06,
+      "loss": 0.5054,
+      "step": 4320
+    },
+    {
+      "epoch": 1.578053679021362,
+      "grad_norm": 1.1160976886749268,
+      "learning_rate": 6.479384620051103e-06,
+      "loss": 0.5472,
+      "step": 4321
+    },
+    {
+      "epoch": 1.5784188424319883,
+      "grad_norm": 0.7750179171562195,
+      "learning_rate": 6.468817523328148e-06,
+      "loss": 0.5519,
+      "step": 4322
+    },
+    {
+      "epoch": 1.5787840058426146,
+      "grad_norm": 0.7808592915534973,
+      "learning_rate": 6.4582573878555975e-06,
+      "loss": 0.5411,
+      "step": 4323
+    },
+    {
+      "epoch": 1.579149169253241,
+      "grad_norm": 1.0584574937820435,
+      "learning_rate": 6.447704219066224e-06,
+      "loss": 0.5027,
+      "step": 4324
+    },
+    {
+      "epoch": 1.5795143326638672,
+      "grad_norm": 1.161388874053955,
+      "learning_rate": 6.437158022389212e-06,
+      "loss": 0.4879,
+      "step": 4325
+    },
+    {
+      "epoch": 1.5798794960744933,
+      "grad_norm": 0.6971124410629272,
+      "learning_rate": 6.4266188032501595e-06,
+      "loss": 0.5498,
+      "step": 4326
+    },
+    {
+      "epoch": 1.5802446594851196,
+      "grad_norm": 0.9522584676742554,
+      "learning_rate": 6.4160865670710605e-06,
+      "loss": 0.5123,
+      "step": 4327
+    },
+    {
+      "epoch": 1.5806098228957457,
+      "grad_norm": 0.8984469175338745,
+      "learning_rate": 6.405561319270335e-06,
+      "loss": 0.4834,
+      "step": 4328
+    },
+    {
+      "epoch": 1.580974986306372,
+      "grad_norm": 0.7459402680397034,
+      "learning_rate": 6.395043065262798e-06,
+      "loss": 0.5795,
+      "step": 4329
+    },
+    {
+      "epoch": 1.5813401497169983,
+      "grad_norm": 0.7355731129646301,
+      "learning_rate": 6.384531810459673e-06,
+      "loss": 0.5507,
+      "step": 4330
+    },
+    {
+      "epoch": 1.5817053131276246,
+      "grad_norm": 0.8336747884750366,
+      "learning_rate": 6.3740275602685835e-06,
+      "loss": 0.4817,
+      "step": 4331
+    },
+    {
+      "epoch": 1.582070476538251,
+      "grad_norm": 1.0634924173355103,
+      "learning_rate": 6.363530320093529e-06,
+      "loss": 0.5401,
+      "step": 4332
+    },
+    {
+      "epoch": 1.5824356399488773,
+      "grad_norm": 0.9377459287643433,
+      "learning_rate": 6.353040095334931e-06,
+      "loss": 0.559,
+      "step": 4333
+    },
+    {
+      "epoch": 1.5828008033595033,
+      "grad_norm": 0.9540988206863403,
+      "learning_rate": 6.342556891389582e-06,
+      "loss": 0.5518,
+      "step": 4334
+    },
+    {
+      "epoch": 1.5831659667701297,
+      "grad_norm": 0.7296430468559265,
+      "learning_rate": 6.332080713650684e-06,
+      "loss": 0.5484,
+      "step": 4335
+    },
+    {
+      "epoch": 1.5835311301807558,
+      "grad_norm": 1.0052680969238281,
+      "learning_rate": 6.321611567507795e-06,
+      "loss": 0.4866,
+      "step": 4336
+    },
+    {
+      "epoch": 1.583896293591382,
+      "grad_norm": 0.9100461006164551,
+      "learning_rate": 6.31114945834687e-06,
+      "loss": 0.513,
+      "step": 4337
+    },
+    {
+      "epoch": 1.5842614570020084,
+      "grad_norm": 0.7573947906494141,
+      "learning_rate": 6.3006943915502506e-06,
+      "loss": 0.5667,
+      "step": 4338
+    },
+    {
+      "epoch": 1.5846266204126347,
+      "grad_norm": 1.0786585807800293,
+      "learning_rate": 6.290246372496646e-06,
+      "loss": 0.5424,
+      "step": 4339
+    },
+    {
+      "epoch": 1.584991783823261,
+      "grad_norm": 0.7700137495994568,
+      "learning_rate": 6.279805406561146e-06,
+      "loss": 0.5446,
+      "step": 4340
+    },
+    {
+      "epoch": 1.5853569472338873,
+      "grad_norm": 1.2569921016693115,
+      "learning_rate": 6.269371499115213e-06,
+      "loss": 0.4965,
+      "step": 4341
+    },
+    {
+      "epoch": 1.5857221106445134,
+      "grad_norm": 0.8147380352020264,
+      "learning_rate": 6.258944655526662e-06,
+      "loss": 0.5451,
+      "step": 4342
+    },
+    {
+      "epoch": 1.5860872740551397,
+      "grad_norm": 1.0732260942459106,
+      "learning_rate": 6.24852488115969e-06,
+      "loss": 0.5136,
+      "step": 4343
+    },
+    {
+      "epoch": 1.5864524374657658,
+      "grad_norm": 0.9572034478187561,
+      "learning_rate": 6.238112181374856e-06,
+      "loss": 0.5621,
+      "step": 4344
+    },
+    {
+      "epoch": 1.5868176008763921,
+      "grad_norm": 0.8697877526283264,
+      "learning_rate": 6.227706561529079e-06,
+      "loss": 0.5513,
+      "step": 4345
+    },
+    {
+      "epoch": 1.5871827642870184,
+      "grad_norm": 0.8382867574691772,
+      "learning_rate": 6.217308026975623e-06,
+      "loss": 0.5509,
+      "step": 4346
+    },
+    {
+      "epoch": 1.5875479276976447,
+      "grad_norm": 0.6258609294891357,
+      "learning_rate": 6.206916583064124e-06,
+      "loss": 0.5468,
+      "step": 4347
+    },
+    {
+      "epoch": 1.587913091108271,
+      "grad_norm": 0.8393055200576782,
+      "learning_rate": 6.196532235140564e-06,
+      "loss": 0.574,
+      "step": 4348
+    },
+    {
+      "epoch": 1.5882782545188974,
+      "grad_norm": 0.9363372325897217,
+      "learning_rate": 6.186154988547266e-06,
+      "loss": 0.5171,
+      "step": 4349
+    },
+    {
+      "epoch": 1.5886434179295235,
+      "grad_norm": 0.6817806363105774,
+      "learning_rate": 6.175784848622913e-06,
+      "loss": 0.5738,
+      "step": 4350
+    },
+    {
+      "epoch": 1.5890085813401496,
+      "grad_norm": 1.1340863704681396,
+      "learning_rate": 6.1654218207025285e-06,
+      "loss": 0.5109,
+      "step": 4351
+    },
+    {
+      "epoch": 1.5893737447507759,
+      "grad_norm": 0.8553128242492676,
+      "learning_rate": 6.155065910117464e-06,
+      "loss": 0.5279,
+      "step": 4352
+    },
+    {
+      "epoch": 1.5897389081614022,
+      "grad_norm": 0.9129517078399658,
+      "learning_rate": 6.144717122195425e-06,
+      "loss": 0.5211,
+      "step": 4353
+    },
+    {
+      "epoch": 1.5901040715720285,
+      "grad_norm": 0.9420212507247925,
+      "learning_rate": 6.134375462260449e-06,
+      "loss": 0.5727,
+      "step": 4354
+    },
+    {
+      "epoch": 1.5904692349826548,
+      "grad_norm": 0.8823181986808777,
+      "learning_rate": 6.124040935632913e-06,
+      "loss": 0.5493,
+      "step": 4355
+    },
+    {
+      "epoch": 1.5908343983932811,
+      "grad_norm": 0.8685112595558167,
+      "learning_rate": 6.113713547629501e-06,
+      "loss": 0.5165,
+      "step": 4356
+    },
+    {
+      "epoch": 1.5911995618039072,
+      "grad_norm": 0.8980146050453186,
+      "learning_rate": 6.103393303563245e-06,
+      "loss": 0.537,
+      "step": 4357
+    },
+    {
+      "epoch": 1.5915647252145335,
+      "grad_norm": 0.8472083210945129,
+      "learning_rate": 6.0930802087435005e-06,
+      "loss": 0.522,
+      "step": 4358
+    },
+    {
+      "epoch": 1.5919298886251596,
+      "grad_norm": 0.8739496469497681,
+      "learning_rate": 6.0827742684759375e-06,
+      "loss": 0.5405,
+      "step": 4359
+    },
+    {
+      "epoch": 1.592295052035786,
+      "grad_norm": 1.0169410705566406,
+      "learning_rate": 6.072475488062557e-06,
+      "loss": 0.5097,
+      "step": 4360
+    },
+    {
+      "epoch": 1.5926602154464122,
+      "grad_norm": 0.8349485397338867,
+      "learning_rate": 6.062183872801662e-06,
+      "loss": 0.5709,
+      "step": 4361
+    },
+    {
+      "epoch": 1.5930253788570385,
+      "grad_norm": 0.8705358505249023,
+      "learning_rate": 6.051899427987866e-06,
+      "loss": 0.5302,
+      "step": 4362
+    },
+    {
+      "epoch": 1.5933905422676649,
+      "grad_norm": 0.832336962223053,
+      "learning_rate": 6.041622158912113e-06,
+      "loss": 0.5589,
+      "step": 4363
+    },
+    {
+      "epoch": 1.5937557056782912,
+      "grad_norm": 1.2144277095794678,
+      "learning_rate": 6.031352070861645e-06,
+      "loss": 0.5204,
+      "step": 4364
+    },
+    {
+      "epoch": 1.5941208690889173,
+      "grad_norm": 1.0942915678024292,
+      "learning_rate": 6.021089169120013e-06,
+      "loss": 0.4832,
+      "step": 4365
+    },
+    {
+      "epoch": 1.5944860324995436,
+      "grad_norm": 0.941108226776123,
+      "learning_rate": 6.010833458967063e-06,
+      "loss": 0.5475,
+      "step": 4366
+    },
+    {
+      "epoch": 1.5948511959101697,
+      "grad_norm": 0.7738760113716125,
+      "learning_rate": 6.000584945678944e-06,
+      "loss": 0.562,
+      "step": 4367
+    },
+    {
+      "epoch": 1.595216359320796,
+      "grad_norm": 0.9169711470603943,
+      "learning_rate": 5.99034363452811e-06,
+      "loss": 0.5416,
+      "step": 4368
+    },
+    {
+      "epoch": 1.5955815227314223,
+      "grad_norm": 0.9156622886657715,
+      "learning_rate": 5.980109530783311e-06,
+      "loss": 0.5087,
+      "step": 4369
+    },
+    {
+      "epoch": 1.5959466861420486,
+      "grad_norm": 0.8764016628265381,
+      "learning_rate": 5.9698826397095676e-06,
+      "loss": 0.4941,
+      "step": 4370
+    },
+    {
+      "epoch": 1.596311849552675,
+      "grad_norm": 0.6216623783111572,
+      "learning_rate": 5.959662966568214e-06,
+      "loss": 0.5913,
+      "step": 4371
+    },
+    {
+      "epoch": 1.5966770129633012,
+      "grad_norm": 1.1773252487182617,
+      "learning_rate": 5.949450516616859e-06,
+      "loss": 0.542,
+      "step": 4372
+    },
+    {
+      "epoch": 1.5970421763739273,
+      "grad_norm": 1.039921522140503,
+      "learning_rate": 5.939245295109401e-06,
+      "loss": 0.5359,
+      "step": 4373
+    },
+    {
+      "epoch": 1.5974073397845536,
+      "grad_norm": 0.9120054244995117,
+      "learning_rate": 5.929047307296023e-06,
+      "loss": 0.5516,
+      "step": 4374
+    },
+    {
+      "epoch": 1.5977725031951797,
+      "grad_norm": 1.2057701349258423,
+      "learning_rate": 5.918856558423171e-06,
+      "loss": 0.531,
+      "step": 4375
+    },
+    {
+      "epoch": 1.598137666605806,
+      "grad_norm": 0.9319344162940979,
+      "learning_rate": 5.908673053733573e-06,
+      "loss": 0.55,
+      "step": 4376
+    },
+    {
+      "epoch": 1.5985028300164323,
+      "grad_norm": 1.064395785331726,
+      "learning_rate": 5.89849679846624e-06,
+      "loss": 0.5352,
+      "step": 4377
+    },
+    {
+      "epoch": 1.5988679934270587,
+      "grad_norm": 0.9176167845726013,
+      "learning_rate": 5.8883277978564434e-06,
+      "loss": 0.5002,
+      "step": 4378
+    },
+    {
+      "epoch": 1.599233156837685,
+      "grad_norm": 0.9245776534080505,
+      "learning_rate": 5.878166057135737e-06,
+      "loss": 0.5517,
+      "step": 4379
+    },
+    {
+      "epoch": 1.5995983202483113,
+      "grad_norm": 0.9759179353713989,
+      "learning_rate": 5.868011581531914e-06,
+      "loss": 0.4983,
+      "step": 4380
+    },
+    {
+      "epoch": 1.5999634836589374,
+      "grad_norm": 0.9405667185783386,
+      "learning_rate": 5.857864376269051e-06,
+      "loss": 0.5289,
+      "step": 4381
+    },
+    {
+      "epoch": 1.6003286470695637,
+      "grad_norm": 0.8187503814697266,
+      "learning_rate": 5.847724446567477e-06,
+      "loss": 0.4938,
+      "step": 4382
+    },
+    {
+      "epoch": 1.6006938104801898,
+      "grad_norm": 0.6253883242607117,
+      "learning_rate": 5.83759179764378e-06,
+      "loss": 0.589,
+      "step": 4383
+    },
+    {
+      "epoch": 1.601058973890816,
+      "grad_norm": 0.6693951487541199,
+      "learning_rate": 5.8274664347108086e-06,
+      "loss": 0.5475,
+      "step": 4384
+    },
+    {
+      "epoch": 1.6014241373014424,
+      "grad_norm": 1.325948715209961,
+      "learning_rate": 5.817348362977642e-06,
+      "loss": 0.487,
+      "step": 4385
+    },
+    {
+      "epoch": 1.6017893007120687,
+      "grad_norm": 0.8186267018318176,
+      "learning_rate": 5.807237587649625e-06,
+      "loss": 0.5538,
+      "step": 4386
+    },
+    {
+      "epoch": 1.602154464122695,
+      "grad_norm": 0.9162182807922363,
+      "learning_rate": 5.7971341139283535e-06,
+      "loss": 0.5602,
+      "step": 4387
+    },
+    {
+      "epoch": 1.6025196275333211,
+      "grad_norm": 1.035686731338501,
+      "learning_rate": 5.787037947011658e-06,
+      "loss": 0.5117,
+      "step": 4388
+    },
+    {
+      "epoch": 1.6028847909439474,
+      "grad_norm": 0.750034511089325,
+      "learning_rate": 5.7769490920936085e-06,
+      "loss": 0.5272,
+      "step": 4389
+    },
+    {
+      "epoch": 1.6032499543545735,
+      "grad_norm": 0.8791800141334534,
+      "learning_rate": 5.76686755436451e-06,
+      "loss": 0.5806,
+      "step": 4390
+    },
+    {
+      "epoch": 1.6036151177651998,
+      "grad_norm": 0.6303296089172363,
+      "learning_rate": 5.756793339010915e-06,
+      "loss": 0.5706,
+      "step": 4391
+    },
+    {
+      "epoch": 1.6039802811758261,
+      "grad_norm": 0.8920360207557678,
+      "learning_rate": 5.746726451215603e-06,
+      "loss": 0.5136,
+      "step": 4392
+    },
+    {
+      "epoch": 1.6043454445864525,
+      "grad_norm": 1.3762799501419067,
+      "learning_rate": 5.7366668961575835e-06,
+      "loss": 0.5153,
+      "step": 4393
+    },
+    {
+      "epoch": 1.6047106079970788,
+      "grad_norm": 1.5691697597503662,
+      "learning_rate": 5.726614679012099e-06,
+      "loss": 0.5258,
+      "step": 4394
+    },
+    {
+      "epoch": 1.605075771407705,
+      "grad_norm": 0.961821973323822,
+      "learning_rate": 5.716569804950603e-06,
+      "loss": 0.5383,
+      "step": 4395
+    },
+    {
+      "epoch": 1.6054409348183312,
+      "grad_norm": 1.1597667932510376,
+      "learning_rate": 5.706532279140782e-06,
+      "loss": 0.502,
+      "step": 4396
+    },
+    {
+      "epoch": 1.6058060982289575,
+      "grad_norm": 1.0048213005065918,
+      "learning_rate": 5.696502106746542e-06,
+      "loss": 0.5543,
+      "step": 4397
+    },
+    {
+      "epoch": 1.6061712616395836,
+      "grad_norm": 1.2439483404159546,
+      "learning_rate": 5.686479292928011e-06,
+      "loss": 0.5152,
+      "step": 4398
+    },
+    {
+      "epoch": 1.6065364250502099,
+      "grad_norm": 0.9458543062210083,
+      "learning_rate": 5.676463842841513e-06,
+      "loss": 0.5099,
+      "step": 4399
+    },
+    {
+      "epoch": 1.6069015884608362,
+      "grad_norm": 1.0276858806610107,
+      "learning_rate": 5.666455761639598e-06,
+      "loss": 0.5485,
+      "step": 4400
+    },
+    {
+      "epoch": 1.6072667518714625,
+      "grad_norm": 0.9634801149368286,
+      "learning_rate": 5.656455054471026e-06,
+      "loss": 0.5549,
+      "step": 4401
+    },
+    {
+      "epoch": 1.6076319152820888,
+      "grad_norm": 0.8646242022514343,
+      "learning_rate": 5.646461726480763e-06,
+      "loss": 0.5114,
+      "step": 4402
+    },
+    {
+      "epoch": 1.6079970786927151,
+      "grad_norm": 5.977184295654297,
+      "learning_rate": 5.636475782809972e-06,
+      "loss": 0.5429,
+      "step": 4403
+    },
+    {
+      "epoch": 1.6083622421033412,
+      "grad_norm": 0.8848747611045837,
+      "learning_rate": 5.6264972285960104e-06,
+      "loss": 0.5088,
+      "step": 4404
+    },
+    {
+      "epoch": 1.6087274055139675,
+      "grad_norm": 0.8562219142913818,
+      "learning_rate": 5.616526068972452e-06,
+      "loss": 0.5502,
+      "step": 4405
+    },
+    {
+      "epoch": 1.6090925689245936,
+      "grad_norm": 0.5673458576202393,
+      "learning_rate": 5.606562309069059e-06,
+      "loss": 0.5854,
+      "step": 4406
+    },
+    {
+      "epoch": 1.60945773233522,
+      "grad_norm": 1.1740148067474365,
+      "learning_rate": 5.596605954011785e-06,
+      "loss": 0.5512,
+      "step": 4407
+    },
+    {
+      "epoch": 1.6098228957458462,
+      "grad_norm": 0.950234055519104,
+      "learning_rate": 5.586657008922782e-06,
+      "loss": 0.4951,
+      "step": 4408
+    },
+    {
+      "epoch": 1.6101880591564726,
+      "grad_norm": 0.7697010040283203,
+      "learning_rate": 5.576715478920369e-06,
+      "loss": 0.5232,
+      "step": 4409
+    },
+    {
+      "epoch": 1.6105532225670989,
+      "grad_norm": 0.8417774438858032,
+      "learning_rate": 5.566781369119072e-06,
+      "loss": 0.5501,
+      "step": 4410
+    },
+    {
+      "epoch": 1.6109183859777252,
+      "grad_norm": 1.1219253540039062,
+      "learning_rate": 5.556854684629593e-06,
+      "loss": 0.5276,
+      "step": 4411
+    },
+    {
+      "epoch": 1.6112835493883513,
+      "grad_norm": 0.8750378489494324,
+      "learning_rate": 5.5469354305588175e-06,
+      "loss": 0.5226,
+      "step": 4412
+    },
+    {
+      "epoch": 1.6116487127989776,
+      "grad_norm": 0.8155550956726074,
+      "learning_rate": 5.537023612009791e-06,
+      "loss": 0.5524,
+      "step": 4413
+    },
+    {
+      "epoch": 1.6120138762096037,
+      "grad_norm": 1.2941229343414307,
+      "learning_rate": 5.527119234081752e-06,
+      "loss": 0.5051,
+      "step": 4414
+    },
+    {
+      "epoch": 1.61237903962023,
+      "grad_norm": 0.8032121658325195,
+      "learning_rate": 5.5172223018701135e-06,
+      "loss": 0.5461,
+      "step": 4415
+    },
+    {
+      "epoch": 1.6127442030308563,
+      "grad_norm": 0.7761710286140442,
+      "learning_rate": 5.5073328204664354e-06,
+      "loss": 0.5779,
+      "step": 4416
+    },
+    {
+      "epoch": 1.6131093664414826,
+      "grad_norm": 0.9179719686508179,
+      "learning_rate": 5.49745079495847e-06,
+      "loss": 0.5251,
+      "step": 4417
+    },
+    {
+      "epoch": 1.613474529852109,
+      "grad_norm": 1.0164270401000977,
+      "learning_rate": 5.487576230430123e-06,
+      "loss": 0.5067,
+      "step": 4418
+    },
+    {
+      "epoch": 1.613839693262735,
+      "grad_norm": 0.854945719242096,
+      "learning_rate": 5.477709131961453e-06,
+      "loss": 0.5369,
+      "step": 4419
+    },
+    {
+      "epoch": 1.6142048566733613,
+      "grad_norm": 0.7561818361282349,
+      "learning_rate": 5.46784950462869e-06,
+      "loss": 0.553,
+      "step": 4420
+    },
+    {
+      "epoch": 1.6145700200839874,
+      "grad_norm": 0.9647868871688843,
+      "learning_rate": 5.457997353504221e-06,
+      "loss": 0.5198,
+      "step": 4421
+    },
+    {
+      "epoch": 1.6149351834946137,
+      "grad_norm": 1.0278557538986206,
+      "learning_rate": 5.448152683656582e-06,
+      "loss": 0.5064,
+      "step": 4422
+    },
+    {
+      "epoch": 1.61530034690524,
+      "grad_norm": 0.8827544450759888,
+      "learning_rate": 5.4383155001504525e-06,
+      "loss": 0.541,
+      "step": 4423
+    },
+    {
+      "epoch": 1.6156655103158664,
+      "grad_norm": 0.7255877256393433,
+      "learning_rate": 5.428485808046677e-06,
+      "loss": 0.5531,
+      "step": 4424
+    },
+    {
+      "epoch": 1.6160306737264927,
+      "grad_norm": 0.8962209224700928,
+      "learning_rate": 5.418663612402233e-06,
+      "loss": 0.542,
+      "step": 4425
+    },
+    {
+      "epoch": 1.616395837137119,
+      "grad_norm": 0.8296344876289368,
+      "learning_rate": 5.408848918270246e-06,
+      "loss": 0.5864,
+      "step": 4426
+    },
+    {
+      "epoch": 1.616761000547745,
+      "grad_norm": 1.0218677520751953,
+      "learning_rate": 5.399041730699992e-06,
+      "loss": 0.5439,
+      "step": 4427
+    },
+    {
+      "epoch": 1.6171261639583714,
+      "grad_norm": 0.7651405930519104,
+      "learning_rate": 5.389242054736867e-06,
+      "loss": 0.5707,
+      "step": 4428
+    },
+    {
+      "epoch": 1.6174913273689975,
+      "grad_norm": 0.7582070231437683,
+      "learning_rate": 5.3794498954224085e-06,
+      "loss": 0.5297,
+      "step": 4429
+    },
+    {
+      "epoch": 1.6178564907796238,
+      "grad_norm": 0.7480289340019226,
+      "learning_rate": 5.36966525779429e-06,
+      "loss": 0.5734,
+      "step": 4430
+    },
+    {
+      "epoch": 1.61822165419025,
+      "grad_norm": 1.042218804359436,
+      "learning_rate": 5.359888146886316e-06,
+      "loss": 0.489,
+      "step": 4431
+    },
+    {
+      "epoch": 1.6185868176008764,
+      "grad_norm": 0.6962181329727173,
+      "learning_rate": 5.350118567728429e-06,
+      "loss": 0.5604,
+      "step": 4432
+    },
+    {
+      "epoch": 1.6189519810115027,
+      "grad_norm": 1.0039385557174683,
+      "learning_rate": 5.340356525346666e-06,
+      "loss": 0.4632,
+      "step": 4433
+    },
+    {
+      "epoch": 1.619317144422129,
+      "grad_norm": 0.6350936889648438,
+      "learning_rate": 5.330602024763218e-06,
+      "loss": 0.572,
+      "step": 4434
+    },
+    {
+      "epoch": 1.6196823078327551,
+      "grad_norm": 0.8618603348731995,
+      "learning_rate": 5.320855070996383e-06,
+      "loss": 0.5159,
+      "step": 4435
+    },
+    {
+      "epoch": 1.6200474712433814,
+      "grad_norm": 0.776425302028656,
+      "learning_rate": 5.311115669060576e-06,
+      "loss": 0.5402,
+      "step": 4436
+    },
+    {
+      "epoch": 1.6204126346540075,
+      "grad_norm": 0.9595008492469788,
+      "learning_rate": 5.3013838239663325e-06,
+      "loss": 0.5249,
+      "step": 4437
+    },
+    {
+      "epoch": 1.6207777980646338,
+      "grad_norm": 1.2430531978607178,
+      "learning_rate": 5.291659540720289e-06,
+      "loss": 0.495,
+      "step": 4438
+    },
+    {
+      "epoch": 1.6211429614752602,
+      "grad_norm": 0.8708037734031677,
+      "learning_rate": 5.281942824325204e-06,
+      "loss": 0.5486,
+      "step": 4439
+    },
+    {
+      "epoch": 1.6215081248858865,
+      "grad_norm": 0.7520557045936584,
+      "learning_rate": 5.272233679779934e-06,
+      "loss": 0.5284,
+      "step": 4440
+    },
+    {
+      "epoch": 1.6218732882965128,
+      "grad_norm": 0.847500205039978,
+      "learning_rate": 5.262532112079455e-06,
+      "loss": 0.5211,
+      "step": 4441
+    },
+    {
+      "epoch": 1.622238451707139,
+      "grad_norm": 0.9707865715026855,
+      "learning_rate": 5.252838126214827e-06,
+      "loss": 0.5611,
+      "step": 4442
+    },
+    {
+      "epoch": 1.6226036151177652,
+      "grad_norm": 0.8329001069068909,
+      "learning_rate": 5.2431517271732106e-06,
+      "loss": 0.5207,
+      "step": 4443
+    },
+    {
+      "epoch": 1.6229687785283915,
+      "grad_norm": 0.8859490752220154,
+      "learning_rate": 5.233472919937872e-06,
+      "loss": 0.5468,
+      "step": 4444
+    },
+    {
+      "epoch": 1.6233339419390176,
+      "grad_norm": 0.9074608087539673,
+      "learning_rate": 5.2238017094881765e-06,
+      "loss": 0.5367,
+      "step": 4445
+    },
+    {
+      "epoch": 1.623699105349644,
+      "grad_norm": 0.6423451900482178,
+      "learning_rate": 5.214138100799573e-06,
+      "loss": 0.5806,
+      "step": 4446
+    },
+    {
+      "epoch": 1.6240642687602702,
+      "grad_norm": 0.9432792067527771,
+      "learning_rate": 5.2044820988435906e-06,
+      "loss": 0.5418,
+      "step": 4447
+    },
+    {
+      "epoch": 1.6244294321708965,
+      "grad_norm": 1.0792163610458374,
+      "learning_rate": 5.194833708587863e-06,
+      "loss": 0.4911,
+      "step": 4448
+    },
+    {
+      "epoch": 1.6247945955815228,
+      "grad_norm": 1.0764150619506836,
+      "learning_rate": 5.185192934996097e-06,
+      "loss": 0.4913,
+      "step": 4449
+    },
+    {
+      "epoch": 1.6251597589921492,
+      "grad_norm": 0.9995002746582031,
+      "learning_rate": 5.175559783028084e-06,
+      "loss": 0.5176,
+      "step": 4450
+    },
+    {
+      "epoch": 1.6255249224027752,
+      "grad_norm": 0.9153430461883545,
+      "learning_rate": 5.165934257639702e-06,
+      "loss": 0.5353,
+      "step": 4451
+    },
+    {
+      "epoch": 1.6258900858134013,
+      "grad_norm": 0.9202440977096558,
+      "learning_rate": 5.156316363782885e-06,
+      "loss": 0.5455,
+      "step": 4452
+    },
+    {
+      "epoch": 1.6262552492240276,
+      "grad_norm": 0.7982866168022156,
+      "learning_rate": 5.146706106405657e-06,
+      "loss": 0.5364,
+      "step": 4453
+    },
+    {
+      "epoch": 1.626620412634654,
+      "grad_norm": 1.1405035257339478,
+      "learning_rate": 5.137103490452113e-06,
+      "loss": 0.4984,
+      "step": 4454
+    },
+    {
+      "epoch": 1.6269855760452803,
+      "grad_norm": 1.1086018085479736,
+      "learning_rate": 5.1275085208624185e-06,
+      "loss": 0.5183,
+      "step": 4455
+    },
+    {
+      "epoch": 1.6273507394559066,
+      "grad_norm": 0.6876553893089294,
+      "learning_rate": 5.1179212025727935e-06,
+      "loss": 0.5603,
+      "step": 4456
+    },
+    {
+      "epoch": 1.627715902866533,
+      "grad_norm": 0.7829571962356567,
+      "learning_rate": 5.108341540515522e-06,
+      "loss": 0.5641,
+      "step": 4457
+    },
+    {
+      "epoch": 1.628081066277159,
+      "grad_norm": 0.8676231503486633,
+      "learning_rate": 5.098769539618964e-06,
+      "loss": 0.5289,
+      "step": 4458
+    },
+    {
+      "epoch": 1.6284462296877853,
+      "grad_norm": 1.0025681257247925,
+      "learning_rate": 5.08920520480753e-06,
+      "loss": 0.5165,
+      "step": 4459
+    },
+    {
+      "epoch": 1.6288113930984114,
+      "grad_norm": 0.9066442847251892,
+      "learning_rate": 5.0796485410016825e-06,
+      "loss": 0.543,
+      "step": 4460
+    },
+    {
+      "epoch": 1.6291765565090377,
+      "grad_norm": 1.3929529190063477,
+      "learning_rate": 5.070099553117953e-06,
+      "loss": 0.5398,
+      "step": 4461
+    },
+    {
+      "epoch": 1.629541719919664,
+      "grad_norm": 1.1188859939575195,
+      "learning_rate": 5.060558246068897e-06,
+      "loss": 0.536,
+      "step": 4462
+    },
+    {
+      "epoch": 1.6299068833302903,
+      "grad_norm": 0.7350136637687683,
+      "learning_rate": 5.0510246247631385e-06,
+      "loss": 0.5649,
+      "step": 4463
+    },
+    {
+      "epoch": 1.6302720467409166,
+      "grad_norm": 1.0863380432128906,
+      "learning_rate": 5.041498694105349e-06,
+      "loss": 0.4687,
+      "step": 4464
+    },
+    {
+      "epoch": 1.630637210151543,
+      "grad_norm": 0.8181613683700562,
+      "learning_rate": 5.03198045899624e-06,
+      "loss": 0.5439,
+      "step": 4465
+    },
+    {
+      "epoch": 1.631002373562169,
+      "grad_norm": 0.9987273216247559,
+      "learning_rate": 5.022469924332547e-06,
+      "loss": 0.5333,
+      "step": 4466
+    },
+    {
+      "epoch": 1.6313675369727954,
+      "grad_norm": 0.9395421147346497,
+      "learning_rate": 5.012967095007068e-06,
+      "loss": 0.5405,
+      "step": 4467
+    },
+    {
+      "epoch": 1.6317327003834214,
+      "grad_norm": 0.8800456523895264,
+      "learning_rate": 5.0034719759086335e-06,
+      "loss": 0.5277,
+      "step": 4468
+    },
+    {
+      "epoch": 1.6320978637940478,
+      "grad_norm": 0.8373021483421326,
+      "learning_rate": 4.993984571922086e-06,
+      "loss": 0.5268,
+      "step": 4469
+    },
+    {
+      "epoch": 1.632463027204674,
+      "grad_norm": 1.0338077545166016,
+      "learning_rate": 4.984504887928325e-06,
+      "loss": 0.5418,
+      "step": 4470
+    },
+    {
+      "epoch": 1.6328281906153004,
+      "grad_norm": 0.9462342262268066,
+      "learning_rate": 4.975032928804269e-06,
+      "loss": 0.5069,
+      "step": 4471
+    },
+    {
+      "epoch": 1.6331933540259267,
+      "grad_norm": 1.029945731163025,
+      "learning_rate": 4.965568699422851e-06,
+      "loss": 0.5525,
+      "step": 4472
+    },
+    {
+      "epoch": 1.633558517436553,
+      "grad_norm": 0.8257748484611511,
+      "learning_rate": 4.956112204653043e-06,
+      "loss": 0.5594,
+      "step": 4473
+    },
+    {
+      "epoch": 1.633923680847179,
+      "grad_norm": 0.9749363660812378,
+      "learning_rate": 4.946663449359834e-06,
+      "loss": 0.4737,
+      "step": 4474
+    },
+    {
+      "epoch": 1.6342888442578054,
+      "grad_norm": 0.9025813341140747,
+      "learning_rate": 4.937222438404232e-06,
+      "loss": 0.5493,
+      "step": 4475
+    },
+    {
+      "epoch": 1.6346540076684315,
+      "grad_norm": 1.1877425909042358,
+      "learning_rate": 4.927789176643247e-06,
+      "loss": 0.5672,
+      "step": 4476
+    },
+    {
+      "epoch": 1.6350191710790578,
+      "grad_norm": 0.843192458152771,
+      "learning_rate": 4.918363668929922e-06,
+      "loss": 0.5676,
+      "step": 4477
+    },
+    {
+      "epoch": 1.6353843344896841,
+      "grad_norm": 0.840177595615387,
+      "learning_rate": 4.908945920113299e-06,
+      "loss": 0.5505,
+      "step": 4478
+    },
+    {
+      "epoch": 1.6357494979003104,
+      "grad_norm": 0.7071192860603333,
+      "learning_rate": 4.899535935038436e-06,
+      "loss": 0.5436,
+      "step": 4479
+    },
+    {
+      "epoch": 1.6361146613109367,
+      "grad_norm": 0.8813037872314453,
+      "learning_rate": 4.890133718546395e-06,
+      "loss": 0.5706,
+      "step": 4480
+    },
+    {
+      "epoch": 1.636479824721563,
+      "grad_norm": 0.8635373711585999,
+      "learning_rate": 4.880739275474229e-06,
+      "loss": 0.5175,
+      "step": 4481
+    },
+    {
+      "epoch": 1.6368449881321891,
+      "grad_norm": 0.9497634768486023,
+      "learning_rate": 4.8713526106550134e-06,
+      "loss": 0.5266,
+      "step": 4482
+    },
+    {
+      "epoch": 1.6372101515428155,
+      "grad_norm": 0.8595808148384094,
+      "learning_rate": 4.861973728917799e-06,
+      "loss": 0.5293,
+      "step": 4483
+    },
+    {
+      "epoch": 1.6375753149534416,
+      "grad_norm": 1.2146114110946655,
+      "learning_rate": 4.852602635087651e-06,
+      "loss": 0.4702,
+      "step": 4484
+    },
+    {
+      "epoch": 1.6379404783640679,
+      "grad_norm": 0.818753182888031,
+      "learning_rate": 4.843239333985625e-06,
+      "loss": 0.5299,
+      "step": 4485
+    },
+    {
+      "epoch": 1.6383056417746942,
+      "grad_norm": 1.060773491859436,
+      "learning_rate": 4.833883830428754e-06,
+      "loss": 0.5176,
+      "step": 4486
+    },
+    {
+      "epoch": 1.6386708051853205,
+      "grad_norm": 0.8212208151817322,
+      "learning_rate": 4.8245361292300705e-06,
+      "loss": 0.5612,
+      "step": 4487
+    },
+    {
+      "epoch": 1.6390359685959468,
+      "grad_norm": 0.8592978119850159,
+      "learning_rate": 4.815196235198598e-06,
+      "loss": 0.5427,
+      "step": 4488
+    },
+    {
+      "epoch": 1.639401132006573,
+      "grad_norm": 1.0273016691207886,
+      "learning_rate": 4.805864153139339e-06,
+      "loss": 0.5248,
+      "step": 4489
+    },
+    {
+      "epoch": 1.6397662954171992,
+      "grad_norm": 0.8871325254440308,
+      "learning_rate": 4.796539887853266e-06,
+      "loss": 0.5452,
+      "step": 4490
+    },
+    {
+      "epoch": 1.6401314588278253,
+      "grad_norm": 0.8996334075927734,
+      "learning_rate": 4.7872234441373434e-06,
+      "loss": 0.5277,
+      "step": 4491
+    },
+    {
+      "epoch": 1.6404966222384516,
+      "grad_norm": 0.8074132204055786,
+      "learning_rate": 4.7779148267845065e-06,
+      "loss": 0.5553,
+      "step": 4492
+    },
+    {
+      "epoch": 1.640861785649078,
+      "grad_norm": 1.0954314470291138,
+      "learning_rate": 4.768614040583668e-06,
+      "loss": 0.5074,
+      "step": 4493
+    },
+    {
+      "epoch": 1.6412269490597042,
+      "grad_norm": 0.7810633778572083,
+      "learning_rate": 4.7593210903197155e-06,
+      "loss": 0.5637,
+      "step": 4494
+    },
+    {
+      "epoch": 1.6415921124703305,
+      "grad_norm": 1.101777195930481,
+      "learning_rate": 4.750035980773488e-06,
+      "loss": 0.5187,
+      "step": 4495
+    },
+    {
+      "epoch": 1.6419572758809569,
+      "grad_norm": 0.7545532584190369,
+      "learning_rate": 4.740758716721803e-06,
+      "loss": 0.5423,
+      "step": 4496
+    },
+    {
+      "epoch": 1.642322439291583,
+      "grad_norm": 0.9766955375671387,
+      "learning_rate": 4.731489302937442e-06,
+      "loss": 0.5436,
+      "step": 4497
+    },
+    {
+      "epoch": 1.6426876027022093,
+      "grad_norm": 0.7398218512535095,
+      "learning_rate": 4.722227744189148e-06,
+      "loss": 0.5809,
+      "step": 4498
+    },
+    {
+      "epoch": 1.6430527661128354,
+      "grad_norm": 0.7854627966880798,
+      "learning_rate": 4.712974045241625e-06,
+      "loss": 0.507,
+      "step": 4499
+    },
+    {
+      "epoch": 1.6434179295234617,
+      "grad_norm": 1.1670691967010498,
+      "learning_rate": 4.70372821085552e-06,
+      "loss": 0.5069,
+      "step": 4500
+    },
+    {
+      "epoch": 1.643783092934088,
+      "grad_norm": 0.7382610440254211,
+      "learning_rate": 4.694490245787451e-06,
+      "loss": 0.5783,
+      "step": 4501
+    },
+    {
+      "epoch": 1.6441482563447143,
+      "grad_norm": 0.8627193570137024,
+      "learning_rate": 4.685260154789979e-06,
+      "loss": 0.53,
+      "step": 4502
+    },
+    {
+      "epoch": 1.6445134197553406,
+      "grad_norm": 1.0902482271194458,
+      "learning_rate": 4.676037942611613e-06,
+      "loss": 0.5189,
+      "step": 4503
+    },
+    {
+      "epoch": 1.644878583165967,
+      "grad_norm": 0.8628425598144531,
+      "learning_rate": 4.6668236139968205e-06,
+      "loss": 0.5364,
+      "step": 4504
+    },
+    {
+      "epoch": 1.645243746576593,
+      "grad_norm": 0.9813662171363831,
+      "learning_rate": 4.657617173685989e-06,
+      "loss": 0.493,
+      "step": 4505
+    },
+    {
+      "epoch": 1.6456089099872193,
+      "grad_norm": 0.723275899887085,
+      "learning_rate": 4.648418626415472e-06,
+      "loss": 0.5497,
+      "step": 4506
+    },
+    {
+      "epoch": 1.6459740733978454,
+      "grad_norm": 0.9946385622024536,
+      "learning_rate": 4.63922797691755e-06,
+      "loss": 0.5436,
+      "step": 4507
+    },
+    {
+      "epoch": 1.6463392368084717,
+      "grad_norm": 1.0270251035690308,
+      "learning_rate": 4.63004522992045e-06,
+      "loss": 0.5001,
+      "step": 4508
+    },
+    {
+      "epoch": 1.646704400219098,
+      "grad_norm": 0.974694013595581,
+      "learning_rate": 4.620870390148322e-06,
+      "loss": 0.5034,
+      "step": 4509
+    },
+    {
+      "epoch": 1.6470695636297243,
+      "grad_norm": 0.8969122767448425,
+      "learning_rate": 4.611703462321246e-06,
+      "loss": 0.5258,
+      "step": 4510
+    },
+    {
+      "epoch": 1.6474347270403507,
+      "grad_norm": 0.906024694442749,
+      "learning_rate": 4.602544451155247e-06,
+      "loss": 0.5309,
+      "step": 4511
+    },
+    {
+      "epoch": 1.647799890450977,
+      "grad_norm": 0.7742956876754761,
+      "learning_rate": 4.593393361362264e-06,
+      "loss": 0.5596,
+      "step": 4512
+    },
+    {
+      "epoch": 1.648165053861603,
+      "grad_norm": 0.8088904023170471,
+      "learning_rate": 4.584250197650169e-06,
+      "loss": 0.5823,
+      "step": 4513
+    },
+    {
+      "epoch": 1.6485302172722294,
+      "grad_norm": 0.8699324727058411,
+      "learning_rate": 4.575114964722758e-06,
+      "loss": 0.5165,
+      "step": 4514
+    },
+    {
+      "epoch": 1.6488953806828555,
+      "grad_norm": 0.6912913918495178,
+      "learning_rate": 4.565987667279728e-06,
+      "loss": 0.5538,
+      "step": 4515
+    },
+    {
+      "epoch": 1.6492605440934818,
+      "grad_norm": 1.3685637712478638,
+      "learning_rate": 4.556868310016715e-06,
+      "loss": 0.5148,
+      "step": 4516
+    },
+    {
+      "epoch": 1.649625707504108,
+      "grad_norm": 1.088139533996582,
+      "learning_rate": 4.547756897625264e-06,
+      "loss": 0.5196,
+      "step": 4517
+    },
+    {
+      "epoch": 1.6499908709147344,
+      "grad_norm": 0.9013568758964539,
+      "learning_rate": 4.538653434792833e-06,
+      "loss": 0.5044,
+      "step": 4518
+    },
+    {
+      "epoch": 1.6503560343253607,
+      "grad_norm": 0.8035263419151306,
+      "learning_rate": 4.529557926202781e-06,
+      "loss": 0.5621,
+      "step": 4519
+    },
+    {
+      "epoch": 1.6507211977359868,
+      "grad_norm": 0.770159900188446,
+      "learning_rate": 4.520470376534385e-06,
+      "loss": 0.5474,
+      "step": 4520
+    },
+    {
+      "epoch": 1.6510863611466131,
+      "grad_norm": 0.6742453575134277,
+      "learning_rate": 4.511390790462824e-06,
+      "loss": 0.5527,
+      "step": 4521
+    },
+    {
+      "epoch": 1.6514515245572392,
+      "grad_norm": 0.9856719374656677,
+      "learning_rate": 4.502319172659189e-06,
+      "loss": 0.5764,
+      "step": 4522
+    },
+    {
+      "epoch": 1.6518166879678655,
+      "grad_norm": 1.1472326517105103,
+      "learning_rate": 4.49325552779045e-06,
+      "loss": 0.5225,
+      "step": 4523
+    },
+    {
+      "epoch": 1.6521818513784918,
+      "grad_norm": 1.3834545612335205,
+      "learning_rate": 4.484199860519502e-06,
+      "loss": 0.5,
+      "step": 4524
+    },
+    {
+      "epoch": 1.6525470147891181,
+      "grad_norm": 1.005962610244751,
+      "learning_rate": 4.475152175505108e-06,
+      "loss": 0.5237,
+      "step": 4525
+    },
+    {
+      "epoch": 1.6529121781997445,
+      "grad_norm": 0.7523673176765442,
+      "learning_rate": 4.466112477401949e-06,
+      "loss": 0.5524,
+      "step": 4526
+    },
+    {
+      "epoch": 1.6532773416103708,
+      "grad_norm": 0.7982600331306458,
+      "learning_rate": 4.4570807708605825e-06,
+      "loss": 0.5645,
+      "step": 4527
+    },
+    {
+      "epoch": 1.6536425050209969,
+      "grad_norm": 0.8677096366882324,
+      "learning_rate": 4.448057060527466e-06,
+      "loss": 0.4809,
+      "step": 4528
+    },
+    {
+      "epoch": 1.6540076684316232,
+      "grad_norm": 0.8809179663658142,
+      "learning_rate": 4.439041351044926e-06,
+      "loss": 0.4971,
+      "step": 4529
+    },
+    {
+      "epoch": 1.6543728318422493,
+      "grad_norm": 0.8608189225196838,
+      "learning_rate": 4.430033647051191e-06,
+      "loss": 0.5663,
+      "step": 4530
+    },
+    {
+      "epoch": 1.6547379952528756,
+      "grad_norm": 1.1307836771011353,
+      "learning_rate": 4.421033953180358e-06,
+      "loss": 0.5461,
+      "step": 4531
+    },
+    {
+      "epoch": 1.6551031586635019,
+      "grad_norm": 0.8868815302848816,
+      "learning_rate": 4.412042274062415e-06,
+      "loss": 0.5142,
+      "step": 4532
+    },
+    {
+      "epoch": 1.6554683220741282,
+      "grad_norm": 1.160722017288208,
+      "learning_rate": 4.4030586143232145e-06,
+      "loss": 0.5339,
+      "step": 4533
+    },
+    {
+      "epoch": 1.6558334854847545,
+      "grad_norm": 0.7950741052627563,
+      "learning_rate": 4.394082978584488e-06,
+      "loss": 0.561,
+      "step": 4534
+    },
+    {
+      "epoch": 1.6561986488953808,
+      "grad_norm": 0.7816884517669678,
+      "learning_rate": 4.38511537146385e-06,
+      "loss": 0.5635,
+      "step": 4535
+    },
+    {
+      "epoch": 1.656563812306007,
+      "grad_norm": 1.2029130458831787,
+      "learning_rate": 4.376155797574761e-06,
+      "loss": 0.5617,
+      "step": 4536
+    },
+    {
+      "epoch": 1.6569289757166332,
+      "grad_norm": 0.8650860786437988,
+      "learning_rate": 4.367204261526568e-06,
+      "loss": 0.4732,
+      "step": 4537
+    },
+    {
+      "epoch": 1.6572941391272593,
+      "grad_norm": 0.8647130131721497,
+      "learning_rate": 4.358260767924482e-06,
+      "loss": 0.4763,
+      "step": 4538
+    },
+    {
+      "epoch": 1.6576593025378856,
+      "grad_norm": 1.0285489559173584,
+      "learning_rate": 4.349325321369564e-06,
+      "loss": 0.4862,
+      "step": 4539
+    },
+    {
+      "epoch": 1.658024465948512,
+      "grad_norm": 1.072596788406372,
+      "learning_rate": 4.340397926458744e-06,
+      "loss": 0.5381,
+      "step": 4540
+    },
+    {
+      "epoch": 1.6583896293591383,
+      "grad_norm": 0.9929602146148682,
+      "learning_rate": 4.331478587784809e-06,
+      "loss": 0.5239,
+      "step": 4541
+    },
+    {
+      "epoch": 1.6587547927697646,
+      "grad_norm": 0.8808034658432007,
+      "learning_rate": 4.32256730993641e-06,
+      "loss": 0.4933,
+      "step": 4542
+    },
+    {
+      "epoch": 1.6591199561803909,
+      "grad_norm": 0.9162267446517944,
+      "learning_rate": 4.313664097498027e-06,
+      "loss": 0.5379,
+      "step": 4543
+    },
+    {
+      "epoch": 1.659485119591017,
+      "grad_norm": 0.8565149903297424,
+      "learning_rate": 4.304768955050014e-06,
+      "loss": 0.5077,
+      "step": 4544
+    },
+    {
+      "epoch": 1.6598502830016433,
+      "grad_norm": 1.352622389793396,
+      "learning_rate": 4.29588188716856e-06,
+      "loss": 0.5193,
+      "step": 4545
+    },
+    {
+      "epoch": 1.6602154464122694,
+      "grad_norm": 1.0091907978057861,
+      "learning_rate": 4.287002898425709e-06,
+      "loss": 0.5691,
+      "step": 4546
+    },
+    {
+      "epoch": 1.6605806098228957,
+      "grad_norm": 0.9868928790092468,
+      "learning_rate": 4.2781319933893515e-06,
+      "loss": 0.5322,
+      "step": 4547
+    },
+    {
+      "epoch": 1.660945773233522,
+      "grad_norm": 0.8290953636169434,
+      "learning_rate": 4.269269176623203e-06,
+      "loss": 0.5705,
+      "step": 4548
+    },
+    {
+      "epoch": 1.6613109366441483,
+      "grad_norm": 0.9042468667030334,
+      "learning_rate": 4.260414452686821e-06,
+      "loss": 0.552,
+      "step": 4549
+    },
+    {
+      "epoch": 1.6616761000547746,
+      "grad_norm": 0.8311589360237122,
+      "learning_rate": 4.251567826135614e-06,
+      "loss": 0.5229,
+      "step": 4550
+    },
+    {
+      "epoch": 1.662041263465401,
+      "grad_norm": 0.7332230806350708,
+      "learning_rate": 4.242729301520816e-06,
+      "loss": 0.539,
+      "step": 4551
+    },
+    {
+      "epoch": 1.662406426876027,
+      "grad_norm": 0.7751251459121704,
+      "learning_rate": 4.233898883389496e-06,
+      "loss": 0.5619,
+      "step": 4552
+    },
+    {
+      "epoch": 1.6627715902866531,
+      "grad_norm": 0.696730375289917,
+      "learning_rate": 4.225076576284541e-06,
+      "loss": 0.5615,
+      "step": 4553
+    },
+    {
+      "epoch": 1.6631367536972794,
+      "grad_norm": 1.0118292570114136,
+      "learning_rate": 4.2162623847446806e-06,
+      "loss": 0.5767,
+      "step": 4554
+    },
+    {
+      "epoch": 1.6635019171079057,
+      "grad_norm": 0.9420759081840515,
+      "learning_rate": 4.207456313304461e-06,
+      "loss": 0.5393,
+      "step": 4555
+    },
+    {
+      "epoch": 1.663867080518532,
+      "grad_norm": 0.7974475622177124,
+      "learning_rate": 4.198658366494252e-06,
+      "loss": 0.5333,
+      "step": 4556
+    },
+    {
+      "epoch": 1.6642322439291584,
+      "grad_norm": 0.912028968334198,
+      "learning_rate": 4.189868548840253e-06,
+      "loss": 0.5154,
+      "step": 4557
+    },
+    {
+      "epoch": 1.6645974073397847,
+      "grad_norm": 1.3836195468902588,
+      "learning_rate": 4.181086864864457e-06,
+      "loss": 0.4968,
+      "step": 4558
+    },
+    {
+      "epoch": 1.6649625707504108,
+      "grad_norm": 1.2628008127212524,
+      "learning_rate": 4.172313319084695e-06,
+      "loss": 0.5497,
+      "step": 4559
+    },
+    {
+      "epoch": 1.665327734161037,
+      "grad_norm": 0.8643813729286194,
+      "learning_rate": 4.163547916014605e-06,
+      "loss": 0.5334,
+      "step": 4560
+    },
+    {
+      "epoch": 1.6656928975716632,
+      "grad_norm": 0.8692271113395691,
+      "learning_rate": 4.154790660163641e-06,
+      "loss": 0.5178,
+      "step": 4561
+    },
+    {
+      "epoch": 1.6660580609822895,
+      "grad_norm": 0.7350249290466309,
+      "learning_rate": 4.1460415560370545e-06,
+      "loss": 0.5498,
+      "step": 4562
+    },
+    {
+      "epoch": 1.6664232243929158,
+      "grad_norm": 0.8751364350318909,
+      "learning_rate": 4.137300608135901e-06,
+      "loss": 0.4749,
+      "step": 4563
+    },
+    {
+      "epoch": 1.666788387803542,
+      "grad_norm": 0.7478228211402893,
+      "learning_rate": 4.128567820957054e-06,
+      "loss": 0.5363,
+      "step": 4564
+    },
+    {
+      "epoch": 1.6671535512141684,
+      "grad_norm": 1.0465996265411377,
+      "learning_rate": 4.119843198993185e-06,
+      "loss": 0.4948,
+      "step": 4565
+    },
+    {
+      "epoch": 1.6675187146247947,
+      "grad_norm": 1.0394774675369263,
+      "learning_rate": 4.111126746732756e-06,
+      "loss": 0.5178,
+      "step": 4566
+    },
+    {
+      "epoch": 1.6678838780354208,
+      "grad_norm": 0.839904248714447,
+      "learning_rate": 4.102418468660041e-06,
+      "loss": 0.5305,
+      "step": 4567
+    },
+    {
+      "epoch": 1.6682490414460471,
+      "grad_norm": 0.7755999565124512,
+      "learning_rate": 4.0937183692550885e-06,
+      "loss": 0.5279,
+      "step": 4568
+    },
+    {
+      "epoch": 1.6686142048566732,
+      "grad_norm": 1.0791418552398682,
+      "learning_rate": 4.0850264529937565e-06,
+      "loss": 0.5492,
+      "step": 4569
+    },
+    {
+      "epoch": 1.6689793682672995,
+      "grad_norm": 0.7471647262573242,
+      "learning_rate": 4.076342724347686e-06,
+      "loss": 0.5443,
+      "step": 4570
+    },
+    {
+      "epoch": 1.6693445316779258,
+      "grad_norm": 1.0776115655899048,
+      "learning_rate": 4.067667187784312e-06,
+      "loss": 0.492,
+      "step": 4571
+    },
+    {
+      "epoch": 1.6697096950885522,
+      "grad_norm": 0.9209566712379456,
+      "learning_rate": 4.0589998477668405e-06,
+      "loss": 0.5322,
+      "step": 4572
+    },
+    {
+      "epoch": 1.6700748584991785,
+      "grad_norm": 0.9458308219909668,
+      "learning_rate": 4.050340708754274e-06,
+      "loss": 0.5481,
+      "step": 4573
+    },
+    {
+      "epoch": 1.6704400219098048,
+      "grad_norm": 0.8129193782806396,
+      "learning_rate": 4.041689775201394e-06,
+      "loss": 0.5698,
+      "step": 4574
+    },
+    {
+      "epoch": 1.6708051853204309,
+      "grad_norm": 0.7496057748794556,
+      "learning_rate": 4.03304705155876e-06,
+      "loss": 0.5632,
+      "step": 4575
+    },
+    {
+      "epoch": 1.6711703487310572,
+      "grad_norm": 0.9546471238136292,
+      "learning_rate": 4.024412542272706e-06,
+      "loss": 0.5239,
+      "step": 4576
+    },
+    {
+      "epoch": 1.6715355121416833,
+      "grad_norm": 0.9323283433914185,
+      "learning_rate": 4.015786251785334e-06,
+      "loss": 0.4836,
+      "step": 4577
+    },
+    {
+      "epoch": 1.6719006755523096,
+      "grad_norm": 0.8260817527770996,
+      "learning_rate": 4.007168184534529e-06,
+      "loss": 0.5256,
+      "step": 4578
+    },
+    {
+      "epoch": 1.672265838962936,
+      "grad_norm": 0.7987567782402039,
+      "learning_rate": 3.99855834495394e-06,
+      "loss": 0.5239,
+      "step": 4579
+    },
+    {
+      "epoch": 1.6726310023735622,
+      "grad_norm": 0.9163694977760315,
+      "learning_rate": 3.989956737472984e-06,
+      "loss": 0.5173,
+      "step": 4580
+    },
+    {
+      "epoch": 1.6729961657841885,
+      "grad_norm": 0.5946126580238342,
+      "learning_rate": 3.98136336651685e-06,
+      "loss": 0.5784,
+      "step": 4581
+    },
+    {
+      "epoch": 1.6733613291948148,
+      "grad_norm": 0.780558168888092,
+      "learning_rate": 3.9727782365064695e-06,
+      "loss": 0.5597,
+      "step": 4582
+    },
+    {
+      "epoch": 1.673726492605441,
+      "grad_norm": 0.9302628040313721,
+      "learning_rate": 3.9642013518585564e-06,
+      "loss": 0.5172,
+      "step": 4583
+    },
+    {
+      "epoch": 1.6740916560160672,
+      "grad_norm": 0.8205109238624573,
+      "learning_rate": 3.955632716985569e-06,
+      "loss": 0.535,
+      "step": 4584
+    },
+    {
+      "epoch": 1.6744568194266933,
+      "grad_norm": 1.1331835985183716,
+      "learning_rate": 3.947072336295734e-06,
+      "loss": 0.5323,
+      "step": 4585
+    },
+    {
+      "epoch": 1.6748219828373196,
+      "grad_norm": 0.8803338408470154,
+      "learning_rate": 3.938520214193014e-06,
+      "loss": 0.5562,
+      "step": 4586
+    },
+    {
+      "epoch": 1.675187146247946,
+      "grad_norm": 0.8349100351333618,
+      "learning_rate": 3.929976355077134e-06,
+      "loss": 0.5216,
+      "step": 4587
+    },
+    {
+      "epoch": 1.6755523096585723,
+      "grad_norm": 1.0220751762390137,
+      "learning_rate": 3.921440763343578e-06,
+      "loss": 0.5576,
+      "step": 4588
+    },
+    {
+      "epoch": 1.6759174730691986,
+      "grad_norm": 0.9241355657577515,
+      "learning_rate": 3.91291344338355e-06,
+      "loss": 0.5306,
+      "step": 4589
+    },
+    {
+      "epoch": 1.6762826364798247,
+      "grad_norm": 0.9509197473526001,
+      "learning_rate": 3.90439439958402e-06,
+      "loss": 0.5306,
+      "step": 4590
+    },
+    {
+      "epoch": 1.676647799890451,
+      "grad_norm": 0.9647827744483948,
+      "learning_rate": 3.8958836363277016e-06,
+      "loss": 0.4928,
+      "step": 4591
+    },
+    {
+      "epoch": 1.677012963301077,
+      "grad_norm": 0.9714269042015076,
+      "learning_rate": 3.887381157993029e-06,
+      "loss": 0.4742,
+      "step": 4592
+    },
+    {
+      "epoch": 1.6773781267117034,
+      "grad_norm": 0.8928794860839844,
+      "learning_rate": 3.87888696895419e-06,
+      "loss": 0.5338,
+      "step": 4593
+    },
+    {
+      "epoch": 1.6777432901223297,
+      "grad_norm": 0.8356234431266785,
+      "learning_rate": 3.870401073581107e-06,
+      "loss": 0.5305,
+      "step": 4594
+    },
+    {
+      "epoch": 1.678108453532956,
+      "grad_norm": 0.7654070854187012,
+      "learning_rate": 3.861923476239435e-06,
+      "loss": 0.5637,
+      "step": 4595
+    },
+    {
+      "epoch": 1.6784736169435823,
+      "grad_norm": 0.6427393555641174,
+      "learning_rate": 3.85345418129055e-06,
+      "loss": 0.5503,
+      "step": 4596
+    },
+    {
+      "epoch": 1.6788387803542086,
+      "grad_norm": 0.7948646545410156,
+      "learning_rate": 3.844993193091568e-06,
+      "loss": 0.525,
+      "step": 4597
+    },
+    {
+      "epoch": 1.6792039437648347,
+      "grad_norm": 0.9657595753669739,
+      "learning_rate": 3.8365405159953265e-06,
+      "loss": 0.5482,
+      "step": 4598
+    },
+    {
+      "epoch": 1.679569107175461,
+      "grad_norm": 0.7931142449378967,
+      "learning_rate": 3.828096154350391e-06,
+      "loss": 0.5335,
+      "step": 4599
+    },
+    {
+      "epoch": 1.6799342705860871,
+      "grad_norm": 0.8278718590736389,
+      "learning_rate": 3.819660112501053e-06,
+      "loss": 0.5414,
+      "step": 4600
+    },
+    {
+      "epoch": 1.6802994339967134,
+      "grad_norm": 0.8467036485671997,
+      "learning_rate": 3.811232394787303e-06,
+      "loss": 0.5532,
+      "step": 4601
+    },
+    {
+      "epoch": 1.6806645974073398,
+      "grad_norm": 0.992175817489624,
+      "learning_rate": 3.802813005544879e-06,
+      "loss": 0.5353,
+      "step": 4602
+    },
+    {
+      "epoch": 1.681029760817966,
+      "grad_norm": 1.0147136449813843,
+      "learning_rate": 3.7944019491052043e-06,
+      "loss": 0.5217,
+      "step": 4603
+    },
+    {
+      "epoch": 1.6813949242285924,
+      "grad_norm": 0.8257679343223572,
+      "learning_rate": 3.7859992297954363e-06,
+      "loss": 0.5546,
+      "step": 4604
+    },
+    {
+      "epoch": 1.6817600876392187,
+      "grad_norm": 0.876408576965332,
+      "learning_rate": 3.7776048519384413e-06,
+      "loss": 0.5532,
+      "step": 4605
+    },
+    {
+      "epoch": 1.6821252510498448,
+      "grad_norm": 0.9336234331130981,
+      "learning_rate": 3.7692188198527822e-06,
+      "loss": 0.5084,
+      "step": 4606
+    },
+    {
+      "epoch": 1.682490414460471,
+      "grad_norm": 0.8870642185211182,
+      "learning_rate": 3.76084113785274e-06,
+      "loss": 0.4638,
+      "step": 4607
+    },
+    {
+      "epoch": 1.6828555778710972,
+      "grad_norm": 0.7216123938560486,
+      "learning_rate": 3.7524718102482947e-06,
+      "loss": 0.5434,
+      "step": 4608
+    },
+    {
+      "epoch": 1.6832207412817235,
+      "grad_norm": 0.9200725555419922,
+      "learning_rate": 3.74411084134513e-06,
+      "loss": 0.5232,
+      "step": 4609
+    },
+    {
+      "epoch": 1.6835859046923498,
+      "grad_norm": 0.7329199314117432,
+      "learning_rate": 3.7357582354446352e-06,
+      "loss": 0.5414,
+      "step": 4610
+    },
+    {
+      "epoch": 1.6839510681029761,
+      "grad_norm": 1.110066294670105,
+      "learning_rate": 3.7274139968438782e-06,
+      "loss": 0.5015,
+      "step": 4611
+    },
+    {
+      "epoch": 1.6843162315136024,
+      "grad_norm": 0.9416675567626953,
+      "learning_rate": 3.7190781298356428e-06,
+      "loss": 0.5309,
+      "step": 4612
+    },
+    {
+      "epoch": 1.6846813949242287,
+      "grad_norm": 1.218187928199768,
+      "learning_rate": 3.710750638708398e-06,
+      "loss": 0.5042,
+      "step": 4613
+    },
+    {
+      "epoch": 1.6850465583348548,
+      "grad_norm": 1.1069811582565308,
+      "learning_rate": 3.7024315277463064e-06,
+      "loss": 0.5276,
+      "step": 4614
+    },
+    {
+      "epoch": 1.6854117217454812,
+      "grad_norm": 0.7823757529258728,
+      "learning_rate": 3.694120801229213e-06,
+      "loss": 0.5285,
+      "step": 4615
+    },
+    {
+      "epoch": 1.6857768851561072,
+      "grad_norm": 0.7776532173156738,
+      "learning_rate": 3.685818463432649e-06,
+      "loss": 0.5238,
+      "step": 4616
+    },
+    {
+      "epoch": 1.6861420485667336,
+      "grad_norm": 0.7341038584709167,
+      "learning_rate": 3.6775245186278375e-06,
+      "loss": 0.5586,
+      "step": 4617
+    },
+    {
+      "epoch": 1.6865072119773599,
+      "grad_norm": 0.9906373620033264,
+      "learning_rate": 3.669238971081681e-06,
+      "loss": 0.568,
+      "step": 4618
+    },
+    {
+      "epoch": 1.6868723753879862,
+      "grad_norm": 1.0237858295440674,
+      "learning_rate": 3.6609618250567657e-06,
+      "loss": 0.5143,
+      "step": 4619
+    },
+    {
+      "epoch": 1.6872375387986125,
+      "grad_norm": 0.6446452140808105,
+      "learning_rate": 3.652693084811343e-06,
+      "loss": 0.5756,
+      "step": 4620
+    },
+    {
+      "epoch": 1.6876027022092386,
+      "grad_norm": 0.7709906697273254,
+      "learning_rate": 3.6444327545993497e-06,
+      "loss": 0.5422,
+      "step": 4621
+    },
+    {
+      "epoch": 1.687967865619865,
+      "grad_norm": 0.8992085456848145,
+      "learning_rate": 3.636180838670398e-06,
+      "loss": 0.5671,
+      "step": 4622
+    },
+    {
+      "epoch": 1.688333029030491,
+      "grad_norm": 0.9192847013473511,
+      "learning_rate": 3.627937341269765e-06,
+      "loss": 0.523,
+      "step": 4623
+    },
+    {
+      "epoch": 1.6886981924411173,
+      "grad_norm": 0.9154180884361267,
+      "learning_rate": 3.619702266638405e-06,
+      "loss": 0.5198,
+      "step": 4624
+    },
+    {
+      "epoch": 1.6890633558517436,
+      "grad_norm": 0.960006594657898,
+      "learning_rate": 3.611475619012923e-06,
+      "loss": 0.5456,
+      "step": 4625
+    },
+    {
+      "epoch": 1.68942851926237,
+      "grad_norm": 1.1184638738632202,
+      "learning_rate": 3.603257402625604e-06,
+      "loss": 0.4909,
+      "step": 4626
+    },
+    {
+      "epoch": 1.6897936826729962,
+      "grad_norm": 1.1449368000030518,
+      "learning_rate": 3.5950476217043928e-06,
+      "loss": 0.5238,
+      "step": 4627
+    },
+    {
+      "epoch": 1.6901588460836225,
+      "grad_norm": 0.9218161702156067,
+      "learning_rate": 3.5868462804728933e-06,
+      "loss": 0.5313,
+      "step": 4628
+    },
+    {
+      "epoch": 1.6905240094942486,
+      "grad_norm": 0.8735278844833374,
+      "learning_rate": 3.5786533831503654e-06,
+      "loss": 0.5478,
+      "step": 4629
+    },
+    {
+      "epoch": 1.690889172904875,
+      "grad_norm": 0.8266289830207825,
+      "learning_rate": 3.5704689339517184e-06,
+      "loss": 0.529,
+      "step": 4630
+    },
+    {
+      "epoch": 1.691254336315501,
+      "grad_norm": 0.8460071682929993,
+      "learning_rate": 3.56229293708753e-06,
+      "loss": 0.5178,
+      "step": 4631
+    },
+    {
+      "epoch": 1.6916194997261274,
+      "grad_norm": 0.9639946818351746,
+      "learning_rate": 3.554125396764021e-06,
+      "loss": 0.508,
+      "step": 4632
+    },
+    {
+      "epoch": 1.6919846631367537,
+      "grad_norm": 0.9099420309066772,
+      "learning_rate": 3.5459663171830626e-06,
+      "loss": 0.5219,
+      "step": 4633
+    },
+    {
+      "epoch": 1.69234982654738,
+      "grad_norm": 0.6585804224014282,
+      "learning_rate": 3.53781570254218e-06,
+      "loss": 0.5629,
+      "step": 4634
+    },
+    {
+      "epoch": 1.6927149899580063,
+      "grad_norm": 0.8587778210639954,
+      "learning_rate": 3.5296735570345276e-06,
+      "loss": 0.5479,
+      "step": 4635
+    },
+    {
+      "epoch": 1.6930801533686326,
+      "grad_norm": 0.7885758280754089,
+      "learning_rate": 3.5215398848489167e-06,
+      "loss": 0.5573,
+      "step": 4636
+    },
+    {
+      "epoch": 1.6934453167792587,
+      "grad_norm": 0.6936434507369995,
+      "learning_rate": 3.513414690169794e-06,
+      "loss": 0.5464,
+      "step": 4637
+    },
+    {
+      "epoch": 1.693810480189885,
+      "grad_norm": 0.9549039006233215,
+      "learning_rate": 3.5052979771772555e-06,
+      "loss": 0.5423,
+      "step": 4638
+    },
+    {
+      "epoch": 1.694175643600511,
+      "grad_norm": 1.0629690885543823,
+      "learning_rate": 3.49718975004701e-06,
+      "loss": 0.4832,
+      "step": 4639
+    },
+    {
+      "epoch": 1.6945408070111374,
+      "grad_norm": 0.7019412517547607,
+      "learning_rate": 3.489090012950422e-06,
+      "loss": 0.5543,
+      "step": 4640
+    },
+    {
+      "epoch": 1.6949059704217637,
+      "grad_norm": 0.8659648299217224,
+      "learning_rate": 3.48099877005448e-06,
+      "loss": 0.5479,
+      "step": 4641
+    },
+    {
+      "epoch": 1.69527113383239,
+      "grad_norm": 0.7017379403114319,
+      "learning_rate": 3.4729160255218107e-06,
+      "loss": 0.5425,
+      "step": 4642
+    },
+    {
+      "epoch": 1.6956362972430163,
+      "grad_norm": 0.6785612106323242,
+      "learning_rate": 3.4648417835106507e-06,
+      "loss": 0.5347,
+      "step": 4643
+    },
+    {
+      "epoch": 1.6960014606536427,
+      "grad_norm": 0.7757022976875305,
+      "learning_rate": 3.4567760481748835e-06,
+      "loss": 0.5406,
+      "step": 4644
+    },
+    {
+      "epoch": 1.6963666240642687,
+      "grad_norm": 0.9227039217948914,
+      "learning_rate": 3.4487188236639966e-06,
+      "loss": 0.5199,
+      "step": 4645
+    },
+    {
+      "epoch": 1.696731787474895,
+      "grad_norm": 0.693849503993988,
+      "learning_rate": 3.4406701141231126e-06,
+      "loss": 0.5845,
+      "step": 4646
+    },
+    {
+      "epoch": 1.6970969508855211,
+      "grad_norm": 0.734327495098114,
+      "learning_rate": 3.4326299236929693e-06,
+      "loss": 0.5612,
+      "step": 4647
+    },
+    {
+      "epoch": 1.6974621142961475,
+      "grad_norm": 0.9029885530471802,
+      "learning_rate": 3.42459825650993e-06,
+      "loss": 0.514,
+      "step": 4648
+    },
+    {
+      "epoch": 1.6978272777067738,
+      "grad_norm": 0.7814059853553772,
+      "learning_rate": 3.416575116705951e-06,
+      "loss": 0.468,
+      "step": 4649
+    },
+    {
+      "epoch": 1.6981924411174,
+      "grad_norm": 0.7431153655052185,
+      "learning_rate": 3.408560508408625e-06,
+      "loss": 0.5677,
+      "step": 4650
+    },
+    {
+      "epoch": 1.6985576045280264,
+      "grad_norm": 1.0016436576843262,
+      "learning_rate": 3.4005544357411433e-06,
+      "loss": 0.5071,
+      "step": 4651
+    },
+    {
+      "epoch": 1.6989227679386527,
+      "grad_norm": 0.9274985194206238,
+      "learning_rate": 3.392556902822313e-06,
+      "loss": 0.4889,
+      "step": 4652
+    },
+    {
+      "epoch": 1.6992879313492788,
+      "grad_norm": 0.9575678706169128,
+      "learning_rate": 3.3845679137665434e-06,
+      "loss": 0.4849,
+      "step": 4653
+    },
+    {
+      "epoch": 1.699653094759905,
+      "grad_norm": 0.766573429107666,
+      "learning_rate": 3.376587472683841e-06,
+      "loss": 0.4888,
+      "step": 4654
+    },
+    {
+      "epoch": 1.7000182581705312,
+      "grad_norm": 0.9481750726699829,
+      "learning_rate": 3.368615583679833e-06,
+      "loss": 0.4871,
+      "step": 4655
+    },
+    {
+      "epoch": 1.7003834215811575,
+      "grad_norm": 0.9773246049880981,
+      "learning_rate": 3.360652250855727e-06,
+      "loss": 0.5846,
+      "step": 4656
+    },
+    {
+      "epoch": 1.7007485849917838,
+      "grad_norm": 0.8178773522377014,
+      "learning_rate": 3.352697478308342e-06,
+      "loss": 0.5229,
+      "step": 4657
+    },
+    {
+      "epoch": 1.7011137484024101,
+      "grad_norm": 0.8280562162399292,
+      "learning_rate": 3.3447512701300956e-06,
+      "loss": 0.5539,
+      "step": 4658
+    },
+    {
+      "epoch": 1.7014789118130365,
+      "grad_norm": 0.7413144111633301,
+      "learning_rate": 3.3368136304089815e-06,
+      "loss": 0.5374,
+      "step": 4659
+    },
+    {
+      "epoch": 1.7018440752236625,
+      "grad_norm": 0.9765322208404541,
+      "learning_rate": 3.328884563228605e-06,
+      "loss": 0.5624,
+      "step": 4660
+    },
+    {
+      "epoch": 1.7022092386342889,
+      "grad_norm": 0.8073693513870239,
+      "learning_rate": 3.320964072668147e-06,
+      "loss": 0.5722,
+      "step": 4661
+    },
+    {
+      "epoch": 1.702574402044915,
+      "grad_norm": 1.0331017971038818,
+      "learning_rate": 3.3130521628023926e-06,
+      "loss": 0.5284,
+      "step": 4662
+    },
+    {
+      "epoch": 1.7029395654555413,
+      "grad_norm": 0.8730890154838562,
+      "learning_rate": 3.3051488377016884e-06,
+      "loss": 0.5876,
+      "step": 4663
+    },
+    {
+      "epoch": 1.7033047288661676,
+      "grad_norm": 1.1102790832519531,
+      "learning_rate": 3.297254101431986e-06,
+      "loss": 0.5334,
+      "step": 4664
+    },
+    {
+      "epoch": 1.7036698922767939,
+      "grad_norm": 0.9046598672866821,
+      "learning_rate": 3.2893679580548075e-06,
+      "loss": 0.5447,
+      "step": 4665
+    },
+    {
+      "epoch": 1.7040350556874202,
+      "grad_norm": 0.6133385300636292,
+      "learning_rate": 3.2814904116272595e-06,
+      "loss": 0.5544,
+      "step": 4666
+    },
+    {
+      "epoch": 1.7044002190980465,
+      "grad_norm": 0.7287967801094055,
+      "learning_rate": 3.2736214662020284e-06,
+      "loss": 0.531,
+      "step": 4667
+    },
+    {
+      "epoch": 1.7047653825086726,
+      "grad_norm": 1.0205227136611938,
+      "learning_rate": 3.2657611258273602e-06,
+      "loss": 0.4849,
+      "step": 4668
+    },
+    {
+      "epoch": 1.705130545919299,
+      "grad_norm": 0.8684663772583008,
+      "learning_rate": 3.257909394547092e-06,
+      "loss": 0.5234,
+      "step": 4669
+    },
+    {
+      "epoch": 1.705495709329925,
+      "grad_norm": 4.236178398132324,
+      "learning_rate": 3.250066276400621e-06,
+      "loss": 0.5249,
+      "step": 4670
+    },
+    {
+      "epoch": 1.7058608727405513,
+      "grad_norm": 1.5054926872253418,
+      "learning_rate": 3.242231775422915e-06,
+      "loss": 0.5444,
+      "step": 4671
+    },
+    {
+      "epoch": 1.7062260361511776,
+      "grad_norm": 0.7978805303573608,
+      "learning_rate": 3.234405895644519e-06,
+      "loss": 0.4977,
+      "step": 4672
+    },
+    {
+      "epoch": 1.706591199561804,
+      "grad_norm": 0.7267588973045349,
+      "learning_rate": 3.2265886410915214e-06,
+      "loss": 0.546,
+      "step": 4673
+    },
+    {
+      "epoch": 1.7069563629724303,
+      "grad_norm": 1.0223290920257568,
+      "learning_rate": 3.2187800157855964e-06,
+      "loss": 0.5465,
+      "step": 4674
+    },
+    {
+      "epoch": 1.7073215263830566,
+      "grad_norm": 0.8201450109481812,
+      "learning_rate": 3.2109800237439616e-06,
+      "loss": 0.5579,
+      "step": 4675
+    },
+    {
+      "epoch": 1.7076866897936827,
+      "grad_norm": 0.7174233794212341,
+      "learning_rate": 3.2031886689794044e-06,
+      "loss": 0.5502,
+      "step": 4676
+    },
+    {
+      "epoch": 1.708051853204309,
+      "grad_norm": 1.003202199935913,
+      "learning_rate": 3.1954059555002683e-06,
+      "loss": 0.5174,
+      "step": 4677
+    },
+    {
+      "epoch": 1.708417016614935,
+      "grad_norm": 0.89548259973526,
+      "learning_rate": 3.1876318873104383e-06,
+      "loss": 0.5293,
+      "step": 4678
+    },
+    {
+      "epoch": 1.7087821800255614,
+      "grad_norm": 0.8657696843147278,
+      "learning_rate": 3.1798664684093606e-06,
+      "loss": 0.5137,
+      "step": 4679
+    },
+    {
+      "epoch": 1.7091473434361877,
+      "grad_norm": 0.7795391082763672,
+      "learning_rate": 3.1721097027920367e-06,
+      "loss": 0.5038,
+      "step": 4680
+    },
+    {
+      "epoch": 1.709512506846814,
+      "grad_norm": 0.8812219500541687,
+      "learning_rate": 3.1643615944490147e-06,
+      "loss": 0.5207,
+      "step": 4681
+    },
+    {
+      "epoch": 1.7098776702574403,
+      "grad_norm": 0.8421750664710999,
+      "learning_rate": 3.1566221473663794e-06,
+      "loss": 0.5429,
+      "step": 4682
+    },
+    {
+      "epoch": 1.7102428336680666,
+      "grad_norm": 0.750450074672699,
+      "learning_rate": 3.1488913655257635e-06,
+      "loss": 0.5621,
+      "step": 4683
+    },
+    {
+      "epoch": 1.7106079970786927,
+      "grad_norm": 0.8702534437179565,
+      "learning_rate": 3.1411692529043457e-06,
+      "loss": 0.5168,
+      "step": 4684
+    },
+    {
+      "epoch": 1.710973160489319,
+      "grad_norm": 1.2278602123260498,
+      "learning_rate": 3.133455813474844e-06,
+      "loss": 0.5337,
+      "step": 4685
+    },
+    {
+      "epoch": 1.7113383238999451,
+      "grad_norm": 0.8343172073364258,
+      "learning_rate": 3.1257510512055145e-06,
+      "loss": 0.5358,
+      "step": 4686
+    },
+    {
+      "epoch": 1.7117034873105714,
+      "grad_norm": 0.66392982006073,
+      "learning_rate": 3.1180549700601535e-06,
+      "loss": 0.5743,
+      "step": 4687
+    },
+    {
+      "epoch": 1.7120686507211977,
+      "grad_norm": 0.9112119078636169,
+      "learning_rate": 3.1103675739980745e-06,
+      "loss": 0.5335,
+      "step": 4688
+    },
+    {
+      "epoch": 1.712433814131824,
+      "grad_norm": 1.051275610923767,
+      "learning_rate": 3.1026888669741396e-06,
+      "loss": 0.5261,
+      "step": 4689
+    },
+    {
+      "epoch": 1.7127989775424504,
+      "grad_norm": 0.9014411568641663,
+      "learning_rate": 3.095018852938736e-06,
+      "loss": 0.5255,
+      "step": 4690
+    },
+    {
+      "epoch": 1.7131641409530765,
+      "grad_norm": 1.0786256790161133,
+      "learning_rate": 3.0873575358377826e-06,
+      "loss": 0.5172,
+      "step": 4691
+    },
+    {
+      "epoch": 1.7135293043637028,
+      "grad_norm": 0.8149101734161377,
+      "learning_rate": 3.0797049196127115e-06,
+      "loss": 0.5845,
+      "step": 4692
+    },
+    {
+      "epoch": 1.7138944677743289,
+      "grad_norm": 0.71403568983078,
+      "learning_rate": 3.0720610082004887e-06,
+      "loss": 0.5392,
+      "step": 4693
+    },
+    {
+      "epoch": 1.7142596311849552,
+      "grad_norm": 0.6756953001022339,
+      "learning_rate": 3.0644258055336017e-06,
+      "loss": 0.5479,
+      "step": 4694
+    },
+    {
+      "epoch": 1.7146247945955815,
+      "grad_norm": 0.9829939007759094,
+      "learning_rate": 3.05679931554006e-06,
+      "loss": 0.5401,
+      "step": 4695
+    },
+    {
+      "epoch": 1.7149899580062078,
+      "grad_norm": 0.918777346611023,
+      "learning_rate": 3.0491815421433825e-06,
+      "loss": 0.5274,
+      "step": 4696
+    },
+    {
+      "epoch": 1.715355121416834,
+      "grad_norm": 0.759799063205719,
+      "learning_rate": 3.041572489262603e-06,
+      "loss": 0.5356,
+      "step": 4697
+    },
+    {
+      "epoch": 1.7157202848274604,
+      "grad_norm": 0.7517125010490417,
+      "learning_rate": 3.0339721608122774e-06,
+      "loss": 0.5537,
+      "step": 4698
+    },
+    {
+      "epoch": 1.7160854482380865,
+      "grad_norm": 1.0564138889312744,
+      "learning_rate": 3.026380560702471e-06,
+      "loss": 0.4696,
+      "step": 4699
+    },
+    {
+      "epoch": 1.7164506116487128,
+      "grad_norm": 1.2564903497695923,
+      "learning_rate": 3.0187976928387573e-06,
+      "loss": 0.5359,
+      "step": 4700
+    },
+    {
+      "epoch": 1.716815775059339,
+      "grad_norm": 0.8190529346466064,
+      "learning_rate": 3.011223561122223e-06,
+      "loss": 0.5407,
+      "step": 4701
+    },
+    {
+      "epoch": 1.7171809384699652,
+      "grad_norm": 1.0427823066711426,
+      "learning_rate": 3.0036581694494436e-06,
+      "loss": 0.5139,
+      "step": 4702
+    },
+    {
+      "epoch": 1.7175461018805915,
+      "grad_norm": 0.9642361998558044,
+      "learning_rate": 2.9961015217125155e-06,
+      "loss": 0.5073,
+      "step": 4703
+    },
+    {
+      "epoch": 1.7179112652912178,
+      "grad_norm": 0.7839714884757996,
+      "learning_rate": 2.988553621799033e-06,
+      "loss": 0.5373,
+      "step": 4704
+    },
+    {
+      "epoch": 1.7182764287018442,
+      "grad_norm": 0.8481120467185974,
+      "learning_rate": 2.9810144735920877e-06,
+      "loss": 0.5574,
+      "step": 4705
+    },
+    {
+      "epoch": 1.7186415921124705,
+      "grad_norm": 0.7420259118080139,
+      "learning_rate": 2.9734840809702613e-06,
+      "loss": 0.5377,
+      "step": 4706
+    },
+    {
+      "epoch": 1.7190067555230966,
+      "grad_norm": 0.9698185920715332,
+      "learning_rate": 2.965962447807644e-06,
+      "loss": 0.5281,
+      "step": 4707
+    },
+    {
+      "epoch": 1.7193719189337229,
+      "grad_norm": 0.7731484770774841,
+      "learning_rate": 2.9584495779738144e-06,
+      "loss": 0.5192,
+      "step": 4708
+    },
+    {
+      "epoch": 1.719737082344349,
+      "grad_norm": 1.15269136428833,
+      "learning_rate": 2.950945475333846e-06,
+      "loss": 0.5162,
+      "step": 4709
+    },
+    {
+      "epoch": 1.7201022457549753,
+      "grad_norm": 0.6507449746131897,
+      "learning_rate": 2.94345014374829e-06,
+      "loss": 0.5638,
+      "step": 4710
+    },
+    {
+      "epoch": 1.7204674091656016,
+      "grad_norm": 0.7738745212554932,
+      "learning_rate": 2.9359635870732028e-06,
+      "loss": 0.5279,
+      "step": 4711
+    },
+    {
+      "epoch": 1.720832572576228,
+      "grad_norm": 0.7932908535003662,
+      "learning_rate": 2.928485809160109e-06,
+      "loss": 0.5308,
+      "step": 4712
+    },
+    {
+      "epoch": 1.7211977359868542,
+      "grad_norm": 0.9706488847732544,
+      "learning_rate": 2.921016813856028e-06,
+      "loss": 0.5123,
+      "step": 4713
+    },
+    {
+      "epoch": 1.7215628993974805,
+      "grad_norm": 0.9505299925804138,
+      "learning_rate": 2.91355660500346e-06,
+      "loss": 0.5431,
+      "step": 4714
+    },
+    {
+      "epoch": 1.7219280628081066,
+      "grad_norm": 0.6139938235282898,
+      "learning_rate": 2.906105186440389e-06,
+      "loss": 0.5527,
+      "step": 4715
+    },
+    {
+      "epoch": 1.722293226218733,
+      "grad_norm": 1.4412078857421875,
+      "learning_rate": 2.8986625620002586e-06,
+      "loss": 0.5151,
+      "step": 4716
+    },
+    {
+      "epoch": 1.722658389629359,
+      "grad_norm": 0.7758844494819641,
+      "learning_rate": 2.891228735512004e-06,
+      "loss": 0.5332,
+      "step": 4717
+    },
+    {
+      "epoch": 1.7230235530399853,
+      "grad_norm": 0.9113196730613708,
+      "learning_rate": 2.883803710800035e-06,
+      "loss": 0.5276,
+      "step": 4718
+    },
+    {
+      "epoch": 1.7233887164506116,
+      "grad_norm": 0.7024984955787659,
+      "learning_rate": 2.876387491684225e-06,
+      "loss": 0.5439,
+      "step": 4719
+    },
+    {
+      "epoch": 1.723753879861238,
+      "grad_norm": 0.7437683343887329,
+      "learning_rate": 2.8689800819799286e-06,
+      "loss": 0.5068,
+      "step": 4720
+    },
+    {
+      "epoch": 1.7241190432718643,
+      "grad_norm": 1.359061598777771,
+      "learning_rate": 2.8615814854979507e-06,
+      "loss": 0.5174,
+      "step": 4721
+    },
+    {
+      "epoch": 1.7244842066824904,
+      "grad_norm": 0.9796109795570374,
+      "learning_rate": 2.85419170604458e-06,
+      "loss": 0.5132,
+      "step": 4722
+    },
+    {
+      "epoch": 1.7248493700931167,
+      "grad_norm": 0.8437585830688477,
+      "learning_rate": 2.846810747421553e-06,
+      "loss": 0.4893,
+      "step": 4723
+    },
+    {
+      "epoch": 1.7252145335037428,
+      "grad_norm": 0.8218032121658325,
+      "learning_rate": 2.8394386134260843e-06,
+      "loss": 0.5335,
+      "step": 4724
+    },
+    {
+      "epoch": 1.725579696914369,
+      "grad_norm": 0.9758670330047607,
+      "learning_rate": 2.832075307850841e-06,
+      "loss": 0.5168,
+      "step": 4725
+    },
+    {
+      "epoch": 1.7259448603249954,
+      "grad_norm": 0.8763068914413452,
+      "learning_rate": 2.8247208344839428e-06,
+      "loss": 0.5391,
+      "step": 4726
+    },
+    {
+      "epoch": 1.7263100237356217,
+      "grad_norm": 0.8063436150550842,
+      "learning_rate": 2.8173751971089734e-06,
+      "loss": 0.5856,
+      "step": 4727
+    },
+    {
+      "epoch": 1.726675187146248,
+      "grad_norm": 0.7337233424186707,
+      "learning_rate": 2.8100383995049687e-06,
+      "loss": 0.58,
+      "step": 4728
+    },
+    {
+      "epoch": 1.7270403505568743,
+      "grad_norm": 1.227530837059021,
+      "learning_rate": 2.8027104454464172e-06,
+      "loss": 0.4644,
+      "step": 4729
+    },
+    {
+      "epoch": 1.7274055139675004,
+      "grad_norm": 1.321012020111084,
+      "learning_rate": 2.795391338703264e-06,
+      "loss": 0.4856,
+      "step": 4730
+    },
+    {
+      "epoch": 1.7277706773781267,
+      "grad_norm": 0.9268075227737427,
+      "learning_rate": 2.7880810830408834e-06,
+      "loss": 0.5202,
+      "step": 4731
+    },
+    {
+      "epoch": 1.7281358407887528,
+      "grad_norm": 0.896336555480957,
+      "learning_rate": 2.7807796822201137e-06,
+      "loss": 0.5237,
+      "step": 4732
+    },
+    {
+      "epoch": 1.7285010041993791,
+      "grad_norm": 0.802261233329773,
+      "learning_rate": 2.773487139997233e-06,
+      "loss": 0.543,
+      "step": 4733
+    },
+    {
+      "epoch": 1.7288661676100054,
+      "grad_norm": 0.7332318425178528,
+      "learning_rate": 2.7662034601239664e-06,
+      "loss": 0.5299,
+      "step": 4734
+    },
+    {
+      "epoch": 1.7292313310206318,
+      "grad_norm": 0.9159461259841919,
+      "learning_rate": 2.7589286463474698e-06,
+      "loss": 0.5172,
+      "step": 4735
+    },
+    {
+      "epoch": 1.729596494431258,
+      "grad_norm": 0.8737739324569702,
+      "learning_rate": 2.7516627024103403e-06,
+      "loss": 0.5149,
+      "step": 4736
+    },
+    {
+      "epoch": 1.7299616578418844,
+      "grad_norm": 0.8592380881309509,
+      "learning_rate": 2.7444056320506175e-06,
+      "loss": 0.5326,
+      "step": 4737
+    },
+    {
+      "epoch": 1.7303268212525105,
+      "grad_norm": 0.8493632078170776,
+      "learning_rate": 2.7371574390017742e-06,
+      "loss": 0.5212,
+      "step": 4738
+    },
+    {
+      "epoch": 1.7306919846631368,
+      "grad_norm": 0.8829220533370972,
+      "learning_rate": 2.7299181269927165e-06,
+      "loss": 0.5368,
+      "step": 4739
+    },
+    {
+      "epoch": 1.7310571480737629,
+      "grad_norm": 0.8817154169082642,
+      "learning_rate": 2.7226876997477723e-06,
+      "loss": 0.5236,
+      "step": 4740
+    },
+    {
+      "epoch": 1.7314223114843892,
+      "grad_norm": 2.187655210494995,
+      "learning_rate": 2.7154661609867126e-06,
+      "loss": 0.5365,
+      "step": 4741
+    },
+    {
+      "epoch": 1.7317874748950155,
+      "grad_norm": 0.9684646725654602,
+      "learning_rate": 2.708253514424728e-06,
+      "loss": 0.4993,
+      "step": 4742
+    },
+    {
+      "epoch": 1.7321526383056418,
+      "grad_norm": 0.9810671210289001,
+      "learning_rate": 2.701049763772434e-06,
+      "loss": 0.5244,
+      "step": 4743
+    },
+    {
+      "epoch": 1.7325178017162681,
+      "grad_norm": 0.9408937692642212,
+      "learning_rate": 2.6938549127358803e-06,
+      "loss": 0.5175,
+      "step": 4744
+    },
+    {
+      "epoch": 1.7328829651268944,
+      "grad_norm": 0.9165375828742981,
+      "learning_rate": 2.6866689650165146e-06,
+      "loss": 0.5063,
+      "step": 4745
+    },
+    {
+      "epoch": 1.7332481285375205,
+      "grad_norm": 0.8951494097709656,
+      "learning_rate": 2.679491924311226e-06,
+      "loss": 0.4961,
+      "step": 4746
+    },
+    {
+      "epoch": 1.7336132919481468,
+      "grad_norm": 0.9282603859901428,
+      "learning_rate": 2.672323794312315e-06,
+      "loss": 0.4906,
+      "step": 4747
+    },
+    {
+      "epoch": 1.733978455358773,
+      "grad_norm": 0.9112241864204407,
+      "learning_rate": 2.6651645787075e-06,
+      "loss": 0.5159,
+      "step": 4748
+    },
+    {
+      "epoch": 1.7343436187693992,
+      "grad_norm": 0.7655164003372192,
+      "learning_rate": 2.6580142811799037e-06,
+      "loss": 0.5637,
+      "step": 4749
+    },
+    {
+      "epoch": 1.7347087821800256,
+      "grad_norm": 1.001355528831482,
+      "learning_rate": 2.6508729054080664e-06,
+      "loss": 0.5419,
+      "step": 4750
+    },
+    {
+      "epoch": 1.7350739455906519,
+      "grad_norm": 0.9860793352127075,
+      "learning_rate": 2.6437404550659416e-06,
+      "loss": 0.5027,
+      "step": 4751
+    },
+    {
+      "epoch": 1.7354391090012782,
+      "grad_norm": 0.9379138946533203,
+      "learning_rate": 2.6366169338228885e-06,
+      "loss": 0.5681,
+      "step": 4752
+    },
+    {
+      "epoch": 1.7358042724119045,
+      "grad_norm": 0.9619566798210144,
+      "learning_rate": 2.629502345343675e-06,
+      "loss": 0.5085,
+      "step": 4753
+    },
+    {
+      "epoch": 1.7361694358225306,
+      "grad_norm": 0.8003615140914917,
+      "learning_rate": 2.622396693288474e-06,
+      "loss": 0.54,
+      "step": 4754
+    },
+    {
+      "epoch": 1.7365345992331567,
+      "grad_norm": 0.8115968108177185,
+      "learning_rate": 2.6152999813128487e-06,
+      "loss": 0.5348,
+      "step": 4755
+    },
+    {
+      "epoch": 1.736899762643783,
+      "grad_norm": 0.6173460483551025,
+      "learning_rate": 2.60821221306778e-06,
+      "loss": 0.5458,
+      "step": 4756
+    },
+    {
+      "epoch": 1.7372649260544093,
+      "grad_norm": 0.7967430949211121,
+      "learning_rate": 2.6011333921996397e-06,
+      "loss": 0.5471,
+      "step": 4757
+    },
+    {
+      "epoch": 1.7376300894650356,
+      "grad_norm": 1.2879356145858765,
+      "learning_rate": 2.5940635223501985e-06,
+      "loss": 0.4722,
+      "step": 4758
+    },
+    {
+      "epoch": 1.737995252875662,
+      "grad_norm": 0.6266947984695435,
+      "learning_rate": 2.5870026071566145e-06,
+      "loss": 0.5734,
+      "step": 4759
+    },
+    {
+      "epoch": 1.7383604162862882,
+      "grad_norm": 0.8225075006484985,
+      "learning_rate": 2.5799506502514504e-06,
+      "loss": 0.5502,
+      "step": 4760
+    },
+    {
+      "epoch": 1.7387255796969143,
+      "grad_norm": 0.9278892874717712,
+      "learning_rate": 2.572907655262653e-06,
+      "loss": 0.5358,
+      "step": 4761
+    },
+    {
+      "epoch": 1.7390907431075406,
+      "grad_norm": 1.1641044616699219,
+      "learning_rate": 2.565873625813564e-06,
+      "loss": 0.5253,
+      "step": 4762
+    },
+    {
+      "epoch": 1.7394559065181667,
+      "grad_norm": 0.6092157959938049,
+      "learning_rate": 2.5588485655229046e-06,
+      "loss": 0.5792,
+      "step": 4763
+    },
+    {
+      "epoch": 1.739821069928793,
+      "grad_norm": 0.8689990043640137,
+      "learning_rate": 2.5518324780047922e-06,
+      "loss": 0.519,
+      "step": 4764
+    },
+    {
+      "epoch": 1.7401862333394194,
+      "grad_norm": 0.952075719833374,
+      "learning_rate": 2.544825366868713e-06,
+      "loss": 0.51,
+      "step": 4765
+    },
+    {
+      "epoch": 1.7405513967500457,
+      "grad_norm": 0.996828019618988,
+      "learning_rate": 2.53782723571955e-06,
+      "loss": 0.5074,
+      "step": 4766
+    },
+    {
+      "epoch": 1.740916560160672,
+      "grad_norm": 1.044696569442749,
+      "learning_rate": 2.5308380881575613e-06,
+      "loss": 0.5215,
+      "step": 4767
+    },
+    {
+      "epoch": 1.7412817235712983,
+      "grad_norm": 0.796924889087677,
+      "learning_rate": 2.523857927778388e-06,
+      "loss": 0.5408,
+      "step": 4768
+    },
+    {
+      "epoch": 1.7416468869819244,
+      "grad_norm": 0.7988172769546509,
+      "learning_rate": 2.5168867581730315e-06,
+      "loss": 0.5623,
+      "step": 4769
+    },
+    {
+      "epoch": 1.7420120503925507,
+      "grad_norm": 1.13484525680542,
+      "learning_rate": 2.509924582927883e-06,
+      "loss": 0.5479,
+      "step": 4770
+    },
+    {
+      "epoch": 1.7423772138031768,
+      "grad_norm": 0.7947938442230225,
+      "learning_rate": 2.502971405624706e-06,
+      "loss": 0.5462,
+      "step": 4771
+    },
+    {
+      "epoch": 1.742742377213803,
+      "grad_norm": 0.9073025584220886,
+      "learning_rate": 2.4960272298406276e-06,
+      "loss": 0.497,
+      "step": 4772
+    },
+    {
+      "epoch": 1.7431075406244294,
+      "grad_norm": 1.380642056465149,
+      "learning_rate": 2.4890920591481525e-06,
+      "loss": 0.5008,
+      "step": 4773
+    },
+    {
+      "epoch": 1.7434727040350557,
+      "grad_norm": 1.0848580598831177,
+      "learning_rate": 2.4821658971151406e-06,
+      "loss": 0.5126,
+      "step": 4774
+    },
+    {
+      "epoch": 1.743837867445682,
+      "grad_norm": 0.920659601688385,
+      "learning_rate": 2.4752487473048327e-06,
+      "loss": 0.5457,
+      "step": 4775
+    },
+    {
+      "epoch": 1.7442030308563083,
+      "grad_norm": 0.8806758522987366,
+      "learning_rate": 2.4683406132758147e-06,
+      "loss": 0.5411,
+      "step": 4776
+    },
+    {
+      "epoch": 1.7445681942669344,
+      "grad_norm": 0.6796289086341858,
+      "learning_rate": 2.461441498582049e-06,
+      "loss": 0.5435,
+      "step": 4777
+    },
+    {
+      "epoch": 1.7449333576775607,
+      "grad_norm": 1.0951828956604004,
+      "learning_rate": 2.454551406772858e-06,
+      "loss": 0.5091,
+      "step": 4778
+    },
+    {
+      "epoch": 1.7452985210881868,
+      "grad_norm": 0.8472336530685425,
+      "learning_rate": 2.447670341392909e-06,
+      "loss": 0.5439,
+      "step": 4779
+    },
+    {
+      "epoch": 1.7456636844988132,
+      "grad_norm": 0.878318190574646,
+      "learning_rate": 2.4407983059822394e-06,
+      "loss": 0.5339,
+      "step": 4780
+    },
+    {
+      "epoch": 1.7460288479094395,
+      "grad_norm": 1.5507493019104004,
+      "learning_rate": 2.4339353040762337e-06,
+      "loss": 0.4882,
+      "step": 4781
+    },
+    {
+      "epoch": 1.7463940113200658,
+      "grad_norm": 0.8883445262908936,
+      "learning_rate": 2.427081339205635e-06,
+      "loss": 0.5267,
+      "step": 4782
+    },
+    {
+      "epoch": 1.746759174730692,
+      "grad_norm": 1.020946741104126,
+      "learning_rate": 2.4202364148965262e-06,
+      "loss": 0.5491,
+      "step": 4783
+    },
+    {
+      "epoch": 1.7471243381413184,
+      "grad_norm": 1.1917569637298584,
+      "learning_rate": 2.4134005346703517e-06,
+      "loss": 0.4731,
+      "step": 4784
+    },
+    {
+      "epoch": 1.7474895015519445,
+      "grad_norm": 0.8810885548591614,
+      "learning_rate": 2.406573702043893e-06,
+      "loss": 0.5244,
+      "step": 4785
+    },
+    {
+      "epoch": 1.7478546649625708,
+      "grad_norm": 0.9442121386528015,
+      "learning_rate": 2.3997559205292877e-06,
+      "loss": 0.5521,
+      "step": 4786
+    },
+    {
+      "epoch": 1.748219828373197,
+      "grad_norm": 0.6434993147850037,
+      "learning_rate": 2.3929471936340075e-06,
+      "loss": 0.5661,
+      "step": 4787
+    },
+    {
+      "epoch": 1.7485849917838232,
+      "grad_norm": 1.0258697271347046,
+      "learning_rate": 2.386147524860869e-06,
+      "loss": 0.5139,
+      "step": 4788
+    },
+    {
+      "epoch": 1.7489501551944495,
+      "grad_norm": 0.7254213094711304,
+      "learning_rate": 2.379356917708031e-06,
+      "loss": 0.5439,
+      "step": 4789
+    },
+    {
+      "epoch": 1.7493153186050758,
+      "grad_norm": 0.9993529915809631,
+      "learning_rate": 2.3725753756689816e-06,
+      "loss": 0.5168,
+      "step": 4790
+    },
+    {
+      "epoch": 1.7496804820157021,
+      "grad_norm": 1.0882890224456787,
+      "learning_rate": 2.365802902232559e-06,
+      "loss": 0.5161,
+      "step": 4791
+    },
+    {
+      "epoch": 1.7500456454263282,
+      "grad_norm": 1.1464070081710815,
+      "learning_rate": 2.3590395008829314e-06,
+      "loss": 0.5489,
+      "step": 4792
+    },
+    {
+      "epoch": 1.7504108088369545,
+      "grad_norm": 0.8863118290901184,
+      "learning_rate": 2.352285175099587e-06,
+      "loss": 0.5322,
+      "step": 4793
+    },
+    {
+      "epoch": 1.7507759722475806,
+      "grad_norm": 0.8491637110710144,
+      "learning_rate": 2.345539928357361e-06,
+      "loss": 0.4818,
+      "step": 4794
+    },
+    {
+      "epoch": 1.751141135658207,
+      "grad_norm": 0.7878150343894958,
+      "learning_rate": 2.338803764126414e-06,
+      "loss": 0.558,
+      "step": 4795
+    },
+    {
+      "epoch": 1.7515062990688333,
+      "grad_norm": 0.8596755862236023,
+      "learning_rate": 2.332076685872231e-06,
+      "loss": 0.5386,
+      "step": 4796
+    },
+    {
+      "epoch": 1.7518714624794596,
+      "grad_norm": 0.6973880529403687,
+      "learning_rate": 2.325358697055626e-06,
+      "loss": 0.5399,
+      "step": 4797
+    },
+    {
+      "epoch": 1.7522366258900859,
+      "grad_norm": 0.9609445333480835,
+      "learning_rate": 2.3186498011327286e-06,
+      "loss": 0.5519,
+      "step": 4798
+    },
+    {
+      "epoch": 1.7526017893007122,
+      "grad_norm": 0.8189699053764343,
+      "learning_rate": 2.3119500015550012e-06,
+      "loss": 0.5047,
+      "step": 4799
+    },
+    {
+      "epoch": 1.7529669527113383,
+      "grad_norm": 0.820521891117096,
+      "learning_rate": 2.3052593017692184e-06,
+      "loss": 0.5433,
+      "step": 4800
+    },
+    {
+      "epoch": 1.7533321161219646,
+      "grad_norm": 0.7218295335769653,
+      "learning_rate": 2.298577705217486e-06,
+      "loss": 0.5228,
+      "step": 4801
+    },
+    {
+      "epoch": 1.7536972795325907,
+      "grad_norm": 0.9512649178504944,
+      "learning_rate": 2.291905215337209e-06,
+      "loss": 0.4886,
+      "step": 4802
+    },
+    {
+      "epoch": 1.754062442943217,
+      "grad_norm": 1.2566230297088623,
+      "learning_rate": 2.285241835561112e-06,
+      "loss": 0.4784,
+      "step": 4803
+    },
+    {
+      "epoch": 1.7544276063538433,
+      "grad_norm": 1.1414672136306763,
+      "learning_rate": 2.2785875693172433e-06,
+      "loss": 0.5043,
+      "step": 4804
+    },
+    {
+      "epoch": 1.7547927697644696,
+      "grad_norm": 1.2826882600784302,
+      "learning_rate": 2.271942420028954e-06,
+      "loss": 0.5812,
+      "step": 4805
+    },
+    {
+      "epoch": 1.755157933175096,
+      "grad_norm": 0.8251254558563232,
+      "learning_rate": 2.2653063911149052e-06,
+      "loss": 0.5404,
+      "step": 4806
+    },
+    {
+      "epoch": 1.7555230965857223,
+      "grad_norm": 0.7828681468963623,
+      "learning_rate": 2.258679485989075e-06,
+      "loss": 0.5413,
+      "step": 4807
+    },
+    {
+      "epoch": 1.7558882599963483,
+      "grad_norm": 0.8482041358947754,
+      "learning_rate": 2.252061708060731e-06,
+      "loss": 0.5275,
+      "step": 4808
+    },
+    {
+      "epoch": 1.7562534234069747,
+      "grad_norm": 1.0937703847885132,
+      "learning_rate": 2.245453060734457e-06,
+      "loss": 0.511,
+      "step": 4809
+    },
+    {
+      "epoch": 1.7566185868176007,
+      "grad_norm": 0.6529538035392761,
+      "learning_rate": 2.238853547410136e-06,
+      "loss": 0.5653,
+      "step": 4810
+    },
+    {
+      "epoch": 1.756983750228227,
+      "grad_norm": 0.8129596710205078,
+      "learning_rate": 2.23226317148296e-06,
+      "loss": 0.561,
+      "step": 4811
+    },
+    {
+      "epoch": 1.7573489136388534,
+      "grad_norm": 0.8450109362602234,
+      "learning_rate": 2.2256819363434048e-06,
+      "loss": 0.4653,
+      "step": 4812
+    },
+    {
+      "epoch": 1.7577140770494797,
+      "grad_norm": 0.7205243706703186,
+      "learning_rate": 2.219109845377252e-06,
+      "loss": 0.5592,
+      "step": 4813
+    },
+    {
+      "epoch": 1.758079240460106,
+      "grad_norm": 0.8044624924659729,
+      "learning_rate": 2.212546901965582e-06,
+      "loss": 0.5601,
+      "step": 4814
+    },
+    {
+      "epoch": 1.7584444038707323,
+      "grad_norm": 1.0224130153656006,
+      "learning_rate": 2.2059931094847676e-06,
+      "loss": 0.511,
+      "step": 4815
+    },
+    {
+      "epoch": 1.7588095672813584,
+      "grad_norm": 0.8828184604644775,
+      "learning_rate": 2.199448471306467e-06,
+      "loss": 0.4883,
+      "step": 4816
+    },
+    {
+      "epoch": 1.7591747306919847,
+      "grad_norm": 0.6978619694709778,
+      "learning_rate": 2.19291299079764e-06,
+      "loss": 0.5757,
+      "step": 4817
+    },
+    {
+      "epoch": 1.7595398941026108,
+      "grad_norm": 0.8363275527954102,
+      "learning_rate": 2.186386671320522e-06,
+      "loss": 0.5259,
+      "step": 4818
+    },
+    {
+      "epoch": 1.7599050575132371,
+      "grad_norm": 0.7557588219642639,
+      "learning_rate": 2.1798695162326444e-06,
+      "loss": 0.5323,
+      "step": 4819
+    },
+    {
+      "epoch": 1.7602702209238634,
+      "grad_norm": 0.9395306706428528,
+      "learning_rate": 2.1733615288868236e-06,
+      "loss": 0.4881,
+      "step": 4820
+    },
+    {
+      "epoch": 1.7606353843344897,
+      "grad_norm": 0.9875335097312927,
+      "learning_rate": 2.1668627126311613e-06,
+      "loss": 0.5258,
+      "step": 4821
+    },
+    {
+      "epoch": 1.761000547745116,
+      "grad_norm": 0.999130368232727,
+      "learning_rate": 2.160373070809032e-06,
+      "loss": 0.5168,
+      "step": 4822
+    },
+    {
+      "epoch": 1.7613657111557421,
+      "grad_norm": 0.8791902661323547,
+      "learning_rate": 2.153892606759096e-06,
+      "loss": 0.5003,
+      "step": 4823
+    },
+    {
+      "epoch": 1.7617308745663685,
+      "grad_norm": 0.8618801832199097,
+      "learning_rate": 2.1474213238152954e-06,
+      "loss": 0.5528,
+      "step": 4824
+    },
+    {
+      "epoch": 1.7620960379769945,
+      "grad_norm": 0.7623894214630127,
+      "learning_rate": 2.1409592253068467e-06,
+      "loss": 0.538,
+      "step": 4825
+    },
+    {
+      "epoch": 1.7624612013876209,
+      "grad_norm": 0.8928764462471008,
+      "learning_rate": 2.1345063145582357e-06,
+      "loss": 0.5246,
+      "step": 4826
+    },
+    {
+      "epoch": 1.7628263647982472,
+      "grad_norm": 0.6507142782211304,
+      "learning_rate": 2.128062594889229e-06,
+      "loss": 0.5662,
+      "step": 4827
+    },
+    {
+      "epoch": 1.7631915282088735,
+      "grad_norm": 0.6624026894569397,
+      "learning_rate": 2.1216280696148585e-06,
+      "loss": 0.5513,
+      "step": 4828
+    },
+    {
+      "epoch": 1.7635566916194998,
+      "grad_norm": 0.8341464400291443,
+      "learning_rate": 2.115202742045437e-06,
+      "loss": 0.5278,
+      "step": 4829
+    },
+    {
+      "epoch": 1.763921855030126,
+      "grad_norm": 1.1716243028640747,
+      "learning_rate": 2.108786615486529e-06,
+      "loss": 0.507,
+      "step": 4830
+    },
+    {
+      "epoch": 1.7642870184407522,
+      "grad_norm": 0.7278099656105042,
+      "learning_rate": 2.1023796932389805e-06,
+      "loss": 0.5435,
+      "step": 4831
+    },
+    {
+      "epoch": 1.7646521818513785,
+      "grad_norm": 1.058091402053833,
+      "learning_rate": 2.0959819785988912e-06,
+      "loss": 0.5063,
+      "step": 4832
+    },
+    {
+      "epoch": 1.7650173452620046,
+      "grad_norm": 0.9359176754951477,
+      "learning_rate": 2.0895934748576273e-06,
+      "loss": 0.5524,
+      "step": 4833
+    },
+    {
+      "epoch": 1.765382508672631,
+      "grad_norm": 0.8545036911964417,
+      "learning_rate": 2.0832141853018227e-06,
+      "loss": 0.5243,
+      "step": 4834
+    },
+    {
+      "epoch": 1.7657476720832572,
+      "grad_norm": 0.8072575330734253,
+      "learning_rate": 2.0768441132133676e-06,
+      "loss": 0.5597,
+      "step": 4835
+    },
+    {
+      "epoch": 1.7661128354938835,
+      "grad_norm": 0.9316344857215881,
+      "learning_rate": 2.0704832618694006e-06,
+      "loss": 0.5033,
+      "step": 4836
+    },
+    {
+      "epoch": 1.7664779989045098,
+      "grad_norm": 1.037131428718567,
+      "learning_rate": 2.0641316345423303e-06,
+      "loss": 0.5213,
+      "step": 4837
+    },
+    {
+      "epoch": 1.7668431623151362,
+      "grad_norm": 0.9110147356987,
+      "learning_rate": 2.0577892344998097e-06,
+      "loss": 0.5194,
+      "step": 4838
+    },
+    {
+      "epoch": 1.7672083257257623,
+      "grad_norm": 1.0251874923706055,
+      "learning_rate": 2.051456065004753e-06,
+      "loss": 0.5171,
+      "step": 4839
+    },
+    {
+      "epoch": 1.7675734891363886,
+      "grad_norm": 0.7813372015953064,
+      "learning_rate": 2.045132129315326e-06,
+      "loss": 0.5572,
+      "step": 4840
+    },
+    {
+      "epoch": 1.7679386525470147,
+      "grad_norm": 0.8873542547225952,
+      "learning_rate": 2.0388174306849297e-06,
+      "loss": 0.5058,
+      "step": 4841
+    },
+    {
+      "epoch": 1.768303815957641,
+      "grad_norm": 0.9328134655952454,
+      "learning_rate": 2.03251197236223e-06,
+      "loss": 0.5251,
+      "step": 4842
+    },
+    {
+      "epoch": 1.7686689793682673,
+      "grad_norm": 0.6912343502044678,
+      "learning_rate": 2.026215757591128e-06,
+      "loss": 0.5766,
+      "step": 4843
+    },
+    {
+      "epoch": 1.7690341427788936,
+      "grad_norm": 1.0498008728027344,
+      "learning_rate": 2.0199287896107743e-06,
+      "loss": 0.4877,
+      "step": 4844
+    },
+    {
+      "epoch": 1.76939930618952,
+      "grad_norm": 0.9849687814712524,
+      "learning_rate": 2.013651071655569e-06,
+      "loss": 0.5003,
+      "step": 4845
+    },
+    {
+      "epoch": 1.7697644696001462,
+      "grad_norm": 0.7693188190460205,
+      "learning_rate": 2.007382606955135e-06,
+      "loss": 0.5307,
+      "step": 4846
+    },
+    {
+      "epoch": 1.7701296330107723,
+      "grad_norm": 0.6070563197135925,
+      "learning_rate": 2.00112339873435e-06,
+      "loss": 0.5847,
+      "step": 4847
+    },
+    {
+      "epoch": 1.7704947964213986,
+      "grad_norm": 1.0467851161956787,
+      "learning_rate": 1.9948734502133284e-06,
+      "loss": 0.4864,
+      "step": 4848
+    },
+    {
+      "epoch": 1.7708599598320247,
+      "grad_norm": 0.8477541208267212,
+      "learning_rate": 1.9886327646074143e-06,
+      "loss": 0.5194,
+      "step": 4849
+    },
+    {
+      "epoch": 1.771225123242651,
+      "grad_norm": 0.8596732020378113,
+      "learning_rate": 1.9824013451271964e-06,
+      "loss": 0.5591,
+      "step": 4850
+    },
+    {
+      "epoch": 1.7715902866532773,
+      "grad_norm": 0.963614284992218,
+      "learning_rate": 1.9761791949784827e-06,
+      "loss": 0.5387,
+      "step": 4851
+    },
+    {
+      "epoch": 1.7719554500639036,
+      "grad_norm": 0.8316560983657837,
+      "learning_rate": 1.9699663173623195e-06,
+      "loss": 0.5331,
+      "step": 4852
+    },
+    {
+      "epoch": 1.77232061347453,
+      "grad_norm": 0.8175371289253235,
+      "learning_rate": 1.9637627154749882e-06,
+      "loss": 0.5364,
+      "step": 4853
+    },
+    {
+      "epoch": 1.7726857768851563,
+      "grad_norm": 1.0496881008148193,
+      "learning_rate": 1.9575683925079913e-06,
+      "loss": 0.5104,
+      "step": 4854
+    },
+    {
+      "epoch": 1.7730509402957824,
+      "grad_norm": 0.9106097221374512,
+      "learning_rate": 1.951383351648057e-06,
+      "loss": 0.4966,
+      "step": 4855
+    },
+    {
+      "epoch": 1.7734161037064085,
+      "grad_norm": 1.158538818359375,
+      "learning_rate": 1.945207596077148e-06,
+      "loss": 0.4819,
+      "step": 4856
+    },
+    {
+      "epoch": 1.7737812671170348,
+      "grad_norm": 0.7018920183181763,
+      "learning_rate": 1.93904112897243e-06,
+      "loss": 0.6078,
+      "step": 4857
+    },
+    {
+      "epoch": 1.774146430527661,
+      "grad_norm": 0.92885422706604,
+      "learning_rate": 1.9328839535063125e-06,
+      "loss": 0.5323,
+      "step": 4858
+    },
+    {
+      "epoch": 1.7745115939382874,
+      "grad_norm": 1.1306513547897339,
+      "learning_rate": 1.9267360728464113e-06,
+      "loss": 0.553,
+      "step": 4859
+    },
+    {
+      "epoch": 1.7748767573489137,
+      "grad_norm": 0.9900164008140564,
+      "learning_rate": 1.920597490155569e-06,
+      "loss": 0.5171,
+      "step": 4860
+    },
+    {
+      "epoch": 1.77524192075954,
+      "grad_norm": 0.9863156676292419,
+      "learning_rate": 1.9144682085918354e-06,
+      "loss": 0.5778,
+      "step": 4861
+    },
+    {
+      "epoch": 1.775607084170166,
+      "grad_norm": 1.0554563999176025,
+      "learning_rate": 1.908348231308479e-06,
+      "loss": 0.5439,
+      "step": 4862
+    },
+    {
+      "epoch": 1.7759722475807924,
+      "grad_norm": 0.9097983241081238,
+      "learning_rate": 1.9022375614539857e-06,
+      "loss": 0.5279,
+      "step": 4863
+    },
+    {
+      "epoch": 1.7763374109914185,
+      "grad_norm": 0.9413915872573853,
+      "learning_rate": 1.896136202172052e-06,
+      "loss": 0.5353,
+      "step": 4864
+    },
+    {
+      "epoch": 1.7767025744020448,
+      "grad_norm": 1.017560601234436,
+      "learning_rate": 1.890044156601576e-06,
+      "loss": 0.5193,
+      "step": 4865
+    },
+    {
+      "epoch": 1.7770677378126711,
+      "grad_norm": 1.0063955783843994,
+      "learning_rate": 1.883961427876675e-06,
+      "loss": 0.5507,
+      "step": 4866
+    },
+    {
+      "epoch": 1.7774329012232974,
+      "grad_norm": 0.7404409646987915,
+      "learning_rate": 1.877888019126668e-06,
+      "loss": 0.5698,
+      "step": 4867
+    },
+    {
+      "epoch": 1.7777980646339238,
+      "grad_norm": 0.9323861002922058,
+      "learning_rate": 1.8718239334760824e-06,
+      "loss": 0.4848,
+      "step": 4868
+    },
+    {
+      "epoch": 1.77816322804455,
+      "grad_norm": 0.8537315130233765,
+      "learning_rate": 1.865769174044647e-06,
+      "loss": 0.5092,
+      "step": 4869
+    },
+    {
+      "epoch": 1.7785283914551762,
+      "grad_norm": 0.7566486597061157,
+      "learning_rate": 1.8597237439472837e-06,
+      "loss": 0.539,
+      "step": 4870
+    },
+    {
+      "epoch": 1.7788935548658025,
+      "grad_norm": 0.7786027789115906,
+      "learning_rate": 1.8536876462941311e-06,
+      "loss": 0.5466,
+      "step": 4871
+    },
+    {
+      "epoch": 1.7792587182764286,
+      "grad_norm": 0.7033721804618835,
+      "learning_rate": 1.8476608841905186e-06,
+      "loss": 0.5851,
+      "step": 4872
+    },
+    {
+      "epoch": 1.7796238816870549,
+      "grad_norm": 0.6912455558776855,
+      "learning_rate": 1.841643460736975e-06,
+      "loss": 0.5567,
+      "step": 4873
+    },
+    {
+      "epoch": 1.7799890450976812,
+      "grad_norm": 0.8948424458503723,
+      "learning_rate": 1.8356353790292237e-06,
+      "loss": 0.5106,
+      "step": 4874
+    },
+    {
+      "epoch": 1.7803542085083075,
+      "grad_norm": 0.7366495728492737,
+      "learning_rate": 1.8296366421581747e-06,
+      "loss": 0.5799,
+      "step": 4875
+    },
+    {
+      "epoch": 1.7807193719189338,
+      "grad_norm": 1.0875005722045898,
+      "learning_rate": 1.8236472532099413e-06,
+      "loss": 0.5114,
+      "step": 4876
+    },
+    {
+      "epoch": 1.7810845353295601,
+      "grad_norm": 1.5355473756790161,
+      "learning_rate": 1.817667215265826e-06,
+      "loss": 0.4878,
+      "step": 4877
+    },
+    {
+      "epoch": 1.7814496987401862,
+      "grad_norm": 0.9621209502220154,
+      "learning_rate": 1.8116965314023205e-06,
+      "loss": 0.5393,
+      "step": 4878
+    },
+    {
+      "epoch": 1.7818148621508125,
+      "grad_norm": 1.0355948209762573,
+      "learning_rate": 1.8057352046910948e-06,
+      "loss": 0.4891,
+      "step": 4879
+    },
+    {
+      "epoch": 1.7821800255614386,
+      "grad_norm": 0.8816014528274536,
+      "learning_rate": 1.7997832381990156e-06,
+      "loss": 0.4552,
+      "step": 4880
+    },
+    {
+      "epoch": 1.782545188972065,
+      "grad_norm": 0.9060640931129456,
+      "learning_rate": 1.793840634988131e-06,
+      "loss": 0.5117,
+      "step": 4881
+    },
+    {
+      "epoch": 1.7829103523826912,
+      "grad_norm": 1.2189513444900513,
+      "learning_rate": 1.787907398115676e-06,
+      "loss": 0.5229,
+      "step": 4882
+    },
+    {
+      "epoch": 1.7832755157933176,
+      "grad_norm": 1.0151749849319458,
+      "learning_rate": 1.781983530634055e-06,
+      "loss": 0.5406,
+      "step": 4883
+    },
+    {
+      "epoch": 1.7836406792039439,
+      "grad_norm": 0.9411863088607788,
+      "learning_rate": 1.7760690355908682e-06,
+      "loss": 0.4972,
+      "step": 4884
+    },
+    {
+      "epoch": 1.7840058426145702,
+      "grad_norm": 1.045339584350586,
+      "learning_rate": 1.7701639160288775e-06,
+      "loss": 0.501,
+      "step": 4885
+    },
+    {
+      "epoch": 1.7843710060251963,
+      "grad_norm": 0.8763211965560913,
+      "learning_rate": 1.7642681749860346e-06,
+      "loss": 0.5034,
+      "step": 4886
+    },
+    {
+      "epoch": 1.7847361694358226,
+      "grad_norm": 1.0296286344528198,
+      "learning_rate": 1.7583818154954602e-06,
+      "loss": 0.4522,
+      "step": 4887
+    },
+    {
+      "epoch": 1.7851013328464487,
+      "grad_norm": 0.8169348239898682,
+      "learning_rate": 1.7525048405854562e-06,
+      "loss": 0.5501,
+      "step": 4888
+    },
+    {
+      "epoch": 1.785466496257075,
+      "grad_norm": 0.9823426008224487,
+      "learning_rate": 1.7466372532794818e-06,
+      "loss": 0.5288,
+      "step": 4889
+    },
+    {
+      "epoch": 1.7858316596677013,
+      "grad_norm": 0.9222946166992188,
+      "learning_rate": 1.740779056596178e-06,
+      "loss": 0.4924,
+      "step": 4890
+    },
+    {
+      "epoch": 1.7861968230783276,
+      "grad_norm": 0.8936170935630798,
+      "learning_rate": 1.7349302535493539e-06,
+      "loss": 0.5334,
+      "step": 4891
+    },
+    {
+      "epoch": 1.786561986488954,
+      "grad_norm": 1.0143158435821533,
+      "learning_rate": 1.7290908471479805e-06,
+      "loss": 0.556,
+      "step": 4892
+    },
+    {
+      "epoch": 1.78692714989958,
+      "grad_norm": 0.8250148296356201,
+      "learning_rate": 1.723260840396206e-06,
+      "loss": 0.5419,
+      "step": 4893
+    },
+    {
+      "epoch": 1.7872923133102063,
+      "grad_norm": 0.8316763639450073,
+      "learning_rate": 1.717440236293324e-06,
+      "loss": 0.5456,
+      "step": 4894
+    },
+    {
+      "epoch": 1.7876574767208324,
+      "grad_norm": 0.874988317489624,
+      "learning_rate": 1.7116290378338085e-06,
+      "loss": 0.5511,
+      "step": 4895
+    },
+    {
+      "epoch": 1.7880226401314587,
+      "grad_norm": 1.303494930267334,
+      "learning_rate": 1.7058272480072879e-06,
+      "loss": 0.5079,
+      "step": 4896
+    },
+    {
+      "epoch": 1.788387803542085,
+      "grad_norm": 0.9614773988723755,
+      "learning_rate": 1.7000348697985481e-06,
+      "loss": 0.4785,
+      "step": 4897
+    },
+    {
+      "epoch": 1.7887529669527114,
+      "grad_norm": 1.2812979221343994,
+      "learning_rate": 1.6942519061875361e-06,
+      "loss": 0.5123,
+      "step": 4898
+    },
+    {
+      "epoch": 1.7891181303633377,
+      "grad_norm": 0.7624049186706543,
+      "learning_rate": 1.6884783601493525e-06,
+      "loss": 0.5734,
+      "step": 4899
+    },
+    {
+      "epoch": 1.789483293773964,
+      "grad_norm": 1.0612908601760864,
+      "learning_rate": 1.682714234654259e-06,
+      "loss": 0.4995,
+      "step": 4900
+    },
+    {
+      "epoch": 1.78984845718459,
+      "grad_norm": 0.931629478931427,
+      "learning_rate": 1.6769595326676614e-06,
+      "loss": 0.5619,
+      "step": 4901
+    },
+    {
+      "epoch": 1.7902136205952164,
+      "grad_norm": 0.9756954908370972,
+      "learning_rate": 1.6712142571501289e-06,
+      "loss": 0.5531,
+      "step": 4902
+    },
+    {
+      "epoch": 1.7905787840058425,
+      "grad_norm": 0.9437235593795776,
+      "learning_rate": 1.6654784110573752e-06,
+      "loss": 0.5197,
+      "step": 4903
+    },
+    {
+      "epoch": 1.7909439474164688,
+      "grad_norm": 0.9321839213371277,
+      "learning_rate": 1.6597519973402576e-06,
+      "loss": 0.539,
+      "step": 4904
+    },
+    {
+      "epoch": 1.791309110827095,
+      "grad_norm": 0.8487227559089661,
+      "learning_rate": 1.6540350189447885e-06,
+      "loss": 0.468,
+      "step": 4905
+    },
+    {
+      "epoch": 1.7916742742377214,
+      "grad_norm": 0.649875819683075,
+      "learning_rate": 1.6483274788121239e-06,
+      "loss": 0.5441,
+      "step": 4906
+    },
+    {
+      "epoch": 1.7920394376483477,
+      "grad_norm": 0.8302043676376343,
+      "learning_rate": 1.6426293798785687e-06,
+      "loss": 0.5531,
+      "step": 4907
+    },
+    {
+      "epoch": 1.792404601058974,
+      "grad_norm": 0.8178344964981079,
+      "learning_rate": 1.6369407250755598e-06,
+      "loss": 0.5449,
+      "step": 4908
+    },
+    {
+      "epoch": 1.7927697644696001,
+      "grad_norm": 0.9554221034049988,
+      "learning_rate": 1.6312615173296853e-06,
+      "loss": 0.4843,
+      "step": 4909
+    },
+    {
+      "epoch": 1.7931349278802264,
+      "grad_norm": 0.7250683307647705,
+      "learning_rate": 1.6255917595626681e-06,
+      "loss": 0.5209,
+      "step": 4910
+    },
+    {
+      "epoch": 1.7935000912908525,
+      "grad_norm": 0.7468822002410889,
+      "learning_rate": 1.619931454691368e-06,
+      "loss": 0.5529,
+      "step": 4911
+    },
+    {
+      "epoch": 1.7938652547014788,
+      "grad_norm": 0.866430401802063,
+      "learning_rate": 1.6142806056277937e-06,
+      "loss": 0.5628,
+      "step": 4912
+    },
+    {
+      "epoch": 1.7942304181121052,
+      "grad_norm": 1.0588103532791138,
+      "learning_rate": 1.6086392152790709e-06,
+      "loss": 0.4854,
+      "step": 4913
+    },
+    {
+      "epoch": 1.7945955815227315,
+      "grad_norm": 0.8642292022705078,
+      "learning_rate": 1.6030072865474733e-06,
+      "loss": 0.5345,
+      "step": 4914
+    },
+    {
+      "epoch": 1.7949607449333578,
+      "grad_norm": 0.8990249633789062,
+      "learning_rate": 1.5973848223304012e-06,
+      "loss": 0.5628,
+      "step": 4915
+    },
+    {
+      "epoch": 1.795325908343984,
+      "grad_norm": 0.7247717976570129,
+      "learning_rate": 1.5917718255203873e-06,
+      "loss": 0.5196,
+      "step": 4916
+    },
+    {
+      "epoch": 1.7956910717546102,
+      "grad_norm": 1.2262150049209595,
+      "learning_rate": 1.5861682990050954e-06,
+      "loss": 0.5317,
+      "step": 4917
+    },
+    {
+      "epoch": 1.7960562351652365,
+      "grad_norm": 0.8901282548904419,
+      "learning_rate": 1.5805742456673101e-06,
+      "loss": 0.524,
+      "step": 4918
+    },
+    {
+      "epoch": 1.7964213985758626,
+      "grad_norm": 0.8625813126564026,
+      "learning_rate": 1.5749896683849474e-06,
+      "loss": 0.5604,
+      "step": 4919
+    },
+    {
+      "epoch": 1.796786561986489,
+      "grad_norm": 0.7180377244949341,
+      "learning_rate": 1.5694145700310536e-06,
+      "loss": 0.5193,
+      "step": 4920
+    },
+    {
+      "epoch": 1.7971517253971152,
+      "grad_norm": 0.7622803449630737,
+      "learning_rate": 1.563848953473792e-06,
+      "loss": 0.5828,
+      "step": 4921
+    },
+    {
+      "epoch": 1.7975168888077415,
+      "grad_norm": 1.0300092697143555,
+      "learning_rate": 1.5582928215764481e-06,
+      "loss": 0.4961,
+      "step": 4922
+    },
+    {
+      "epoch": 1.7978820522183678,
+      "grad_norm": 0.8336998820304871,
+      "learning_rate": 1.552746177197424e-06,
+      "loss": 0.4831,
+      "step": 4923
+    },
+    {
+      "epoch": 1.798247215628994,
+      "grad_norm": 1.1535495519638062,
+      "learning_rate": 1.5472090231902504e-06,
+      "loss": 0.5051,
+      "step": 4924
+    },
+    {
+      "epoch": 1.7986123790396202,
+      "grad_norm": 0.9202743172645569,
+      "learning_rate": 1.5416813624035688e-06,
+      "loss": 0.5196,
+      "step": 4925
+    },
+    {
+      "epoch": 1.7989775424502463,
+      "grad_norm": 0.8379017114639282,
+      "learning_rate": 1.5361631976811397e-06,
+      "loss": 0.5328,
+      "step": 4926
+    },
+    {
+      "epoch": 1.7993427058608726,
+      "grad_norm": 1.020539402961731,
+      "learning_rate": 1.5306545318618437e-06,
+      "loss": 0.5235,
+      "step": 4927
+    },
+    {
+      "epoch": 1.799707869271499,
+      "grad_norm": 0.7980867028236389,
+      "learning_rate": 1.525155367779656e-06,
+      "loss": 0.5013,
+      "step": 4928
+    },
+    {
+      "epoch": 1.8000730326821253,
+      "grad_norm": 0.8660412430763245,
+      "learning_rate": 1.5196657082636845e-06,
+      "loss": 0.5398,
+      "step": 4929
+    },
+    {
+      "epoch": 1.8004381960927516,
+      "grad_norm": 0.7763341069221497,
+      "learning_rate": 1.5141855561381347e-06,
+      "loss": 0.5747,
+      "step": 4930
+    },
+    {
+      "epoch": 1.8008033595033779,
+      "grad_norm": 0.7591754198074341,
+      "learning_rate": 1.5087149142223313e-06,
+      "loss": 0.5247,
+      "step": 4931
+    },
+    {
+      "epoch": 1.801168522914004,
+      "grad_norm": 0.8435109257698059,
+      "learning_rate": 1.5032537853306917e-06,
+      "loss": 0.5799,
+      "step": 4932
+    },
+    {
+      "epoch": 1.8015336863246303,
+      "grad_norm": 0.8091663122177124,
+      "learning_rate": 1.4978021722727509e-06,
+      "loss": 0.5258,
+      "step": 4933
+    },
+    {
+      "epoch": 1.8018988497352564,
+      "grad_norm": 0.7816550731658936,
+      "learning_rate": 1.4923600778531456e-06,
+      "loss": 0.5383,
+      "step": 4934
+    },
+    {
+      "epoch": 1.8022640131458827,
+      "grad_norm": 0.7945713996887207,
+      "learning_rate": 1.486927504871616e-06,
+      "loss": 0.5642,
+      "step": 4935
+    },
+    {
+      "epoch": 1.802629176556509,
+      "grad_norm": 1.025600552558899,
+      "learning_rate": 1.481504456123004e-06,
+      "loss": 0.4561,
+      "step": 4936
+    },
+    {
+      "epoch": 1.8029943399671353,
+      "grad_norm": 0.870495617389679,
+      "learning_rate": 1.4760909343972473e-06,
+      "loss": 0.5342,
+      "step": 4937
+    },
+    {
+      "epoch": 1.8033595033777616,
+      "grad_norm": 0.7982380390167236,
+      "learning_rate": 1.4706869424793847e-06,
+      "loss": 0.5445,
+      "step": 4938
+    },
+    {
+      "epoch": 1.803724666788388,
+      "grad_norm": 1.338091492652893,
+      "learning_rate": 1.4652924831495563e-06,
+      "loss": 0.5234,
+      "step": 4939
+    },
+    {
+      "epoch": 1.804089830199014,
+      "grad_norm": 1.0072129964828491,
+      "learning_rate": 1.4599075591829915e-06,
+      "loss": 0.5588,
+      "step": 4940
+    },
+    {
+      "epoch": 1.8044549936096403,
+      "grad_norm": 1.072298288345337,
+      "learning_rate": 1.454532173350025e-06,
+      "loss": 0.5201,
+      "step": 4941
+    },
+    {
+      "epoch": 1.8048201570202664,
+      "grad_norm": 1.1159089803695679,
+      "learning_rate": 1.4491663284160694e-06,
+      "loss": 0.5026,
+      "step": 4942
+    },
+    {
+      "epoch": 1.8051853204308927,
+      "grad_norm": 0.9787336587905884,
+      "learning_rate": 1.4438100271416367e-06,
+      "loss": 0.5298,
+      "step": 4943
+    },
+    {
+      "epoch": 1.805550483841519,
+      "grad_norm": 0.9807587265968323,
+      "learning_rate": 1.4384632722823333e-06,
+      "loss": 0.5079,
+      "step": 4944
+    },
+    {
+      "epoch": 1.8059156472521454,
+      "grad_norm": 0.7864483594894409,
+      "learning_rate": 1.433126066588848e-06,
+      "loss": 0.561,
+      "step": 4945
+    },
+    {
+      "epoch": 1.8062808106627717,
+      "grad_norm": 0.9076181054115295,
+      "learning_rate": 1.4277984128069622e-06,
+      "loss": 0.5233,
+      "step": 4946
+    },
+    {
+      "epoch": 1.806645974073398,
+      "grad_norm": 1.0766689777374268,
+      "learning_rate": 1.4224803136775323e-06,
+      "loss": 0.4543,
+      "step": 4947
+    },
+    {
+      "epoch": 1.807011137484024,
+      "grad_norm": 0.8053579330444336,
+      "learning_rate": 1.417171771936514e-06,
+      "loss": 0.5475,
+      "step": 4948
+    },
+    {
+      "epoch": 1.8073763008946504,
+      "grad_norm": 0.856950581073761,
+      "learning_rate": 1.4118727903149387e-06,
+      "loss": 0.5093,
+      "step": 4949
+    },
+    {
+      "epoch": 1.8077414643052765,
+      "grad_norm": 1.2885442972183228,
+      "learning_rate": 1.4065833715389143e-06,
+      "loss": 0.433,
+      "step": 4950
+    },
+    {
+      "epoch": 1.8081066277159028,
+      "grad_norm": 0.9208719730377197,
+      "learning_rate": 1.401303518329642e-06,
+      "loss": 0.5619,
+      "step": 4951
+    },
+    {
+      "epoch": 1.8084717911265291,
+      "grad_norm": 0.6822531223297119,
+      "learning_rate": 1.3960332334033844e-06,
+      "loss": 0.5667,
+      "step": 4952
+    },
+    {
+      "epoch": 1.8088369545371554,
+      "grad_norm": 0.9479916095733643,
+      "learning_rate": 1.3907725194714994e-06,
+      "loss": 0.5314,
+      "step": 4953
+    },
+    {
+      "epoch": 1.8092021179477817,
+      "grad_norm": 0.8681054711341858,
+      "learning_rate": 1.3855213792404132e-06,
+      "loss": 0.5623,
+      "step": 4954
+    },
+    {
+      "epoch": 1.809567281358408,
+      "grad_norm": 0.8470045328140259,
+      "learning_rate": 1.3802798154116249e-06,
+      "loss": 0.5168,
+      "step": 4955
+    },
+    {
+      "epoch": 1.8099324447690341,
+      "grad_norm": 1.1664899587631226,
+      "learning_rate": 1.3750478306817082e-06,
+      "loss": 0.4947,
+      "step": 4956
+    },
+    {
+      "epoch": 1.8102976081796602,
+      "grad_norm": 1.0370721817016602,
+      "learning_rate": 1.3698254277423083e-06,
+      "loss": 0.536,
+      "step": 4957
+    },
+    {
+      "epoch": 1.8106627715902865,
+      "grad_norm": 0.7912999987602234,
+      "learning_rate": 1.3646126092801425e-06,
+      "loss": 0.5617,
+      "step": 4958
+    },
+    {
+      "epoch": 1.8110279350009129,
+      "grad_norm": 1.4642152786254883,
+      "learning_rate": 1.359409377976999e-06,
+      "loss": 0.5126,
+      "step": 4959
+    },
+    {
+      "epoch": 1.8113930984115392,
+      "grad_norm": 0.799483060836792,
+      "learning_rate": 1.354215736509734e-06,
+      "loss": 0.5196,
+      "step": 4960
+    },
+    {
+      "epoch": 1.8117582618221655,
+      "grad_norm": 0.8305796980857849,
+      "learning_rate": 1.3490316875502597e-06,
+      "loss": 0.5394,
+      "step": 4961
+    },
+    {
+      "epoch": 1.8121234252327918,
+      "grad_norm": 0.8141010999679565,
+      "learning_rate": 1.3438572337655686e-06,
+      "loss": 0.5229,
+      "step": 4962
+    },
+    {
+      "epoch": 1.8124885886434179,
+      "grad_norm": 0.9301368594169617,
+      "learning_rate": 1.338692377817703e-06,
+      "loss": 0.4989,
+      "step": 4963
+    },
+    {
+      "epoch": 1.8128537520540442,
+      "grad_norm": 0.9604087471961975,
+      "learning_rate": 1.3335371223637772e-06,
+      "loss": 0.5339,
+      "step": 4964
+    },
+    {
+      "epoch": 1.8132189154646703,
+      "grad_norm": 0.8913207650184631,
+      "learning_rate": 1.3283914700559675e-06,
+      "loss": 0.5143,
+      "step": 4965
+    },
+    {
+      "epoch": 1.8135840788752966,
+      "grad_norm": 0.7561500072479248,
+      "learning_rate": 1.3232554235414985e-06,
+      "loss": 0.5594,
+      "step": 4966
+    },
+    {
+      "epoch": 1.813949242285923,
+      "grad_norm": 1.0631812810897827,
+      "learning_rate": 1.3181289854626633e-06,
+      "loss": 0.5388,
+      "step": 4967
+    },
+    {
+      "epoch": 1.8143144056965492,
+      "grad_norm": 0.9004719853401184,
+      "learning_rate": 1.3130121584568055e-06,
+      "loss": 0.5388,
+      "step": 4968
+    },
+    {
+      "epoch": 1.8146795691071755,
+      "grad_norm": 0.777157723903656,
+      "learning_rate": 1.3079049451563331e-06,
+      "loss": 0.5369,
+      "step": 4969
+    },
+    {
+      "epoch": 1.8150447325178019,
+      "grad_norm": 0.816108763217926,
+      "learning_rate": 1.3028073481887016e-06,
+      "loss": 0.5091,
+      "step": 4970
+    },
+    {
+      "epoch": 1.815409895928428,
+      "grad_norm": 1.2835263013839722,
+      "learning_rate": 1.2977193701764135e-06,
+      "loss": 0.515,
+      "step": 4971
+    },
+    {
+      "epoch": 1.8157750593390543,
+      "grad_norm": 0.7208719253540039,
+      "learning_rate": 1.2926410137370348e-06,
+      "loss": 0.5374,
+      "step": 4972
+    },
+    {
+      "epoch": 1.8161402227496803,
+      "grad_norm": 0.7209171652793884,
+      "learning_rate": 1.2875722814831737e-06,
+      "loss": 0.554,
+      "step": 4973
+    },
+    {
+      "epoch": 1.8165053861603067,
+      "grad_norm": 0.847134530544281,
+      "learning_rate": 1.2825131760224952e-06,
+      "loss": 0.5379,
+      "step": 4974
+    },
+    {
+      "epoch": 1.816870549570933,
+      "grad_norm": 0.874386191368103,
+      "learning_rate": 1.2774636999576995e-06,
+      "loss": 0.5429,
+      "step": 4975
+    },
+    {
+      "epoch": 1.8172357129815593,
+      "grad_norm": 0.86895352602005,
+      "learning_rate": 1.272423855886542e-06,
+      "loss": 0.5114,
+      "step": 4976
+    },
+    {
+      "epoch": 1.8176008763921856,
+      "grad_norm": 0.8646639585494995,
+      "learning_rate": 1.26739364640182e-06,
+      "loss": 0.5099,
+      "step": 4977
+    },
+    {
+      "epoch": 1.817966039802812,
+      "grad_norm": 0.9552761912345886,
+      "learning_rate": 1.262373074091372e-06,
+      "loss": 0.477,
+      "step": 4978
+    },
+    {
+      "epoch": 1.818331203213438,
+      "grad_norm": 0.8524230718612671,
+      "learning_rate": 1.2573621415380832e-06,
+      "loss": 0.5237,
+      "step": 4979
+    },
+    {
+      "epoch": 1.8186963666240643,
+      "grad_norm": 0.9472958445549011,
+      "learning_rate": 1.2523608513198803e-06,
+      "loss": 0.492,
+      "step": 4980
+    },
+    {
+      "epoch": 1.8190615300346904,
+      "grad_norm": 1.036932349205017,
+      "learning_rate": 1.247369206009721e-06,
+      "loss": 0.5062,
+      "step": 4981
+    },
+    {
+      "epoch": 1.8194266934453167,
+      "grad_norm": 1.1887929439544678,
+      "learning_rate": 1.2423872081756106e-06,
+      "loss": 0.5118,
+      "step": 4982
+    },
+    {
+      "epoch": 1.819791856855943,
+      "grad_norm": 1.0288493633270264,
+      "learning_rate": 1.2374148603805835e-06,
+      "loss": 0.5062,
+      "step": 4983
+    },
+    {
+      "epoch": 1.8201570202665693,
+      "grad_norm": 2.3531882762908936,
+      "learning_rate": 1.2324521651827182e-06,
+      "loss": 0.5302,
+      "step": 4984
+    },
+    {
+      "epoch": 1.8205221836771956,
+      "grad_norm": 0.8009319305419922,
+      "learning_rate": 1.2274991251351166e-06,
+      "loss": 0.5413,
+      "step": 4985
+    },
+    {
+      "epoch": 1.820887347087822,
+      "grad_norm": 0.6363532543182373,
+      "learning_rate": 1.2225557427859203e-06,
+      "loss": 0.5743,
+      "step": 4986
+    },
+    {
+      "epoch": 1.821252510498448,
+      "grad_norm": 0.9401905536651611,
+      "learning_rate": 1.217622020678304e-06,
+      "loss": 0.5066,
+      "step": 4987
+    },
+    {
+      "epoch": 1.8216176739090744,
+      "grad_norm": 0.9579744935035706,
+      "learning_rate": 1.2126979613504664e-06,
+      "loss": 0.5391,
+      "step": 4988
+    },
+    {
+      "epoch": 1.8219828373197005,
+      "grad_norm": 0.7629438042640686,
+      "learning_rate": 1.2077835673356454e-06,
+      "loss": 0.5311,
+      "step": 4989
+    },
+    {
+      "epoch": 1.8223480007303268,
+      "grad_norm": 0.8086284399032593,
+      "learning_rate": 1.202878841162094e-06,
+      "loss": 0.5348,
+      "step": 4990
+    },
+    {
+      "epoch": 1.822713164140953,
+      "grad_norm": 1.3739471435546875,
+      "learning_rate": 1.197983785353094e-06,
+      "loss": 0.5128,
+      "step": 4991
+    },
+    {
+      "epoch": 1.8230783275515794,
+      "grad_norm": 0.6952402591705322,
+      "learning_rate": 1.1930984024269575e-06,
+      "loss": 0.5634,
+      "step": 4992
+    },
+    {
+      "epoch": 1.8234434909622057,
+      "grad_norm": 0.8957251906394958,
+      "learning_rate": 1.1882226948970188e-06,
+      "loss": 0.531,
+      "step": 4993
+    },
+    {
+      "epoch": 1.8238086543728318,
+      "grad_norm": 1.115297555923462,
+      "learning_rate": 1.1833566652716378e-06,
+      "loss": 0.506,
+      "step": 4994
+    },
+    {
+      "epoch": 1.824173817783458,
+      "grad_norm": 0.8088821172714233,
+      "learning_rate": 1.1785003160541852e-06,
+      "loss": 0.5461,
+      "step": 4995
+    },
+    {
+      "epoch": 1.8245389811940842,
+      "grad_norm": 0.9462229013442993,
+      "learning_rate": 1.1736536497430584e-06,
+      "loss": 0.5522,
+      "step": 4996
+    },
+    {
+      "epoch": 1.8249041446047105,
+      "grad_norm": 1.1374109983444214,
+      "learning_rate": 1.168816668831676e-06,
+      "loss": 0.5076,
+      "step": 4997
+    },
+    {
+      "epoch": 1.8252693080153368,
+      "grad_norm": 1.011764407157898,
+      "learning_rate": 1.1639893758084719e-06,
+      "loss": 0.5021,
+      "step": 4998
+    },
+    {
+      "epoch": 1.8256344714259631,
+      "grad_norm": 0.6716031432151794,
+      "learning_rate": 1.1591717731568909e-06,
+      "loss": 0.5431,
+      "step": 4999
+    },
+    {
+      "epoch": 1.8259996348365894,
+      "grad_norm": 1.1675262451171875,
+      "learning_rate": 1.1543638633553945e-06,
+      "loss": 0.5005,
+      "step": 5000
+    },
+    {
+      "epoch": 1.8263647982472158,
+      "grad_norm": 1.0584003925323486,
+      "learning_rate": 1.149565648877462e-06,
+      "loss": 0.54,
+      "step": 5001
+    },
+    {
+      "epoch": 1.8267299616578418,
+      "grad_norm": 0.7157919406890869,
+      "learning_rate": 1.144777132191588e-06,
+      "loss": 0.5447,
+      "step": 5002
+    },
+    {
+      "epoch": 1.8270951250684682,
+      "grad_norm": 1.2633295059204102,
+      "learning_rate": 1.1399983157612616e-06,
+      "loss": 0.5209,
+      "step": 5003
+    },
+    {
+      "epoch": 1.8274602884790943,
+      "grad_norm": 0.8516579270362854,
+      "learning_rate": 1.1352292020449984e-06,
+      "loss": 0.5413,
+      "step": 5004
+    },
+    {
+      "epoch": 1.8278254518897206,
+      "grad_norm": 0.7476287484169006,
+      "learning_rate": 1.130469793496314e-06,
+      "loss": 0.5039,
+      "step": 5005
+    },
+    {
+      "epoch": 1.8281906153003469,
+      "grad_norm": 0.7512426376342773,
+      "learning_rate": 1.1257200925637336e-06,
+      "loss": 0.5333,
+      "step": 5006
+    },
+    {
+      "epoch": 1.8285557787109732,
+      "grad_norm": 0.8339123725891113,
+      "learning_rate": 1.1209801016907872e-06,
+      "loss": 0.5235,
+      "step": 5007
+    },
+    {
+      "epoch": 1.8289209421215995,
+      "grad_norm": 0.8975398540496826,
+      "learning_rate": 1.1162498233160136e-06,
+      "loss": 0.5236,
+      "step": 5008
+    },
+    {
+      "epoch": 1.8292861055322258,
+      "grad_norm": 0.7915427684783936,
+      "learning_rate": 1.1115292598729454e-06,
+      "loss": 0.5349,
+      "step": 5009
+    },
+    {
+      "epoch": 1.829651268942852,
+      "grad_norm": 0.797041118144989,
+      "learning_rate": 1.106818413790125e-06,
+      "loss": 0.5127,
+      "step": 5010
+    },
+    {
+      "epoch": 1.8300164323534782,
+      "grad_norm": 0.9191210269927979,
+      "learning_rate": 1.1021172874910957e-06,
+      "loss": 0.5061,
+      "step": 5011
+    },
+    {
+      "epoch": 1.8303815957641043,
+      "grad_norm": 0.8843085765838623,
+      "learning_rate": 1.0974258833943985e-06,
+      "loss": 0.5046,
+      "step": 5012
+    },
+    {
+      "epoch": 1.8307467591747306,
+      "grad_norm": 0.8699705600738525,
+      "learning_rate": 1.0927442039135717e-06,
+      "loss": 0.517,
+      "step": 5013
+    },
+    {
+      "epoch": 1.831111922585357,
+      "grad_norm": 0.602509081363678,
+      "learning_rate": 1.0880722514571484e-06,
+      "loss": 0.5811,
+      "step": 5014
+    },
+    {
+      "epoch": 1.8314770859959832,
+      "grad_norm": 0.8212730884552002,
+      "learning_rate": 1.0834100284286641e-06,
+      "loss": 0.5353,
+      "step": 5015
+    },
+    {
+      "epoch": 1.8318422494066096,
+      "grad_norm": 1.0749436616897583,
+      "learning_rate": 1.0787575372266467e-06,
+      "loss": 0.4735,
+      "step": 5016
+    },
+    {
+      "epoch": 1.8322074128172359,
+      "grad_norm": 0.844359815120697,
+      "learning_rate": 1.0741147802446128e-06,
+      "loss": 0.5265,
+      "step": 5017
+    },
+    {
+      "epoch": 1.832572576227862,
+      "grad_norm": 0.7892604470252991,
+      "learning_rate": 1.0694817598710782e-06,
+      "loss": 0.5421,
+      "step": 5018
+    },
+    {
+      "epoch": 1.8329377396384883,
+      "grad_norm": 0.9948855638504028,
+      "learning_rate": 1.0648584784895411e-06,
+      "loss": 0.4838,
+      "step": 5019
+    },
+    {
+      "epoch": 1.8333029030491144,
+      "grad_norm": 0.8588843941688538,
+      "learning_rate": 1.0602449384784963e-06,
+      "loss": 0.5593,
+      "step": 5020
+    },
+    {
+      "epoch": 1.8336680664597407,
+      "grad_norm": 0.8757759928703308,
+      "learning_rate": 1.0556411422114254e-06,
+      "loss": 0.5088,
+      "step": 5021
+    },
+    {
+      "epoch": 1.834033229870367,
+      "grad_norm": 0.8504748344421387,
+      "learning_rate": 1.0510470920567983e-06,
+      "loss": 0.5199,
+      "step": 5022
+    },
+    {
+      "epoch": 1.8343983932809933,
+      "grad_norm": 0.8831689953804016,
+      "learning_rate": 1.0464627903780689e-06,
+      "loss": 0.5128,
+      "step": 5023
+    },
+    {
+      "epoch": 1.8347635566916196,
+      "grad_norm": 1.101543664932251,
+      "learning_rate": 1.041888239533675e-06,
+      "loss": 0.5398,
+      "step": 5024
+    },
+    {
+      "epoch": 1.8351287201022457,
+      "grad_norm": 1.0294517278671265,
+      "learning_rate": 1.0373234418770385e-06,
+      "loss": 0.4874,
+      "step": 5025
+    },
+    {
+      "epoch": 1.835493883512872,
+      "grad_norm": 0.9753843545913696,
+      "learning_rate": 1.0327683997565674e-06,
+      "loss": 0.5183,
+      "step": 5026
+    },
+    {
+      "epoch": 1.835859046923498,
+      "grad_norm": 0.6896717548370361,
+      "learning_rate": 1.0282231155156498e-06,
+      "loss": 0.5707,
+      "step": 5027
+    },
+    {
+      "epoch": 1.8362242103341244,
+      "grad_norm": 0.9963521957397461,
+      "learning_rate": 1.0236875914926458e-06,
+      "loss": 0.5245,
+      "step": 5028
+    },
+    {
+      "epoch": 1.8365893737447507,
+      "grad_norm": 0.898601770401001,
+      "learning_rate": 1.0191618300209094e-06,
+      "loss": 0.4901,
+      "step": 5029
+    },
+    {
+      "epoch": 1.836954537155377,
+      "grad_norm": 0.8382347822189331,
+      "learning_rate": 1.0146458334287513e-06,
+      "loss": 0.5373,
+      "step": 5030
+    },
+    {
+      "epoch": 1.8373197005660034,
+      "grad_norm": 1.4509795904159546,
+      "learning_rate": 1.0101396040394795e-06,
+      "loss": 0.5228,
+      "step": 5031
+    },
+    {
+      "epoch": 1.8376848639766297,
+      "grad_norm": 0.8937653303146362,
+      "learning_rate": 1.0056431441713643e-06,
+      "loss": 0.5381,
+      "step": 5032
+    },
+    {
+      "epoch": 1.8380500273872558,
+      "grad_norm": 1.0046682357788086,
+      "learning_rate": 1.0011564561376596e-06,
+      "loss": 0.4879,
+      "step": 5033
+    },
+    {
+      "epoch": 1.838415190797882,
+      "grad_norm": 1.0494298934936523,
+      "learning_rate": 9.966795422465792e-07,
+      "loss": 0.5415,
+      "step": 5034
+    },
+    {
+      "epoch": 1.8387803542085082,
+      "grad_norm": 0.838736355304718,
+      "learning_rate": 9.922124048013183e-07,
+      "loss": 0.5259,
+      "step": 5035
+    },
+    {
+      "epoch": 1.8391455176191345,
+      "grad_norm": 1.297999382019043,
+      "learning_rate": 9.877550461000385e-07,
+      "loss": 0.5302,
+      "step": 5036
+    },
+    {
+      "epoch": 1.8395106810297608,
+      "grad_norm": 0.5948071479797363,
+      "learning_rate": 9.833074684358768e-07,
+      "loss": 0.5594,
+      "step": 5037
+    },
+    {
+      "epoch": 1.839875844440387,
+      "grad_norm": 0.86806321144104,
+      "learning_rate": 9.788696740969295e-07,
+      "loss": 0.5304,
+      "step": 5038
+    },
+    {
+      "epoch": 1.8402410078510134,
+      "grad_norm": 1.3624813556671143,
+      "learning_rate": 9.744416653662636e-07,
+      "loss": 0.5087,
+      "step": 5039
+    },
+    {
+      "epoch": 1.8406061712616397,
+      "grad_norm": 1.0623867511749268,
+      "learning_rate": 9.700234445219126e-07,
+      "loss": 0.5437,
+      "step": 5040
+    },
+    {
+      "epoch": 1.8409713346722658,
+      "grad_norm": 0.8289335370063782,
+      "learning_rate": 9.656150138368758e-07,
+      "loss": 0.5636,
+      "step": 5041
+    },
+    {
+      "epoch": 1.8413364980828921,
+      "grad_norm": 0.932661235332489,
+      "learning_rate": 9.612163755791105e-07,
+      "loss": 0.5368,
+      "step": 5042
+    },
+    {
+      "epoch": 1.8417016614935182,
+      "grad_norm": 1.0001920461654663,
+      "learning_rate": 9.568275320115438e-07,
+      "loss": 0.5366,
+      "step": 5043
+    },
+    {
+      "epoch": 1.8420668249041445,
+      "grad_norm": 0.7952094674110413,
+      "learning_rate": 9.524484853920524e-07,
+      "loss": 0.5441,
+      "step": 5044
+    },
+    {
+      "epoch": 1.8424319883147708,
+      "grad_norm": 0.7896952629089355,
+      "learning_rate": 9.480792379734871e-07,
+      "loss": 0.5292,
+      "step": 5045
+    },
+    {
+      "epoch": 1.8427971517253972,
+      "grad_norm": 0.8495169878005981,
+      "learning_rate": 9.437197920036456e-07,
+      "loss": 0.4935,
+      "step": 5046
+    },
+    {
+      "epoch": 1.8431623151360235,
+      "grad_norm": 0.7229299545288086,
+      "learning_rate": 9.393701497252939e-07,
+      "loss": 0.5603,
+      "step": 5047
+    },
+    {
+      "epoch": 1.8435274785466498,
+      "grad_norm": 0.7918698191642761,
+      "learning_rate": 9.35030313376144e-07,
+      "loss": 0.5757,
+      "step": 5048
+    },
+    {
+      "epoch": 1.8438926419572759,
+      "grad_norm": 0.7822374701499939,
+      "learning_rate": 9.307002851888658e-07,
+      "loss": 0.5241,
+      "step": 5049
+    },
+    {
+      "epoch": 1.8442578053679022,
+      "grad_norm": 0.8543692231178284,
+      "learning_rate": 9.263800673910883e-07,
+      "loss": 0.5047,
+      "step": 5050
+    },
+    {
+      "epoch": 1.8446229687785283,
+      "grad_norm": 0.9035658836364746,
+      "learning_rate": 9.220696622053915e-07,
+      "loss": 0.5197,
+      "step": 5051
+    },
+    {
+      "epoch": 1.8449881321891546,
+      "grad_norm": 0.8677619695663452,
+      "learning_rate": 9.177690718493016e-07,
+      "loss": 0.5373,
+      "step": 5052
+    },
+    {
+      "epoch": 1.845353295599781,
+      "grad_norm": 0.9196531772613525,
+      "learning_rate": 9.134782985353019e-07,
+      "loss": 0.4809,
+      "step": 5053
+    },
+    {
+      "epoch": 1.8457184590104072,
+      "grad_norm": 0.7895047068595886,
+      "learning_rate": 9.091973444708247e-07,
+      "loss": 0.5539,
+      "step": 5054
+    },
+    {
+      "epoch": 1.8460836224210335,
+      "grad_norm": 0.9806033372879028,
+      "learning_rate": 9.049262118582458e-07,
+      "loss": 0.4776,
+      "step": 5055
+    },
+    {
+      "epoch": 1.8464487858316598,
+      "grad_norm": 0.6609007716178894,
+      "learning_rate": 9.006649028948966e-07,
+      "loss": 0.5833,
+      "step": 5056
+    },
+    {
+      "epoch": 1.846813949242286,
+      "grad_norm": 1.074385643005371,
+      "learning_rate": 8.964134197730457e-07,
+      "loss": 0.5267,
+      "step": 5057
+    },
+    {
+      "epoch": 1.847179112652912,
+      "grad_norm": 0.837611734867096,
+      "learning_rate": 8.921717646799077e-07,
+      "loss": 0.5246,
+      "step": 5058
+    },
+    {
+      "epoch": 1.8475442760635383,
+      "grad_norm": 0.9103251099586487,
+      "learning_rate": 8.879399397976484e-07,
+      "loss": 0.5136,
+      "step": 5059
+    },
+    {
+      "epoch": 1.8479094394741646,
+      "grad_norm": 0.8506982922554016,
+      "learning_rate": 8.83717947303373e-07,
+      "loss": 0.5287,
+      "step": 5060
+    },
+    {
+      "epoch": 1.848274602884791,
+      "grad_norm": 0.9154410362243652,
+      "learning_rate": 8.795057893691239e-07,
+      "loss": 0.494,
+      "step": 5061
+    },
+    {
+      "epoch": 1.8486397662954173,
+      "grad_norm": 0.8932507634162903,
+      "learning_rate": 8.753034681618877e-07,
+      "loss": 0.5208,
+      "step": 5062
+    },
+    {
+      "epoch": 1.8490049297060436,
+      "grad_norm": 0.7752853631973267,
+      "learning_rate": 8.711109858435907e-07,
+      "loss": 0.5314,
+      "step": 5063
+    },
+    {
+      "epoch": 1.8493700931166697,
+      "grad_norm": 0.837082028388977,
+      "learning_rate": 8.669283445710985e-07,
+      "loss": 0.5121,
+      "step": 5064
+    },
+    {
+      "epoch": 1.849735256527296,
+      "grad_norm": 0.6760733723640442,
+      "learning_rate": 8.627555464962078e-07,
+      "loss": 0.5282,
+      "step": 5065
+    },
+    {
+      "epoch": 1.850100419937922,
+      "grad_norm": 0.8442636728286743,
+      "learning_rate": 8.585925937656636e-07,
+      "loss": 0.5213,
+      "step": 5066
+    },
+    {
+      "epoch": 1.8504655833485484,
+      "grad_norm": 0.9905248880386353,
+      "learning_rate": 8.544394885211305e-07,
+      "loss": 0.5267,
+      "step": 5067
+    },
+    {
+      "epoch": 1.8508307467591747,
+      "grad_norm": 0.6257760524749756,
+      "learning_rate": 8.502962328992149e-07,
+      "loss": 0.5668,
+      "step": 5068
+    },
+    {
+      "epoch": 1.851195910169801,
+      "grad_norm": 0.9485304355621338,
+      "learning_rate": 8.461628290314605e-07,
+      "loss": 0.5191,
+      "step": 5069
+    },
+    {
+      "epoch": 1.8515610735804273,
+      "grad_norm": 0.8224905729293823,
+      "learning_rate": 8.420392790443332e-07,
+      "loss": 0.5058,
+      "step": 5070
+    },
+    {
+      "epoch": 1.8519262369910536,
+      "grad_norm": 1.0173250436782837,
+      "learning_rate": 8.379255850592404e-07,
+      "loss": 0.5284,
+      "step": 5071
+    },
+    {
+      "epoch": 1.8522914004016797,
+      "grad_norm": 0.8787329792976379,
+      "learning_rate": 8.338217491925027e-07,
+      "loss": 0.5308,
+      "step": 5072
+    },
+    {
+      "epoch": 1.852656563812306,
+      "grad_norm": 1.0824297666549683,
+      "learning_rate": 8.297277735553844e-07,
+      "loss": 0.5091,
+      "step": 5073
+    },
+    {
+      "epoch": 1.8530217272229321,
+      "grad_norm": 0.9079974889755249,
+      "learning_rate": 8.256436602540718e-07,
+      "loss": 0.5069,
+      "step": 5074
+    },
+    {
+      "epoch": 1.8533868906335584,
+      "grad_norm": 0.6514511108398438,
+      "learning_rate": 8.215694113896777e-07,
+      "loss": 0.5233,
+      "step": 5075
+    },
+    {
+      "epoch": 1.8537520540441847,
+      "grad_norm": 0.8281962275505066,
+      "learning_rate": 8.17505029058241e-07,
+      "loss": 0.4781,
+      "step": 5076
+    },
+    {
+      "epoch": 1.854117217454811,
+      "grad_norm": 0.8062742352485657,
+      "learning_rate": 8.134505153507177e-07,
+      "loss": 0.4791,
+      "step": 5077
+    },
+    {
+      "epoch": 1.8544823808654374,
+      "grad_norm": 0.782093346118927,
+      "learning_rate": 8.094058723529974e-07,
+      "loss": 0.5082,
+      "step": 5078
+    },
+    {
+      "epoch": 1.8548475442760637,
+      "grad_norm": 0.9898812770843506,
+      "learning_rate": 8.053711021458843e-07,
+      "loss": 0.5347,
+      "step": 5079
+    },
+    {
+      "epoch": 1.8552127076866898,
+      "grad_norm": 0.9271273612976074,
+      "learning_rate": 8.013462068051092e-07,
+      "loss": 0.5262,
+      "step": 5080
+    },
+    {
+      "epoch": 1.855577871097316,
+      "grad_norm": 0.9929766058921814,
+      "learning_rate": 7.973311884013158e-07,
+      "loss": 0.4949,
+      "step": 5081
+    },
+    {
+      "epoch": 1.8559430345079422,
+      "grad_norm": 0.8740081787109375,
+      "learning_rate": 7.933260490000694e-07,
+      "loss": 0.5351,
+      "step": 5082
+    },
+    {
+      "epoch": 1.8563081979185685,
+      "grad_norm": 1.054709553718567,
+      "learning_rate": 7.893307906618575e-07,
+      "loss": 0.5242,
+      "step": 5083
+    },
+    {
+      "epoch": 1.8566733613291948,
+      "grad_norm": 1.1594157218933105,
+      "learning_rate": 7.853454154420758e-07,
+      "loss": 0.4914,
+      "step": 5084
+    },
+    {
+      "epoch": 1.8570385247398211,
+      "grad_norm": 0.7112655639648438,
+      "learning_rate": 7.813699253910423e-07,
+      "loss": 0.5694,
+      "step": 5085
+    },
+    {
+      "epoch": 1.8574036881504474,
+      "grad_norm": 0.8234994411468506,
+      "learning_rate": 7.774043225539874e-07,
+      "loss": 0.5347,
+      "step": 5086
+    },
+    {
+      "epoch": 1.8577688515610737,
+      "grad_norm": 1.0197539329528809,
+      "learning_rate": 7.734486089710502e-07,
+      "loss": 0.5295,
+      "step": 5087
+    },
+    {
+      "epoch": 1.8581340149716998,
+      "grad_norm": 1.031860113143921,
+      "learning_rate": 7.695027866772919e-07,
+      "loss": 0.5211,
+      "step": 5088
+    },
+    {
+      "epoch": 1.8584991783823261,
+      "grad_norm": 0.9370394349098206,
+      "learning_rate": 7.655668577026798e-07,
+      "loss": 0.4739,
+      "step": 5089
+    },
+    {
+      "epoch": 1.8588643417929522,
+      "grad_norm": 0.6947592496871948,
+      "learning_rate": 7.616408240720896e-07,
+      "loss": 0.5311,
+      "step": 5090
+    },
+    {
+      "epoch": 1.8592295052035785,
+      "grad_norm": 0.9864943027496338,
+      "learning_rate": 7.577246878053057e-07,
+      "loss": 0.5118,
+      "step": 5091
+    },
+    {
+      "epoch": 1.8595946686142049,
+      "grad_norm": 0.8007437586784363,
+      "learning_rate": 7.538184509170276e-07,
+      "loss": 0.5092,
+      "step": 5092
+    },
+    {
+      "epoch": 1.8599598320248312,
+      "grad_norm": 0.9444168210029602,
+      "learning_rate": 7.499221154168545e-07,
+      "loss": 0.5236,
+      "step": 5093
+    },
+    {
+      "epoch": 1.8603249954354575,
+      "grad_norm": 0.8404966592788696,
+      "learning_rate": 7.460356833092963e-07,
+      "loss": 0.5286,
+      "step": 5094
+    },
+    {
+      "epoch": 1.8606901588460836,
+      "grad_norm": 0.7893446683883667,
+      "learning_rate": 7.421591565937647e-07,
+      "loss": 0.5523,
+      "step": 5095
+    },
+    {
+      "epoch": 1.8610553222567099,
+      "grad_norm": 0.8035286664962769,
+      "learning_rate": 7.3829253726458e-07,
+      "loss": 0.5595,
+      "step": 5096
+    },
+    {
+      "epoch": 1.861420485667336,
+      "grad_norm": 0.709818422794342,
+      "learning_rate": 7.344358273109575e-07,
+      "loss": 0.5464,
+      "step": 5097
+    },
+    {
+      "epoch": 1.8617856490779623,
+      "grad_norm": 0.8071439862251282,
+      "learning_rate": 7.305890287170236e-07,
+      "loss": 0.5075,
+      "step": 5098
+    },
+    {
+      "epoch": 1.8621508124885886,
+      "grad_norm": 1.1296306848526,
+      "learning_rate": 7.267521434618018e-07,
+      "loss": 0.5022,
+      "step": 5099
+    },
+    {
+      "epoch": 1.862515975899215,
+      "grad_norm": 1.1329855918884277,
+      "learning_rate": 7.229251735192178e-07,
+      "loss": 0.5441,
+      "step": 5100
+    },
+    {
+      "epoch": 1.8628811393098412,
+      "grad_norm": 1.3989837169647217,
+      "learning_rate": 7.191081208580874e-07,
+      "loss": 0.498,
+      "step": 5101
+    },
+    {
+      "epoch": 1.8632463027204675,
+      "grad_norm": 1.5248597860336304,
+      "learning_rate": 7.153009874421357e-07,
+      "loss": 0.4777,
+      "step": 5102
+    },
+    {
+      "epoch": 1.8636114661310936,
+      "grad_norm": 0.9464701414108276,
+      "learning_rate": 7.11503775229978e-07,
+      "loss": 0.5377,
+      "step": 5103
+    },
+    {
+      "epoch": 1.86397662954172,
+      "grad_norm": 0.874332845211029,
+      "learning_rate": 7.077164861751318e-07,
+      "loss": 0.4953,
+      "step": 5104
+    },
+    {
+      "epoch": 1.864341792952346,
+      "grad_norm": 0.9725618362426758,
+      "learning_rate": 7.039391222260005e-07,
+      "loss": 0.5128,
+      "step": 5105
+    },
+    {
+      "epoch": 1.8647069563629723,
+      "grad_norm": 0.7084645628929138,
+      "learning_rate": 7.001716853258877e-07,
+      "loss": 0.5797,
+      "step": 5106
+    },
+    {
+      "epoch": 1.8650721197735987,
+      "grad_norm": 0.9991376399993896,
+      "learning_rate": 6.964141774129873e-07,
+      "loss": 0.53,
+      "step": 5107
+    },
+    {
+      "epoch": 1.865437283184225,
+      "grad_norm": 1.4431263208389282,
+      "learning_rate": 6.926666004203908e-07,
+      "loss": 0.5262,
+      "step": 5108
+    },
+    {
+      "epoch": 1.8658024465948513,
+      "grad_norm": 2.061160087585449,
+      "learning_rate": 6.889289562760738e-07,
+      "loss": 0.551,
+      "step": 5109
+    },
+    {
+      "epoch": 1.8661676100054776,
+      "grad_norm": 0.9618075489997864,
+      "learning_rate": 6.852012469029046e-07,
+      "loss": 0.5092,
+      "step": 5110
+    },
+    {
+      "epoch": 1.8665327734161037,
+      "grad_norm": 1.071441888809204,
+      "learning_rate": 6.814834742186361e-07,
+      "loss": 0.5159,
+      "step": 5111
+    },
+    {
+      "epoch": 1.86689793682673,
+      "grad_norm": 1.1133548021316528,
+      "learning_rate": 6.777756401359159e-07,
+      "loss": 0.5017,
+      "step": 5112
+    },
+    {
+      "epoch": 1.867263100237356,
+      "grad_norm": 1.0975421667099,
+      "learning_rate": 6.740777465622784e-07,
+      "loss": 0.5208,
+      "step": 5113
+    },
+    {
+      "epoch": 1.8676282636479824,
+      "grad_norm": 0.9239022135734558,
+      "learning_rate": 6.703897954001392e-07,
+      "loss": 0.5184,
+      "step": 5114
+    },
+    {
+      "epoch": 1.8679934270586087,
+      "grad_norm": 0.9000834226608276,
+      "learning_rate": 6.667117885468011e-07,
+      "loss": 0.5036,
+      "step": 5115
+    },
+    {
+      "epoch": 1.868358590469235,
+      "grad_norm": 0.8452208638191223,
+      "learning_rate": 6.630437278944501e-07,
+      "loss": 0.5719,
+      "step": 5116
+    },
+    {
+      "epoch": 1.8687237538798613,
+      "grad_norm": 1.1362179517745972,
+      "learning_rate": 6.59385615330157e-07,
+      "loss": 0.5406,
+      "step": 5117
+    },
+    {
+      "epoch": 1.8690889172904877,
+      "grad_norm": 0.7628581523895264,
+      "learning_rate": 6.557374527358762e-07,
+      "loss": 0.5428,
+      "step": 5118
+    },
+    {
+      "epoch": 1.8694540807011137,
+      "grad_norm": 0.836574912071228,
+      "learning_rate": 6.520992419884398e-07,
+      "loss": 0.551,
+      "step": 5119
+    },
+    {
+      "epoch": 1.86981924411174,
+      "grad_norm": 1.2260987758636475,
+      "learning_rate": 6.484709849595572e-07,
+      "loss": 0.4869,
+      "step": 5120
+    },
+    {
+      "epoch": 1.8701844075223661,
+      "grad_norm": 0.948838472366333,
+      "learning_rate": 6.448526835158264e-07,
+      "loss": 0.5066,
+      "step": 5121
+    },
+    {
+      "epoch": 1.8705495709329925,
+      "grad_norm": 0.7823359966278076,
+      "learning_rate": 6.41244339518714e-07,
+      "loss": 0.5254,
+      "step": 5122
+    },
+    {
+      "epoch": 1.8709147343436188,
+      "grad_norm": 1.5459065437316895,
+      "learning_rate": 6.37645954824575e-07,
+      "loss": 0.5058,
+      "step": 5123
+    },
+    {
+      "epoch": 1.871279897754245,
+      "grad_norm": 1.0922521352767944,
+      "learning_rate": 6.340575312846287e-07,
+      "loss": 0.5208,
+      "step": 5124
+    },
+    {
+      "epoch": 1.8716450611648714,
+      "grad_norm": 0.7133572697639465,
+      "learning_rate": 6.304790707449738e-07,
+      "loss": 0.5318,
+      "step": 5125
+    },
+    {
+      "epoch": 1.8720102245754975,
+      "grad_norm": 0.7148061990737915,
+      "learning_rate": 6.269105750465843e-07,
+      "loss": 0.5628,
+      "step": 5126
+    },
+    {
+      "epoch": 1.8723753879861238,
+      "grad_norm": 1.031386137008667,
+      "learning_rate": 6.233520460253117e-07,
+      "loss": 0.5663,
+      "step": 5127
+    },
+    {
+      "epoch": 1.8727405513967499,
+      "grad_norm": 0.8790083527565002,
+      "learning_rate": 6.198034855118784e-07,
+      "loss": 0.5196,
+      "step": 5128
+    },
+    {
+      "epoch": 1.8731057148073762,
+      "grad_norm": 0.8304159641265869,
+      "learning_rate": 6.162648953318684e-07,
+      "loss": 0.5443,
+      "step": 5129
+    },
+    {
+      "epoch": 1.8734708782180025,
+      "grad_norm": 1.0308268070220947,
+      "learning_rate": 6.1273627730575e-07,
+      "loss": 0.5205,
+      "step": 5130
+    },
+    {
+      "epoch": 1.8738360416286288,
+      "grad_norm": 0.9887058734893799,
+      "learning_rate": 6.092176332488553e-07,
+      "loss": 0.5578,
+      "step": 5131
+    },
+    {
+      "epoch": 1.8742012050392551,
+      "grad_norm": 0.7946871519088745,
+      "learning_rate": 6.057089649713832e-07,
+      "loss": 0.5652,
+      "step": 5132
+    },
+    {
+      "epoch": 1.8745663684498814,
+      "grad_norm": 1.2616994380950928,
+      "learning_rate": 6.022102742784075e-07,
+      "loss": 0.5202,
+      "step": 5133
+    },
+    {
+      "epoch": 1.8749315318605075,
+      "grad_norm": 1.0364519357681274,
+      "learning_rate": 5.987215629698595e-07,
+      "loss": 0.4841,
+      "step": 5134
+    },
+    {
+      "epoch": 1.8752966952711339,
+      "grad_norm": 0.867057740688324,
+      "learning_rate": 5.952428328405413e-07,
+      "loss": 0.5433,
+      "step": 5135
+    },
+    {
+      "epoch": 1.87566185868176,
+      "grad_norm": 0.9206724166870117,
+      "learning_rate": 5.917740856801235e-07,
+      "loss": 0.4821,
+      "step": 5136
+    },
+    {
+      "epoch": 1.8760270220923863,
+      "grad_norm": 0.737504243850708,
+      "learning_rate": 5.88315323273132e-07,
+      "loss": 0.5388,
+      "step": 5137
+    },
+    {
+      "epoch": 1.8763921855030126,
+      "grad_norm": 0.9644716382026672,
+      "learning_rate": 5.848665473989679e-07,
+      "loss": 0.5231,
+      "step": 5138
+    },
+    {
+      "epoch": 1.8767573489136389,
+      "grad_norm": 0.8571596741676331,
+      "learning_rate": 5.814277598318808e-07,
+      "loss": 0.5346,
+      "step": 5139
+    },
+    {
+      "epoch": 1.8771225123242652,
+      "grad_norm": 0.8317295908927917,
+      "learning_rate": 5.779989623409932e-07,
+      "loss": 0.5537,
+      "step": 5140
+    },
+    {
+      "epoch": 1.8774876757348915,
+      "grad_norm": 0.9167125225067139,
+      "learning_rate": 5.745801566902831e-07,
+      "loss": 0.5292,
+      "step": 5141
+    },
+    {
+      "epoch": 1.8778528391455176,
+      "grad_norm": 1.2240817546844482,
+      "learning_rate": 5.71171344638588e-07,
+      "loss": 0.4914,
+      "step": 5142
+    },
+    {
+      "epoch": 1.878218002556144,
+      "grad_norm": 0.9029586911201477,
+      "learning_rate": 5.677725279396096e-07,
+      "loss": 0.5697,
+      "step": 5143
+    },
+    {
+      "epoch": 1.87858316596677,
+      "grad_norm": 0.9542390704154968,
+      "learning_rate": 5.643837083418957e-07,
+      "loss": 0.5768,
+      "step": 5144
+    },
+    {
+      "epoch": 1.8789483293773963,
+      "grad_norm": 0.8885607719421387,
+      "learning_rate": 5.610048875888607e-07,
+      "loss": 0.5492,
+      "step": 5145
+    },
+    {
+      "epoch": 1.8793134927880226,
+      "grad_norm": 1.0386658906936646,
+      "learning_rate": 5.57636067418772e-07,
+      "loss": 0.5406,
+      "step": 5146
+    },
+    {
+      "epoch": 1.879678656198649,
+      "grad_norm": 0.7747313380241394,
+      "learning_rate": 5.542772495647563e-07,
+      "loss": 0.5815,
+      "step": 5147
+    },
+    {
+      "epoch": 1.8800438196092752,
+      "grad_norm": 1.0311294794082642,
+      "learning_rate": 5.509284357547873e-07,
+      "loss": 0.4989,
+      "step": 5148
+    },
+    {
+      "epoch": 1.8804089830199016,
+      "grad_norm": 0.6967086791992188,
+      "learning_rate": 5.475896277116954e-07,
+      "loss": 0.57,
+      "step": 5149
+    },
+    {
+      "epoch": 1.8807741464305276,
+      "grad_norm": 0.9049924612045288,
+      "learning_rate": 5.442608271531602e-07,
+      "loss": 0.5278,
+      "step": 5150
+    },
+    {
+      "epoch": 1.881139309841154,
+      "grad_norm": 0.9503490328788757,
+      "learning_rate": 5.409420357917205e-07,
+      "loss": 0.5592,
+      "step": 5151
+    },
+    {
+      "epoch": 1.88150447325178,
+      "grad_norm": 0.7792781591415405,
+      "learning_rate": 5.376332553347618e-07,
+      "loss": 0.5201,
+      "step": 5152
+    },
+    {
+      "epoch": 1.8818696366624064,
+      "grad_norm": 0.9307024478912354,
+      "learning_rate": 5.34334487484518e-07,
+      "loss": 0.4548,
+      "step": 5153
+    },
+    {
+      "epoch": 1.8822348000730327,
+      "grad_norm": 0.7584879398345947,
+      "learning_rate": 5.310457339380693e-07,
+      "loss": 0.5837,
+      "step": 5154
+    },
+    {
+      "epoch": 1.882599963483659,
+      "grad_norm": 0.6833974123001099,
+      "learning_rate": 5.277669963873489e-07,
+      "loss": 0.579,
+      "step": 5155
+    },
+    {
+      "epoch": 1.8829651268942853,
+      "grad_norm": 0.7575476169586182,
+      "learning_rate": 5.244982765191387e-07,
+      "loss": 0.5337,
+      "step": 5156
+    },
+    {
+      "epoch": 1.8833302903049116,
+      "grad_norm": 1.2384319305419922,
+      "learning_rate": 5.212395760150623e-07,
+      "loss": 0.5113,
+      "step": 5157
+    },
+    {
+      "epoch": 1.8836954537155377,
+      "grad_norm": 0.811079204082489,
+      "learning_rate": 5.1799089655159e-07,
+      "loss": 0.556,
+      "step": 5158
+    },
+    {
+      "epoch": 1.8840606171261638,
+      "grad_norm": 0.8482411503791809,
+      "learning_rate": 5.14752239800036e-07,
+      "loss": 0.5125,
+      "step": 5159
+    },
+    {
+      "epoch": 1.88442578053679,
+      "grad_norm": 1.042563557624817,
+      "learning_rate": 5.115236074265606e-07,
+      "loss": 0.4683,
+      "step": 5160
+    },
+    {
+      "epoch": 1.8847909439474164,
+      "grad_norm": 1.1850470304489136,
+      "learning_rate": 5.083050010921642e-07,
+      "loss": 0.5096,
+      "step": 5161
+    },
+    {
+      "epoch": 1.8851561073580427,
+      "grad_norm": 0.8976897597312927,
+      "learning_rate": 5.050964224526956e-07,
+      "loss": 0.5477,
+      "step": 5162
+    },
+    {
+      "epoch": 1.885521270768669,
+      "grad_norm": 1.336876392364502,
+      "learning_rate": 5.018978731588342e-07,
+      "loss": 0.4924,
+      "step": 5163
+    },
+    {
+      "epoch": 1.8858864341792954,
+      "grad_norm": 0.8368218541145325,
+      "learning_rate": 4.987093548561062e-07,
+      "loss": 0.506,
+      "step": 5164
+    },
+    {
+      "epoch": 1.8862515975899214,
+      "grad_norm": 0.8556219935417175,
+      "learning_rate": 4.95530869184877e-07,
+      "loss": 0.5411,
+      "step": 5165
+    },
+    {
+      "epoch": 1.8866167610005478,
+      "grad_norm": 1.0150723457336426,
+      "learning_rate": 4.923624177803498e-07,
+      "loss": 0.5634,
+      "step": 5166
+    },
+    {
+      "epoch": 1.8869819244111738,
+      "grad_norm": 0.8867017030715942,
+      "learning_rate": 4.892040022725675e-07,
+      "loss": 0.5142,
+      "step": 5167
+    },
+    {
+      "epoch": 1.8873470878218002,
+      "grad_norm": 0.9332976937294006,
+      "learning_rate": 4.860556242864034e-07,
+      "loss": 0.5117,
+      "step": 5168
+    },
+    {
+      "epoch": 1.8877122512324265,
+      "grad_norm": 3.3063771724700928,
+      "learning_rate": 4.829172854415775e-07,
+      "loss": 0.5136,
+      "step": 5169
+    },
+    {
+      "epoch": 1.8880774146430528,
+      "grad_norm": 0.8885701298713684,
+      "learning_rate": 4.79788987352634e-07,
+      "loss": 0.5219,
+      "step": 5170
+    },
+    {
+      "epoch": 1.888442578053679,
+      "grad_norm": 1.0955491065979004,
+      "learning_rate": 4.7667073162896315e-07,
+      "loss": 0.504,
+      "step": 5171
+    },
+    {
+      "epoch": 1.8888077414643054,
+      "grad_norm": 1.0627782344818115,
+      "learning_rate": 4.7356251987477507e-07,
+      "loss": 0.555,
+      "step": 5172
+    },
+    {
+      "epoch": 1.8891729048749315,
+      "grad_norm": 0.6565744876861572,
+      "learning_rate": 4.7046435368912404e-07,
+      "loss": 0.5494,
+      "step": 5173
+    },
+    {
+      "epoch": 1.8895380682855578,
+      "grad_norm": 0.8472015857696533,
+      "learning_rate": 4.6737623466589055e-07,
+      "loss": 0.5099,
+      "step": 5174
+    },
+    {
+      "epoch": 1.889903231696184,
+      "grad_norm": 1.0763630867004395,
+      "learning_rate": 4.642981643937905e-07,
+      "loss": 0.5105,
+      "step": 5175
+    },
+    {
+      "epoch": 1.8902683951068102,
+      "grad_norm": 0.6906781196594238,
+      "learning_rate": 4.6123014445636605e-07,
+      "loss": 0.5608,
+      "step": 5176
+    },
+    {
+      "epoch": 1.8906335585174365,
+      "grad_norm": 0.9713335633277893,
+      "learning_rate": 4.581721764319924e-07,
+      "loss": 0.515,
+      "step": 5177
+    },
+    {
+      "epoch": 1.8909987219280628,
+      "grad_norm": 1.1435459852218628,
+      "learning_rate": 4.5512426189386674e-07,
+      "loss": 0.5635,
+      "step": 5178
+    },
+    {
+      "epoch": 1.8913638853386892,
+      "grad_norm": 0.9018896818161011,
+      "learning_rate": 4.520864024100191e-07,
+      "loss": 0.5465,
+      "step": 5179
+    },
+    {
+      "epoch": 1.8917290487493155,
+      "grad_norm": 0.7561548948287964,
+      "learning_rate": 4.4905859954331057e-07,
+      "loss": 0.5474,
+      "step": 5180
+    },
+    {
+      "epoch": 1.8920942121599416,
+      "grad_norm": 1.0169223546981812,
+      "learning_rate": 4.460408548514239e-07,
+      "loss": 0.5194,
+      "step": 5181
+    },
+    {
+      "epoch": 1.8924593755705679,
+      "grad_norm": 1.0248956680297852,
+      "learning_rate": 4.4303316988686396e-07,
+      "loss": 0.5539,
+      "step": 5182
+    },
+    {
+      "epoch": 1.892824538981194,
+      "grad_norm": 0.9140022993087769,
+      "learning_rate": 4.400355461969663e-07,
+      "loss": 0.5504,
+      "step": 5183
+    },
+    {
+      "epoch": 1.8931897023918203,
+      "grad_norm": 0.8542696833610535,
+      "learning_rate": 4.3704798532388624e-07,
+      "loss": 0.5561,
+      "step": 5184
+    },
+    {
+      "epoch": 1.8935548658024466,
+      "grad_norm": 1.0125644207000732,
+      "learning_rate": 4.3407048880460765e-07,
+      "loss": 0.5264,
+      "step": 5185
+    },
+    {
+      "epoch": 1.893920029213073,
+      "grad_norm": 1.2030428647994995,
+      "learning_rate": 4.311030581709297e-07,
+      "loss": 0.535,
+      "step": 5186
+    },
+    {
+      "epoch": 1.8942851926236992,
+      "grad_norm": 2.7859158515930176,
+      "learning_rate": 4.281456949494778e-07,
+      "loss": 0.4829,
+      "step": 5187
+    },
+    {
+      "epoch": 1.8946503560343255,
+      "grad_norm": 0.9519990086555481,
+      "learning_rate": 4.2519840066169493e-07,
+      "loss": 0.5758,
+      "step": 5188
+    },
+    {
+      "epoch": 1.8950155194449516,
+      "grad_norm": 0.8662074208259583,
+      "learning_rate": 4.222611768238505e-07,
+      "loss": 0.5784,
+      "step": 5189
+    },
+    {
+      "epoch": 1.895380682855578,
+      "grad_norm": 0.9285888075828552,
+      "learning_rate": 4.1933402494702235e-07,
+      "loss": 0.4911,
+      "step": 5190
+    },
+    {
+      "epoch": 1.895745846266204,
+      "grad_norm": 1.0755295753479004,
+      "learning_rate": 4.164169465371148e-07,
+      "loss": 0.4974,
+      "step": 5191
+    },
+    {
+      "epoch": 1.8961110096768303,
+      "grad_norm": 0.8906064629554749,
+      "learning_rate": 4.135099430948475e-07,
+      "loss": 0.5153,
+      "step": 5192
+    },
+    {
+      "epoch": 1.8964761730874566,
+      "grad_norm": 1.133794903755188,
+      "learning_rate": 4.106130161157595e-07,
+      "loss": 0.5216,
+      "step": 5193
+    },
+    {
+      "epoch": 1.896841336498083,
+      "grad_norm": 1.1784135103225708,
+      "learning_rate": 4.077261670901989e-07,
+      "loss": 0.525,
+      "step": 5194
+    },
+    {
+      "epoch": 1.8972064999087093,
+      "grad_norm": 0.8882606029510498,
+      "learning_rate": 4.0484939750333743e-07,
+      "loss": 0.5609,
+      "step": 5195
+    },
+    {
+      "epoch": 1.8975716633193354,
+      "grad_norm": 0.8074411153793335,
+      "learning_rate": 4.01982708835158e-07,
+      "loss": 0.5096,
+      "step": 5196
+    },
+    {
+      "epoch": 1.8979368267299617,
+      "grad_norm": 0.8147451281547546,
+      "learning_rate": 3.991261025604543e-07,
+      "loss": 0.5411,
+      "step": 5197
+    },
+    {
+      "epoch": 1.8983019901405878,
+      "grad_norm": 1.0759518146514893,
+      "learning_rate": 3.9627958014883725e-07,
+      "loss": 0.5085,
+      "step": 5198
+    },
+    {
+      "epoch": 1.898667153551214,
+      "grad_norm": 0.7098891139030457,
+      "learning_rate": 3.9344314306472674e-07,
+      "loss": 0.5232,
+      "step": 5199
+    },
+    {
+      "epoch": 1.8990323169618404,
+      "grad_norm": 1.0414230823516846,
+      "learning_rate": 3.9061679276735986e-07,
+      "loss": 0.4982,
+      "step": 5200
+    },
+    {
+      "epoch": 1.8993974803724667,
+      "grad_norm": 0.7669789791107178,
+      "learning_rate": 3.878005307107735e-07,
+      "loss": 0.5128,
+      "step": 5201
+    },
+    {
+      "epoch": 1.899762643783093,
+      "grad_norm": 0.6906653642654419,
+      "learning_rate": 3.849943583438287e-07,
+      "loss": 0.5851,
+      "step": 5202
+    },
+    {
+      "epoch": 1.9001278071937193,
+      "grad_norm": 0.7735735177993774,
+      "learning_rate": 3.8219827711018397e-07,
+      "loss": 0.5872,
+      "step": 5203
+    },
+    {
+      "epoch": 1.9004929706043454,
+      "grad_norm": 0.9029932618141174,
+      "learning_rate": 3.794122884483131e-07,
+      "loss": 0.5286,
+      "step": 5204
+    },
+    {
+      "epoch": 1.9008581340149717,
+      "grad_norm": 1.4787055253982544,
+      "learning_rate": 3.7663639379149406e-07,
+      "loss": 0.5121,
+      "step": 5205
+    },
+    {
+      "epoch": 1.9012232974255978,
+      "grad_norm": 0.8257893919944763,
+      "learning_rate": 3.738705945678134e-07,
+      "loss": 0.5512,
+      "step": 5206
+    },
+    {
+      "epoch": 1.9015884608362241,
+      "grad_norm": 1.0745633840560913,
+      "learning_rate": 3.7111489220016617e-07,
+      "loss": 0.4923,
+      "step": 5207
+    },
+    {
+      "epoch": 1.9019536242468504,
+      "grad_norm": 1.098933458328247,
+      "learning_rate": 3.6836928810624506e-07,
+      "loss": 0.5365,
+      "step": 5208
+    },
+    {
+      "epoch": 1.9023187876574768,
+      "grad_norm": 1.2179603576660156,
+      "learning_rate": 3.656337836985602e-07,
+      "loss": 0.5296,
+      "step": 5209
+    },
+    {
+      "epoch": 1.902683951068103,
+      "grad_norm": 1.182113528251648,
+      "learning_rate": 3.629083803844147e-07,
+      "loss": 0.483,
+      "step": 5210
+    },
+    {
+      "epoch": 1.9030491144787294,
+      "grad_norm": 1.2880009412765503,
+      "learning_rate": 3.6019307956592034e-07,
+      "loss": 0.5182,
+      "step": 5211
+    },
+    {
+      "epoch": 1.9034142778893555,
+      "grad_norm": 0.8011050224304199,
+      "learning_rate": 3.5748788263998855e-07,
+      "loss": 0.5498,
+      "step": 5212
+    },
+    {
+      "epoch": 1.9037794412999818,
+      "grad_norm": 1.1082504987716675,
+      "learning_rate": 3.547927909983373e-07,
+      "loss": 0.564,
+      "step": 5213
+    },
+    {
+      "epoch": 1.9041446047106079,
+      "grad_norm": 0.8854731321334839,
+      "learning_rate": 3.521078060274841e-07,
+      "loss": 0.5016,
+      "step": 5214
+    },
+    {
+      "epoch": 1.9045097681212342,
+      "grad_norm": 0.816261351108551,
+      "learning_rate": 3.4943292910874173e-07,
+      "loss": 0.5451,
+      "step": 5215
+    },
+    {
+      "epoch": 1.9048749315318605,
+      "grad_norm": 0.8715974688529968,
+      "learning_rate": 3.4676816161822947e-07,
+      "loss": 0.485,
+      "step": 5216
+    },
+    {
+      "epoch": 1.9052400949424868,
+      "grad_norm": 0.9326238036155701,
+      "learning_rate": 3.4411350492686404e-07,
+      "loss": 0.5318,
+      "step": 5217
+    },
+    {
+      "epoch": 1.9056052583531131,
+      "grad_norm": 1.1741446256637573,
+      "learning_rate": 3.4146896040035514e-07,
+      "loss": 0.4912,
+      "step": 5218
+    },
+    {
+      "epoch": 1.9059704217637394,
+      "grad_norm": 0.9231522083282471,
+      "learning_rate": 3.3883452939922123e-07,
+      "loss": 0.4973,
+      "step": 5219
+    },
+    {
+      "epoch": 1.9063355851743655,
+      "grad_norm": 0.7504870295524597,
+      "learning_rate": 3.3621021327876923e-07,
+      "loss": 0.5522,
+      "step": 5220
+    },
+    {
+      "epoch": 1.9067007485849918,
+      "grad_norm": 0.8062773942947388,
+      "learning_rate": 3.3359601338910143e-07,
+      "loss": 0.5267,
+      "step": 5221
+    },
+    {
+      "epoch": 1.907065911995618,
+      "grad_norm": 0.8056736588478088,
+      "learning_rate": 3.3099193107512197e-07,
+      "loss": 0.5195,
+      "step": 5222
+    },
+    {
+      "epoch": 1.9074310754062442,
+      "grad_norm": 0.86921226978302,
+      "learning_rate": 3.283979676765259e-07,
+      "loss": 0.5514,
+      "step": 5223
+    },
+    {
+      "epoch": 1.9077962388168705,
+      "grad_norm": 0.8477928042411804,
+      "learning_rate": 3.258141245278057e-07,
+      "loss": 0.5361,
+      "step": 5224
+    },
+    {
+      "epoch": 1.9081614022274969,
+      "grad_norm": 0.6933594346046448,
+      "learning_rate": 3.2324040295824033e-07,
+      "loss": 0.556,
+      "step": 5225
+    },
+    {
+      "epoch": 1.9085265656381232,
+      "grad_norm": 0.7396385073661804,
+      "learning_rate": 3.2067680429190617e-07,
+      "loss": 0.6044,
+      "step": 5226
+    },
+    {
+      "epoch": 1.9088917290487493,
+      "grad_norm": 0.843963623046875,
+      "learning_rate": 3.181233298476771e-07,
+      "loss": 0.5346,
+      "step": 5227
+    },
+    {
+      "epoch": 1.9092568924593756,
+      "grad_norm": 0.7993393540382385,
+      "learning_rate": 3.1557998093920904e-07,
+      "loss": 0.5474,
+      "step": 5228
+    },
+    {
+      "epoch": 1.9096220558700017,
+      "grad_norm": 1.0452121496200562,
+      "learning_rate": 3.130467588749553e-07,
+      "loss": 0.5049,
+      "step": 5229
+    },
+    {
+      "epoch": 1.909987219280628,
+      "grad_norm": 0.920778214931488,
+      "learning_rate": 3.105236649581556e-07,
+      "loss": 0.5573,
+      "step": 5230
+    },
+    {
+      "epoch": 1.9103523826912543,
+      "grad_norm": 0.6959015727043152,
+      "learning_rate": 3.0801070048684046e-07,
+      "loss": 0.5451,
+      "step": 5231
+    },
+    {
+      "epoch": 1.9107175461018806,
+      "grad_norm": 0.9710888862609863,
+      "learning_rate": 3.055078667538292e-07,
+      "loss": 0.5067,
+      "step": 5232
+    },
+    {
+      "epoch": 1.911082709512507,
+      "grad_norm": 0.9164549112319946,
+      "learning_rate": 3.0301516504672944e-07,
+      "loss": 0.5367,
+      "step": 5233
+    },
+    {
+      "epoch": 1.9114478729231332,
+      "grad_norm": 0.9263083338737488,
+      "learning_rate": 3.0053259664793997e-07,
+      "loss": 0.5489,
+      "step": 5234
+    },
+    {
+      "epoch": 1.9118130363337593,
+      "grad_norm": 1.0650849342346191,
+      "learning_rate": 2.980601628346347e-07,
+      "loss": 0.5145,
+      "step": 5235
+    },
+    {
+      "epoch": 1.9121781997443856,
+      "grad_norm": 0.8558089733123779,
+      "learning_rate": 2.9559786487878716e-07,
+      "loss": 0.5443,
+      "step": 5236
+    },
+    {
+      "epoch": 1.9125433631550117,
+      "grad_norm": 0.9622271656990051,
+      "learning_rate": 2.931457040471508e-07,
+      "loss": 0.5552,
+      "step": 5237
+    },
+    {
+      "epoch": 1.912908526565638,
+      "grad_norm": 1.4848085641860962,
+      "learning_rate": 2.907036816012609e-07,
+      "loss": 0.5295,
+      "step": 5238
+    },
+    {
+      "epoch": 1.9132736899762643,
+      "grad_norm": 0.7723660469055176,
+      "learning_rate": 2.882717987974437e-07,
+      "loss": 0.5229,
+      "step": 5239
+    },
+    {
+      "epoch": 1.9136388533868907,
+      "grad_norm": 0.8148236274719238,
+      "learning_rate": 2.85850056886805e-07,
+      "loss": 0.5573,
+      "step": 5240
+    },
+    {
+      "epoch": 1.914004016797517,
+      "grad_norm": 0.6501472592353821,
+      "learning_rate": 2.834384571152282e-07,
+      "loss": 0.5251,
+      "step": 5241
+    },
+    {
+      "epoch": 1.9143691802081433,
+      "grad_norm": 1.1822669506072998,
+      "learning_rate": 2.8103700072339203e-07,
+      "loss": 0.5328,
+      "step": 5242
+    },
+    {
+      "epoch": 1.9147343436187694,
+      "grad_norm": 0.7121161222457886,
+      "learning_rate": 2.7864568894674593e-07,
+      "loss": 0.5369,
+      "step": 5243
+    },
+    {
+      "epoch": 1.9150995070293957,
+      "grad_norm": 0.8092657327651978,
+      "learning_rate": 2.7626452301552586e-07,
+      "loss": 0.5597,
+      "step": 5244
+    },
+    {
+      "epoch": 1.9154646704400218,
+      "grad_norm": 0.7837401032447815,
+      "learning_rate": 2.7389350415474305e-07,
+      "loss": 0.5346,
+      "step": 5245
+    },
+    {
+      "epoch": 1.915829833850648,
+      "grad_norm": 0.9305130243301392,
+      "learning_rate": 2.715326335841906e-07,
+      "loss": 0.5218,
+      "step": 5246
+    },
+    {
+      "epoch": 1.9161949972612744,
+      "grad_norm": 0.9589447975158691,
+      "learning_rate": 2.691819125184458e-07,
+      "loss": 0.5131,
+      "step": 5247
+    },
+    {
+      "epoch": 1.9165601606719007,
+      "grad_norm": 1.0760564804077148,
+      "learning_rate": 2.668413421668592e-07,
+      "loss": 0.5446,
+      "step": 5248
+    },
+    {
+      "epoch": 1.916925324082527,
+      "grad_norm": 0.7688742280006409,
+      "learning_rate": 2.645109237335608e-07,
+      "loss": 0.5536,
+      "step": 5249
+    },
+    {
+      "epoch": 1.9172904874931533,
+      "grad_norm": 0.8162276744842529,
+      "learning_rate": 2.6219065841745383e-07,
+      "loss": 0.5089,
+      "step": 5250
+    },
+    {
+      "epoch": 1.9176556509037794,
+      "grad_norm": 0.7043295502662659,
+      "learning_rate": 2.5988054741222345e-07,
+      "loss": 0.5647,
+      "step": 5251
+    },
+    {
+      "epoch": 1.9180208143144057,
+      "grad_norm": 0.8645861744880676,
+      "learning_rate": 2.5758059190633233e-07,
+      "loss": 0.5314,
+      "step": 5252
+    },
+    {
+      "epoch": 1.9183859777250318,
+      "grad_norm": 0.6989781260490417,
+      "learning_rate": 2.5529079308301174e-07,
+      "loss": 0.53,
+      "step": 5253
+    },
+    {
+      "epoch": 1.9187511411356581,
+      "grad_norm": 0.7858476638793945,
+      "learning_rate": 2.530111521202727e-07,
+      "loss": 0.5229,
+      "step": 5254
+    },
+    {
+      "epoch": 1.9191163045462845,
+      "grad_norm": 0.8875958919525146,
+      "learning_rate": 2.5074167019089714e-07,
+      "loss": 0.5294,
+      "step": 5255
+    },
+    {
+      "epoch": 1.9194814679569108,
+      "grad_norm": 0.9924106001853943,
+      "learning_rate": 2.484823484624466e-07,
+      "loss": 0.5052,
+      "step": 5256
+    },
+    {
+      "epoch": 1.919846631367537,
+      "grad_norm": 0.8700651526451111,
+      "learning_rate": 2.462331880972468e-07,
+      "loss": 0.5251,
+      "step": 5257
+    },
+    {
+      "epoch": 1.9202117947781634,
+      "grad_norm": 1.09417724609375,
+      "learning_rate": 2.4399419025240344e-07,
+      "loss": 0.4907,
+      "step": 5258
+    },
+    {
+      "epoch": 1.9205769581887895,
+      "grad_norm": 0.8320743441581726,
+      "learning_rate": 2.4176535607978835e-07,
+      "loss": 0.4963,
+      "step": 5259
+    },
+    {
+      "epoch": 1.9209421215994156,
+      "grad_norm": 0.9074090123176575,
+      "learning_rate": 2.3954668672604874e-07,
+      "loss": 0.528,
+      "step": 5260
+    },
+    {
+      "epoch": 1.9213072850100419,
+      "grad_norm": 0.9395208954811096,
+      "learning_rate": 2.373381833326027e-07,
+      "loss": 0.4746,
+      "step": 5261
+    },
+    {
+      "epoch": 1.9216724484206682,
+      "grad_norm": 0.9719210267066956,
+      "learning_rate": 2.3513984703563476e-07,
+      "loss": 0.533,
+      "step": 5262
+    },
+    {
+      "epoch": 1.9220376118312945,
+      "grad_norm": 0.9111335277557373,
+      "learning_rate": 2.3295167896610016e-07,
+      "loss": 0.5145,
+      "step": 5263
+    },
+    {
+      "epoch": 1.9224027752419208,
+      "grad_norm": 0.6564895510673523,
+      "learning_rate": 2.3077368024972514e-07,
+      "loss": 0.5348,
+      "step": 5264
+    },
+    {
+      "epoch": 1.9227679386525471,
+      "grad_norm": 0.8111744523048401,
+      "learning_rate": 2.2860585200700226e-07,
+      "loss": 0.5407,
+      "step": 5265
+    },
+    {
+      "epoch": 1.9231331020631732,
+      "grad_norm": 0.8375979661941528,
+      "learning_rate": 2.2644819535319051e-07,
+      "loss": 0.5314,
+      "step": 5266
+    },
+    {
+      "epoch": 1.9234982654737995,
+      "grad_norm": 1.4540085792541504,
+      "learning_rate": 2.2430071139832198e-07,
+      "loss": 0.5235,
+      "step": 5267
+    },
+    {
+      "epoch": 1.9238634288844256,
+      "grad_norm": 0.8702842593193054,
+      "learning_rate": 2.2216340124718626e-07,
+      "loss": 0.4963,
+      "step": 5268
+    },
+    {
+      "epoch": 1.924228592295052,
+      "grad_norm": 0.8858942985534668,
+      "learning_rate": 2.2003626599934602e-07,
+      "loss": 0.5234,
+      "step": 5269
+    },
+    {
+      "epoch": 1.9245937557056783,
+      "grad_norm": 0.7983465194702148,
+      "learning_rate": 2.1791930674912587e-07,
+      "loss": 0.4987,
+      "step": 5270
+    },
+    {
+      "epoch": 1.9249589191163046,
+      "grad_norm": 0.9610999822616577,
+      "learning_rate": 2.1581252458561684e-07,
+      "loss": 0.4934,
+      "step": 5271
+    },
+    {
+      "epoch": 1.9253240825269309,
+      "grad_norm": 0.8956314921379089,
+      "learning_rate": 2.137159205926742e-07,
+      "loss": 0.5299,
+      "step": 5272
+    },
+    {
+      "epoch": 1.9256892459375572,
+      "grad_norm": 0.9054751396179199,
+      "learning_rate": 2.1162949584891512e-07,
+      "loss": 0.588,
+      "step": 5273
+    },
+    {
+      "epoch": 1.9260544093481833,
+      "grad_norm": 0.7674873471260071,
+      "learning_rate": 2.095532514277232e-07,
+      "loss": 0.5629,
+      "step": 5274
+    },
+    {
+      "epoch": 1.9264195727588096,
+      "grad_norm": 0.8046696782112122,
+      "learning_rate": 2.0748718839724403e-07,
+      "loss": 0.4931,
+      "step": 5275
+    },
+    {
+      "epoch": 1.9267847361694357,
+      "grad_norm": 1.0259958505630493,
+      "learning_rate": 2.0543130782037845e-07,
+      "loss": 0.5065,
+      "step": 5276
+    },
+    {
+      "epoch": 1.927149899580062,
+      "grad_norm": 0.9353147745132446,
+      "learning_rate": 2.0338561075480269e-07,
+      "loss": 0.5138,
+      "step": 5277
+    },
+    {
+      "epoch": 1.9275150629906883,
+      "grad_norm": 0.8617286682128906,
+      "learning_rate": 2.0135009825293928e-07,
+      "loss": 0.5269,
+      "step": 5278
+    },
+    {
+      "epoch": 1.9278802264013146,
+      "grad_norm": 0.8524954915046692,
+      "learning_rate": 1.9932477136197949e-07,
+      "loss": 0.5413,
+      "step": 5279
+    },
+    {
+      "epoch": 1.928245389811941,
+      "grad_norm": 0.9233561754226685,
+      "learning_rate": 1.9730963112387425e-07,
+      "loss": 0.5186,
+      "step": 5280
+    },
+    {
+      "epoch": 1.9286105532225672,
+      "grad_norm": 0.7263604402542114,
+      "learning_rate": 1.9530467857532986e-07,
+      "loss": 0.5771,
+      "step": 5281
+    },
+    {
+      "epoch": 1.9289757166331933,
+      "grad_norm": 0.7217230796813965,
+      "learning_rate": 1.93309914747819e-07,
+      "loss": 0.5419,
+      "step": 5282
+    },
+    {
+      "epoch": 1.9293408800438197,
+      "grad_norm": 0.9343338012695312,
+      "learning_rate": 1.9132534066756304e-07,
+      "loss": 0.5175,
+      "step": 5283
+    },
+    {
+      "epoch": 1.9297060434544457,
+      "grad_norm": 0.7075455784797668,
+      "learning_rate": 1.8935095735554522e-07,
+      "loss": 0.5212,
+      "step": 5284
+    },
+    {
+      "epoch": 1.930071206865072,
+      "grad_norm": 0.7148510217666626,
+      "learning_rate": 1.8738676582750638e-07,
+      "loss": 0.5789,
+      "step": 5285
+    },
+    {
+      "epoch": 1.9304363702756984,
+      "grad_norm": 1.1724885702133179,
+      "learning_rate": 1.854327670939471e-07,
+      "loss": 0.5462,
+      "step": 5286
+    },
+    {
+      "epoch": 1.9308015336863247,
+      "grad_norm": 0.9485020637512207,
+      "learning_rate": 1.8348896216012102e-07,
+      "loss": 0.509,
+      "step": 5287
+    },
+    {
+      "epoch": 1.931166697096951,
+      "grad_norm": 1.7501707077026367,
+      "learning_rate": 1.8155535202603712e-07,
+      "loss": 0.5125,
+      "step": 5288
+    },
+    {
+      "epoch": 1.9315318605075773,
+      "grad_norm": 1.040753960609436,
+      "learning_rate": 1.796319376864597e-07,
+      "loss": 0.4997,
+      "step": 5289
+    },
+    {
+      "epoch": 1.9318970239182034,
+      "grad_norm": 1.0983657836914062,
+      "learning_rate": 1.7771872013090608e-07,
+      "loss": 0.5547,
+      "step": 5290
+    },
+    {
+      "epoch": 1.9322621873288297,
+      "grad_norm": 1.2510367631912231,
+      "learning_rate": 1.7581570034365557e-07,
+      "loss": 0.5208,
+      "step": 5291
+    },
+    {
+      "epoch": 1.9326273507394558,
+      "grad_norm": 1.0194956064224243,
+      "learning_rate": 1.7392287930373175e-07,
+      "loss": 0.5549,
+      "step": 5292
+    },
+    {
+      "epoch": 1.932992514150082,
+      "grad_norm": 0.9619592428207397,
+      "learning_rate": 1.7204025798491342e-07,
+      "loss": 0.5373,
+      "step": 5293
+    },
+    {
+      "epoch": 1.9333576775607084,
+      "grad_norm": 0.7508404850959778,
+      "learning_rate": 1.7016783735573693e-07,
+      "loss": 0.5258,
+      "step": 5294
+    },
+    {
+      "epoch": 1.9337228409713347,
+      "grad_norm": 1.070589542388916,
+      "learning_rate": 1.6830561837948735e-07,
+      "loss": 0.5208,
+      "step": 5295
+    },
+    {
+      "epoch": 1.934088004381961,
+      "grad_norm": 0.9107255339622498,
+      "learning_rate": 1.6645360201420046e-07,
+      "loss": 0.5433,
+      "step": 5296
+    },
+    {
+      "epoch": 1.9344531677925871,
+      "grad_norm": 0.9078409075737,
+      "learning_rate": 1.646117892126653e-07,
+      "loss": 0.5283,
+      "step": 5297
+    },
+    {
+      "epoch": 1.9348183312032134,
+      "grad_norm": 0.9286045432090759,
+      "learning_rate": 1.6278018092241943e-07,
+      "loss": 0.5377,
+      "step": 5298
+    },
+    {
+      "epoch": 1.9351834946138395,
+      "grad_norm": 0.6777305603027344,
+      "learning_rate": 1.6095877808575133e-07,
+      "loss": 0.5591,
+      "step": 5299
+    },
+    {
+      "epoch": 1.9355486580244659,
+      "grad_norm": 0.869694173336029,
+      "learning_rate": 1.5914758163970033e-07,
+      "loss": 0.5428,
+      "step": 5300
+    },
+    {
+      "epoch": 1.9359138214350922,
+      "grad_norm": 1.2617483139038086,
+      "learning_rate": 1.5734659251605666e-07,
+      "loss": 0.5204,
+      "step": 5301
+    },
+    {
+      "epoch": 1.9362789848457185,
+      "grad_norm": 0.9180070161819458,
+      "learning_rate": 1.5555581164135468e-07,
+      "loss": 0.5474,
+      "step": 5302
+    },
+    {
+      "epoch": 1.9366441482563448,
+      "grad_norm": 0.7689355611801147,
+      "learning_rate": 1.537752399368797e-07,
+      "loss": 0.5398,
+      "step": 5303
+    },
+    {
+      "epoch": 1.937009311666971,
+      "grad_norm": 0.9898827075958252,
+      "learning_rate": 1.520048783186634e-07,
+      "loss": 0.51,
+      "step": 5304
+    },
+    {
+      "epoch": 1.9373744750775972,
+      "grad_norm": 0.7218833565711975,
+      "learning_rate": 1.502447276974861e-07,
+      "loss": 0.5702,
+      "step": 5305
+    },
+    {
+      "epoch": 1.9377396384882235,
+      "grad_norm": 0.7822741866111755,
+      "learning_rate": 1.484947889788768e-07,
+      "loss": 0.5656,
+      "step": 5306
+    },
+    {
+      "epoch": 1.9381048018988496,
+      "grad_norm": 0.9823201894760132,
+      "learning_rate": 1.4675506306310873e-07,
+      "loss": 0.5863,
+      "step": 5307
+    },
+    {
+      "epoch": 1.938469965309476,
+      "grad_norm": 1.0413130521774292,
+      "learning_rate": 1.4502555084519698e-07,
+      "loss": 0.5248,
+      "step": 5308
+    },
+    {
+      "epoch": 1.9388351287201022,
+      "grad_norm": 1.0553734302520752,
+      "learning_rate": 1.4330625321490988e-07,
+      "loss": 0.4869,
+      "step": 5309
+    },
+    {
+      "epoch": 1.9392002921307285,
+      "grad_norm": 1.0827317237854004,
+      "learning_rate": 1.4159717105675542e-07,
+      "loss": 0.5398,
+      "step": 5310
+    },
+    {
+      "epoch": 1.9395654555413548,
+      "grad_norm": 1.0234954357147217,
+      "learning_rate": 1.3989830524999025e-07,
+      "loss": 0.5092,
+      "step": 5311
+    },
+    {
+      "epoch": 1.9399306189519812,
+      "grad_norm": 1.1130499839782715,
+      "learning_rate": 1.3820965666860865e-07,
+      "loss": 0.5089,
+      "step": 5312
+    },
+    {
+      "epoch": 1.9402957823626072,
+      "grad_norm": 0.9848282337188721,
+      "learning_rate": 1.3653122618135562e-07,
+      "loss": 0.5146,
+      "step": 5313
+    },
+    {
+      "epoch": 1.9406609457732336,
+      "grad_norm": 1.0807373523712158,
+      "learning_rate": 1.348630146517138e-07,
+      "loss": 0.527,
+      "step": 5314
+    },
+    {
+      "epoch": 1.9410261091838596,
+      "grad_norm": 0.9990309476852417,
+      "learning_rate": 1.3320502293791448e-07,
+      "loss": 0.5344,
+      "step": 5315
+    },
+    {
+      "epoch": 1.941391272594486,
+      "grad_norm": 1.0408899784088135,
+      "learning_rate": 1.3155725189292646e-07,
+      "loss": 0.5235,
+      "step": 5316
+    },
+    {
+      "epoch": 1.9417564360051123,
+      "grad_norm": 0.9949876666069031,
+      "learning_rate": 1.2991970236445828e-07,
+      "loss": 0.5322,
+      "step": 5317
+    },
+    {
+      "epoch": 1.9421215994157386,
+      "grad_norm": 0.745749294757843,
+      "learning_rate": 1.282923751949694e-07,
+      "loss": 0.5435,
+      "step": 5318
+    },
+    {
+      "epoch": 1.942486762826365,
+      "grad_norm": 0.7982057929039001,
+      "learning_rate": 1.2667527122165014e-07,
+      "loss": 0.5428,
+      "step": 5319
+    },
+    {
+      "epoch": 1.9428519262369912,
+      "grad_norm": 0.9611827731132507,
+      "learning_rate": 1.2506839127643943e-07,
+      "loss": 0.5186,
+      "step": 5320
+    },
+    {
+      "epoch": 1.9432170896476173,
+      "grad_norm": 0.8420519828796387,
+      "learning_rate": 1.2347173618600717e-07,
+      "loss": 0.5284,
+      "step": 5321
+    },
+    {
+      "epoch": 1.9435822530582436,
+      "grad_norm": 1.0904260873794556,
+      "learning_rate": 1.218853067717718e-07,
+      "loss": 0.4985,
+      "step": 5322
+    },
+    {
+      "epoch": 1.9439474164688697,
+      "grad_norm": 2.987138509750366,
+      "learning_rate": 1.2030910384988716e-07,
+      "loss": 0.5245,
+      "step": 5323
+    },
+    {
+      "epoch": 1.944312579879496,
+      "grad_norm": 0.9676865339279175,
+      "learning_rate": 1.1874312823124678e-07,
+      "loss": 0.5267,
+      "step": 5324
+    },
+    {
+      "epoch": 1.9446777432901223,
+      "grad_norm": 0.6485067009925842,
+      "learning_rate": 1.1718738072148184e-07,
+      "loss": 0.5727,
+      "step": 5325
+    },
+    {
+      "epoch": 1.9450429067007486,
+      "grad_norm": 1.0290840864181519,
+      "learning_rate": 1.1564186212096317e-07,
+      "loss": 0.5164,
+      "step": 5326
+    },
+    {
+      "epoch": 1.945408070111375,
+      "grad_norm": 0.7069109082221985,
+      "learning_rate": 1.1410657322479479e-07,
+      "loss": 0.5349,
+      "step": 5327
+    },
+    {
+      "epoch": 1.945773233522001,
+      "grad_norm": 0.7101110219955444,
+      "learning_rate": 1.1258151482282265e-07,
+      "loss": 0.5997,
+      "step": 5328
+    },
+    {
+      "epoch": 1.9461383969326274,
+      "grad_norm": 0.8071128726005554,
+      "learning_rate": 1.1106668769963025e-07,
+      "loss": 0.5282,
+      "step": 5329
+    },
+    {
+      "epoch": 1.9465035603432534,
+      "grad_norm": 0.7718363404273987,
+      "learning_rate": 1.0956209263453421e-07,
+      "loss": 0.5654,
+      "step": 5330
+    },
+    {
+      "epoch": 1.9468687237538798,
+      "grad_norm": 0.9163565635681152,
+      "learning_rate": 1.0806773040158647e-07,
+      "loss": 0.4722,
+      "step": 5331
+    },
+    {
+      "epoch": 1.947233887164506,
+      "grad_norm": 0.6575088500976562,
+      "learning_rate": 1.0658360176957871e-07,
+      "loss": 0.5298,
+      "step": 5332
+    },
+    {
+      "epoch": 1.9475990505751324,
+      "grad_norm": 0.9368276000022888,
+      "learning_rate": 1.0510970750203353e-07,
+      "loss": 0.4665,
+      "step": 5333
+    },
+    {
+      "epoch": 1.9479642139857587,
+      "grad_norm": 0.8309523463249207,
+      "learning_rate": 1.0364604835721325e-07,
+      "loss": 0.5255,
+      "step": 5334
+    },
+    {
+      "epoch": 1.948329377396385,
+      "grad_norm": 0.9614001512527466,
+      "learning_rate": 1.021926250881089e-07,
+      "loss": 0.5179,
+      "step": 5335
+    },
+    {
+      "epoch": 1.948694540807011,
+      "grad_norm": 0.9124447703361511,
+      "learning_rate": 1.0074943844245122e-07,
+      "loss": 0.5408,
+      "step": 5336
+    },
+    {
+      "epoch": 1.9490597042176374,
+      "grad_norm": 0.887027382850647,
+      "learning_rate": 9.931648916269965e-08,
+      "loss": 0.5436,
+      "step": 5337
+    },
+    {
+      "epoch": 1.9494248676282635,
+      "grad_norm": 0.9393956661224365,
+      "learning_rate": 9.789377798604894e-08,
+      "loss": 0.5057,
+      "step": 5338
+    },
+    {
+      "epoch": 1.9497900310388898,
+      "grad_norm": 0.8401802778244019,
+      "learning_rate": 9.648130564442915e-08,
+      "loss": 0.5381,
+      "step": 5339
+    },
+    {
+      "epoch": 1.9501551944495161,
+      "grad_norm": 1.0334709882736206,
+      "learning_rate": 9.507907286449903e-08,
+      "loss": 0.4995,
+      "step": 5340
+    },
+    {
+      "epoch": 1.9505203578601424,
+      "grad_norm": 0.9041337370872498,
+      "learning_rate": 9.368708036764818e-08,
+      "loss": 0.5391,
+      "step": 5341
+    },
+    {
+      "epoch": 1.9508855212707688,
+      "grad_norm": 1.1579806804656982,
+      "learning_rate": 9.230532887000598e-08,
+      "loss": 0.5238,
+      "step": 5342
+    },
+    {
+      "epoch": 1.951250684681395,
+      "grad_norm": 1.1014970541000366,
+      "learning_rate": 9.093381908242605e-08,
+      "loss": 0.5234,
+      "step": 5343
+    },
+    {
+      "epoch": 1.9516158480920212,
+      "grad_norm": 0.9048789739608765,
+      "learning_rate": 8.957255171049506e-08,
+      "loss": 0.5793,
+      "step": 5344
+    },
+    {
+      "epoch": 1.9519810115026475,
+      "grad_norm": 0.8284463286399841,
+      "learning_rate": 8.822152745453061e-08,
+      "loss": 0.5453,
+      "step": 5345
+    },
+    {
+      "epoch": 1.9523461749132736,
+      "grad_norm": 0.9747271537780762,
+      "learning_rate": 8.688074700958115e-08,
+      "loss": 0.4893,
+      "step": 5346
+    },
+    {
+      "epoch": 1.9527113383238999,
+      "grad_norm": 0.9729907512664795,
+      "learning_rate": 8.55502110654216e-08,
+      "loss": 0.5385,
+      "step": 5347
+    },
+    {
+      "epoch": 1.9530765017345262,
+      "grad_norm": 0.7343084216117859,
+      "learning_rate": 8.422992030656218e-08,
+      "loss": 0.5416,
+      "step": 5348
+    },
+    {
+      "epoch": 1.9534416651451525,
+      "grad_norm": 0.9176368713378906,
+      "learning_rate": 8.291987541223955e-08,
+      "loss": 0.5335,
+      "step": 5349
+    },
+    {
+      "epoch": 1.9538068285557788,
+      "grad_norm": 0.9123088121414185,
+      "learning_rate": 8.162007705641905e-08,
+      "loss": 0.5065,
+      "step": 5350
+    },
+    {
+      "epoch": 1.9541719919664051,
+      "grad_norm": 1.1829543113708496,
+      "learning_rate": 8.033052590779245e-08,
+      "loss": 0.5075,
+      "step": 5351
+    },
+    {
+      "epoch": 1.9545371553770312,
+      "grad_norm": 0.9668950438499451,
+      "learning_rate": 7.90512226297846e-08,
+      "loss": 0.4947,
+      "step": 5352
+    },
+    {
+      "epoch": 1.9549023187876575,
+      "grad_norm": 1.1358628273010254,
+      "learning_rate": 7.77821678805446e-08,
+      "loss": 0.5368,
+      "step": 5353
+    },
+    {
+      "epoch": 1.9552674821982836,
+      "grad_norm": 0.9124475121498108,
+      "learning_rate": 7.652336231295021e-08,
+      "loss": 0.4982,
+      "step": 5354
+    },
+    {
+      "epoch": 1.95563264560891,
+      "grad_norm": 0.8891622424125671,
+      "learning_rate": 7.527480657460562e-08,
+      "loss": 0.4711,
+      "step": 5355
+    },
+    {
+      "epoch": 1.9559978090195362,
+      "grad_norm": 2.195021629333496,
+      "learning_rate": 7.403650130784368e-08,
+      "loss": 0.5155,
+      "step": 5356
+    },
+    {
+      "epoch": 1.9563629724301626,
+      "grad_norm": 0.8266718983650208,
+      "learning_rate": 7.280844714972368e-08,
+      "loss": 0.5251,
+      "step": 5357
+    },
+    {
+      "epoch": 1.9567281358407889,
+      "grad_norm": 0.7959644794464111,
+      "learning_rate": 7.159064473202914e-08,
+      "loss": 0.5285,
+      "step": 5358
+    },
+    {
+      "epoch": 1.957093299251415,
+      "grad_norm": 0.9930187463760376,
+      "learning_rate": 7.038309468127225e-08,
+      "loss": 0.4782,
+      "step": 5359
+    },
+    {
+      "epoch": 1.9574584626620413,
+      "grad_norm": 1.3513890504837036,
+      "learning_rate": 6.918579761868493e-08,
+      "loss": 0.4846,
+      "step": 5360
+    },
+    {
+      "epoch": 1.9578236260726674,
+      "grad_norm": 0.8065437078475952,
+      "learning_rate": 6.799875416023005e-08,
+      "loss": 0.51,
+      "step": 5361
+    },
+    {
+      "epoch": 1.9581887894832937,
+      "grad_norm": 0.9978945851325989,
+      "learning_rate": 6.682196491659687e-08,
+      "loss": 0.4882,
+      "step": 5362
+    },
+    {
+      "epoch": 1.95855395289392,
+      "grad_norm": 0.8857904076576233,
+      "learning_rate": 6.565543049319445e-08,
+      "loss": 0.5433,
+      "step": 5363
+    },
+    {
+      "epoch": 1.9589191163045463,
+      "grad_norm": 0.8729395270347595,
+      "learning_rate": 6.449915149015828e-08,
+      "loss": 0.5188,
+      "step": 5364
+    },
+    {
+      "epoch": 1.9592842797151726,
+      "grad_norm": 1.0837210416793823,
+      "learning_rate": 6.335312850234365e-08,
+      "loss": 0.5128,
+      "step": 5365
+    },
+    {
+      "epoch": 1.959649443125799,
+      "grad_norm": 0.758825421333313,
+      "learning_rate": 6.221736211933893e-08,
+      "loss": 0.5235,
+      "step": 5366
+    },
+    {
+      "epoch": 1.960014606536425,
+      "grad_norm": 0.9562605619430542,
+      "learning_rate": 6.109185292544784e-08,
+      "loss": 0.5296,
+      "step": 5367
+    },
+    {
+      "epoch": 1.9603797699470513,
+      "grad_norm": 1.308916687965393,
+      "learning_rate": 5.99766014996983e-08,
+      "loss": 0.5097,
+      "step": 5368
+    },
+    {
+      "epoch": 1.9607449333576774,
+      "grad_norm": 1.2315113544464111,
+      "learning_rate": 5.887160841584472e-08,
+      "loss": 0.521,
+      "step": 5369
+    },
+    {
+      "epoch": 1.9611100967683037,
+      "grad_norm": 0.6987513899803162,
+      "learning_rate": 5.777687424236123e-08,
+      "loss": 0.5208,
+      "step": 5370
+    },
+    {
+      "epoch": 1.96147526017893,
+      "grad_norm": 0.820620059967041,
+      "learning_rate": 5.669239954244399e-08,
+      "loss": 0.5549,
+      "step": 5371
+    },
+    {
+      "epoch": 1.9618404235895563,
+      "grad_norm": 0.9430913925170898,
+      "learning_rate": 5.561818487401338e-08,
+      "loss": 0.5097,
+      "step": 5372
+    },
+    {
+      "epoch": 1.9622055870001827,
+      "grad_norm": 0.9805968999862671,
+      "learning_rate": 5.455423078970734e-08,
+      "loss": 0.5195,
+      "step": 5373
+    },
+    {
+      "epoch": 1.962570750410809,
+      "grad_norm": 1.0393964052200317,
+      "learning_rate": 5.350053783689024e-08,
+      "loss": 0.5257,
+      "step": 5374
+    },
+    {
+      "epoch": 1.962935913821435,
+      "grad_norm": 0.6415253281593323,
+      "learning_rate": 5.2457106557641803e-08,
+      "loss": 0.5567,
+      "step": 5375
+    },
+    {
+      "epoch": 1.9633010772320614,
+      "grad_norm": 0.987578809261322,
+      "learning_rate": 5.142393748876595e-08,
+      "loss": 0.5237,
+      "step": 5376
+    },
+    {
+      "epoch": 1.9636662406426875,
+      "grad_norm": 1.112883448600769,
+      "learning_rate": 5.040103116178863e-08,
+      "loss": 0.5076,
+      "step": 5377
+    },
+    {
+      "epoch": 1.9640314040533138,
+      "grad_norm": 0.9256656169891357,
+      "learning_rate": 4.938838810295554e-08,
+      "loss": 0.5236,
+      "step": 5378
+    },
+    {
+      "epoch": 1.96439656746394,
+      "grad_norm": 0.9695578217506409,
+      "learning_rate": 4.8386008833225526e-08,
+      "loss": 0.5601,
+      "step": 5379
+    },
+    {
+      "epoch": 1.9647617308745664,
+      "grad_norm": 0.7416371703147888,
+      "learning_rate": 4.739389386828608e-08,
+      "loss": 0.5699,
+      "step": 5380
+    },
+    {
+      "epoch": 1.9651268942851927,
+      "grad_norm": 1.8474271297454834,
+      "learning_rate": 4.6412043718540024e-08,
+      "loss": 0.5528,
+      "step": 5381
+    },
+    {
+      "epoch": 1.965492057695819,
+      "grad_norm": 1.0167008638381958,
+      "learning_rate": 4.544045888910997e-08,
+      "loss": 0.5311,
+      "step": 5382
+    },
+    {
+      "epoch": 1.9658572211064451,
+      "grad_norm": 0.8540982604026794,
+      "learning_rate": 4.447913987983832e-08,
+      "loss": 0.5136,
+      "step": 5383
+    },
+    {
+      "epoch": 1.9662223845170714,
+      "grad_norm": 0.8130719065666199,
+      "learning_rate": 4.352808718528279e-08,
+      "loss": 0.5834,
+      "step": 5384
+    },
+    {
+      "epoch": 1.9665875479276975,
+      "grad_norm": 2.0834383964538574,
+      "learning_rate": 4.2587301294723105e-08,
+      "loss": 0.5196,
+      "step": 5385
+    },
+    {
+      "epoch": 1.9669527113383238,
+      "grad_norm": 1.1500024795532227,
+      "learning_rate": 4.165678269215656e-08,
+      "loss": 0.5363,
+      "step": 5386
+    },
+    {
+      "epoch": 1.9673178747489501,
+      "grad_norm": 1.5207279920578003,
+      "learning_rate": 4.073653185629578e-08,
+      "loss": 0.5292,
+      "step": 5387
+    },
+    {
+      "epoch": 1.9676830381595765,
+      "grad_norm": 0.7735336422920227,
+      "learning_rate": 3.982654926057539e-08,
+      "loss": 0.5164,
+      "step": 5388
+    },
+    {
+      "epoch": 1.9680482015702028,
+      "grad_norm": 0.9119135141372681,
+      "learning_rate": 3.8926835373143125e-08,
+      "loss": 0.5292,
+      "step": 5389
+    },
+    {
+      "epoch": 1.968413364980829,
+      "grad_norm": 0.876446545124054,
+      "learning_rate": 3.803739065686651e-08,
+      "loss": 0.5468,
+      "step": 5390
+    },
+    {
+      "epoch": 1.9687785283914552,
+      "grad_norm": 0.9272382259368896,
+      "learning_rate": 3.7158215569326194e-08,
+      "loss": 0.5084,
+      "step": 5391
+    },
+    {
+      "epoch": 1.9691436918020815,
+      "grad_norm": 0.9327825307846069,
+      "learning_rate": 3.628931056282703e-08,
+      "loss": 0.5548,
+      "step": 5392
+    },
+    {
+      "epoch": 1.9695088552127076,
+      "grad_norm": 1.1887319087982178,
+      "learning_rate": 3.5430676084384775e-08,
+      "loss": 0.5042,
+      "step": 5393
+    },
+    {
+      "epoch": 1.9698740186233339,
+      "grad_norm": 1.0209521055221558,
+      "learning_rate": 3.4582312575728306e-08,
+      "loss": 0.4762,
+      "step": 5394
+    },
+    {
+      "epoch": 1.9702391820339602,
+      "grad_norm": 1.0543705224990845,
+      "learning_rate": 3.3744220473312937e-08,
+      "loss": 0.5107,
+      "step": 5395
+    },
+    {
+      "epoch": 1.9706043454445865,
+      "grad_norm": 0.7981095314025879,
+      "learning_rate": 3.291640020829823e-08,
+      "loss": 0.5123,
+      "step": 5396
+    },
+    {
+      "epoch": 1.9709695088552128,
+      "grad_norm": 0.9885851144790649,
+      "learning_rate": 3.2098852206567944e-08,
+      "loss": 0.5234,
+      "step": 5397
+    },
+    {
+      "epoch": 1.971334672265839,
+      "grad_norm": 0.9228920936584473,
+      "learning_rate": 3.1291576888714536e-08,
+      "loss": 0.5839,
+      "step": 5398
+    },
+    {
+      "epoch": 1.9716998356764652,
+      "grad_norm": 1.0322990417480469,
+      "learning_rate": 3.0494574670050236e-08,
+      "loss": 0.5578,
+      "step": 5399
+    },
+    {
+      "epoch": 1.9720649990870913,
+      "grad_norm": 1.2793372869491577,
+      "learning_rate": 2.970784596060261e-08,
+      "loss": 0.4935,
+      "step": 5400
+    },
+    {
+      "epoch": 1.9724301624977176,
+      "grad_norm": 0.8219887614250183,
+      "learning_rate": 2.8931391165107902e-08,
+      "loss": 0.5135,
+      "step": 5401
+    },
+    {
+      "epoch": 1.972795325908344,
+      "grad_norm": 1.2734088897705078,
+      "learning_rate": 2.816521068302658e-08,
+      "loss": 0.5017,
+      "step": 5402
+    },
+    {
+      "epoch": 1.9731604893189703,
+      "grad_norm": 0.7988722920417786,
+      "learning_rate": 2.740930490852334e-08,
+      "loss": 0.5737,
+      "step": 5403
+    },
+    {
+      "epoch": 1.9735256527295966,
+      "grad_norm": 1.2488044500350952,
+      "learning_rate": 2.6663674230482663e-08,
+      "loss": 0.5502,
+      "step": 5404
+    },
+    {
+      "epoch": 1.9738908161402229,
+      "grad_norm": 0.9395003318786621,
+      "learning_rate": 2.5928319032499928e-08,
+      "loss": 0.4857,
+      "step": 5405
+    },
+    {
+      "epoch": 1.974255979550849,
+      "grad_norm": 0.849014937877655,
+      "learning_rate": 2.520323969288807e-08,
+      "loss": 0.5157,
+      "step": 5406
+    },
+    {
+      "epoch": 1.9746211429614753,
+      "grad_norm": 0.9843391180038452,
+      "learning_rate": 2.4488436584670928e-08,
+      "loss": 0.521,
+      "step": 5407
+    },
+    {
+      "epoch": 1.9749863063721014,
+      "grad_norm": 1.1458913087844849,
+      "learning_rate": 2.378391007558767e-08,
+      "loss": 0.5219,
+      "step": 5408
+    },
+    {
+      "epoch": 1.9753514697827277,
+      "grad_norm": 1.136871099472046,
+      "learning_rate": 2.3089660528083923e-08,
+      "loss": 0.475,
+      "step": 5409
+    },
+    {
+      "epoch": 1.975716633193354,
+      "grad_norm": 0.814587414264679,
+      "learning_rate": 2.240568829932732e-08,
+      "loss": 0.5254,
+      "step": 5410
+    },
+    {
+      "epoch": 1.9760817966039803,
+      "grad_norm": 0.8695433139801025,
+      "learning_rate": 2.173199374119417e-08,
+      "loss": 0.5297,
+      "step": 5411
+    },
+    {
+      "epoch": 1.9764469600146066,
+      "grad_norm": 0.9298883676528931,
+      "learning_rate": 2.106857720027167e-08,
+      "loss": 0.5593,
+      "step": 5412
+    },
+    {
+      "epoch": 1.976812123425233,
+      "grad_norm": 0.7862787842750549,
+      "learning_rate": 2.041543901786236e-08,
+      "loss": 0.5323,
+      "step": 5413
+    },
+    {
+      "epoch": 1.977177286835859,
+      "grad_norm": 0.9040335416793823,
+      "learning_rate": 1.9772579529977463e-08,
+      "loss": 0.537,
+      "step": 5414
+    },
+    {
+      "epoch": 1.9775424502464853,
+      "grad_norm": 0.9744546413421631,
+      "learning_rate": 1.913999906734354e-08,
+      "loss": 0.4876,
+      "step": 5415
+    },
+    {
+      "epoch": 1.9779076136571114,
+      "grad_norm": 1.0703387260437012,
+      "learning_rate": 1.851769795540026e-08,
+      "loss": 0.5482,
+      "step": 5416
+    },
+    {
+      "epoch": 1.9782727770677377,
+      "grad_norm": 0.7839832305908203,
+      "learning_rate": 1.7905676514293757e-08,
+      "loss": 0.5452,
+      "step": 5417
+    },
+    {
+      "epoch": 1.978637940478364,
+      "grad_norm": 0.7852559685707092,
+      "learning_rate": 1.7303935058885502e-08,
+      "loss": 0.5621,
+      "step": 5418
+    },
+    {
+      "epoch": 1.9790031038889904,
+      "grad_norm": 0.9149523973464966,
+      "learning_rate": 1.6712473898745642e-08,
+      "loss": 0.5286,
+      "step": 5419
+    },
+    {
+      "epoch": 1.9793682672996167,
+      "grad_norm": 0.8258424997329712,
+      "learning_rate": 1.6131293338157438e-08,
+      "loss": 0.5724,
+      "step": 5420
+    },
+    {
+      "epoch": 1.979733430710243,
+      "grad_norm": 0.9903010129928589,
+      "learning_rate": 1.55603936761195e-08,
+      "loss": 0.5513,
+      "step": 5421
+    },
+    {
+      "epoch": 1.980098594120869,
+      "grad_norm": 1.2193019390106201,
+      "learning_rate": 1.4999775206330224e-08,
+      "loss": 0.4595,
+      "step": 5422
+    },
+    {
+      "epoch": 1.9804637575314954,
+      "grad_norm": 1.0448379516601562,
+      "learning_rate": 1.444943821721001e-08,
+      "loss": 0.5557,
+      "step": 5423
+    },
+    {
+      "epoch": 1.9808289209421215,
+      "grad_norm": 0.9651296138763428,
+      "learning_rate": 1.390938299188349e-08,
+      "loss": 0.5552,
+      "step": 5424
+    },
+    {
+      "epoch": 1.9811940843527478,
+      "grad_norm": 1.150664210319519,
+      "learning_rate": 1.337960980818842e-08,
+      "loss": 0.5406,
+      "step": 5425
+    },
+    {
+      "epoch": 1.981559247763374,
+      "grad_norm": 0.8297967910766602,
+      "learning_rate": 1.2860118938669008e-08,
+      "loss": 0.5714,
+      "step": 5426
+    },
+    {
+      "epoch": 1.9819244111740004,
+      "grad_norm": 0.8567715287208557,
+      "learning_rate": 1.2350910650587022e-08,
+      "loss": 0.5518,
+      "step": 5427
+    },
+    {
+      "epoch": 1.9822895745846267,
+      "grad_norm": 0.8794555068016052,
+      "learning_rate": 1.1851985205904026e-08,
+      "loss": 0.5276,
+      "step": 5428
+    },
+    {
+      "epoch": 1.9826547379952528,
+      "grad_norm": 0.735395073890686,
+      "learning_rate": 1.1363342861301363e-08,
+      "loss": 0.5508,
+      "step": 5429
+    },
+    {
+      "epoch": 1.9830199014058791,
+      "grad_norm": 0.86039799451828,
+      "learning_rate": 1.0884983868166832e-08,
+      "loss": 0.4865,
+      "step": 5430
+    },
+    {
+      "epoch": 1.9833850648165052,
+      "grad_norm": 1.0929391384124756,
+      "learning_rate": 1.0416908472592468e-08,
+      "loss": 0.473,
+      "step": 5431
+    },
+    {
+      "epoch": 1.9837502282271315,
+      "grad_norm": 0.8757457137107849,
+      "learning_rate": 9.959116915387868e-09,
+      "loss": 0.5257,
+      "step": 5432
+    },
+    {
+      "epoch": 1.9841153916377579,
+      "grad_norm": 1.0023913383483887,
+      "learning_rate": 9.51160943206686e-09,
+      "loss": 0.533,
+      "step": 5433
+    },
+    {
+      "epoch": 1.9844805550483842,
+      "grad_norm": 0.9719913601875305,
+      "learning_rate": 9.074386252854172e-09,
+      "loss": 0.4845,
+      "step": 5434
+    },
+    {
+      "epoch": 1.9848457184590105,
+      "grad_norm": 0.8217945098876953,
+      "learning_rate": 8.647447602683212e-09,
+      "loss": 0.5406,
+      "step": 5435
+    },
+    {
+      "epoch": 1.9852108818696368,
+      "grad_norm": 1.0762484073638916,
+      "learning_rate": 8.23079370119828e-09,
+      "loss": 0.4913,
+      "step": 5436
+    },
+    {
+      "epoch": 1.9855760452802629,
+      "grad_norm": 0.937466561794281,
+      "learning_rate": 7.824424762750137e-09,
+      "loss": 0.5121,
+      "step": 5437
+    },
+    {
+      "epoch": 1.9859412086908892,
+      "grad_norm": 0.7162231802940369,
+      "learning_rate": 7.428340996400441e-09,
+      "loss": 0.5219,
+      "step": 5438
+    },
+    {
+      "epoch": 1.9863063721015153,
+      "grad_norm": 0.8674814105033875,
+      "learning_rate": 7.042542605915081e-09,
+      "loss": 0.5443,
+      "step": 5439
+    },
+    {
+      "epoch": 1.9866715355121416,
+      "grad_norm": 0.9962090849876404,
+      "learning_rate": 6.667029789775292e-09,
+      "loss": 0.5136,
+      "step": 5440
+    },
+    {
+      "epoch": 1.987036698922768,
+      "grad_norm": 0.9544779658317566,
+      "learning_rate": 6.301802741166541e-09,
+      "loss": 0.5163,
+      "step": 5441
+    },
+    {
+      "epoch": 1.9874018623333942,
+      "grad_norm": 1.1398656368255615,
+      "learning_rate": 5.946861647982971e-09,
+      "loss": 0.5375,
+      "step": 5442
+    },
+    {
+      "epoch": 1.9877670257440205,
+      "grad_norm": 0.8499765396118164,
+      "learning_rate": 5.602206692827405e-09,
+      "loss": 0.4969,
+      "step": 5443
+    },
+    {
+      "epoch": 1.9881321891546468,
+      "grad_norm": 0.8699721097946167,
+      "learning_rate": 5.267838053011343e-09,
+      "loss": 0.5186,
+      "step": 5444
+    },
+    {
+      "epoch": 1.988497352565273,
+      "grad_norm": 0.5777058601379395,
+      "learning_rate": 4.943755900554958e-09,
+      "loss": 0.5463,
+      "step": 5445
+    },
+    {
+      "epoch": 1.9888625159758992,
+      "grad_norm": 0.7768506407737732,
+      "learning_rate": 4.629960402182665e-09,
+      "loss": 0.5792,
+      "step": 5446
+    },
+    {
+      "epoch": 1.9892276793865253,
+      "grad_norm": 0.8641973733901978,
+      "learning_rate": 4.326451719334213e-09,
+      "loss": 0.5445,
+      "step": 5447
+    },
+    {
+      "epoch": 1.9895928427971517,
+      "grad_norm": 1.0852634906768799,
+      "learning_rate": 4.033230008146927e-09,
+      "loss": 0.5142,
+      "step": 5448
+    },
+    {
+      "epoch": 1.989958006207778,
+      "grad_norm": 0.8240854740142822,
+      "learning_rate": 3.750295419475692e-09,
+      "loss": 0.5847,
+      "step": 5449
+    },
+    {
+      "epoch": 1.9903231696184043,
+      "grad_norm": 1.0229144096374512,
+      "learning_rate": 3.477648098879627e-09,
+      "loss": 0.5779,
+      "step": 5450
+    },
+    {
+      "epoch": 1.9906883330290306,
+      "grad_norm": 0.9161554574966431,
+      "learning_rate": 3.2152881866198695e-09,
+      "loss": 0.5316,
+      "step": 5451
+    },
+    {
+      "epoch": 1.991053496439657,
+      "grad_norm": 1.0170706510543823,
+      "learning_rate": 2.9632158176751134e-09,
+      "loss": 0.5254,
+      "step": 5452
+    },
+    {
+      "epoch": 1.991418659850283,
+      "grad_norm": 0.9410620927810669,
+      "learning_rate": 2.72143112172607e-09,
+      "loss": 0.548,
+      "step": 5453
+    },
+    {
+      "epoch": 1.9917838232609093,
+      "grad_norm": 1.0087649822235107,
+      "learning_rate": 2.489934223157686e-09,
+      "loss": 0.5626,
+      "step": 5454
+    },
+    {
+      "epoch": 1.9921489866715354,
+      "grad_norm": 0.9506940841674805,
+      "learning_rate": 2.268725241068026e-09,
+      "loss": 0.5505,
+      "step": 5455
+    },
+    {
+      "epoch": 1.9925141500821617,
+      "grad_norm": 0.9841387867927551,
+      "learning_rate": 2.0578042892616114e-09,
+      "loss": 0.523,
+      "step": 5456
+    },
+    {
+      "epoch": 1.992879313492788,
+      "grad_norm": 0.7301874756813049,
+      "learning_rate": 1.8571714762471993e-09,
+      "loss": 0.5695,
+      "step": 5457
+    },
+    {
+      "epoch": 1.9932444769034143,
+      "grad_norm": 1.060308575630188,
+      "learning_rate": 1.6668269052422248e-09,
+      "loss": 0.5644,
+      "step": 5458
+    },
+    {
+      "epoch": 1.9936096403140406,
+      "grad_norm": 0.9705191254615784,
+      "learning_rate": 1.4867706741727994e-09,
+      "loss": 0.5026,
+      "step": 5459
+    },
+    {
+      "epoch": 1.9939748037246667,
+      "grad_norm": 1.1569206714630127,
+      "learning_rate": 1.3170028756670506e-09,
+      "loss": 0.5347,
+      "step": 5460
+    },
+    {
+      "epoch": 1.994339967135293,
+      "grad_norm": 1.3227145671844482,
+      "learning_rate": 1.157523597068444e-09,
+      "loss": 0.5019,
+      "step": 5461
+    },
+    {
+      "epoch": 1.9947051305459191,
+      "grad_norm": 1.1881853342056274,
+      "learning_rate": 1.0083329204180204e-09,
+      "loss": 0.5177,
+      "step": 5462
+    },
+    {
+      "epoch": 1.9950702939565454,
+      "grad_norm": 0.8623504638671875,
+      "learning_rate": 8.694309224721586e-10,
+      "loss": 0.5263,
+      "step": 5463
+    },
+    {
+      "epoch": 1.9954354573671718,
+      "grad_norm": 0.7508077025413513,
+      "learning_rate": 7.408176746892537e-10,
+      "loss": 0.5711,
+      "step": 5464
+    },
+    {
+      "epoch": 1.995800620777798,
+      "grad_norm": 1.054730772972107,
+      "learning_rate": 6.224932432363773e-10,
+      "loss": 0.5274,
+      "step": 5465
+    },
+    {
+      "epoch": 1.9961657841884244,
+      "grad_norm": 0.898906946182251,
+      "learning_rate": 5.144576889826169e-10,
+      "loss": 0.5164,
+      "step": 5466
+    },
+    {
+      "epoch": 1.9965309475990507,
+      "grad_norm": 0.8694043159484863,
+      "learning_rate": 4.1671106751239866e-10,
+      "loss": 0.4893,
+      "step": 5467
+    },
+    {
+      "epoch": 1.9968961110096768,
+      "grad_norm": 0.9791451096534729,
+      "learning_rate": 3.2925342911216405e-10,
+      "loss": 0.5417,
+      "step": 5468
+    },
+    {
+      "epoch": 1.997261274420303,
+      "grad_norm": 0.8434933423995972,
+      "learning_rate": 2.5208481877259103e-10,
+      "loss": 0.554,
+      "step": 5469
+    },
+    {
+      "epoch": 1.9976264378309292,
+      "grad_norm": 0.9651261568069458,
+      "learning_rate": 1.8520527619747543e-10,
+      "loss": 0.542,
+      "step": 5470
+    },
+    {
+      "epoch": 1.9979916012415555,
+      "grad_norm": 0.9461495280265808,
+      "learning_rate": 1.2861483579040824e-10,
+      "loss": 0.5278,
+      "step": 5471
+    },
+    {
+      "epoch": 1.9983567646521818,
+      "grad_norm": 1.4563730955123901,
+      "learning_rate": 8.231352666587811e-11,
+      "loss": 0.5299,
+      "step": 5472
+    },
+    {
+      "epoch": 1.9987219280628081,
+      "grad_norm": 0.9893348217010498,
+      "learning_rate": 4.630137264483026e-11,
+      "loss": 0.5204,
+      "step": 5473
+    },
+    {
+      "epoch": 1.9990870914734344,
+      "grad_norm": 0.9429623484611511,
+      "learning_rate": 2.0578392252446066e-11,
+      "loss": 0.5097,
+      "step": 5474
+    },
+    {
+      "epoch": 1.9994522548840608,
+      "grad_norm": 0.7210927605628967,
+      "learning_rate": 5.1445987248044394e-12,
+      "loss": 0.5456,
+      "step": 5475
+    },
+    {
+      "epoch": 1.9998174182946868,
+      "grad_norm": 1.0174647569656372,
+      "learning_rate": 0.0,
+      "loss": 0.4995,
+      "step": 5476
+    },
+    {
+      "epoch": 1.9998174182946868,
+      "step": 5476,
+      "total_flos": 0.0,
+      "train_loss": 0.11873858405169646,
+      "train_runtime": 35260.5195,
+      "train_samples_per_second": 4.971,
+      "train_steps_per_second": 0.155
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5476,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}