diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9597 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.99835255354201,
+  "eval_steps": 500,
+  "global_step": 1365,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002196595277320154,
+      "grad_norm": 2467772.75,
+      "learning_rate": 0.0,
+      "loss": 3.8094,
+      "step": 1
+    },
+    {
+      "epoch": 0.004393190554640308,
+      "grad_norm": 8924863.0,
+      "learning_rate": 3.6496350364963505e-07,
+      "loss": 2.7355,
+      "step": 2
+    },
+    {
+      "epoch": 0.006589785831960461,
+      "grad_norm": 6245380.5,
+      "learning_rate": 7.299270072992701e-07,
+      "loss": 2.1256,
+      "step": 3
+    },
+    {
+      "epoch": 0.008786381109280615,
+      "grad_norm": 5445801.5,
+      "learning_rate": 1.0948905109489052e-06,
+      "loss": 2.7073,
+      "step": 4
+    },
+    {
+      "epoch": 0.010982976386600769,
+      "grad_norm": 6924310.0,
+      "learning_rate": 1.4598540145985402e-06,
+      "loss": 2.689,
+      "step": 5
+    },
+    {
+      "epoch": 0.013179571663920923,
+      "grad_norm": 3737947.75,
+      "learning_rate": 1.824817518248175e-06,
+      "loss": 2.6055,
+      "step": 6
+    },
+    {
+      "epoch": 0.015376166941241077,
+      "grad_norm": 1185103360.0,
+      "learning_rate": 2.1897810218978103e-06,
+      "loss": 3.0763,
+      "step": 7
+    },
+    {
+      "epoch": 0.01757276221856123,
+      "grad_norm": 4184358.5,
+      "learning_rate": 2.5547445255474454e-06,
+      "loss": 2.3816,
+      "step": 8
+    },
+    {
+      "epoch": 0.019769357495881382,
+      "grad_norm": 9564488.0,
+      "learning_rate": 2.9197080291970804e-06,
+      "loss": 2.7307,
+      "step": 9
+    },
+    {
+      "epoch": 0.021965952773201538,
+      "grad_norm": 3663107.0,
+      "learning_rate": 3.2846715328467155e-06,
+      "loss": 2.5384,
+      "step": 10
+    },
+    {
+      "epoch": 0.02416254805052169,
+      "grad_norm": 3062231.0,
+      "learning_rate": 3.64963503649635e-06,
+      "loss": 2.5786,
+      "step": 11
+    },
+    {
+      "epoch": 0.026359143327841845,
+      "grad_norm": 6068510.0,
+      "learning_rate": 4.014598540145985e-06,
+      "loss": 2.513,
+      "step": 12
+    },
+    {
+      "epoch": 0.028555738605161998,
+      "grad_norm": 5020339.0,
+      "learning_rate": 4.379562043795621e-06,
+      "loss": 2.8808,
+      "step": 13
+    },
+    {
+      "epoch": 0.030752333882482153,
+      "grad_norm": 5829525.5,
+      "learning_rate": 4.744525547445255e-06,
+      "loss": 2.3627,
+      "step": 14
+    },
+    {
+      "epoch": 0.032948929159802305,
+      "grad_norm": 8788993.0,
+      "learning_rate": 5.109489051094891e-06,
+      "loss": 3.0105,
+      "step": 15
+    },
+    {
+      "epoch": 0.03514552443712246,
+      "grad_norm": 3496011.75,
+      "learning_rate": 5.474452554744526e-06,
+      "loss": 2.6785,
+      "step": 16
+    },
+    {
+      "epoch": 0.037342119714442616,
+      "grad_norm": 7778505.5,
+      "learning_rate": 5.839416058394161e-06,
+      "loss": 2.3278,
+      "step": 17
+    },
+    {
+      "epoch": 0.039538714991762765,
+      "grad_norm": 59079588.0,
+      "learning_rate": 6.204379562043796e-06,
+      "loss": 2.3634,
+      "step": 18
+    },
+    {
+      "epoch": 0.04173531026908292,
+      "grad_norm": 10647139.0,
+      "learning_rate": 6.569343065693431e-06,
+      "loss": 3.1693,
+      "step": 19
+    },
+    {
+      "epoch": 0.043931905546403076,
+      "grad_norm": 10929726.0,
+      "learning_rate": 6.9343065693430655e-06,
+      "loss": 2.6225,
+      "step": 20
+    },
+    {
+      "epoch": 0.04612850082372323,
+      "grad_norm": 5365907.0,
+      "learning_rate": 7.2992700729927e-06,
+      "loss": 2.699,
+      "step": 21
+    },
+    {
+      "epoch": 0.04832509610104338,
+      "grad_norm": 13093565.0,
+      "learning_rate": 7.664233576642336e-06,
+      "loss": 3.5333,
+      "step": 22
+    },
+    {
+      "epoch": 0.050521691378363535,
+      "grad_norm": 4448815.5,
+      "learning_rate": 8.02919708029197e-06,
+      "loss": 2.7296,
+      "step": 23
+    },
+    {
+      "epoch": 0.05271828665568369,
+      "grad_norm": 4401016.0,
+      "learning_rate": 8.394160583941606e-06,
+      "loss": 2.6338,
+      "step": 24
+    },
+    {
+      "epoch": 0.054914881933003847,
+      "grad_norm": 8596703.0,
+      "learning_rate": 8.759124087591241e-06,
+      "loss": 3.0268,
+      "step": 25
+    },
+    {
+      "epoch": 0.057111477210323995,
+      "grad_norm": 5117410.0,
+      "learning_rate": 9.124087591240877e-06,
+      "loss": 2.4458,
+      "step": 26
+    },
+    {
+      "epoch": 0.05930807248764415,
+      "grad_norm": 7494324.0,
+      "learning_rate": 9.48905109489051e-06,
+      "loss": 2.4891,
+      "step": 27
+    },
+    {
+      "epoch": 0.061504667764964306,
+      "grad_norm": 25724502.0,
+      "learning_rate": 9.854014598540148e-06,
+      "loss": 2.5657,
+      "step": 28
+    },
+    {
+      "epoch": 0.06370126304228446,
+      "grad_norm": 47939308.0,
+      "learning_rate": 1.0218978102189781e-05,
+      "loss": 2.4896,
+      "step": 29
+    },
+    {
+      "epoch": 0.06589785831960461,
+      "grad_norm": 5460008.5,
+      "learning_rate": 1.0583941605839417e-05,
+      "loss": 3.0905,
+      "step": 30
+    },
+    {
+      "epoch": 0.06809445359692477,
+      "grad_norm": 2835397.0,
+      "learning_rate": 1.0948905109489052e-05,
+      "loss": 2.4545,
+      "step": 31
+    },
+    {
+      "epoch": 0.07029104887424492,
+      "grad_norm": 3232029.25,
+      "learning_rate": 1.1313868613138686e-05,
+      "loss": 2.362,
+      "step": 32
+    },
+    {
+      "epoch": 0.07248764415156507,
+      "grad_norm": 2468841.75,
+      "learning_rate": 1.1678832116788322e-05,
+      "loss": 2.8764,
+      "step": 33
+    },
+    {
+      "epoch": 0.07468423942888523,
+      "grad_norm": 2014460.25,
+      "learning_rate": 1.2043795620437957e-05,
+      "loss": 2.0141,
+      "step": 34
+    },
+    {
+      "epoch": 0.07688083470620538,
+      "grad_norm": 3662961.75,
+      "learning_rate": 1.2408759124087593e-05,
+      "loss": 2.6523,
+      "step": 35
+    },
+    {
+      "epoch": 0.07907742998352553,
+      "grad_norm": 6703157.0,
+      "learning_rate": 1.2773722627737228e-05,
+      "loss": 2.5558,
+      "step": 36
+    },
+    {
+      "epoch": 0.08127402526084569,
+      "grad_norm": 3148150.25,
+      "learning_rate": 1.3138686131386862e-05,
+      "loss": 2.036,
+      "step": 37
+    },
+    {
+      "epoch": 0.08347062053816584,
+      "grad_norm": 5571218.0,
+      "learning_rate": 1.3503649635036497e-05,
+      "loss": 1.9416,
+      "step": 38
+    },
+    {
+      "epoch": 0.085667215815486,
+      "grad_norm": 1321473.0,
+      "learning_rate": 1.3868613138686131e-05,
+      "loss": 2.6681,
+      "step": 39
+    },
+    {
+      "epoch": 0.08786381109280615,
+      "grad_norm": 4135731.25,
+      "learning_rate": 1.4233576642335767e-05,
+      "loss": 1.8817,
+      "step": 40
+    },
+    {
+      "epoch": 0.0900604063701263,
+      "grad_norm": 7996901.0,
+      "learning_rate": 1.45985401459854e-05,
+      "loss": 3.2942,
+      "step": 41
+    },
+    {
+      "epoch": 0.09225700164744646,
+      "grad_norm": 4043322.25,
+      "learning_rate": 1.496350364963504e-05,
+      "loss": 2.206,
+      "step": 42
+    },
+    {
+      "epoch": 0.09445359692476661,
+      "grad_norm": 7796392.0,
+      "learning_rate": 1.5328467153284673e-05,
+      "loss": 2.4073,
+      "step": 43
+    },
+    {
+      "epoch": 0.09665019220208676,
+      "grad_norm": 3322222.0,
+      "learning_rate": 1.569343065693431e-05,
+      "loss": 2.4769,
+      "step": 44
+    },
+    {
+      "epoch": 0.09884678747940692,
+      "grad_norm": 2561443.0,
+      "learning_rate": 1.605839416058394e-05,
+      "loss": 2.5182,
+      "step": 45
+    },
+    {
+      "epoch": 0.10104338275672707,
+      "grad_norm": 4878629.5,
+      "learning_rate": 1.6423357664233576e-05,
+      "loss": 2.3746,
+      "step": 46
+    },
+    {
+      "epoch": 0.10323997803404723,
+      "grad_norm": 1863335.625,
+      "learning_rate": 1.678832116788321e-05,
+      "loss": 2.2903,
+      "step": 47
+    },
+    {
+      "epoch": 0.10543657331136738,
+      "grad_norm": 5373952.5,
+      "learning_rate": 1.715328467153285e-05,
+      "loss": 2.7225,
+      "step": 48
+    },
+    {
+      "epoch": 0.10763316858868753,
+      "grad_norm": 2341575.75,
+      "learning_rate": 1.7518248175182482e-05,
+      "loss": 3.2183,
+      "step": 49
+    },
+    {
+      "epoch": 0.10982976386600769,
+      "grad_norm": 2184809.75,
+      "learning_rate": 1.7883211678832118e-05,
+      "loss": 2.0685,
+      "step": 50
+    },
+    {
+      "epoch": 0.11202635914332784,
+      "grad_norm": 4236161.5,
+      "learning_rate": 1.8248175182481753e-05,
+      "loss": 2.2825,
+      "step": 51
+    },
+    {
+      "epoch": 0.11422295442064799,
+      "grad_norm": 3270804.5,
+      "learning_rate": 1.861313868613139e-05,
+      "loss": 2.148,
+      "step": 52
+    },
+    {
+      "epoch": 0.11641954969796815,
+      "grad_norm": 7266757.5,
+      "learning_rate": 1.897810218978102e-05,
+      "loss": 2.2311,
+      "step": 53
+    },
+    {
+      "epoch": 0.1186161449752883,
+      "grad_norm": 2214112.0,
+      "learning_rate": 1.934306569343066e-05,
+      "loss": 2.1008,
+      "step": 54
+    },
+    {
+      "epoch": 0.12081274025260846,
+      "grad_norm": 2281926.5,
+      "learning_rate": 1.9708029197080295e-05,
+      "loss": 2.3559,
+      "step": 55
+    },
+    {
+      "epoch": 0.12300933552992861,
+      "grad_norm": 7432281.0,
+      "learning_rate": 2.0072992700729927e-05,
+      "loss": 2.1465,
+      "step": 56
+    },
+    {
+      "epoch": 0.12520593080724876,
+      "grad_norm": 5022644.5,
+      "learning_rate": 2.0437956204379563e-05,
+      "loss": 1.8344,
+      "step": 57
+    },
+    {
+      "epoch": 0.12740252608456892,
+      "grad_norm": 5462314.0,
+      "learning_rate": 2.08029197080292e-05,
+      "loss": 2.4385,
+      "step": 58
+    },
+    {
+      "epoch": 0.12959912136188906,
+      "grad_norm": 2613603.25,
+      "learning_rate": 2.1167883211678834e-05,
+      "loss": 2.1183,
+      "step": 59
+    },
+    {
+      "epoch": 0.13179571663920922,
+      "grad_norm": 7366409.5,
+      "learning_rate": 2.1532846715328466e-05,
+      "loss": 1.9459,
+      "step": 60
+    },
+    {
+      "epoch": 0.13399231191652938,
+      "grad_norm": 7063570.0,
+      "learning_rate": 2.1897810218978105e-05,
+      "loss": 2.1052,
+      "step": 61
+    },
+    {
+      "epoch": 0.13618890719384955,
+      "grad_norm": 6508024.5,
+      "learning_rate": 2.226277372262774e-05,
+      "loss": 1.8749,
+      "step": 62
+    },
+    {
+      "epoch": 0.13838550247116968,
+      "grad_norm": 6069544.5,
+      "learning_rate": 2.2627737226277372e-05,
+      "loss": 1.623,
+      "step": 63
+    },
+    {
+      "epoch": 0.14058209774848984,
+      "grad_norm": 9291386.0,
+      "learning_rate": 2.2992700729927008e-05,
+      "loss": 1.8306,
+      "step": 64
+    },
+    {
+      "epoch": 0.14277869302581,
+      "grad_norm": 2420929.25,
+      "learning_rate": 2.3357664233576643e-05,
+      "loss": 2.1262,
+      "step": 65
+    },
+    {
+      "epoch": 0.14497528830313014,
+      "grad_norm": 3247623.0,
+      "learning_rate": 2.372262773722628e-05,
+      "loss": 1.4933,
+      "step": 66
+    },
+    {
+      "epoch": 0.1471718835804503,
+      "grad_norm": 4324912.0,
+      "learning_rate": 2.4087591240875914e-05,
+      "loss": 1.7588,
+      "step": 67
+    },
+    {
+      "epoch": 0.14936847885777046,
+      "grad_norm": 3899299.0,
+      "learning_rate": 2.445255474452555e-05,
+      "loss": 1.5728,
+      "step": 68
+    },
+    {
+      "epoch": 0.1515650741350906,
+      "grad_norm": 4147348.0,
+      "learning_rate": 2.4817518248175185e-05,
+      "loss": 1.913,
+      "step": 69
+    },
+    {
+      "epoch": 0.15376166941241076,
+      "grad_norm": 2594638.0,
+      "learning_rate": 2.518248175182482e-05,
+      "loss": 1.6764,
+      "step": 70
+    },
+    {
+      "epoch": 0.15595826468973092,
+      "grad_norm": 2653024.75,
+      "learning_rate": 2.5547445255474456e-05,
+      "loss": 2.0145,
+      "step": 71
+    },
+    {
+      "epoch": 0.15815485996705106,
+      "grad_norm": 3380332.5,
+      "learning_rate": 2.591240875912409e-05,
+      "loss": 1.685,
+      "step": 72
+    },
+    {
+      "epoch": 0.16035145524437122,
+      "grad_norm": 7767693.0,
+      "learning_rate": 2.6277372262773724e-05,
+      "loss": 1.9886,
+      "step": 73
+    },
+    {
+      "epoch": 0.16254805052169138,
+      "grad_norm": 3916485.0,
+      "learning_rate": 2.664233576642336e-05,
+      "loss": 1.7766,
+      "step": 74
+    },
+    {
+      "epoch": 0.16474464579901152,
+      "grad_norm": 3991656.0,
+      "learning_rate": 2.7007299270072995e-05,
+      "loss": 1.781,
+      "step": 75
+    },
+    {
+      "epoch": 0.16694124107633168,
+      "grad_norm": 6003922.0,
+      "learning_rate": 2.737226277372263e-05,
+      "loss": 1.8282,
+      "step": 76
+    },
+    {
+      "epoch": 0.16913783635365184,
+      "grad_norm": 4023165.0,
+      "learning_rate": 2.7737226277372262e-05,
+      "loss": 1.8054,
+      "step": 77
+    },
+    {
+      "epoch": 0.171334431630972,
+      "grad_norm": 2995137.75,
+      "learning_rate": 2.8102189781021898e-05,
+      "loss": 1.3594,
+      "step": 78
+    },
+    {
+      "epoch": 0.17353102690829214,
+      "grad_norm": 2938031.25,
+      "learning_rate": 2.8467153284671533e-05,
+      "loss": 1.6536,
+      "step": 79
+    },
+    {
+      "epoch": 0.1757276221856123,
+      "grad_norm": 7882325.0,
+      "learning_rate": 2.883211678832117e-05,
+      "loss": 2.185,
+      "step": 80
+    },
+    {
+      "epoch": 0.17792421746293247,
+      "grad_norm": 6382671.0,
+      "learning_rate": 2.91970802919708e-05,
+      "loss": 1.4565,
+      "step": 81
+    },
+    {
+      "epoch": 0.1801208127402526,
+      "grad_norm": 3907469.5,
+      "learning_rate": 2.9562043795620443e-05,
+      "loss": 1.472,
+      "step": 82
+    },
+    {
+      "epoch": 0.18231740801757276,
+      "grad_norm": 4005602.25,
+      "learning_rate": 2.992700729927008e-05,
+      "loss": 1.8337,
+      "step": 83
+    },
+    {
+      "epoch": 0.18451400329489293,
+      "grad_norm": 2590928.25,
+      "learning_rate": 3.029197080291971e-05,
+      "loss": 1.8806,
+      "step": 84
+    },
+    {
+      "epoch": 0.18671059857221306,
+      "grad_norm": 7677667.5,
+      "learning_rate": 3.0656934306569346e-05,
+      "loss": 1.6737,
+      "step": 85
+    },
+    {
+      "epoch": 0.18890719384953322,
+      "grad_norm": 4423182.0,
+      "learning_rate": 3.102189781021898e-05,
+      "loss": 1.7264,
+      "step": 86
+    },
+    {
+      "epoch": 0.19110378912685339,
+      "grad_norm": 6856657.5,
+      "learning_rate": 3.138686131386862e-05,
+      "loss": 1.7004,
+      "step": 87
+    },
+    {
+      "epoch": 0.19330038440417352,
+      "grad_norm": 5635655.5,
+      "learning_rate": 3.175182481751825e-05,
+      "loss": 1.6449,
+      "step": 88
+    },
+    {
+      "epoch": 0.19549697968149368,
+      "grad_norm": 4043275.5,
+      "learning_rate": 3.211678832116788e-05,
+      "loss": 1.4042,
+      "step": 89
+    },
+    {
+      "epoch": 0.19769357495881384,
+      "grad_norm": 2330061.0,
+      "learning_rate": 3.248175182481752e-05,
+      "loss": 1.9056,
+      "step": 90
+    },
+    {
+      "epoch": 0.19989017023613398,
+      "grad_norm": 6289549.5,
+      "learning_rate": 3.284671532846715e-05,
+      "loss": 1.6793,
+      "step": 91
+    },
+    {
+      "epoch": 0.20208676551345414,
+      "grad_norm": 4284888.5,
+      "learning_rate": 3.321167883211679e-05,
+      "loss": 1.422,
+      "step": 92
+    },
+    {
+      "epoch": 0.2042833607907743,
+      "grad_norm": 3451985.75,
+      "learning_rate": 3.357664233576642e-05,
+      "loss": 1.6215,
+      "step": 93
+    },
+    {
+      "epoch": 0.20647995606809447,
+      "grad_norm": 2298407.75,
+      "learning_rate": 3.3941605839416055e-05,
+      "loss": 1.8692,
+      "step": 94
+    },
+    {
+      "epoch": 0.2086765513454146,
+      "grad_norm": 2612601.5,
+      "learning_rate": 3.43065693430657e-05,
+      "loss": 1.8131,
+      "step": 95
+    },
+    {
+      "epoch": 0.21087314662273476,
+      "grad_norm": 1575659.625,
+      "learning_rate": 3.467153284671533e-05,
+      "loss": 1.6159,
+      "step": 96
+    },
+    {
+      "epoch": 0.21306974190005493,
+      "grad_norm": 3489119.75,
+      "learning_rate": 3.5036496350364965e-05,
+      "loss": 1.4007,
+      "step": 97
+    },
+    {
+      "epoch": 0.21526633717737506,
+      "grad_norm": 4576117.5,
+      "learning_rate": 3.5401459854014604e-05,
+      "loss": 1.8109,
+      "step": 98
+    },
+    {
+      "epoch": 0.21746293245469522,
+      "grad_norm": 2598908.0,
+      "learning_rate": 3.5766423357664236e-05,
+      "loss": 1.7825,
+      "step": 99
+    },
+    {
+      "epoch": 0.21965952773201539,
+      "grad_norm": 7803835.5,
+      "learning_rate": 3.613138686131387e-05,
+      "loss": 1.9972,
+      "step": 100
+    },
+    {
+      "epoch": 0.22185612300933552,
+      "grad_norm": 3021372.75,
+      "learning_rate": 3.649635036496351e-05,
+      "loss": 1.8425,
+      "step": 101
+    },
+    {
+      "epoch": 0.22405271828665568,
+      "grad_norm": 6223615.0,
+      "learning_rate": 3.686131386861314e-05,
+      "loss": 1.63,
+      "step": 102
+    },
+    {
+      "epoch": 0.22624931356397585,
+      "grad_norm": 5016269.5,
+      "learning_rate": 3.722627737226278e-05,
+      "loss": 1.7721,
+      "step": 103
+    },
+    {
+      "epoch": 0.22844590884129598,
+      "grad_norm": 3646410.75,
+      "learning_rate": 3.759124087591241e-05,
+      "loss": 1.556,
+      "step": 104
+    },
+    {
+      "epoch": 0.23064250411861614,
+      "grad_norm": 5376572.0,
+      "learning_rate": 3.795620437956204e-05,
+      "loss": 1.6136,
+      "step": 105
+    },
+    {
+      "epoch": 0.2328390993959363,
+      "grad_norm": 2385716.5,
+      "learning_rate": 3.832116788321168e-05,
+      "loss": 1.6672,
+      "step": 106
+    },
+    {
+      "epoch": 0.23503569467325644,
+      "grad_norm": 18639212.0,
+      "learning_rate": 3.868613138686132e-05,
+      "loss": 1.4671,
+      "step": 107
+    },
+    {
+      "epoch": 0.2372322899505766,
+      "grad_norm": 1677811.25,
+      "learning_rate": 3.905109489051095e-05,
+      "loss": 1.7063,
+      "step": 108
+    },
+    {
+      "epoch": 0.23942888522789676,
+      "grad_norm": 2199827.25,
+      "learning_rate": 3.941605839416059e-05,
+      "loss": 1.5712,
+      "step": 109
+    },
+    {
+      "epoch": 0.24162548050521693,
+      "grad_norm": 2521098.5,
+      "learning_rate": 3.978102189781022e-05,
+      "loss": 1.5903,
+      "step": 110
+    },
+    {
+      "epoch": 0.24382207578253706,
+      "grad_norm": 2247344.5,
+      "learning_rate": 4.0145985401459855e-05,
+      "loss": 1.4015,
+      "step": 111
+    },
+    {
+      "epoch": 0.24601867105985722,
+      "grad_norm": 8935620.0,
+      "learning_rate": 4.0510948905109494e-05,
+      "loss": 1.868,
+      "step": 112
+    },
+    {
+      "epoch": 0.2482152663371774,
+      "grad_norm": 1877222.0,
+      "learning_rate": 4.0875912408759126e-05,
+      "loss": 1.356,
+      "step": 113
+    },
+    {
+      "epoch": 0.2504118616144975,
+      "grad_norm": 2704155.25,
+      "learning_rate": 4.124087591240876e-05,
+      "loss": 1.4311,
+      "step": 114
+    },
+    {
+      "epoch": 0.25260845689181766,
+      "grad_norm": 5727084.5,
+      "learning_rate": 4.16058394160584e-05,
+      "loss": 1.3517,
+      "step": 115
+    },
+    {
+      "epoch": 0.25480505216913785,
+      "grad_norm": 2100491.75,
+      "learning_rate": 4.197080291970803e-05,
+      "loss": 1.5282,
+      "step": 116
+    },
+    {
+      "epoch": 0.257001647446458,
+      "grad_norm": 1965230.125,
+      "learning_rate": 4.233576642335767e-05,
+      "loss": 1.6464,
+      "step": 117
+    },
+    {
+      "epoch": 0.2591982427237781,
+      "grad_norm": 2276479.5,
+      "learning_rate": 4.27007299270073e-05,
+      "loss": 1.4084,
+      "step": 118
+    },
+    {
+      "epoch": 0.2613948380010983,
+      "grad_norm": 3081904.5,
+      "learning_rate": 4.306569343065693e-05,
+      "loss": 1.5393,
+      "step": 119
+    },
+    {
+      "epoch": 0.26359143327841844,
+      "grad_norm": 68824408.0,
+      "learning_rate": 4.343065693430657e-05,
+      "loss": 1.218,
+      "step": 120
+    },
+    {
+      "epoch": 0.26578802855573863,
+      "grad_norm": 3422234.75,
+      "learning_rate": 4.379562043795621e-05,
+      "loss": 1.502,
+      "step": 121
+    },
+    {
+      "epoch": 0.26798462383305877,
+      "grad_norm": 5766482.0,
+      "learning_rate": 4.416058394160584e-05,
+      "loss": 1.3889,
+      "step": 122
+    },
+    {
+      "epoch": 0.2701812191103789,
+      "grad_norm": 6983521.5,
+      "learning_rate": 4.452554744525548e-05,
+      "loss": 1.2038,
+      "step": 123
+    },
+    {
+      "epoch": 0.2723778143876991,
+      "grad_norm": 2714528.5,
+      "learning_rate": 4.489051094890511e-05,
+      "loss": 1.2669,
+      "step": 124
+    },
+    {
+      "epoch": 0.2745744096650192,
+      "grad_norm": 2568518.0,
+      "learning_rate": 4.5255474452554745e-05,
+      "loss": 1.4763,
+      "step": 125
+    },
+    {
+      "epoch": 0.27677100494233936,
+      "grad_norm": 2769107.0,
+      "learning_rate": 4.5620437956204383e-05,
+      "loss": 1.2843,
+      "step": 126
+    },
+    {
+      "epoch": 0.27896760021965955,
+      "grad_norm": 2451091.5,
+      "learning_rate": 4.5985401459854016e-05,
+      "loss": 1.4485,
+      "step": 127
+    },
+    {
+      "epoch": 0.2811641954969797,
+      "grad_norm": 4845333.5,
+      "learning_rate": 4.635036496350365e-05,
+      "loss": 1.7274,
+      "step": 128
+    },
+    {
+      "epoch": 0.2833607907742998,
+      "grad_norm": 4105798.0,
+      "learning_rate": 4.6715328467153287e-05,
+      "loss": 1.5568,
+      "step": 129
+    },
+    {
+      "epoch": 0.28555738605162,
+      "grad_norm": 4509214.5,
+      "learning_rate": 4.708029197080292e-05,
+      "loss": 1.3009,
+      "step": 130
+    },
+    {
+      "epoch": 0.28775398132894014,
+      "grad_norm": 9591719.0,
+      "learning_rate": 4.744525547445256e-05,
+      "loss": 1.2464,
+      "step": 131
+    },
+    {
+      "epoch": 0.2899505766062603,
+      "grad_norm": 4519166.5,
+      "learning_rate": 4.7810218978102196e-05,
+      "loss": 1.2914,
+      "step": 132
+    },
+    {
+      "epoch": 0.29214717188358047,
+      "grad_norm": 1731388.25,
+      "learning_rate": 4.817518248175183e-05,
+      "loss": 1.2892,
+      "step": 133
+    },
+    {
+      "epoch": 0.2943437671609006,
+      "grad_norm": 5530516.5,
+      "learning_rate": 4.854014598540147e-05,
+      "loss": 1.2947,
+      "step": 134
+    },
+    {
+      "epoch": 0.29654036243822074,
+      "grad_norm": 4922165.5,
+      "learning_rate": 4.89051094890511e-05,
+      "loss": 1.3001,
+      "step": 135
+    },
+    {
+      "epoch": 0.29873695771554093,
+      "grad_norm": 25799972.0,
+      "learning_rate": 4.927007299270073e-05,
+      "loss": 1.4416,
+      "step": 136
+    },
+    {
+      "epoch": 0.30093355299286106,
+      "grad_norm": 5071725.5,
+      "learning_rate": 4.963503649635037e-05,
+      "loss": 1.244,
+      "step": 137
+    },
+    {
+      "epoch": 0.3031301482701812,
+      "grad_norm": 9090560.0,
+      "learning_rate": 5e-05,
+      "loss": 1.3367,
+      "step": 138
+    },
+    {
+      "epoch": 0.3053267435475014,
+      "grad_norm": 1505044.125,
+      "learning_rate": 4.995928338762215e-05,
+      "loss": 1.2862,
+      "step": 139
+    },
+    {
+      "epoch": 0.3075233388248215,
+      "grad_norm": 64422448.0,
+      "learning_rate": 4.99185667752443e-05,
+      "loss": 1.3202,
+      "step": 140
+    },
+    {
+      "epoch": 0.30971993410214166,
+      "grad_norm": 2528250.0,
+      "learning_rate": 4.9877850162866454e-05,
+      "loss": 1.395,
+      "step": 141
+    },
+    {
+      "epoch": 0.31191652937946185,
+      "grad_norm": 4403366.5,
+      "learning_rate": 4.9837133550488604e-05,
+      "loss": 1.3303,
+      "step": 142
+    },
+    {
+      "epoch": 0.314113124656782,
+      "grad_norm": 1430878.5,
+      "learning_rate": 4.9796416938110755e-05,
+      "loss": 1.3883,
+      "step": 143
+    },
+    {
+      "epoch": 0.3163097199341021,
+      "grad_norm": 2634957.75,
+      "learning_rate": 4.97557003257329e-05,
+      "loss": 1.5322,
+      "step": 144
+    },
+    {
+      "epoch": 0.3185063152114223,
+      "grad_norm": 3262163.75,
+      "learning_rate": 4.971498371335505e-05,
+      "loss": 1.5423,
+      "step": 145
+    },
+    {
+      "epoch": 0.32070291048874244,
+      "grad_norm": 5317616.5,
+      "learning_rate": 4.96742671009772e-05,
+      "loss": 1.4146,
+      "step": 146
+    },
+    {
+      "epoch": 0.3228995057660626,
+      "grad_norm": 2778870.75,
+      "learning_rate": 4.963355048859935e-05,
+      "loss": 1.2832,
+      "step": 147
+    },
+    {
+      "epoch": 0.32509610104338277,
+      "grad_norm": 3236960.75,
+      "learning_rate": 4.95928338762215e-05,
+      "loss": 1.3334,
+      "step": 148
+    },
+    {
+      "epoch": 0.3272926963207029,
+      "grad_norm": 1861854.875,
+      "learning_rate": 4.955211726384365e-05,
+      "loss": 1.455,
+      "step": 149
+    },
+    {
+      "epoch": 0.32948929159802304,
+      "grad_norm": 2268978.5,
+      "learning_rate": 4.95114006514658e-05,
+      "loss": 1.7726,
+      "step": 150
+    },
+    {
+      "epoch": 0.3316858868753432,
+      "grad_norm": 2783583.75,
+      "learning_rate": 4.947068403908795e-05,
+      "loss": 1.726,
+      "step": 151
+    },
+    {
+      "epoch": 0.33388248215266336,
+      "grad_norm": 5952947.0,
+      "learning_rate": 4.94299674267101e-05,
+      "loss": 1.5935,
+      "step": 152
+    },
+    {
+      "epoch": 0.33607907742998355,
+      "grad_norm": 3580445.0,
+      "learning_rate": 4.938925081433225e-05,
+      "loss": 1.2819,
+      "step": 153
+    },
+    {
+      "epoch": 0.3382756727073037,
+      "grad_norm": 3056661.25,
+      "learning_rate": 4.9348534201954396e-05,
+      "loss": 1.1594,
+      "step": 154
+    },
+    {
+      "epoch": 0.3404722679846238,
+      "grad_norm": 5228316.5,
+      "learning_rate": 4.930781758957655e-05,
+      "loss": 1.259,
+      "step": 155
+    },
+    {
+      "epoch": 0.342668863261944,
+      "grad_norm": 8552230.0,
+      "learning_rate": 4.92671009771987e-05,
+      "loss": 1.2736,
+      "step": 156
+    },
+    {
+      "epoch": 0.34486545853926415,
+      "grad_norm": 2874161.25,
+      "learning_rate": 4.922638436482085e-05,
+      "loss": 1.3053,
+      "step": 157
+    },
+    {
+      "epoch": 0.3470620538165843,
+      "grad_norm": 3153549.0,
+      "learning_rate": 4.9185667752443e-05,
+      "loss": 1.0651,
+      "step": 158
+    },
+    {
+      "epoch": 0.34925864909390447,
+      "grad_norm": 2861551.75,
+      "learning_rate": 4.914495114006515e-05,
+      "loss": 1.3791,
+      "step": 159
+    },
+    {
+      "epoch": 0.3514552443712246,
+      "grad_norm": 11226787.0,
+      "learning_rate": 4.91042345276873e-05,
+      "loss": 1.1513,
+      "step": 160
+    },
+    {
+      "epoch": 0.35365183964854474,
+      "grad_norm": 8956181.0,
+      "learning_rate": 4.906351791530945e-05,
+      "loss": 1.3938,
+      "step": 161
+    },
+    {
+      "epoch": 0.35584843492586493,
+      "grad_norm": 1086894.25,
+      "learning_rate": 4.90228013029316e-05,
+      "loss": 1.3423,
+      "step": 162
+    },
+    {
+      "epoch": 0.35804503020318507,
+      "grad_norm": 1803490.5,
+      "learning_rate": 4.898208469055375e-05,
+      "loss": 1.0421,
+      "step": 163
+    },
+    {
+      "epoch": 0.3602416254805052,
+      "grad_norm": 3009170.5,
+      "learning_rate": 4.89413680781759e-05,
+      "loss": 1.1531,
+      "step": 164
+    },
+    {
+      "epoch": 0.3624382207578254,
+      "grad_norm": 2426892.5,
+      "learning_rate": 4.8900651465798044e-05,
+      "loss": 1.2423,
+      "step": 165
+    },
+    {
+      "epoch": 0.3646348160351455,
+      "grad_norm": 3730131.75,
+      "learning_rate": 4.8859934853420195e-05,
+      "loss": 1.1871,
+      "step": 166
+    },
+    {
+      "epoch": 0.36683141131246566,
+      "grad_norm": 1975670.0,
+      "learning_rate": 4.8819218241042345e-05,
+      "loss": 1.0772,
+      "step": 167
+    },
+    {
+      "epoch": 0.36902800658978585,
+      "grad_norm": 3969347.0,
+      "learning_rate": 4.8778501628664496e-05,
+      "loss": 1.2244,
+      "step": 168
+    },
+    {
+      "epoch": 0.371224601867106,
+      "grad_norm": 5956313.0,
+      "learning_rate": 4.8737785016286646e-05,
+      "loss": 1.5129,
+      "step": 169
+    },
+    {
+      "epoch": 0.3734211971444261,
+      "grad_norm": 11064708.0,
+      "learning_rate": 4.86970684039088e-05,
+      "loss": 1.5668,
+      "step": 170
+    },
+    {
+      "epoch": 0.3756177924217463,
+      "grad_norm": 9480759.0,
+      "learning_rate": 4.865635179153095e-05,
+      "loss": 1.0575,
+      "step": 171
+    },
+    {
+      "epoch": 0.37781438769906645,
+      "grad_norm": 1367503.125,
+      "learning_rate": 4.86156351791531e-05,
+      "loss": 1.1315,
+      "step": 172
+    },
+    {
+      "epoch": 0.3800109829763866,
+      "grad_norm": 1330542.25,
+      "learning_rate": 4.857491856677525e-05,
+      "loss": 1.4121,
+      "step": 173
+    },
+    {
+      "epoch": 0.38220757825370677,
+      "grad_norm": 1408385.75,
+      "learning_rate": 4.85342019543974e-05,
+      "loss": 1.4364,
+      "step": 174
+    },
+    {
+      "epoch": 0.3844041735310269,
+      "grad_norm": 4672034.5,
+      "learning_rate": 4.849348534201954e-05,
+      "loss": 1.0354,
+      "step": 175
+    },
+    {
+      "epoch": 0.38660076880834704,
+      "grad_norm": 1851254.75,
+      "learning_rate": 4.845276872964169e-05,
+      "loss": 1.5363,
+      "step": 176
+    },
+    {
+      "epoch": 0.38879736408566723,
+      "grad_norm": 13708761.0,
+      "learning_rate": 4.841205211726384e-05,
+      "loss": 0.8119,
+      "step": 177
+    },
+    {
+      "epoch": 0.39099395936298736,
+      "grad_norm": 5870485.5,
+      "learning_rate": 4.8371335504885994e-05,
+      "loss": 1.532,
+      "step": 178
+    },
+    {
+      "epoch": 0.3931905546403075,
+      "grad_norm": 2743023.25,
+      "learning_rate": 4.8330618892508144e-05,
+      "loss": 1.2041,
+      "step": 179
+    },
+    {
+      "epoch": 0.3953871499176277,
+      "grad_norm": 7092243.5,
+      "learning_rate": 4.8289902280130295e-05,
+      "loss": 1.0993,
+      "step": 180
+    },
+    {
+      "epoch": 0.3975837451949478,
+      "grad_norm": 43953068.0,
+      "learning_rate": 4.8249185667752445e-05,
+      "loss": 1.1332,
+      "step": 181
+    },
+    {
+      "epoch": 0.39978034047226796,
+      "grad_norm": 5665686.0,
+      "learning_rate": 4.8208469055374595e-05,
+      "loss": 1.005,
+      "step": 182
+    },
+    {
+      "epoch": 0.40197693574958815,
+      "grad_norm": 3115273.75,
+      "learning_rate": 4.8167752442996746e-05,
+      "loss": 1.2214,
+      "step": 183
+    },
+    {
+      "epoch": 0.4041735310269083,
+      "grad_norm": 3682000.25,
+      "learning_rate": 4.8127035830618896e-05,
+      "loss": 1.7004,
+      "step": 184
+    },
+    {
+      "epoch": 0.4063701263042284,
+      "grad_norm": 6770059.0,
+      "learning_rate": 4.808631921824105e-05,
+      "loss": 1.3679,
+      "step": 185
+    },
+    {
+      "epoch": 0.4085667215815486,
+      "grad_norm": 2806108.25,
+      "learning_rate": 4.804560260586319e-05,
+      "loss": 1.1225,
+      "step": 186
+    },
+    {
+      "epoch": 0.41076331685886874,
+      "grad_norm": 19982294.0,
+      "learning_rate": 4.800488599348534e-05,
+      "loss": 1.2016,
+      "step": 187
+    },
+    {
+      "epoch": 0.41295991213618893,
+      "grad_norm": 3720764.5,
+      "learning_rate": 4.796416938110749e-05,
+      "loss": 1.2192,
+      "step": 188
+    },
+    {
+      "epoch": 0.41515650741350907,
+      "grad_norm": 5204978.5,
+      "learning_rate": 4.792345276872964e-05,
+      "loss": 1.2992,
+      "step": 189
+    },
+    {
+      "epoch": 0.4173531026908292,
+      "grad_norm": 3860083.5,
+      "learning_rate": 4.788273615635179e-05,
+      "loss": 1.1482,
+      "step": 190
+    },
+    {
+      "epoch": 0.4195496979681494,
+      "grad_norm": 9096174.0,
+      "learning_rate": 4.784201954397394e-05,
+      "loss": 1.0143,
+      "step": 191
+    },
+    {
+      "epoch": 0.42174629324546953,
+      "grad_norm": 2424713.75,
+      "learning_rate": 4.780130293159609e-05,
+      "loss": 1.324,
+      "step": 192
+    },
+    {
+      "epoch": 0.42394288852278966,
+      "grad_norm": 9070478.0,
+      "learning_rate": 4.7760586319218244e-05,
+      "loss": 1.4039,
+      "step": 193
+    },
+    {
+      "epoch": 0.42613948380010985,
+      "grad_norm": 4270321.0,
+      "learning_rate": 4.7719869706840394e-05,
+      "loss": 1.3143,
+      "step": 194
+    },
+    {
+      "epoch": 0.42833607907743,
+      "grad_norm": 6623606.0,
+      "learning_rate": 4.7679153094462545e-05,
+      "loss": 1.5661,
+      "step": 195
+    },
+    {
+      "epoch": 0.4305326743547501,
+      "grad_norm": 2961344.5,
+      "learning_rate": 4.7638436482084695e-05,
+      "loss": 1.6329,
+      "step": 196
+    },
+    {
+      "epoch": 0.4327292696320703,
+      "grad_norm": 8572697.0,
+      "learning_rate": 4.759771986970684e-05,
+      "loss": 1.1109,
+      "step": 197
+    },
+    {
+      "epoch": 0.43492586490939045,
+      "grad_norm": 1814927.125,
+      "learning_rate": 4.755700325732899e-05,
+      "loss": 1.0376,
+      "step": 198
+    },
+    {
+      "epoch": 0.4371224601867106,
+      "grad_norm": 3639497.25,
+      "learning_rate": 4.751628664495114e-05,
+      "loss": 1.5448,
+      "step": 199
+    },
+    {
+      "epoch": 0.43931905546403077,
+      "grad_norm": 5656754.5,
+      "learning_rate": 4.747557003257329e-05,
+      "loss": 1.3913,
+      "step": 200
+    },
+    {
+      "epoch": 0.4415156507413509,
+      "grad_norm": 4670615.5,
+      "learning_rate": 4.743485342019544e-05,
+      "loss": 1.0906,
+      "step": 201
+    },
+    {
+      "epoch": 0.44371224601867104,
+      "grad_norm": 4905259.5,
+      "learning_rate": 4.739413680781759e-05,
+      "loss": 1.3009,
+      "step": 202
+    },
+    {
+      "epoch": 0.44590884129599123,
+      "grad_norm": 1286014.125,
+      "learning_rate": 4.735342019543974e-05,
+      "loss": 1.3623,
+      "step": 203
+    },
+    {
+      "epoch": 0.44810543657331137,
+      "grad_norm": 6117930.0,
+      "learning_rate": 4.731270358306189e-05,
+      "loss": 1.4801,
+      "step": 204
+    },
+    {
+      "epoch": 0.4503020318506315,
+      "grad_norm": 3035536.75,
+      "learning_rate": 4.727198697068404e-05,
+      "loss": 1.3136,
+      "step": 205
+    },
+    {
+      "epoch": 0.4524986271279517,
+      "grad_norm": 3008669.25,
+      "learning_rate": 4.723127035830619e-05,
+      "loss": 1.4152,
+      "step": 206
+    },
+    {
+      "epoch": 0.4546952224052718,
+      "grad_norm": 2453379.25,
+      "learning_rate": 4.719055374592834e-05,
+      "loss": 1.1594,
+      "step": 207
+    },
+    {
+      "epoch": 0.45689181768259196,
+      "grad_norm": 814823.8125,
+      "learning_rate": 4.714983713355049e-05,
+      "loss": 1.1467,
+      "step": 208
+    },
+    {
+      "epoch": 0.45908841295991215,
+      "grad_norm": 2359125.5,
+      "learning_rate": 4.710912052117264e-05,
+      "loss": 1.2623,
+      "step": 209
+    },
+    {
+      "epoch": 0.4612850082372323,
+      "grad_norm": 4699321.5,
+      "learning_rate": 4.706840390879479e-05,
+      "loss": 1.4737,
+      "step": 210
+    },
+    {
+      "epoch": 0.4634816035145524,
+      "grad_norm": 2087632.5,
+      "learning_rate": 4.702768729641694e-05,
+      "loss": 1.3978,
+      "step": 211
+    },
+    {
+      "epoch": 0.4656781987918726,
+      "grad_norm": 2167067.0,
+      "learning_rate": 4.698697068403909e-05,
+      "loss": 1.2355,
+      "step": 212
+    },
+    {
+      "epoch": 0.46787479406919275,
+      "grad_norm": 3234822.75,
+      "learning_rate": 4.694625407166124e-05,
+      "loss": 0.9401,
+      "step": 213
+    },
+    {
+      "epoch": 0.4700713893465129,
+      "grad_norm": 2579396.75,
+      "learning_rate": 4.690553745928339e-05,
+      "loss": 1.1238,
+      "step": 214
+    },
+    {
+      "epoch": 0.47226798462383307,
+      "grad_norm": 5746735.5,
+      "learning_rate": 4.686482084690554e-05,
+      "loss": 1.3909,
+      "step": 215
+    },
+    {
+      "epoch": 0.4744645799011532,
+      "grad_norm": 3341654.25,
+      "learning_rate": 4.682410423452769e-05,
+      "loss": 1.3506,
+      "step": 216
+    },
+    {
+      "epoch": 0.47666117517847334,
+      "grad_norm": 4109063.25,
+      "learning_rate": 4.678338762214984e-05,
+      "loss": 1.1583,
+      "step": 217
+    },
+    {
+      "epoch": 0.47885777045579353,
+      "grad_norm": 4749338.0,
+      "learning_rate": 4.6742671009771985e-05,
+      "loss": 1.2787,
+      "step": 218
+    },
+    {
+      "epoch": 0.48105436573311366,
+      "grad_norm": 2185890.25,
+      "learning_rate": 4.6701954397394135e-05,
+      "loss": 1.1011,
+      "step": 219
+    },
+    {
+      "epoch": 0.48325096101043385,
+      "grad_norm": 4453409.5,
+      "learning_rate": 4.6661237785016286e-05,
+      "loss": 1.3357,
+      "step": 220
+    },
+    {
+      "epoch": 0.485447556287754,
+      "grad_norm": 928414.4375,
+      "learning_rate": 4.6620521172638436e-05,
+      "loss": 1.1144,
+      "step": 221
+    },
+    {
+      "epoch": 0.4876441515650741,
+      "grad_norm": 1783463.0,
+      "learning_rate": 4.657980456026059e-05,
+      "loss": 0.9974,
+      "step": 222
+    },
+    {
+      "epoch": 0.4898407468423943,
+      "grad_norm": 3937690.5,
+      "learning_rate": 4.653908794788274e-05,
+      "loss": 1.1442,
+      "step": 223
+    },
+    {
+      "epoch": 0.49203734211971445,
+      "grad_norm": 24161426.0,
+      "learning_rate": 4.649837133550489e-05,
+      "loss": 1.2237,
+      "step": 224
+    },
+    {
+      "epoch": 0.4942339373970346,
+      "grad_norm": 18245274.0,
+      "learning_rate": 4.645765472312704e-05,
+      "loss": 0.8382,
+      "step": 225
+    },
+    {
+      "epoch": 0.4964305326743548,
+      "grad_norm": 2946201.25,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 1.5405,
+      "step": 226
+    },
+    {
+      "epoch": 0.4986271279516749,
+      "grad_norm": 1755778.875,
+      "learning_rate": 4.637622149837134e-05,
+      "loss": 1.0148,
+      "step": 227
+    },
+    {
+      "epoch": 0.500823723228995,
+      "grad_norm": 1576456.125,
+      "learning_rate": 4.633550488599348e-05,
+      "loss": 1.0725,
+      "step": 228
+    },
+    {
+      "epoch": 0.5030203185063152,
+      "grad_norm": 3363727.0,
+      "learning_rate": 4.629478827361563e-05,
+      "loss": 1.2198,
+      "step": 229
+    },
+    {
+      "epoch": 0.5052169137836353,
+      "grad_norm": 2715469.0,
+      "learning_rate": 4.6254071661237784e-05,
+      "loss": 1.3515,
+      "step": 230
+    },
+    {
+      "epoch": 0.5074135090609555,
+      "grad_norm": 2468360.5,
+      "learning_rate": 4.6213355048859934e-05,
+      "loss": 1.1817,
+      "step": 231
+    },
+    {
+      "epoch": 0.5096101043382757,
+      "grad_norm": 3551493.5,
+      "learning_rate": 4.6172638436482085e-05,
+      "loss": 1.3529,
+      "step": 232
+    },
+    {
+      "epoch": 0.5118066996155958,
+      "grad_norm": 5423752.0,
+      "learning_rate": 4.6131921824104235e-05,
+      "loss": 1.2535,
+      "step": 233
+    },
+    {
+      "epoch": 0.514003294892916,
+      "grad_norm": 2281067.25,
+      "learning_rate": 4.6091205211726385e-05,
+      "loss": 1.1795,
+      "step": 234
+    },
+    {
+      "epoch": 0.5161998901702362,
+      "grad_norm": 5830553.0,
+      "learning_rate": 4.6050488599348536e-05,
+      "loss": 1.2728,
+      "step": 235
+    },
+    {
+      "epoch": 0.5183964854475562,
+      "grad_norm": 2158330.0,
+      "learning_rate": 4.6009771986970686e-05,
+      "loss": 1.1708,
+      "step": 236
+    },
+    {
+      "epoch": 0.5205930807248764,
+      "grad_norm": 2404838.25,
+      "learning_rate": 4.596905537459284e-05,
+      "loss": 1.5318,
+      "step": 237
+    },
+    {
+      "epoch": 0.5227896760021966,
+      "grad_norm": 2930560.0,
+      "learning_rate": 4.592833876221499e-05,
+      "loss": 1.2926,
+      "step": 238
+    },
+    {
+      "epoch": 0.5249862712795168,
+      "grad_norm": 3127808.25,
+      "learning_rate": 4.588762214983713e-05,
+      "loss": 1.1212,
+      "step": 239
+    },
+    {
+      "epoch": 0.5271828665568369,
+      "grad_norm": 4783648.5,
+      "learning_rate": 4.584690553745928e-05,
+      "loss": 1.2263,
+      "step": 240
+    },
+    {
+      "epoch": 0.5293794618341571,
+      "grad_norm": 2675607.25,
+      "learning_rate": 4.580618892508143e-05,
+      "loss": 1.2268,
+      "step": 241
+    },
+    {
+      "epoch": 0.5315760571114773,
+      "grad_norm": 2888610.5,
+      "learning_rate": 4.576547231270358e-05,
+      "loss": 1.036,
+      "step": 242
+    },
+    {
+      "epoch": 0.5337726523887973,
+      "grad_norm": 3417312.0,
+      "learning_rate": 4.572475570032573e-05,
+      "loss": 1.0593,
+      "step": 243
+    },
+    {
+      "epoch": 0.5359692476661175,
+      "grad_norm": 1233640.5,
+      "learning_rate": 4.568403908794788e-05,
+      "loss": 1.186,
+      "step": 244
+    },
+    {
+      "epoch": 0.5381658429434377,
+      "grad_norm": 3559585.5,
+      "learning_rate": 4.5643322475570034e-05,
+      "loss": 1.4095,
+      "step": 245
+    },
+    {
+      "epoch": 0.5403624382207578,
+      "grad_norm": 1482595.25,
+      "learning_rate": 4.5602605863192184e-05,
+      "loss": 1.1919,
+      "step": 246
+    },
+    {
+      "epoch": 0.542559033498078,
+      "grad_norm": 2399361.25,
+      "learning_rate": 4.5561889250814335e-05,
+      "loss": 1.0821,
+      "step": 247
+    },
+    {
+      "epoch": 0.5447556287753982,
+      "grad_norm": 2945265.75,
+      "learning_rate": 4.5521172638436485e-05,
+      "loss": 1.189,
+      "step": 248
+    },
+    {
+      "epoch": 0.5469522240527183,
+      "grad_norm": 6141536.5,
+      "learning_rate": 4.548045602605863e-05,
+      "loss": 1.1893,
+      "step": 249
+    },
+    {
+      "epoch": 0.5491488193300385,
+      "grad_norm": 11363059.0,
+      "learning_rate": 4.543973941368078e-05,
+      "loss": 1.1137,
+      "step": 250
+    },
+    {
+      "epoch": 0.5513454146073586,
+      "grad_norm": 1562360.625,
+      "learning_rate": 4.539902280130293e-05,
+      "loss": 1.1162,
+      "step": 251
+    },
+    {
+      "epoch": 0.5535420098846787,
+      "grad_norm": 2415007.5,
+      "learning_rate": 4.535830618892508e-05,
+      "loss": 1.1728,
+      "step": 252
+    },
+    {
+      "epoch": 0.5557386051619989,
+      "grad_norm": 4460530.5,
+      "learning_rate": 4.531758957654723e-05,
+      "loss": 1.4081,
+      "step": 253
+    },
+    {
+      "epoch": 0.5579352004393191,
+      "grad_norm": 1058439.125,
+      "learning_rate": 4.527687296416938e-05,
+      "loss": 1.0889,
+      "step": 254
+    },
+    {
+      "epoch": 0.5601317957166392,
+      "grad_norm": 4184706.5,
+      "learning_rate": 4.523615635179153e-05,
+      "loss": 1.3016,
+      "step": 255
+    },
+    {
+      "epoch": 0.5623283909939594,
+      "grad_norm": 3175304.5,
+      "learning_rate": 4.519543973941368e-05,
+      "loss": 1.1958,
+      "step": 256
+    },
+    {
+      "epoch": 0.5645249862712796,
+      "grad_norm": 1754757.875,
+      "learning_rate": 4.515472312703583e-05,
+      "loss": 0.9709,
+      "step": 257
+    },
+    {
+      "epoch": 0.5667215815485996,
+      "grad_norm": 4480082.5,
+      "learning_rate": 4.511400651465798e-05,
+      "loss": 1.2385,
+      "step": 258
+    },
+    {
+      "epoch": 0.5689181768259198,
+      "grad_norm": 4317042.5,
+      "learning_rate": 4.507328990228013e-05,
+      "loss": 1.2541,
+      "step": 259
+    },
+    {
+      "epoch": 0.57111477210324,
+      "grad_norm": 2381945.5,
+      "learning_rate": 4.503257328990228e-05,
+      "loss": 1.2689,
+      "step": 260
+    },
+    {
+      "epoch": 0.5733113673805601,
+      "grad_norm": 2813218.25,
+      "learning_rate": 4.499185667752443e-05,
+      "loss": 1.2936,
+      "step": 261
+    },
+    {
+      "epoch": 0.5755079626578803,
+      "grad_norm": 101527104.0,
+      "learning_rate": 4.495114006514658e-05,
+      "loss": 1.3709,
+      "step": 262
+    },
+    {
+      "epoch": 0.5777045579352005,
+      "grad_norm": 2624337.25,
+      "learning_rate": 4.491042345276873e-05,
+      "loss": 1.0851,
+      "step": 263
+    },
+    {
+      "epoch": 0.5799011532125206,
+      "grad_norm": 3705165.5,
+      "learning_rate": 4.486970684039088e-05,
+      "loss": 1.2091,
+      "step": 264
+    },
+    {
+      "epoch": 0.5820977484898407,
+      "grad_norm": 4012303.5,
+      "learning_rate": 4.482899022801303e-05,
+      "loss": 1.6748,
+      "step": 265
+    },
+    {
+      "epoch": 0.5842943437671609,
+      "grad_norm": 1273650.25,
+      "learning_rate": 4.478827361563518e-05,
+      "loss": 1.2646,
+      "step": 266
+    },
+    {
+      "epoch": 0.586490939044481,
+      "grad_norm": 3251308.25,
+      "learning_rate": 4.474755700325733e-05,
+      "loss": 1.1402,
+      "step": 267
+    },
+    {
+      "epoch": 0.5886875343218012,
+      "grad_norm": 3377360.5,
+      "learning_rate": 4.470684039087948e-05,
+      "loss": 1.2919,
+      "step": 268
+    },
+    {
+      "epoch": 0.5908841295991214,
+      "grad_norm": 8047918.5,
+      "learning_rate": 4.466612377850163e-05,
+      "loss": 1.3234,
+      "step": 269
+    },
+    {
+      "epoch": 0.5930807248764415,
+      "grad_norm": 3880224.5,
+      "learning_rate": 4.462540716612378e-05,
+      "loss": 1.5569,
+      "step": 270
+    },
+    {
+      "epoch": 0.5952773201537617,
+      "grad_norm": 1401508.75,
+      "learning_rate": 4.4584690553745925e-05,
+      "loss": 1.0261,
+      "step": 271
+    },
+    {
+      "epoch": 0.5974739154310819,
+      "grad_norm": 2840255.0,
+      "learning_rate": 4.4543973941368076e-05,
+      "loss": 1.0601,
+      "step": 272
+    },
+    {
+      "epoch": 0.5996705107084019,
+      "grad_norm": 1783275.375,
+      "learning_rate": 4.4503257328990226e-05,
+      "loss": 1.0173,
+      "step": 273
+    },
+    {
+      "epoch": 0.6018671059857221,
+      "grad_norm": 3187320.75,
+      "learning_rate": 4.446254071661238e-05,
+      "loss": 1.1253,
+      "step": 274
+    },
+    {
+      "epoch": 0.6040637012630423,
+      "grad_norm": 3507886.5,
+      "learning_rate": 4.442182410423453e-05,
+      "loss": 1.0899,
+      "step": 275
+    },
+    {
+      "epoch": 0.6062602965403624,
+      "grad_norm": 21071718.0,
+      "learning_rate": 4.438110749185668e-05,
+      "loss": 0.8644,
+      "step": 276
+    },
+    {
+      "epoch": 0.6084568918176826,
+      "grad_norm": 3260922.25,
+      "learning_rate": 4.434039087947883e-05,
+      "loss": 1.2001,
+      "step": 277
+    },
+    {
+      "epoch": 0.6106534870950028,
+      "grad_norm": 1558948.5,
+      "learning_rate": 4.429967426710098e-05,
+      "loss": 1.3372,
+      "step": 278
+    },
+    {
+      "epoch": 0.6128500823723229,
+      "grad_norm": 9367641.0,
+      "learning_rate": 4.425895765472313e-05,
+      "loss": 1.1057,
+      "step": 279
+    },
+    {
+      "epoch": 0.615046677649643,
+      "grad_norm": 4195653.0,
+      "learning_rate": 4.421824104234528e-05,
+      "loss": 1.1144,
+      "step": 280
+    },
+    {
+      "epoch": 0.6172432729269632,
+      "grad_norm": 2504132.5,
+      "learning_rate": 4.417752442996742e-05,
+      "loss": 0.9134,
+      "step": 281
+    },
+    {
+      "epoch": 0.6194398682042833,
+      "grad_norm": 1837744.25,
+      "learning_rate": 4.4136807817589574e-05,
+      "loss": 1.2113,
+      "step": 282
+    },
+    {
+      "epoch": 0.6216364634816035,
+      "grad_norm": 8317643.0,
+      "learning_rate": 4.4096091205211724e-05,
+      "loss": 1.3014,
+      "step": 283
+    },
+    {
+      "epoch": 0.6238330587589237,
+      "grad_norm": 5956817.5,
+      "learning_rate": 4.4055374592833875e-05,
+      "loss": 0.9833,
+      "step": 284
+    },
+    {
+      "epoch": 0.6260296540362438,
+      "grad_norm": 3230775.25,
+      "learning_rate": 4.4014657980456025e-05,
+      "loss": 1.118,
+      "step": 285
+    },
+    {
+      "epoch": 0.628226249313564,
+      "grad_norm": 11376350.0,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 1.1328,
+      "step": 286
+    },
+    {
+      "epoch": 0.6304228445908842,
+      "grad_norm": 1427589.75,
+      "learning_rate": 4.3933224755700326e-05,
+      "loss": 1.1948,
+      "step": 287
+    },
+    {
+      "epoch": 0.6326194398682042,
+      "grad_norm": 3415921.75,
+      "learning_rate": 4.3892508143322476e-05,
+      "loss": 1.1323,
+      "step": 288
+    },
+    {
+      "epoch": 0.6348160351455244,
+      "grad_norm": 3035935.25,
+      "learning_rate": 4.385179153094463e-05,
+      "loss": 0.8915,
+      "step": 289
+    },
+    {
+      "epoch": 0.6370126304228446,
+      "grad_norm": 2917160.25,
+      "learning_rate": 4.381107491856678e-05,
+      "loss": 0.9789,
+      "step": 290
+    },
+    {
+      "epoch": 0.6392092257001647,
+      "grad_norm": 3809564.0,
+      "learning_rate": 4.377035830618893e-05,
+      "loss": 1.314,
+      "step": 291
+    },
+    {
+      "epoch": 0.6414058209774849,
+      "grad_norm": 19103740.0,
+      "learning_rate": 4.372964169381108e-05,
+      "loss": 1.1236,
+      "step": 292
+    },
+    {
+      "epoch": 0.6436024162548051,
+      "grad_norm": 3117975.5,
+      "learning_rate": 4.368892508143323e-05,
+      "loss": 1.2374,
+      "step": 293
+    },
+    {
+      "epoch": 0.6457990115321252,
+      "grad_norm": 3985447.75,
+      "learning_rate": 4.364820846905538e-05,
+      "loss": 1.1908,
+      "step": 294
+    },
+    {
+      "epoch": 0.6479956068094453,
+      "grad_norm": 2318079.5,
+      "learning_rate": 4.360749185667753e-05,
+      "loss": 1.0371,
+      "step": 295
+    },
+    {
+      "epoch": 0.6501922020867655,
+      "grad_norm": 6203933.0,
+      "learning_rate": 4.356677524429968e-05,
+      "loss": 1.2372,
+      "step": 296
+    },
+    {
+      "epoch": 0.6523887973640856,
+      "grad_norm": 16002894.0,
+      "learning_rate": 4.352605863192183e-05,
+      "loss": 1.0551,
+      "step": 297
+    },
+    {
+      "epoch": 0.6545853926414058,
+      "grad_norm": 2405322.5,
+      "learning_rate": 4.348534201954398e-05,
+      "loss": 1.1988,
+      "step": 298
+    },
+    {
+      "epoch": 0.656781987918726,
+      "grad_norm": 3708790.25,
+      "learning_rate": 4.344462540716613e-05,
+      "loss": 1.115,
+      "step": 299
+    },
+    {
+      "epoch": 0.6589785831960461,
+      "grad_norm": 1741446.875,
+      "learning_rate": 4.3403908794788275e-05,
+      "loss": 1.1241,
+      "step": 300
+    },
+    {
+      "epoch": 0.6611751784733663,
+      "grad_norm": 2228662.0,
+      "learning_rate": 4.3363192182410426e-05,
+      "loss": 1.2623,
+      "step": 301
+    },
+    {
+      "epoch": 0.6633717737506865,
+      "grad_norm": 3416909.0,
+      "learning_rate": 4.3322475570032576e-05,
+      "loss": 1.2562,
+      "step": 302
+    },
+    {
+      "epoch": 0.6655683690280065,
+      "grad_norm": 8822010.0,
+      "learning_rate": 4.3281758957654726e-05,
+      "loss": 1.0034,
+      "step": 303
+    },
+    {
+      "epoch": 0.6677649643053267,
+      "grad_norm": 6467941.5,
+      "learning_rate": 4.324104234527688e-05,
+      "loss": 1.1656,
+      "step": 304
+    },
+    {
+      "epoch": 0.6699615595826469,
+      "grad_norm": 6514345.0,
+      "learning_rate": 4.320032573289903e-05,
+      "loss": 1.1037,
+      "step": 305
+    },
+    {
+      "epoch": 0.6721581548599671,
+      "grad_norm": 82065680.0,
+      "learning_rate": 4.315960912052118e-05,
+      "loss": 1.1626,
+      "step": 306
+    },
+    {
+      "epoch": 0.6743547501372872,
+      "grad_norm": 12891722.0,
+      "learning_rate": 4.311889250814333e-05,
+      "loss": 1.2572,
+      "step": 307
+    },
+    {
+      "epoch": 0.6765513454146074,
+      "grad_norm": 6390971.5,
+      "learning_rate": 4.307817589576548e-05,
+      "loss": 1.2007,
+      "step": 308
+    },
+    {
+      "epoch": 0.6787479406919276,
+      "grad_norm": 3507993.0,
+      "learning_rate": 4.303745928338763e-05,
+      "loss": 1.219,
+      "step": 309
+    },
+    {
+      "epoch": 0.6809445359692476,
+      "grad_norm": 4870334.5,
+      "learning_rate": 4.299674267100978e-05,
+      "loss": 1.2233,
+      "step": 310
+    },
+    {
+      "epoch": 0.6831411312465678,
+      "grad_norm": 9724761.0,
+      "learning_rate": 4.295602605863192e-05,
+      "loss": 0.9479,
+      "step": 311
+    },
+    {
+      "epoch": 0.685337726523888,
+      "grad_norm": 3013746.75,
+      "learning_rate": 4.2915309446254074e-05,
+      "loss": 1.1903,
+      "step": 312
+    },
+    {
+      "epoch": 0.6875343218012081,
+      "grad_norm": 14367541.0,
+      "learning_rate": 4.2874592833876224e-05,
+      "loss": 1.0991,
+      "step": 313
+    },
+    {
+      "epoch": 0.6897309170785283,
+      "grad_norm": 14006771.0,
+      "learning_rate": 4.2833876221498375e-05,
+      "loss": 0.9843,
+      "step": 314
+    },
+    {
+      "epoch": 0.6919275123558485,
+      "grad_norm": 3380090.25,
+      "learning_rate": 4.2793159609120525e-05,
+      "loss": 0.9319,
+      "step": 315
+    },
+    {
+      "epoch": 0.6941241076331686,
+      "grad_norm": 3652418.0,
+      "learning_rate": 4.2752442996742676e-05,
+      "loss": 1.1692,
+      "step": 316
+    },
+    {
+      "epoch": 0.6963207029104888,
+      "grad_norm": 4915524.5,
+      "learning_rate": 4.2711726384364826e-05,
+      "loss": 1.2286,
+      "step": 317
+    },
+    {
+      "epoch": 0.6985172981878089,
+      "grad_norm": 28981046.0,
+      "learning_rate": 4.2671009771986977e-05,
+      "loss": 1.3796,
+      "step": 318
+    },
+    {
+      "epoch": 0.700713893465129,
+      "grad_norm": 2808999.75,
+      "learning_rate": 4.263029315960913e-05,
+      "loss": 1.1535,
+      "step": 319
+    },
+    {
+      "epoch": 0.7029104887424492,
+      "grad_norm": 13118838.0,
+      "learning_rate": 4.258957654723128e-05,
+      "loss": 1.1525,
+      "step": 320
+    },
+    {
+      "epoch": 0.7051070840197694,
+      "grad_norm": 5745464.5,
+      "learning_rate": 4.254885993485342e-05,
+      "loss": 1.1361,
+      "step": 321
+    },
+    {
+      "epoch": 0.7073036792970895,
+      "grad_norm": 42897660.0,
+      "learning_rate": 4.250814332247557e-05,
+      "loss": 1.2996,
+      "step": 322
+    },
+    {
+      "epoch": 0.7095002745744097,
+      "grad_norm": 3062572.75,
+      "learning_rate": 4.246742671009772e-05,
+      "loss": 0.9858,
+      "step": 323
+    },
+    {
+      "epoch": 0.7116968698517299,
+      "grad_norm": 4770826.0,
+      "learning_rate": 4.242671009771987e-05,
+      "loss": 1.273,
+      "step": 324
+    },
+    {
+      "epoch": 0.7138934651290499,
+      "grad_norm": 4174559.0,
+      "learning_rate": 4.238599348534202e-05,
+      "loss": 1.0551,
+      "step": 325
+    },
+    {
+      "epoch": 0.7160900604063701,
+      "grad_norm": 375559296.0,
+      "learning_rate": 4.2345276872964173e-05,
+      "loss": 1.0587,
+      "step": 326
+    },
+    {
+      "epoch": 0.7182866556836903,
+      "grad_norm": 2159214.5,
+      "learning_rate": 4.2304560260586324e-05,
+      "loss": 1.1698,
+      "step": 327
+    },
+    {
+      "epoch": 0.7204832509610104,
+      "grad_norm": 3752315.5,
+      "learning_rate": 4.2263843648208474e-05,
+      "loss": 1.3348,
+      "step": 328
+    },
+    {
+      "epoch": 0.7226798462383306,
+      "grad_norm": 7605865.5,
+      "learning_rate": 4.2223127035830625e-05,
+      "loss": 1.1477,
+      "step": 329
+    },
+    {
+      "epoch": 0.7248764415156508,
+      "grad_norm": 3472608.25,
+      "learning_rate": 4.2182410423452775e-05,
+      "loss": 1.1208,
+      "step": 330
+    },
+    {
+      "epoch": 0.7270730367929709,
+      "grad_norm": 6209556.0,
+      "learning_rate": 4.2141693811074926e-05,
+      "loss": 1.2667,
+      "step": 331
+    },
+    {
+      "epoch": 0.729269632070291,
+      "grad_norm": 5633322.5,
+      "learning_rate": 4.210097719869707e-05,
+      "loss": 1.1284,
+      "step": 332
+    },
+    {
+      "epoch": 0.7314662273476112,
+      "grad_norm": 3484536.0,
+      "learning_rate": 4.206026058631922e-05,
+      "loss": 1.1714,
+      "step": 333
+    },
+    {
+      "epoch": 0.7336628226249313,
+      "grad_norm": 3105687.5,
+      "learning_rate": 4.201954397394137e-05,
+      "loss": 1.3313,
+      "step": 334
+    },
+    {
+      "epoch": 0.7358594179022515,
+      "grad_norm": 2914641.5,
+      "learning_rate": 4.197882736156352e-05,
+      "loss": 1.1196,
+      "step": 335
+    },
+    {
+      "epoch": 0.7380560131795717,
+      "grad_norm": 1880664.125,
+      "learning_rate": 4.193811074918567e-05,
+      "loss": 1.0404,
+      "step": 336
+    },
+    {
+      "epoch": 0.7402526084568918,
+      "grad_norm": 4303799.5,
+      "learning_rate": 4.189739413680782e-05,
+      "loss": 1.3731,
+      "step": 337
+    },
+    {
+      "epoch": 0.742449203734212,
+      "grad_norm": 4548311.0,
+      "learning_rate": 4.185667752442997e-05,
+      "loss": 1.1474,
+      "step": 338
+    },
+    {
+      "epoch": 0.7446457990115322,
+      "grad_norm": 2373950.5,
+      "learning_rate": 4.181596091205212e-05,
+      "loss": 1.1208,
+      "step": 339
+    },
+    {
+      "epoch": 0.7468423942888522,
+      "grad_norm": 2740215.0,
+      "learning_rate": 4.177524429967427e-05,
+      "loss": 1.1219,
+      "step": 340
+    },
+    {
+      "epoch": 0.7490389895661724,
+      "grad_norm": 5622567.0,
+      "learning_rate": 4.1734527687296424e-05,
+      "loss": 1.1398,
+      "step": 341
+    },
+    {
+      "epoch": 0.7512355848434926,
+      "grad_norm": 7509274.5,
+      "learning_rate": 4.1693811074918574e-05,
+      "loss": 1.1063,
+      "step": 342
+    },
+    {
+      "epoch": 0.7534321801208127,
+      "grad_norm": 2636867.5,
+      "learning_rate": 4.165309446254072e-05,
+      "loss": 0.9971,
+      "step": 343
+    },
+    {
+      "epoch": 0.7556287753981329,
+      "grad_norm": 7815588.0,
+      "learning_rate": 4.161237785016287e-05,
+      "loss": 1.2763,
+      "step": 344
+    },
+    {
+      "epoch": 0.7578253706754531,
+      "grad_norm": 3008378.25,
+      "learning_rate": 4.157166123778502e-05,
+      "loss": 1.0572,
+      "step": 345
+    },
+    {
+      "epoch": 0.7600219659527732,
+      "grad_norm": 3505830.75,
+      "learning_rate": 4.153094462540717e-05,
+      "loss": 1.1214,
+      "step": 346
+    },
+    {
+      "epoch": 0.7622185612300933,
+      "grad_norm": 5544672.5,
+      "learning_rate": 4.149022801302932e-05,
+      "loss": 1.2631,
+      "step": 347
+    },
+    {
+      "epoch": 0.7644151565074135,
+      "grad_norm": 9404032.0,
+      "learning_rate": 4.144951140065147e-05,
+      "loss": 1.2632,
+      "step": 348
+    },
+    {
+      "epoch": 0.7666117517847336,
+      "grad_norm": 6564583.5,
+      "learning_rate": 4.140879478827362e-05,
+      "loss": 0.9037,
+      "step": 349
+    },
+    {
+      "epoch": 0.7688083470620538,
+      "grad_norm": 4087187.0,
+      "learning_rate": 4.136807817589577e-05,
+      "loss": 1.4075,
+      "step": 350
+    },
+    {
+      "epoch": 0.771004942339374,
+      "grad_norm": 2820918.25,
+      "learning_rate": 4.132736156351792e-05,
+      "loss": 1.4276,
+      "step": 351
+    },
+    {
+      "epoch": 0.7732015376166941,
+      "grad_norm": 4639506.5,
+      "learning_rate": 4.128664495114007e-05,
+      "loss": 1.1662,
+      "step": 352
+    },
+    {
+      "epoch": 0.7753981328940143,
+      "grad_norm": 1841177.875,
+      "learning_rate": 4.1245928338762215e-05,
+      "loss": 0.9323,
+      "step": 353
+    },
+    {
+      "epoch": 0.7775947281713345,
+      "grad_norm": 5765610.5,
+      "learning_rate": 4.1205211726384366e-05,
+      "loss": 1.2905,
+      "step": 354
+    },
+    {
+      "epoch": 0.7797913234486545,
+      "grad_norm": 4056324.5,
+      "learning_rate": 4.1164495114006516e-05,
+      "loss": 0.8223,
+      "step": 355
+    },
+    {
+      "epoch": 0.7819879187259747,
+      "grad_norm": 9869166.0,
+      "learning_rate": 4.112377850162867e-05,
+      "loss": 1.062,
+      "step": 356
+    },
+    {
+      "epoch": 0.7841845140032949,
+      "grad_norm": 2870571.25,
+      "learning_rate": 4.108306188925082e-05,
+      "loss": 1.1062,
+      "step": 357
+    },
+    {
+      "epoch": 0.786381109280615,
+      "grad_norm": 5351715.5,
+      "learning_rate": 4.104234527687297e-05,
+      "loss": 1.0764,
+      "step": 358
+    },
+    {
+      "epoch": 0.7885777045579352,
+      "grad_norm": 5763360.0,
+      "learning_rate": 4.100162866449512e-05,
+      "loss": 1.0767,
+      "step": 359
+    },
+    {
+      "epoch": 0.7907742998352554,
+      "grad_norm": 2308774.25,
+      "learning_rate": 4.096091205211727e-05,
+      "loss": 0.9574,
+      "step": 360
+    },
+    {
+      "epoch": 0.7929708951125755,
+      "grad_norm": 12431558.0,
+      "learning_rate": 4.092019543973942e-05,
+      "loss": 1.2565,
+      "step": 361
+    },
+    {
+      "epoch": 0.7951674903898956,
+      "grad_norm": 6977140.5,
+      "learning_rate": 4.087947882736157e-05,
+      "loss": 1.2038,
+      "step": 362
+    },
+    {
+      "epoch": 0.7973640856672158,
+      "grad_norm": 5525465.5,
+      "learning_rate": 4.083876221498372e-05,
+      "loss": 1.0123,
+      "step": 363
+    },
+    {
+      "epoch": 0.7995606809445359,
+      "grad_norm": 8538183.0,
+      "learning_rate": 4.0798045602605864e-05,
+      "loss": 1.1986,
+      "step": 364
+    },
+    {
+      "epoch": 0.8017572762218561,
+      "grad_norm": 2110193.75,
+      "learning_rate": 4.0757328990228014e-05,
+      "loss": 0.8521,
+      "step": 365
+    },
+    {
+      "epoch": 0.8039538714991763,
+      "grad_norm": 3741184.0,
+      "learning_rate": 4.0716612377850165e-05,
+      "loss": 1.2286,
+      "step": 366
+    },
+    {
+      "epoch": 0.8061504667764964,
+      "grad_norm": 23761130.0,
+      "learning_rate": 4.0675895765472315e-05,
+      "loss": 0.9898,
+      "step": 367
+    },
+    {
+      "epoch": 0.8083470620538166,
+      "grad_norm": 1594879.5,
+      "learning_rate": 4.0635179153094466e-05,
+      "loss": 0.9077,
+      "step": 368
+    },
+    {
+      "epoch": 0.8105436573311368,
+      "grad_norm": 2658981.0,
+      "learning_rate": 4.0594462540716616e-05,
+      "loss": 1.0001,
+      "step": 369
+    },
+    {
+      "epoch": 0.8127402526084568,
+      "grad_norm": 2533469.0,
+      "learning_rate": 4.0553745928338767e-05,
+      "loss": 0.796,
+      "step": 370
+    },
+    {
+      "epoch": 0.814936847885777,
+      "grad_norm": 2179804.5,
+      "learning_rate": 4.051302931596092e-05,
+      "loss": 1.2468,
+      "step": 371
+    },
+    {
+      "epoch": 0.8171334431630972,
+      "grad_norm": 1274361.625,
+      "learning_rate": 4.047231270358307e-05,
+      "loss": 1.0357,
+      "step": 372
+    },
+    {
+      "epoch": 0.8193300384404174,
+      "grad_norm": 5907040.0,
+      "learning_rate": 4.043159609120522e-05,
+      "loss": 1.1956,
+      "step": 373
+    },
+    {
+      "epoch": 0.8215266337177375,
+      "grad_norm": 9624350.0,
+      "learning_rate": 4.039087947882736e-05,
+      "loss": 1.0426,
+      "step": 374
+    },
+    {
+      "epoch": 0.8237232289950577,
+      "grad_norm": 4473219.5,
+      "learning_rate": 4.035016286644951e-05,
+      "loss": 1.1102,
+      "step": 375
+    },
+    {
+      "epoch": 0.8259198242723779,
+      "grad_norm": 3905218.25,
+      "learning_rate": 4.030944625407166e-05,
+      "loss": 1.2309,
+      "step": 376
+    },
+    {
+      "epoch": 0.828116419549698,
+      "grad_norm": 2925692.5,
+      "learning_rate": 4.026872964169381e-05,
+      "loss": 1.144,
+      "step": 377
+    },
+    {
+      "epoch": 0.8303130148270181,
+      "grad_norm": 2821423.0,
+      "learning_rate": 4.0228013029315963e-05,
+      "loss": 1.174,
+      "step": 378
+    },
+    {
+      "epoch": 0.8325096101043383,
+      "grad_norm": 3189121.25,
+      "learning_rate": 4.0187296416938114e-05,
+      "loss": 0.8828,
+      "step": 379
+    },
+    {
+      "epoch": 0.8347062053816584,
+      "grad_norm": 8460780.0,
+      "learning_rate": 4.0146579804560264e-05,
+      "loss": 1.2649,
+      "step": 380
+    },
+    {
+      "epoch": 0.8369028006589786,
+      "grad_norm": 699227.3125,
+      "learning_rate": 4.0105863192182415e-05,
+      "loss": 1.1021,
+      "step": 381
+    },
+    {
+      "epoch": 0.8390993959362988,
+      "grad_norm": 4438103.5,
+      "learning_rate": 4.0065146579804565e-05,
+      "loss": 1.1471,
+      "step": 382
+    },
+    {
+      "epoch": 0.8412959912136189,
+      "grad_norm": 7009095.5,
+      "learning_rate": 4.0024429967426716e-05,
+      "loss": 1.0706,
+      "step": 383
+    },
+    {
+      "epoch": 0.8434925864909391,
+      "grad_norm": 13305493.0,
+      "learning_rate": 3.9983713355048866e-05,
+      "loss": 1.1409,
+      "step": 384
+    },
+    {
+      "epoch": 0.8456891817682592,
+      "grad_norm": 1076842.125,
+      "learning_rate": 3.994299674267101e-05,
+      "loss": 1.262,
+      "step": 385
+    },
+    {
+      "epoch": 0.8478857770455793,
+      "grad_norm": 6116212.5,
+      "learning_rate": 3.990228013029316e-05,
+      "loss": 1.2028,
+      "step": 386
+    },
+    {
+      "epoch": 0.8500823723228995,
+      "grad_norm": 1569048.75,
+      "learning_rate": 3.986156351791531e-05,
+      "loss": 1.1047,
+      "step": 387
+    },
+    {
+      "epoch": 0.8522789676002197,
+      "grad_norm": 3176028.5,
+      "learning_rate": 3.982084690553746e-05,
+      "loss": 1.2642,
+      "step": 388
+    },
+    {
+      "epoch": 0.8544755628775398,
+      "grad_norm": 1386197.125,
+      "learning_rate": 3.978013029315961e-05,
+      "loss": 0.856,
+      "step": 389
+    },
+    {
+      "epoch": 0.85667215815486,
+      "grad_norm": 8060207.0,
+      "learning_rate": 3.973941368078176e-05,
+      "loss": 1.0036,
+      "step": 390
+    },
+    {
+      "epoch": 0.8588687534321802,
+      "grad_norm": 2319377.25,
+      "learning_rate": 3.969869706840391e-05,
+      "loss": 1.0639,
+      "step": 391
+    },
+    {
+      "epoch": 0.8610653487095002,
+      "grad_norm": 3404097.5,
+      "learning_rate": 3.965798045602606e-05,
+      "loss": 1.3316,
+      "step": 392
+    },
+    {
+      "epoch": 0.8632619439868204,
+      "grad_norm": 3128828.75,
+      "learning_rate": 3.9617263843648214e-05,
+      "loss": 1.4384,
+      "step": 393
+    },
+    {
+      "epoch": 0.8654585392641406,
+      "grad_norm": 6596271.5,
+      "learning_rate": 3.9576547231270364e-05,
+      "loss": 1.0946,
+      "step": 394
+    },
+    {
+      "epoch": 0.8676551345414607,
+      "grad_norm": 9033444.0,
+      "learning_rate": 3.953583061889251e-05,
+      "loss": 1.0786,
+      "step": 395
+    },
+    {
+      "epoch": 0.8698517298187809,
+      "grad_norm": 1401995.125,
+      "learning_rate": 3.949511400651466e-05,
+      "loss": 1.0008,
+      "step": 396
+    },
+    {
+      "epoch": 0.8720483250961011,
+      "grad_norm": 2229208.0,
+      "learning_rate": 3.945439739413681e-05,
+      "loss": 1.255,
+      "step": 397
+    },
+    {
+      "epoch": 0.8742449203734212,
+      "grad_norm": 3185306.75,
+      "learning_rate": 3.941368078175896e-05,
+      "loss": 0.9867,
+      "step": 398
+    },
+    {
+      "epoch": 0.8764415156507414,
+      "grad_norm": 5660074.5,
+      "learning_rate": 3.937296416938111e-05,
+      "loss": 1.077,
+      "step": 399
+    },
+    {
+      "epoch": 0.8786381109280615,
+      "grad_norm": 4436234.5,
+      "learning_rate": 3.933224755700326e-05,
+      "loss": 1.0976,
+      "step": 400
+    },
+    {
+      "epoch": 0.8808347062053816,
+      "grad_norm": 8006974.5,
+      "learning_rate": 3.929153094462541e-05,
+      "loss": 1.0351,
+      "step": 401
+    },
+    {
+      "epoch": 0.8830313014827018,
+      "grad_norm": 2672763.5,
+      "learning_rate": 3.925081433224756e-05,
+      "loss": 1.193,
+      "step": 402
+    },
+    {
+      "epoch": 0.885227896760022,
+      "grad_norm": 1554474.25,
+      "learning_rate": 3.921009771986971e-05,
+      "loss": 1.2548,
+      "step": 403
+    },
+    {
+      "epoch": 0.8874244920373421,
+      "grad_norm": 11707526.0,
+      "learning_rate": 3.916938110749186e-05,
+      "loss": 1.2995,
+      "step": 404
+    },
+    {
+      "epoch": 0.8896210873146623,
+      "grad_norm": 13413093.0,
+      "learning_rate": 3.912866449511401e-05,
+      "loss": 0.9584,
+      "step": 405
+    },
+    {
+      "epoch": 0.8918176825919825,
+      "grad_norm": 1175855.25,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 0.9755,
+      "step": 406
+    },
+    {
+      "epoch": 0.8940142778693025,
+      "grad_norm": 5227610.5,
+      "learning_rate": 3.9047231270358306e-05,
+      "loss": 1.0731,
+      "step": 407
+    },
+    {
+      "epoch": 0.8962108731466227,
+      "grad_norm": 3051889.0,
+      "learning_rate": 3.900651465798046e-05,
+      "loss": 1.0617,
+      "step": 408
+    },
+    {
+      "epoch": 0.8984074684239429,
+      "grad_norm": 1597198.75,
+      "learning_rate": 3.896579804560261e-05,
+      "loss": 0.9874,
+      "step": 409
+    },
+    {
+      "epoch": 0.900604063701263,
+      "grad_norm": 5104410.0,
+      "learning_rate": 3.892508143322476e-05,
+      "loss": 1.198,
+      "step": 410
+    },
+    {
+      "epoch": 0.9028006589785832,
+      "grad_norm": 3480690.75,
+      "learning_rate": 3.888436482084691e-05,
+      "loss": 0.9726,
+      "step": 411
+    },
+    {
+      "epoch": 0.9049972542559034,
+      "grad_norm": 2612652.25,
+      "learning_rate": 3.884364820846906e-05,
+      "loss": 0.9347,
+      "step": 412
+    },
+    {
+      "epoch": 0.9071938495332235,
+      "grad_norm": 1552913.125,
+      "learning_rate": 3.880293159609121e-05,
+      "loss": 1.1616,
+      "step": 413
+    },
+    {
+      "epoch": 0.9093904448105437,
+      "grad_norm": 3294075.75,
+      "learning_rate": 3.876221498371336e-05,
+      "loss": 1.1164,
+      "step": 414
+    },
+    {
+      "epoch": 0.9115870400878638,
+      "grad_norm": 8075532.0,
+      "learning_rate": 3.872149837133551e-05,
+      "loss": 0.9965,
+      "step": 415
+    },
+    {
+      "epoch": 0.9137836353651839,
+      "grad_norm": 5656122.5,
+      "learning_rate": 3.8680781758957654e-05,
+      "loss": 1.1889,
+      "step": 416
+    },
+    {
+      "epoch": 0.9159802306425041,
+      "grad_norm": 4610150.5,
+      "learning_rate": 3.8640065146579804e-05,
+      "loss": 1.0172,
+      "step": 417
+    },
+    {
+      "epoch": 0.9181768259198243,
+      "grad_norm": 2036415.125,
+      "learning_rate": 3.8599348534201955e-05,
+      "loss": 0.9757,
+      "step": 418
+    },
+    {
+      "epoch": 0.9203734211971444,
+      "grad_norm": 3243812.5,
+      "learning_rate": 3.8558631921824105e-05,
+      "loss": 0.965,
+      "step": 419
+    },
+    {
+      "epoch": 0.9225700164744646,
+      "grad_norm": 927049.8125,
+      "learning_rate": 3.8517915309446256e-05,
+      "loss": 1.1806,
+      "step": 420
+    },
+    {
+      "epoch": 0.9247666117517848,
+      "grad_norm": 8733473.0,
+      "learning_rate": 3.8477198697068406e-05,
+      "loss": 0.9931,
+      "step": 421
+    },
+    {
+      "epoch": 0.9269632070291048,
+      "grad_norm": 44334868.0,
+      "learning_rate": 3.8436482084690556e-05,
+      "loss": 1.0207,
+      "step": 422
+    },
+    {
+      "epoch": 0.929159802306425,
+      "grad_norm": 2465587.25,
+      "learning_rate": 3.839576547231271e-05,
+      "loss": 1.2748,
+      "step": 423
+    },
+    {
+      "epoch": 0.9313563975837452,
+      "grad_norm": 1078204.625,
+      "learning_rate": 3.835504885993486e-05,
+      "loss": 1.0208,
+      "step": 424
+    },
+    {
+      "epoch": 0.9335529928610653,
+      "grad_norm": 4282955.5,
+      "learning_rate": 3.831433224755701e-05,
+      "loss": 0.9353,
+      "step": 425
+    },
+    {
+      "epoch": 0.9357495881383855,
+      "grad_norm": 8485153.0,
+      "learning_rate": 3.827361563517916e-05,
+      "loss": 1.0844,
+      "step": 426
+    },
+    {
+      "epoch": 0.9379461834157057,
+      "grad_norm": 2435590.0,
+      "learning_rate": 3.82328990228013e-05,
+      "loss": 1.0136,
+      "step": 427
+    },
+    {
+      "epoch": 0.9401427786930258,
+      "grad_norm": 817770.375,
+      "learning_rate": 3.819218241042345e-05,
+      "loss": 0.8423,
+      "step": 428
+    },
+    {
+      "epoch": 0.942339373970346,
+      "grad_norm": 2028978.375,
+      "learning_rate": 3.81514657980456e-05,
+      "loss": 0.8444,
+      "step": 429
+    },
+    {
+      "epoch": 0.9445359692476661,
+      "grad_norm": 1814217.75,
+      "learning_rate": 3.811074918566775e-05,
+      "loss": 1.088,
+      "step": 430
+    },
+    {
+      "epoch": 0.9467325645249862,
+      "grad_norm": 2696938.25,
+      "learning_rate": 3.8070032573289904e-05,
+      "loss": 0.9114,
+      "step": 431
+    },
+    {
+      "epoch": 0.9489291598023064,
+      "grad_norm": 2703236.5,
+      "learning_rate": 3.8029315960912054e-05,
+      "loss": 1.2709,
+      "step": 432
+    },
+    {
+      "epoch": 0.9511257550796266,
+      "grad_norm": 2063091.625,
+      "learning_rate": 3.7988599348534205e-05,
+      "loss": 1.1004,
+      "step": 433
+    },
+    {
+      "epoch": 0.9533223503569467,
+      "grad_norm": 1677998.875,
+      "learning_rate": 3.7947882736156355e-05,
+      "loss": 0.8937,
+      "step": 434
+    },
+    {
+      "epoch": 0.9555189456342669,
+      "grad_norm": 1331143.0,
+      "learning_rate": 3.7907166123778506e-05,
+      "loss": 0.9216,
+      "step": 435
+    },
+    {
+      "epoch": 0.9577155409115871,
+      "grad_norm": 2384296.75,
+      "learning_rate": 3.7866449511400656e-05,
+      "loss": 0.906,
+      "step": 436
+    },
+    {
+      "epoch": 0.9599121361889071,
+      "grad_norm": 3361354.25,
+      "learning_rate": 3.7825732899022807e-05,
+      "loss": 1.0728,
+      "step": 437
+    },
+    {
+      "epoch": 0.9621087314662273,
+      "grad_norm": 1775946.25,
+      "learning_rate": 3.778501628664495e-05,
+      "loss": 0.9348,
+      "step": 438
+    },
+    {
+      "epoch": 0.9643053267435475,
+      "grad_norm": 2055518.75,
+      "learning_rate": 3.77442996742671e-05,
+      "loss": 1.3997,
+      "step": 439
+    },
+    {
+      "epoch": 0.9665019220208677,
+      "grad_norm": 2835310.75,
+      "learning_rate": 3.770358306188925e-05,
+      "loss": 1.1209,
+      "step": 440
+    },
+    {
+      "epoch": 0.9686985172981878,
+      "grad_norm": 2031844.625,
+      "learning_rate": 3.76628664495114e-05,
+      "loss": 0.9846,
+      "step": 441
+    },
+    {
+      "epoch": 0.970895112575508,
+      "grad_norm": 3366107.25,
+      "learning_rate": 3.762214983713355e-05,
+      "loss": 0.904,
+      "step": 442
+    },
+    {
+      "epoch": 0.9730917078528282,
+      "grad_norm": 1179491.5,
+      "learning_rate": 3.75814332247557e-05,
+      "loss": 1.0191,
+      "step": 443
+    },
+    {
+      "epoch": 0.9752883031301482,
+      "grad_norm": 2321940.25,
+      "learning_rate": 3.754071661237785e-05,
+      "loss": 1.1444,
+      "step": 444
+    },
+    {
+      "epoch": 0.9774848984074684,
+      "grad_norm": 5195643.5,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.0713,
+      "step": 445
+    },
+    {
+      "epoch": 0.9796814936847886,
+      "grad_norm": 2047747.625,
+      "learning_rate": 3.7459283387622154e-05,
+      "loss": 1.2122,
+      "step": 446
+    },
+    {
+      "epoch": 0.9818780889621087,
+      "grad_norm": 6513645.5,
+      "learning_rate": 3.7418566775244304e-05,
+      "loss": 1.3172,
+      "step": 447
+    },
+    {
+      "epoch": 0.9840746842394289,
+      "grad_norm": 701463.4375,
+      "learning_rate": 3.737785016286645e-05,
+      "loss": 1.0555,
+      "step": 448
+    },
+    {
+      "epoch": 0.9862712795167491,
+      "grad_norm": 2300114.5,
+      "learning_rate": 3.73371335504886e-05,
+      "loss": 1.1332,
+      "step": 449
+    },
+    {
+      "epoch": 0.9884678747940692,
+      "grad_norm": 3373970.25,
+      "learning_rate": 3.729641693811075e-05,
+      "loss": 1.0195,
+      "step": 450
+    },
+    {
+      "epoch": 0.9906644700713894,
+      "grad_norm": 3842199.5,
+      "learning_rate": 3.72557003257329e-05,
+      "loss": 1.1882,
+      "step": 451
+    },
+    {
+      "epoch": 0.9928610653487095,
+      "grad_norm": 4639401.5,
+      "learning_rate": 3.721498371335505e-05,
+      "loss": 1.1116,
+      "step": 452
+    },
+    {
+      "epoch": 0.9950576606260296,
+      "grad_norm": 985649.0,
+      "learning_rate": 3.71742671009772e-05,
+      "loss": 1.0859,
+      "step": 453
+    },
+    {
+      "epoch": 0.9972542559033498,
+      "grad_norm": 2646145.25,
+      "learning_rate": 3.713355048859935e-05,
+      "loss": 0.9918,
+      "step": 454
+    },
+    {
+      "epoch": 0.99945085118067,
+      "grad_norm": 4004655.5,
+      "learning_rate": 3.70928338762215e-05,
+      "loss": 1.0025,
+      "step": 455
+    },
+    {
+      "epoch": 1.00164744645799,
+      "grad_norm": 1429614.75,
+      "learning_rate": 3.705211726384365e-05,
+      "loss": 1.2421,
+      "step": 456
+    },
+    {
+      "epoch": 1.0038440417353103,
+      "grad_norm": 1839059.375,
+      "learning_rate": 3.70114006514658e-05,
+      "loss": 1.189,
+      "step": 457
+    },
+    {
+      "epoch": 1.0060406370126305,
+      "grad_norm": 1606543.875,
+      "learning_rate": 3.697068403908795e-05,
+      "loss": 0.9676,
+      "step": 458
+    },
+    {
+      "epoch": 1.0082372322899507,
+      "grad_norm": 2606808.25,
+      "learning_rate": 3.6929967426710096e-05,
+      "loss": 0.9147,
+      "step": 459
+    },
+    {
+      "epoch": 1.0104338275672706,
+      "grad_norm": 5435780.5,
+      "learning_rate": 3.688925081433225e-05,
+      "loss": 0.9722,
+      "step": 460
+    },
+    {
+      "epoch": 1.0126304228445908,
+      "grad_norm": 784391.0,
+      "learning_rate": 3.68485342019544e-05,
+      "loss": 0.7687,
+      "step": 461
+    },
+    {
+      "epoch": 1.014827018121911,
+      "grad_norm": 2984975.0,
+      "learning_rate": 3.680781758957655e-05,
+      "loss": 1.3204,
+      "step": 462
+    },
+    {
+      "epoch": 1.0170236133992312,
+      "grad_norm": 1064107.75,
+      "learning_rate": 3.67671009771987e-05,
+      "loss": 1.2093,
+      "step": 463
+    },
+    {
+      "epoch": 1.0192202086765514,
+      "grad_norm": 2042366.0,
+      "learning_rate": 3.672638436482085e-05,
+      "loss": 0.9512,
+      "step": 464
+    },
+    {
+      "epoch": 1.0214168039538716,
+      "grad_norm": 2486868.75,
+      "learning_rate": 3.6685667752443e-05,
+      "loss": 1.2525,
+      "step": 465
+    },
+    {
+      "epoch": 1.0236133992311915,
+      "grad_norm": 975525.5,
+      "learning_rate": 3.664495114006515e-05,
+      "loss": 1.201,
+      "step": 466
+    },
+    {
+      "epoch": 1.0258099945085117,
+      "grad_norm": 3376055.75,
+      "learning_rate": 3.66042345276873e-05,
+      "loss": 0.9612,
+      "step": 467
+    },
+    {
+      "epoch": 1.028006589785832,
+      "grad_norm": 3357101.5,
+      "learning_rate": 3.656351791530945e-05,
+      "loss": 0.9997,
+      "step": 468
+    },
+    {
+      "epoch": 1.0302031850631521,
+      "grad_norm": 4590952.0,
+      "learning_rate": 3.6522801302931594e-05,
+      "loss": 1.121,
+      "step": 469
+    },
+    {
+      "epoch": 1.0323997803404723,
+      "grad_norm": 1144018.5,
+      "learning_rate": 3.6482084690553745e-05,
+      "loss": 1.1968,
+      "step": 470
+    },
+    {
+      "epoch": 1.0345963756177925,
+      "grad_norm": 2809369.25,
+      "learning_rate": 3.6441368078175895e-05,
+      "loss": 0.8149,
+      "step": 471
+    },
+    {
+      "epoch": 1.0367929708951125,
+      "grad_norm": 3959089.0,
+      "learning_rate": 3.6400651465798046e-05,
+      "loss": 1.0397,
+      "step": 472
+    },
+    {
+      "epoch": 1.0389895661724327,
+      "grad_norm": 10863800.0,
+      "learning_rate": 3.6359934853420196e-05,
+      "loss": 1.202,
+      "step": 473
+    },
+    {
+      "epoch": 1.0411861614497528,
+      "grad_norm": 7142044.0,
+      "learning_rate": 3.6319218241042346e-05,
+      "loss": 1.0329,
+      "step": 474
+    },
+    {
+      "epoch": 1.043382756727073,
+      "grad_norm": 2808248.25,
+      "learning_rate": 3.62785016286645e-05,
+      "loss": 1.0679,
+      "step": 475
+    },
+    {
+      "epoch": 1.0455793520043932,
+      "grad_norm": 3557599.75,
+      "learning_rate": 3.623778501628665e-05,
+      "loss": 1.3316,
+      "step": 476
+    },
+    {
+      "epoch": 1.0477759472817134,
+      "grad_norm": 3720504.25,
+      "learning_rate": 3.61970684039088e-05,
+      "loss": 0.9171,
+      "step": 477
+    },
+    {
+      "epoch": 1.0499725425590336,
+      "grad_norm": 1874587.875,
+      "learning_rate": 3.615635179153095e-05,
+      "loss": 1.1792,
+      "step": 478
+    },
+    {
+      "epoch": 1.0521691378363536,
+      "grad_norm": 2560153.5,
+      "learning_rate": 3.61156351791531e-05,
+      "loss": 1.122,
+      "step": 479
+    },
+    {
+      "epoch": 1.0543657331136738,
+      "grad_norm": 1494178.875,
+      "learning_rate": 3.607491856677524e-05,
+      "loss": 0.841,
+      "step": 480
+    },
+    {
+      "epoch": 1.056562328390994,
+      "grad_norm": 1684792.25,
+      "learning_rate": 3.603420195439739e-05,
+      "loss": 0.9895,
+      "step": 481
+    },
+    {
+      "epoch": 1.0587589236683141,
+      "grad_norm": 1400023.625,
+      "learning_rate": 3.599348534201954e-05,
+      "loss": 0.8756,
+      "step": 482
+    },
+    {
+      "epoch": 1.0609555189456343,
+      "grad_norm": 1907544.625,
+      "learning_rate": 3.5952768729641694e-05,
+      "loss": 0.9737,
+      "step": 483
+    },
+    {
+      "epoch": 1.0631521142229543,
+      "grad_norm": 1084423.125,
+      "learning_rate": 3.5912052117263844e-05,
+      "loss": 1.0923,
+      "step": 484
+    },
+    {
+      "epoch": 1.0653487095002745,
+      "grad_norm": 1053309.5,
+      "learning_rate": 3.5871335504885995e-05,
+      "loss": 0.9703,
+      "step": 485
+    },
+    {
+      "epoch": 1.0675453047775947,
+      "grad_norm": 4440510.0,
+      "learning_rate": 3.5830618892508145e-05,
+      "loss": 1.0654,
+      "step": 486
+    },
+    {
+      "epoch": 1.0697419000549149,
+      "grad_norm": 33323360.0,
+      "learning_rate": 3.5789902280130296e-05,
+      "loss": 0.984,
+      "step": 487
+    },
+    {
+      "epoch": 1.071938495332235,
+      "grad_norm": 2328885.5,
+      "learning_rate": 3.5749185667752446e-05,
+      "loss": 1.1061,
+      "step": 488
+    },
+    {
+      "epoch": 1.0741350906095553,
+      "grad_norm": 1977401.375,
+      "learning_rate": 3.5708469055374597e-05,
+      "loss": 0.9013,
+      "step": 489
+    },
+    {
+      "epoch": 1.0763316858868754,
+      "grad_norm": 1513479.0,
+      "learning_rate": 3.566775244299674e-05,
+      "loss": 0.8682,
+      "step": 490
+    },
+    {
+      "epoch": 1.0785282811641954,
+      "grad_norm": 3539828.0,
+      "learning_rate": 3.562703583061889e-05,
+      "loss": 1.207,
+      "step": 491
+    },
+    {
+      "epoch": 1.0807248764415156,
+      "grad_norm": 2879739.5,
+      "learning_rate": 3.558631921824104e-05,
+      "loss": 0.9897,
+      "step": 492
+    },
+    {
+      "epoch": 1.0829214717188358,
+      "grad_norm": 4785177.5,
+      "learning_rate": 3.554560260586319e-05,
+      "loss": 1.0769,
+      "step": 493
+    },
+    {
+      "epoch": 1.085118066996156,
+      "grad_norm": 1834819.875,
+      "learning_rate": 3.550488599348534e-05,
+      "loss": 1.0115,
+      "step": 494
+    },
+    {
+      "epoch": 1.0873146622734762,
+      "grad_norm": 10225133.0,
+      "learning_rate": 3.546416938110749e-05,
+      "loss": 0.9557,
+      "step": 495
+    },
+    {
+      "epoch": 1.0895112575507964,
+      "grad_norm": 3666309.5,
+      "learning_rate": 3.542345276872964e-05,
+      "loss": 0.8644,
+      "step": 496
+    },
+    {
+      "epoch": 1.0917078528281163,
+      "grad_norm": 1728965.0,
+      "learning_rate": 3.5382736156351793e-05,
+      "loss": 0.9809,
+      "step": 497
+    },
+    {
+      "epoch": 1.0939044481054365,
+      "grad_norm": 1135593.0,
+      "learning_rate": 3.5342019543973944e-05,
+      "loss": 0.9188,
+      "step": 498
+    },
+    {
+      "epoch": 1.0961010433827567,
+      "grad_norm": 1068885.625,
+      "learning_rate": 3.5301302931596094e-05,
+      "loss": 1.116,
+      "step": 499
+    },
+    {
+      "epoch": 1.098297638660077,
+      "grad_norm": 1417097.5,
+      "learning_rate": 3.5260586319218245e-05,
+      "loss": 1.0794,
+      "step": 500
+    },
+    {
+      "epoch": 1.100494233937397,
+      "grad_norm": 863194.6875,
+      "learning_rate": 3.521986970684039e-05,
+      "loss": 1.3076,
+      "step": 501
+    },
+    {
+      "epoch": 1.1026908292147173,
+      "grad_norm": 1992718.25,
+      "learning_rate": 3.517915309446254e-05,
+      "loss": 0.8904,
+      "step": 502
+    },
+    {
+      "epoch": 1.1048874244920373,
+      "grad_norm": 3217854.5,
+      "learning_rate": 3.513843648208469e-05,
+      "loss": 1.0706,
+      "step": 503
+    },
+    {
+      "epoch": 1.1070840197693574,
+      "grad_norm": 3231949.0,
+      "learning_rate": 3.509771986970684e-05,
+      "loss": 0.8292,
+      "step": 504
+    },
+    {
+      "epoch": 1.1092806150466776,
+      "grad_norm": 2536457.5,
+      "learning_rate": 3.505700325732899e-05,
+      "loss": 1.2546,
+      "step": 505
+    },
+    {
+      "epoch": 1.1114772103239978,
+      "grad_norm": 1454219.5,
+      "learning_rate": 3.501628664495114e-05,
+      "loss": 1.0541,
+      "step": 506
+    },
+    {
+      "epoch": 1.113673805601318,
+      "grad_norm": 2954982.5,
+      "learning_rate": 3.497557003257329e-05,
+      "loss": 1.2105,
+      "step": 507
+    },
+    {
+      "epoch": 1.1158704008786382,
+      "grad_norm": 2043719.625,
+      "learning_rate": 3.493485342019544e-05,
+      "loss": 1.1649,
+      "step": 508
+    },
+    {
+      "epoch": 1.1180669961559582,
+      "grad_norm": 2218879.5,
+      "learning_rate": 3.489413680781759e-05,
+      "loss": 1.0053,
+      "step": 509
+    },
+    {
+      "epoch": 1.1202635914332784,
+      "grad_norm": 1429378.375,
+      "learning_rate": 3.485342019543974e-05,
+      "loss": 1.0528,
+      "step": 510
+    },
+    {
+      "epoch": 1.1224601867105986,
+      "grad_norm": 1704472.375,
+      "learning_rate": 3.481270358306189e-05,
+      "loss": 0.9153,
+      "step": 511
+    },
+    {
+      "epoch": 1.1246567819879187,
+      "grad_norm": 2308783.75,
+      "learning_rate": 3.477198697068404e-05,
+      "loss": 1.0339,
+      "step": 512
+    },
+    {
+      "epoch": 1.126853377265239,
+      "grad_norm": 1174828.25,
+      "learning_rate": 3.473127035830619e-05,
+      "loss": 1.1144,
+      "step": 513
+    },
+    {
+      "epoch": 1.1290499725425591,
+      "grad_norm": 837354.5625,
+      "learning_rate": 3.469055374592834e-05,
+      "loss": 0.9514,
+      "step": 514
+    },
+    {
+      "epoch": 1.131246567819879,
+      "grad_norm": 12757209.0,
+      "learning_rate": 3.464983713355049e-05,
+      "loss": 1.0184,
+      "step": 515
+    },
+    {
+      "epoch": 1.1334431630971993,
+      "grad_norm": 1239238.375,
+      "learning_rate": 3.460912052117264e-05,
+      "loss": 1.065,
+      "step": 516
+    },
+    {
+      "epoch": 1.1356397583745195,
+      "grad_norm": 2364006.25,
+      "learning_rate": 3.456840390879479e-05,
+      "loss": 1.1727,
+      "step": 517
+    },
+    {
+      "epoch": 1.1378363536518397,
+      "grad_norm": 2163078.25,
+      "learning_rate": 3.452768729641694e-05,
+      "loss": 0.7989,
+      "step": 518
+    },
+    {
+      "epoch": 1.1400329489291599,
+      "grad_norm": 2489831.0,
+      "learning_rate": 3.448697068403909e-05,
+      "loss": 0.8195,
+      "step": 519
+    },
+    {
+      "epoch": 1.14222954420648,
+      "grad_norm": 5182464.5,
+      "learning_rate": 3.444625407166124e-05,
+      "loss": 0.9715,
+      "step": 520
+    },
+    {
+      "epoch": 1.1444261394838002,
+      "grad_norm": 1047131.3125,
+      "learning_rate": 3.440553745928339e-05,
+      "loss": 0.9709,
+      "step": 521
+    },
+    {
+      "epoch": 1.1466227347611202,
+      "grad_norm": 2147539.75,
+      "learning_rate": 3.4364820846905535e-05,
+      "loss": 0.9364,
+      "step": 522
+    },
+    {
+      "epoch": 1.1488193300384404,
+      "grad_norm": 3809388.0,
+      "learning_rate": 3.4324104234527685e-05,
+      "loss": 0.88,
+      "step": 523
+    },
+    {
+      "epoch": 1.1510159253157606,
+      "grad_norm": 2460255.75,
+      "learning_rate": 3.4283387622149836e-05,
+      "loss": 1.1045,
+      "step": 524
+    },
+    {
+      "epoch": 1.1532125205930808,
+      "grad_norm": 1872048.5,
+      "learning_rate": 3.4242671009771986e-05,
+      "loss": 0.9671,
+      "step": 525
+    },
+    {
+      "epoch": 1.155409115870401,
+      "grad_norm": 2828754.0,
+      "learning_rate": 3.4201954397394136e-05,
+      "loss": 1.2421,
+      "step": 526
+    },
+    {
+      "epoch": 1.157605711147721,
+      "grad_norm": 922732.625,
+      "learning_rate": 3.416123778501629e-05,
+      "loss": 0.8052,
+      "step": 527
+    },
+    {
+      "epoch": 1.1598023064250411,
+      "grad_norm": 2906426.25,
+      "learning_rate": 3.412052117263844e-05,
+      "loss": 1.0421,
+      "step": 528
+    },
+    {
+      "epoch": 1.1619989017023613,
+      "grad_norm": 1269072.125,
+      "learning_rate": 3.407980456026059e-05,
+      "loss": 0.8174,
+      "step": 529
+    },
+    {
+      "epoch": 1.1641954969796815,
+      "grad_norm": 3359562.0,
+      "learning_rate": 3.403908794788274e-05,
+      "loss": 0.8345,
+      "step": 530
+    },
+    {
+      "epoch": 1.1663920922570017,
+      "grad_norm": 2199189.5,
+      "learning_rate": 3.399837133550489e-05,
+      "loss": 0.9069,
+      "step": 531
+    },
+    {
+      "epoch": 1.1685886875343219,
+      "grad_norm": 2671526.5,
+      "learning_rate": 3.395765472312704e-05,
+      "loss": 1.0736,
+      "step": 532
+    },
+    {
+      "epoch": 1.170785282811642,
+      "grad_norm": 2958884.25,
+      "learning_rate": 3.391693811074918e-05,
+      "loss": 1.1218,
+      "step": 533
+    },
+    {
+      "epoch": 1.172981878088962,
+      "grad_norm": 3142275.5,
+      "learning_rate": 3.387622149837133e-05,
+      "loss": 1.1607,
+      "step": 534
+    },
+    {
+      "epoch": 1.1751784733662822,
+      "grad_norm": 3029684.25,
+      "learning_rate": 3.3835504885993484e-05,
+      "loss": 0.9749,
+      "step": 535
+    },
+    {
+      "epoch": 1.1773750686436024,
+      "grad_norm": 5081185.5,
+      "learning_rate": 3.3794788273615634e-05,
+      "loss": 0.8919,
+      "step": 536
+    },
+    {
+      "epoch": 1.1795716639209226,
+      "grad_norm": 3074030.5,
+      "learning_rate": 3.3754071661237785e-05,
+      "loss": 0.9555,
+      "step": 537
+    },
+    {
+      "epoch": 1.1817682591982428,
+      "grad_norm": 2502342.25,
+      "learning_rate": 3.3713355048859935e-05,
+      "loss": 0.853,
+      "step": 538
+    },
+    {
+      "epoch": 1.1839648544755628,
+      "grad_norm": 1463895.5,
+      "learning_rate": 3.3672638436482086e-05,
+      "loss": 1.1891,
+      "step": 539
+    },
+    {
+      "epoch": 1.186161449752883,
+      "grad_norm": 10479990.0,
+      "learning_rate": 3.3631921824104236e-05,
+      "loss": 0.8144,
+      "step": 540
+    },
+    {
+      "epoch": 1.1883580450302031,
+      "grad_norm": 1236805.125,
+      "learning_rate": 3.3591205211726387e-05,
+      "loss": 0.9896,
+      "step": 541
+    },
+    {
+      "epoch": 1.1905546403075233,
+      "grad_norm": 6678930.0,
+      "learning_rate": 3.355048859934854e-05,
+      "loss": 0.9032,
+      "step": 542
+    },
+    {
+      "epoch": 1.1927512355848435,
+      "grad_norm": 979997.25,
+      "learning_rate": 3.350977198697068e-05,
+      "loss": 1.0415,
+      "step": 543
+    },
+    {
+      "epoch": 1.1949478308621637,
+      "grad_norm": 2850193.75,
+      "learning_rate": 3.346905537459283e-05,
+      "loss": 1.0752,
+      "step": 544
+    },
+    {
+      "epoch": 1.197144426139484,
+      "grad_norm": 1898114.375,
+      "learning_rate": 3.342833876221498e-05,
+      "loss": 1.0866,
+      "step": 545
+    },
+    {
+      "epoch": 1.1993410214168039,
+      "grad_norm": 4640011.5,
+      "learning_rate": 3.338762214983713e-05,
+      "loss": 0.981,
+      "step": 546
+    },
+    {
+      "epoch": 1.201537616694124,
+      "grad_norm": 2855971.0,
+      "learning_rate": 3.334690553745928e-05,
+      "loss": 1.3749,
+      "step": 547
+    },
+    {
+      "epoch": 1.2037342119714443,
+      "grad_norm": 6875571.5,
+      "learning_rate": 3.330618892508143e-05,
+      "loss": 1.0454,
+      "step": 548
+    },
+    {
+      "epoch": 1.2059308072487644,
+      "grad_norm": 1109115.75,
+      "learning_rate": 3.3265472312703583e-05,
+      "loss": 1.3001,
+      "step": 549
+    },
+    {
+      "epoch": 1.2081274025260846,
+      "grad_norm": 1510378.875,
+      "learning_rate": 3.3224755700325734e-05,
+      "loss": 1.1756,
+      "step": 550
+    },
+    {
+      "epoch": 1.2103239978034046,
+      "grad_norm": 7556440.0,
+      "learning_rate": 3.3184039087947884e-05,
+      "loss": 0.9424,
+      "step": 551
+    },
+    {
+      "epoch": 1.2125205930807248,
+      "grad_norm": 2186743.0,
+      "learning_rate": 3.3143322475570035e-05,
+      "loss": 1.195,
+      "step": 552
+    },
+    {
+      "epoch": 1.214717188358045,
+      "grad_norm": 1276715.75,
+      "learning_rate": 3.3102605863192185e-05,
+      "loss": 1.065,
+      "step": 553
+    },
+    {
+      "epoch": 1.2169137836353652,
+      "grad_norm": 2144975.0,
+      "learning_rate": 3.306188925081433e-05,
+      "loss": 1.0202,
+      "step": 554
+    },
+    {
+      "epoch": 1.2191103789126854,
+      "grad_norm": 1065460.625,
+      "learning_rate": 3.302117263843648e-05,
+      "loss": 1.1498,
+      "step": 555
+    },
+    {
+      "epoch": 1.2213069741900056,
+      "grad_norm": 1640937.375,
+      "learning_rate": 3.298045602605863e-05,
+      "loss": 0.9547,
+      "step": 556
+    },
+    {
+      "epoch": 1.2235035694673257,
+      "grad_norm": 4626417.5,
+      "learning_rate": 3.293973941368078e-05,
+      "loss": 1.0908,
+      "step": 557
+    },
+    {
+      "epoch": 1.2257001647446457,
+      "grad_norm": 4708184.5,
+      "learning_rate": 3.289902280130293e-05,
+      "loss": 1.0227,
+      "step": 558
+    },
+    {
+      "epoch": 1.227896760021966,
+      "grad_norm": 6358211.0,
+      "learning_rate": 3.285830618892508e-05,
+      "loss": 1.3205,
+      "step": 559
+    },
+    {
+      "epoch": 1.230093355299286,
+      "grad_norm": 2445859.75,
+      "learning_rate": 3.281758957654723e-05,
+      "loss": 1.0514,
+      "step": 560
+    },
+    {
+      "epoch": 1.2322899505766063,
+      "grad_norm": 1775960.375,
+      "learning_rate": 3.277687296416938e-05,
+      "loss": 1.0132,
+      "step": 561
+    },
+    {
+      "epoch": 1.2344865458539265,
+      "grad_norm": 1763435.875,
+      "learning_rate": 3.273615635179153e-05,
+      "loss": 1.022,
+      "step": 562
+    },
+    {
+      "epoch": 1.2366831411312464,
+      "grad_norm": 1846549.875,
+      "learning_rate": 3.269543973941368e-05,
+      "loss": 1.1575,
+      "step": 563
+    },
+    {
+      "epoch": 1.2388797364085666,
+      "grad_norm": 2215039.0,
+      "learning_rate": 3.265472312703583e-05,
+      "loss": 0.8991,
+      "step": 564
+    },
+    {
+      "epoch": 1.2410763316858868,
+      "grad_norm": 2253913.75,
+      "learning_rate": 3.261400651465798e-05,
+      "loss": 1.1167,
+      "step": 565
+    },
+    {
+      "epoch": 1.243272926963207,
+      "grad_norm": 1932530.125,
+      "learning_rate": 3.257328990228013e-05,
+      "loss": 0.9793,
+      "step": 566
+    },
+    {
+      "epoch": 1.2454695222405272,
+      "grad_norm": 7755891.0,
+      "learning_rate": 3.253257328990228e-05,
+      "loss": 0.9557,
+      "step": 567
+    },
+    {
+      "epoch": 1.2476661175178474,
+      "grad_norm": 2169052.25,
+      "learning_rate": 3.249185667752443e-05,
+      "loss": 1.1117,
+      "step": 568
+    },
+    {
+      "epoch": 1.2498627127951676,
+      "grad_norm": 902500.3125,
+      "learning_rate": 3.245114006514658e-05,
+      "loss": 0.8676,
+      "step": 569
+    },
+    {
+      "epoch": 1.2520593080724876,
+      "grad_norm": 1681763.75,
+      "learning_rate": 3.241042345276873e-05,
+      "loss": 0.7691,
+      "step": 570
+    },
+    {
+      "epoch": 1.2542559033498077,
+      "grad_norm": 2140215.5,
+      "learning_rate": 3.236970684039088e-05,
+      "loss": 1.255,
+      "step": 571
+    },
+    {
+      "epoch": 1.256452498627128,
+      "grad_norm": 2489232.5,
+      "learning_rate": 3.232899022801303e-05,
+      "loss": 1.2292,
+      "step": 572
+    },
+    {
+      "epoch": 1.2586490939044481,
+      "grad_norm": 1778908.125,
+      "learning_rate": 3.228827361563518e-05,
+      "loss": 1.1078,
+      "step": 573
+    },
+    {
+      "epoch": 1.2608456891817683,
+      "grad_norm": 1682346.375,
+      "learning_rate": 3.224755700325733e-05,
+      "loss": 1.0804,
+      "step": 574
+    },
+    {
+      "epoch": 1.2630422844590883,
+      "grad_norm": 6326409.0,
+      "learning_rate": 3.2206840390879475e-05,
+      "loss": 1.0588,
+      "step": 575
+    },
+    {
+      "epoch": 1.2652388797364087,
+      "grad_norm": 1429248.875,
+      "learning_rate": 3.2166123778501625e-05,
+      "loss": 0.9257,
+      "step": 576
+    },
+    {
+      "epoch": 1.2674354750137287,
+      "grad_norm": 3434332.5,
+      "learning_rate": 3.2125407166123776e-05,
+      "loss": 1.212,
+      "step": 577
+    },
+    {
+      "epoch": 1.2696320702910489,
+      "grad_norm": 7180025.5,
+      "learning_rate": 3.2084690553745926e-05,
+      "loss": 0.7371,
+      "step": 578
+    },
+    {
+      "epoch": 1.271828665568369,
+      "grad_norm": 1106258.75,
+      "learning_rate": 3.204397394136808e-05,
+      "loss": 0.7993,
+      "step": 579
+    },
+    {
+      "epoch": 1.2740252608456892,
+      "grad_norm": 1581180.125,
+      "learning_rate": 3.200325732899023e-05,
+      "loss": 0.8851,
+      "step": 580
+    },
+    {
+      "epoch": 1.2762218561230094,
+      "grad_norm": 2324048.5,
+      "learning_rate": 3.196254071661238e-05,
+      "loss": 1.0372,
+      "step": 581
+    },
+    {
+      "epoch": 1.2784184514003294,
+      "grad_norm": 6947504.5,
+      "learning_rate": 3.192182410423453e-05,
+      "loss": 0.9574,
+      "step": 582
+    },
+    {
+      "epoch": 1.2806150466776496,
+      "grad_norm": 1462683.0,
+      "learning_rate": 3.188110749185668e-05,
+      "loss": 0.9216,
+      "step": 583
+    },
+    {
+      "epoch": 1.2828116419549698,
+      "grad_norm": 2913410.75,
+      "learning_rate": 3.184039087947883e-05,
+      "loss": 0.8608,
+      "step": 584
+    },
+    {
+      "epoch": 1.28500823723229,
+      "grad_norm": 4090048.25,
+      "learning_rate": 3.179967426710097e-05,
+      "loss": 1.106,
+      "step": 585
+    },
+    {
+      "epoch": 1.2872048325096102,
+      "grad_norm": 3561927.25,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 0.9929,
+      "step": 586
+    },
+    {
+      "epoch": 1.2894014277869301,
+      "grad_norm": 3616566.25,
+      "learning_rate": 3.1718241042345274e-05,
+      "loss": 0.9794,
+      "step": 587
+    },
+    {
+      "epoch": 1.2915980230642505,
+      "grad_norm": 2062823.625,
+      "learning_rate": 3.1677524429967424e-05,
+      "loss": 1.0174,
+      "step": 588
+    },
+    {
+      "epoch": 1.2937946183415705,
+      "grad_norm": 2298692.5,
+      "learning_rate": 3.1636807817589575e-05,
+      "loss": 1.0413,
+      "step": 589
+    },
+    {
+      "epoch": 1.2959912136188907,
+      "grad_norm": 3364183.75,
+      "learning_rate": 3.1596091205211725e-05,
+      "loss": 0.9379,
+      "step": 590
+    },
+    {
+      "epoch": 1.2981878088962109,
+      "grad_norm": 3687931.5,
+      "learning_rate": 3.1555374592833876e-05,
+      "loss": 0.8064,
+      "step": 591
+    },
+    {
+      "epoch": 1.300384404173531,
+      "grad_norm": 2602794.75,
+      "learning_rate": 3.1514657980456026e-05,
+      "loss": 0.9446,
+      "step": 592
+    },
+    {
+      "epoch": 1.3025809994508513,
+      "grad_norm": 1864590.25,
+      "learning_rate": 3.1473941368078177e-05,
+      "loss": 0.8386,
+      "step": 593
+    },
+    {
+      "epoch": 1.3047775947281712,
+      "grad_norm": 3516272.75,
+      "learning_rate": 3.143322475570033e-05,
+      "loss": 1.0291,
+      "step": 594
+    },
+    {
+      "epoch": 1.3069741900054914,
+      "grad_norm": 3399038.5,
+      "learning_rate": 3.139250814332248e-05,
+      "loss": 1.1659,
+      "step": 595
+    },
+    {
+      "epoch": 1.3091707852828116,
+      "grad_norm": 1190284.125,
+      "learning_rate": 3.135179153094462e-05,
+      "loss": 0.9057,
+      "step": 596
+    },
+    {
+      "epoch": 1.3113673805601318,
+      "grad_norm": 1566911.625,
+      "learning_rate": 3.131107491856677e-05,
+      "loss": 0.9039,
+      "step": 597
+    },
+    {
+      "epoch": 1.313563975837452,
+      "grad_norm": 3796274.75,
+      "learning_rate": 3.127035830618892e-05,
+      "loss": 0.8109,
+      "step": 598
+    },
+    {
+      "epoch": 1.3157605711147722,
+      "grad_norm": 1320055.125,
+      "learning_rate": 3.122964169381108e-05,
+      "loss": 0.8848,
+      "step": 599
+    },
+    {
+      "epoch": 1.3179571663920924,
+      "grad_norm": 1807005.875,
+      "learning_rate": 3.118892508143323e-05,
+      "loss": 1.1493,
+      "step": 600
+    },
+    {
+      "epoch": 1.3201537616694123,
+      "grad_norm": 4024277.25,
+      "learning_rate": 3.114820846905538e-05,
+      "loss": 0.9482,
+      "step": 601
+    },
+    {
+      "epoch": 1.3223503569467325,
+      "grad_norm": 2139763.5,
+      "learning_rate": 3.110749185667753e-05,
+      "loss": 0.9137,
+      "step": 602
+    },
+    {
+      "epoch": 1.3245469522240527,
+      "grad_norm": 803565.6875,
+      "learning_rate": 3.106677524429968e-05,
+      "loss": 1.113,
+      "step": 603
+    },
+    {
+      "epoch": 1.326743547501373,
+      "grad_norm": 4807849.5,
+      "learning_rate": 3.102605863192183e-05,
+      "loss": 0.9969,
+      "step": 604
+    },
+    {
+      "epoch": 1.328940142778693,
+      "grad_norm": 7463085.5,
+      "learning_rate": 3.0985342019543975e-05,
+      "loss": 1.1558,
+      "step": 605
+    },
+    {
+      "epoch": 1.331136738056013,
+      "grad_norm": 1131262.0,
+      "learning_rate": 3.0944625407166126e-05,
+      "loss": 0.8634,
+      "step": 606
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 5142421.5,
+      "learning_rate": 3.0903908794788276e-05,
+      "loss": 0.9501,
+      "step": 607
+    },
+    {
+      "epoch": 1.3355299286106534,
+      "grad_norm": 5606728.5,
+      "learning_rate": 3.086319218241043e-05,
+      "loss": 1.0221,
+      "step": 608
+    },
+    {
+      "epoch": 1.3377265238879736,
+      "grad_norm": 1429995.375,
+      "learning_rate": 3.082247557003258e-05,
+      "loss": 1.0901,
+      "step": 609
+    },
+    {
+      "epoch": 1.3399231191652938,
+      "grad_norm": 2226860.75,
+      "learning_rate": 3.078175895765473e-05,
+      "loss": 1.0867,
+      "step": 610
+    },
+    {
+      "epoch": 1.342119714442614,
+      "grad_norm": 2064716.5,
+      "learning_rate": 3.074104234527688e-05,
+      "loss": 1.0293,
+      "step": 611
+    },
+    {
+      "epoch": 1.3443163097199342,
+      "grad_norm": 1999541.0,
+      "learning_rate": 3.070032573289903e-05,
+      "loss": 0.933,
+      "step": 612
+    },
+    {
+      "epoch": 1.3465129049972542,
+      "grad_norm": 1612567.25,
+      "learning_rate": 3.065960912052118e-05,
+      "loss": 1.1263,
+      "step": 613
+    },
+    {
+      "epoch": 1.3487095002745744,
+      "grad_norm": 2965922.0,
+      "learning_rate": 3.061889250814333e-05,
+      "loss": 0.8896,
+      "step": 614
+    },
+    {
+      "epoch": 1.3509060955518946,
+      "grad_norm": 4770825.0,
+      "learning_rate": 3.057817589576547e-05,
+      "loss": 1.1418,
+      "step": 615
+    },
+    {
+      "epoch": 1.3531026908292147,
+      "grad_norm": 1483664.375,
+      "learning_rate": 3.0537459283387624e-05,
+      "loss": 1.1124,
+      "step": 616
+    },
+    {
+      "epoch": 1.355299286106535,
+      "grad_norm": 3633950.75,
+      "learning_rate": 3.0496742671009777e-05,
+      "loss": 0.9104,
+      "step": 617
+    },
+    {
+      "epoch": 1.357495881383855,
+      "grad_norm": 7052970.5,
+      "learning_rate": 3.0456026058631924e-05,
+      "loss": 1.2539,
+      "step": 618
+    },
+    {
+      "epoch": 1.359692476661175,
+      "grad_norm": 2404985.25,
+      "learning_rate": 3.0415309446254075e-05,
+      "loss": 0.8026,
+      "step": 619
+    },
+    {
+      "epoch": 1.3618890719384953,
+      "grad_norm": 4617832.0,
+      "learning_rate": 3.0374592833876225e-05,
+      "loss": 0.886,
+      "step": 620
+    },
+    {
+      "epoch": 1.3640856672158155,
+      "grad_norm": 2048281.25,
+      "learning_rate": 3.0333876221498376e-05,
+      "loss": 1.2382,
+      "step": 621
+    },
+    {
+      "epoch": 1.3662822624931357,
+      "grad_norm": 22741952.0,
+      "learning_rate": 3.0293159609120526e-05,
+      "loss": 0.8907,
+      "step": 622
+    },
+    {
+      "epoch": 1.3684788577704559,
+      "grad_norm": 8022245.5,
+      "learning_rate": 3.0252442996742673e-05,
+      "loss": 0.9153,
+      "step": 623
+    },
+    {
+      "epoch": 1.370675453047776,
+      "grad_norm": 6066003.0,
+      "learning_rate": 3.0211726384364824e-05,
+      "loss": 0.9589,
+      "step": 624
+    },
+    {
+      "epoch": 1.372872048325096,
+      "grad_norm": 1827867.375,
+      "learning_rate": 3.0171009771986974e-05,
+      "loss": 1.0236,
+      "step": 625
+    },
+    {
+      "epoch": 1.3750686436024162,
+      "grad_norm": 1424068.75,
+      "learning_rate": 3.0130293159609125e-05,
+      "loss": 0.8737,
+      "step": 626
+    },
+    {
+      "epoch": 1.3772652388797364,
+      "grad_norm": 930774.1875,
+      "learning_rate": 3.0089576547231275e-05,
+      "loss": 0.8848,
+      "step": 627
+    },
+    {
+      "epoch": 1.3794618341570566,
+      "grad_norm": 926491.5625,
+      "learning_rate": 3.0048859934853422e-05,
+      "loss": 0.7191,
+      "step": 628
+    },
+    {
+      "epoch": 1.3816584294343768,
+      "grad_norm": 2670064.25,
+      "learning_rate": 3.0008143322475573e-05,
+      "loss": 1.0093,
+      "step": 629
+    },
+    {
+      "epoch": 1.3838550247116967,
+      "grad_norm": 2301365.5,
+      "learning_rate": 2.9967426710097723e-05,
+      "loss": 0.8947,
+      "step": 630
+    },
+    {
+      "epoch": 1.3860516199890172,
+      "grad_norm": 4394828.0,
+      "learning_rate": 2.9926710097719874e-05,
+      "loss": 0.9056,
+      "step": 631
+    },
+    {
+      "epoch": 1.3882482152663371,
+      "grad_norm": 2985291.5,
+      "learning_rate": 2.9885993485342024e-05,
+      "loss": 1.1915,
+      "step": 632
+    },
+    {
+      "epoch": 1.3904448105436573,
+      "grad_norm": 2978567.75,
+      "learning_rate": 2.9845276872964175e-05,
+      "loss": 0.9916,
+      "step": 633
+    },
+    {
+      "epoch": 1.3926414058209775,
+      "grad_norm": 2449312.0,
+      "learning_rate": 2.980456026058632e-05,
+      "loss": 0.9344,
+      "step": 634
+    },
+    {
+      "epoch": 1.3948380010982977,
+      "grad_norm": 1105509.875,
+      "learning_rate": 2.9763843648208472e-05,
+      "loss": 0.9923,
+      "step": 635
+    },
+    {
+      "epoch": 1.3970345963756179,
+      "grad_norm": 1652851.25,
+      "learning_rate": 2.9723127035830623e-05,
+      "loss": 1.0319,
+      "step": 636
+    },
+    {
+      "epoch": 1.3992311916529379,
+      "grad_norm": 3967650.25,
+      "learning_rate": 2.9682410423452773e-05,
+      "loss": 0.953,
+      "step": 637
+    },
+    {
+      "epoch": 1.401427786930258,
+      "grad_norm": 2996345.5,
+      "learning_rate": 2.9641693811074923e-05,
+      "loss": 0.8407,
+      "step": 638
+    },
+    {
+      "epoch": 1.4036243822075782,
+      "grad_norm": 2338844.25,
+      "learning_rate": 2.960097719869707e-05,
+      "loss": 0.8501,
+      "step": 639
+    },
+    {
+      "epoch": 1.4058209774848984,
+      "grad_norm": 43624384.0,
+      "learning_rate": 2.956026058631922e-05,
+      "loss": 1.2485,
+      "step": 640
+    },
+    {
+      "epoch": 1.4080175727622186,
+      "grad_norm": 1716135.875,
+      "learning_rate": 2.951954397394137e-05,
+      "loss": 0.8539,
+      "step": 641
+    },
+    {
+      "epoch": 1.4102141680395386,
+      "grad_norm": 2625146.25,
+      "learning_rate": 2.9478827361563522e-05,
+      "loss": 1.1601,
+      "step": 642
+    },
+    {
+      "epoch": 1.412410763316859,
+      "grad_norm": 3038696.5,
+      "learning_rate": 2.9438110749185672e-05,
+      "loss": 0.9097,
+      "step": 643
+    },
+    {
+      "epoch": 1.414607358594179,
+      "grad_norm": 2740381.25,
+      "learning_rate": 2.939739413680782e-05,
+      "loss": 0.8644,
+      "step": 644
+    },
+    {
+      "epoch": 1.4168039538714992,
+      "grad_norm": 7278885.0,
+      "learning_rate": 2.935667752442997e-05,
+      "loss": 1.0384,
+      "step": 645
+    },
+    {
+      "epoch": 1.4190005491488193,
+      "grad_norm": 1225708.75,
+      "learning_rate": 2.931596091205212e-05,
+      "loss": 0.9877,
+      "step": 646
+    },
+    {
+      "epoch": 1.4211971444261395,
+      "grad_norm": 5216531.0,
+      "learning_rate": 2.927524429967427e-05,
+      "loss": 0.9479,
+      "step": 647
+    },
+    {
+      "epoch": 1.4233937397034597,
+      "grad_norm": 3480651.0,
+      "learning_rate": 2.923452768729642e-05,
+      "loss": 0.9442,
+      "step": 648
+    },
+    {
+      "epoch": 1.4255903349807797,
+      "grad_norm": 2862786.75,
+      "learning_rate": 2.9193811074918572e-05,
+      "loss": 1.0182,
+      "step": 649
+    },
+    {
+      "epoch": 1.4277869302580999,
+      "grad_norm": 2844908.25,
+      "learning_rate": 2.915309446254072e-05,
+      "loss": 0.7705,
+      "step": 650
+    },
+    {
+      "epoch": 1.42998352553542,
+      "grad_norm": 66916792.0,
+      "learning_rate": 2.911237785016287e-05,
+      "loss": 0.8414,
+      "step": 651
+    },
+    {
+      "epoch": 1.4321801208127403,
+      "grad_norm": 6270535.5,
+      "learning_rate": 2.907166123778502e-05,
+      "loss": 1.0292,
+      "step": 652
+    },
+    {
+      "epoch": 1.4343767160900605,
+      "grad_norm": 1667101.625,
+      "learning_rate": 2.903094462540717e-05,
+      "loss": 0.9927,
+      "step": 653
+    },
+    {
+      "epoch": 1.4365733113673804,
+      "grad_norm": 4820360.5,
+      "learning_rate": 2.899022801302932e-05,
+      "loss": 0.9602,
+      "step": 654
+    },
+    {
+      "epoch": 1.4387699066447008,
+      "grad_norm": 3476306.0,
+      "learning_rate": 2.8949511400651468e-05,
+      "loss": 0.8817,
+      "step": 655
+    },
+    {
+      "epoch": 1.4409665019220208,
+      "grad_norm": 4604823.0,
+      "learning_rate": 2.8908794788273618e-05,
+      "loss": 1.0817,
+      "step": 656
+    },
+    {
+      "epoch": 1.443163097199341,
+      "grad_norm": 2913780.0,
+      "learning_rate": 2.886807817589577e-05,
+      "loss": 1.1313,
+      "step": 657
+    },
+    {
+      "epoch": 1.4453596924766612,
+      "grad_norm": 2701957.25,
+      "learning_rate": 2.882736156351792e-05,
+      "loss": 1.1245,
+      "step": 658
+    },
+    {
+      "epoch": 1.4475562877539814,
+      "grad_norm": 4105184.75,
+      "learning_rate": 2.878664495114007e-05,
+      "loss": 1.1811,
+      "step": 659
+    },
+    {
+      "epoch": 1.4497528830313016,
+      "grad_norm": 1852256.375,
+      "learning_rate": 2.8745928338762217e-05,
+      "loss": 0.9299,
+      "step": 660
+    },
+    {
+      "epoch": 1.4519494783086215,
+      "grad_norm": 1916288.875,
+      "learning_rate": 2.8705211726384367e-05,
+      "loss": 0.8732,
+      "step": 661
+    },
+    {
+      "epoch": 1.4541460735859417,
+      "grad_norm": 2566700.25,
+      "learning_rate": 2.8664495114006518e-05,
+      "loss": 0.946,
+      "step": 662
+    },
+    {
+      "epoch": 1.456342668863262,
+      "grad_norm": 4365289.5,
+      "learning_rate": 2.8623778501628668e-05,
+      "loss": 1.0269,
+      "step": 663
+    },
+    {
+      "epoch": 1.458539264140582,
+      "grad_norm": 1857104.625,
+      "learning_rate": 2.858306188925082e-05,
+      "loss": 0.8999,
+      "step": 664
+    },
+    {
+      "epoch": 1.4607358594179023,
+      "grad_norm": 4009175.75,
+      "learning_rate": 2.8542345276872965e-05,
+      "loss": 1.175,
+      "step": 665
+    },
+    {
+      "epoch": 1.4629324546952225,
+      "grad_norm": 4724137.5,
+      "learning_rate": 2.8501628664495116e-05,
+      "loss": 1.2009,
+      "step": 666
+    },
+    {
+      "epoch": 1.4651290499725427,
+      "grad_norm": 1823235.375,
+      "learning_rate": 2.8460912052117266e-05,
+      "loss": 0.9044,
+      "step": 667
+    },
+    {
+      "epoch": 1.4673256452498626,
+      "grad_norm": 2893977.25,
+      "learning_rate": 2.8420195439739417e-05,
+      "loss": 1.1381,
+      "step": 668
+    },
+    {
+      "epoch": 1.4695222405271828,
+      "grad_norm": 1991740.5,
+      "learning_rate": 2.8379478827361567e-05,
+      "loss": 0.8651,
+      "step": 669
+    },
+    {
+      "epoch": 1.471718835804503,
+      "grad_norm": 1454860.5,
+      "learning_rate": 2.8338762214983718e-05,
+      "loss": 0.9528,
+      "step": 670
+    },
+    {
+      "epoch": 1.4739154310818232,
+      "grad_norm": 3296265.25,
+      "learning_rate": 2.8298045602605865e-05,
+      "loss": 1.1168,
+      "step": 671
+    },
+    {
+      "epoch": 1.4761120263591434,
+      "grad_norm": 15473088.0,
+      "learning_rate": 2.8257328990228015e-05,
+      "loss": 1.019,
+      "step": 672
+    },
+    {
+      "epoch": 1.4783086216364634,
+      "grad_norm": 5723508.0,
+      "learning_rate": 2.8216612377850166e-05,
+      "loss": 0.9077,
+      "step": 673
+    },
+    {
+      "epoch": 1.4805052169137836,
+      "grad_norm": 4135130.25,
+      "learning_rate": 2.8175895765472316e-05,
+      "loss": 1.1502,
+      "step": 674
+    },
+    {
+      "epoch": 1.4827018121911038,
+      "grad_norm": 2929789.25,
+      "learning_rate": 2.8135179153094467e-05,
+      "loss": 0.8526,
+      "step": 675
+    },
+    {
+      "epoch": 1.484898407468424,
+      "grad_norm": 2066490.875,
+      "learning_rate": 2.8094462540716614e-05,
+      "loss": 0.9804,
+      "step": 676
+    },
+    {
+      "epoch": 1.4870950027457441,
+      "grad_norm": 29482206.0,
+      "learning_rate": 2.8053745928338764e-05,
+      "loss": 0.9839,
+      "step": 677
+    },
+    {
+      "epoch": 1.4892915980230643,
+      "grad_norm": 3601211.0,
+      "learning_rate": 2.8013029315960915e-05,
+      "loss": 1.1429,
+      "step": 678
+    },
+    {
+      "epoch": 1.4914881933003845,
+      "grad_norm": 1365523.375,
+      "learning_rate": 2.7972312703583065e-05,
+      "loss": 1.0505,
+      "step": 679
+    },
+    {
+      "epoch": 1.4936847885777045,
+      "grad_norm": 4186217.0,
+      "learning_rate": 2.7931596091205216e-05,
+      "loss": 1.197,
+      "step": 680
+    },
+    {
+      "epoch": 1.4958813838550247,
+      "grad_norm": 3639623.75,
+      "learning_rate": 2.7890879478827363e-05,
+      "loss": 0.9932,
+      "step": 681
+    },
+    {
+      "epoch": 1.4980779791323449,
+      "grad_norm": 2904028.75,
+      "learning_rate": 2.7850162866449513e-05,
+      "loss": 1.1189,
+      "step": 682
+    },
+    {
+      "epoch": 1.500274574409665,
+      "grad_norm": 1233690.875,
+      "learning_rate": 2.7809446254071664e-05,
+      "loss": 0.6865,
+      "step": 683
+    },
+    {
+      "epoch": 1.5024711696869852,
+      "grad_norm": 2961116.5,
+      "learning_rate": 2.7768729641693814e-05,
+      "loss": 0.8095,
+      "step": 684
+    },
+    {
+      "epoch": 1.5046677649643052,
+      "grad_norm": 4957673.5,
+      "learning_rate": 2.7728013029315965e-05,
+      "loss": 1.0933,
+      "step": 685
+    },
+    {
+      "epoch": 1.5068643602416256,
+      "grad_norm": 2345709.25,
+      "learning_rate": 2.768729641693811e-05,
+      "loss": 1.212,
+      "step": 686
+    },
+    {
+      "epoch": 1.5090609555189456,
+      "grad_norm": 6608750.0,
+      "learning_rate": 2.7646579804560262e-05,
+      "loss": 0.998,
+      "step": 687
+    },
+    {
+      "epoch": 1.5112575507962658,
+      "grad_norm": 962245.5625,
+      "learning_rate": 2.7605863192182412e-05,
+      "loss": 1.0163,
+      "step": 688
+    },
+    {
+      "epoch": 1.513454146073586,
+      "grad_norm": 3030468.0,
+      "learning_rate": 2.7565146579804563e-05,
+      "loss": 0.8987,
+      "step": 689
+    },
+    {
+      "epoch": 1.515650741350906,
+      "grad_norm": 1804417.5,
+      "learning_rate": 2.7524429967426713e-05,
+      "loss": 0.8992,
+      "step": 690
+    },
+    {
+      "epoch": 1.5178473366282264,
+      "grad_norm": 3727933.25,
+      "learning_rate": 2.7483713355048864e-05,
+      "loss": 1.0917,
+      "step": 691
+    },
+    {
+      "epoch": 1.5200439319055463,
+      "grad_norm": 1532706.5,
+      "learning_rate": 2.744299674267101e-05,
+      "loss": 1.1269,
+      "step": 692
+    },
+    {
+      "epoch": 1.5222405271828665,
+      "grad_norm": 1438356.125,
+      "learning_rate": 2.740228013029316e-05,
+      "loss": 0.9677,
+      "step": 693
+    },
+    {
+      "epoch": 1.5244371224601867,
+      "grad_norm": 1788669.75,
+      "learning_rate": 2.7361563517915312e-05,
+      "loss": 0.9762,
+      "step": 694
+    },
+    {
+      "epoch": 1.526633717737507,
+      "grad_norm": 1809670.625,
+      "learning_rate": 2.7320846905537462e-05,
+      "loss": 0.9647,
+      "step": 695
+    },
+    {
+      "epoch": 1.528830313014827,
+      "grad_norm": 2086256.0,
+      "learning_rate": 2.7280130293159613e-05,
+      "loss": 0.9269,
+      "step": 696
+    },
+    {
+      "epoch": 1.531026908292147,
+      "grad_norm": 2125394.0,
+      "learning_rate": 2.723941368078176e-05,
+      "loss": 1.0228,
+      "step": 697
+    },
+    {
+      "epoch": 1.5332235035694675,
+      "grad_norm": 2275858.0,
+      "learning_rate": 2.719869706840391e-05,
+      "loss": 0.908,
+      "step": 698
+    },
+    {
+      "epoch": 1.5354200988467874,
+      "grad_norm": 2501676.5,
+      "learning_rate": 2.715798045602606e-05,
+      "loss": 1.0269,
+      "step": 699
+    },
+    {
+      "epoch": 1.5376166941241076,
+      "grad_norm": 1784091.375,
+      "learning_rate": 2.711726384364821e-05,
+      "loss": 1.0795,
+      "step": 700
+    },
+    {
+      "epoch": 1.5398132894014278,
+      "grad_norm": 5035097.5,
+      "learning_rate": 2.707654723127036e-05,
+      "loss": 0.949,
+      "step": 701
+    },
+    {
+      "epoch": 1.5420098846787478,
+      "grad_norm": 862841.75,
+      "learning_rate": 2.703583061889251e-05,
+      "loss": 1.0089,
+      "step": 702
+    },
+    {
+      "epoch": 1.5442064799560682,
+      "grad_norm": 1456520.0,
+      "learning_rate": 2.699511400651466e-05,
+      "loss": 0.8402,
+      "step": 703
+    },
+    {
+      "epoch": 1.5464030752333882,
+      "grad_norm": 1621797.5,
+      "learning_rate": 2.695439739413681e-05,
+      "loss": 0.8711,
+      "step": 704
+    },
+    {
+      "epoch": 1.5485996705107083,
+      "grad_norm": 1343480.25,
+      "learning_rate": 2.691368078175896e-05,
+      "loss": 0.954,
+      "step": 705
+    },
+    {
+      "epoch": 1.5507962657880285,
+      "grad_norm": 5710876.5,
+      "learning_rate": 2.687296416938111e-05,
+      "loss": 0.9478,
+      "step": 706
+    },
+    {
+      "epoch": 1.5529928610653487,
+      "grad_norm": 1701437.125,
+      "learning_rate": 2.683224755700326e-05,
+      "loss": 1.015,
+      "step": 707
+    },
+    {
+      "epoch": 1.555189456342669,
+      "grad_norm": 2259651.75,
+      "learning_rate": 2.6791530944625408e-05,
+      "loss": 1.0208,
+      "step": 708
+    },
+    {
+      "epoch": 1.5573860516199889,
+      "grad_norm": 14428314.0,
+      "learning_rate": 2.675081433224756e-05,
+      "loss": 1.1434,
+      "step": 709
+    },
+    {
+      "epoch": 1.5595826468973093,
+      "grad_norm": 4960932.0,
+      "learning_rate": 2.671009771986971e-05,
+      "loss": 0.7325,
+      "step": 710
+    },
+    {
+      "epoch": 1.5617792421746293,
+      "grad_norm": 3129015.5,
+      "learning_rate": 2.666938110749186e-05,
+      "loss": 1.1014,
+      "step": 711
+    },
+    {
+      "epoch": 1.5639758374519495,
+      "grad_norm": 1832791.125,
+      "learning_rate": 2.662866449511401e-05,
+      "loss": 0.7567,
+      "step": 712
+    },
+    {
+      "epoch": 1.5661724327292696,
+      "grad_norm": 6786697.5,
+      "learning_rate": 2.6587947882736157e-05,
+      "loss": 0.7495,
+      "step": 713
+    },
+    {
+      "epoch": 1.5683690280065898,
+      "grad_norm": 1282327.375,
+      "learning_rate": 2.6547231270358307e-05,
+      "loss": 0.8281,
+      "step": 714
+    },
+    {
+      "epoch": 1.57056562328391,
+      "grad_norm": 1363727.125,
+      "learning_rate": 2.6506514657980458e-05,
+      "loss": 0.9672,
+      "step": 715
+    },
+    {
+      "epoch": 1.57276221856123,
+      "grad_norm": 3522238.5,
+      "learning_rate": 2.646579804560261e-05,
+      "loss": 1.141,
+      "step": 716
+    },
+    {
+      "epoch": 1.5749588138385504,
+      "grad_norm": 16011782.0,
+      "learning_rate": 2.642508143322476e-05,
+      "loss": 1.0315,
+      "step": 717
+    },
+    {
+      "epoch": 1.5771554091158704,
+      "grad_norm": 7736844.0,
+      "learning_rate": 2.6384364820846906e-05,
+      "loss": 0.898,
+      "step": 718
+    },
+    {
+      "epoch": 1.5793520043931906,
+      "grad_norm": 2005041.375,
+      "learning_rate": 2.6343648208469056e-05,
+      "loss": 0.9615,
+      "step": 719
+    },
+    {
+      "epoch": 1.5815485996705108,
+      "grad_norm": 1287283.0,
+      "learning_rate": 2.6302931596091207e-05,
+      "loss": 0.8312,
+      "step": 720
+    },
+    {
+      "epoch": 1.5837451949478307,
+      "grad_norm": 1586717.5,
+      "learning_rate": 2.6262214983713357e-05,
+      "loss": 1.0735,
+      "step": 721
+    },
+    {
+      "epoch": 1.5859417902251511,
+      "grad_norm": 2030226.375,
+      "learning_rate": 2.6221498371335508e-05,
+      "loss": 1.0678,
+      "step": 722
+    },
+    {
+      "epoch": 1.588138385502471,
+      "grad_norm": 1790424.875,
+      "learning_rate": 2.6180781758957655e-05,
+      "loss": 0.9037,
+      "step": 723
+    },
+    {
+      "epoch": 1.5903349807797913,
+      "grad_norm": 8990346.0,
+      "learning_rate": 2.6140065146579805e-05,
+      "loss": 0.846,
+      "step": 724
+    },
+    {
+      "epoch": 1.5925315760571115,
+      "grad_norm": 2293356.25,
+      "learning_rate": 2.6099348534201956e-05,
+      "loss": 1.0865,
+      "step": 725
+    },
+    {
+      "epoch": 1.5947281713344317,
+      "grad_norm": 1875239.5,
+      "learning_rate": 2.6058631921824106e-05,
+      "loss": 0.9228,
+      "step": 726
+    },
+    {
+      "epoch": 1.5969247666117519,
+      "grad_norm": 6825329.5,
+      "learning_rate": 2.6017915309446257e-05,
+      "loss": 0.9518,
+      "step": 727
+    },
+    {
+      "epoch": 1.5991213618890718,
+      "grad_norm": 4187193.5,
+      "learning_rate": 2.5977198697068407e-05,
+      "loss": 1.0072,
+      "step": 728
+    },
+    {
+      "epoch": 1.6013179571663922,
+      "grad_norm": 3281487.25,
+      "learning_rate": 2.5936482084690554e-05,
+      "loss": 0.903,
+      "step": 729
+    },
+    {
+      "epoch": 1.6035145524437122,
+      "grad_norm": 1101845.25,
+      "learning_rate": 2.5895765472312705e-05,
+      "loss": 0.9399,
+      "step": 730
+    },
+    {
+      "epoch": 1.6057111477210324,
+      "grad_norm": 1902750.5,
+      "learning_rate": 2.5855048859934855e-05,
+      "loss": 0.7314,
+      "step": 731
+    },
+    {
+      "epoch": 1.6079077429983526,
+      "grad_norm": 3878078.25,
+      "learning_rate": 2.5814332247557006e-05,
+      "loss": 0.9636,
+      "step": 732
+    },
+    {
+      "epoch": 1.6101043382756726,
+      "grad_norm": 3645371.75,
+      "learning_rate": 2.5773615635179156e-05,
+      "loss": 1.1867,
+      "step": 733
+    },
+    {
+      "epoch": 1.612300933552993,
+      "grad_norm": 3396171.5,
+      "learning_rate": 2.5732899022801303e-05,
+      "loss": 1.0462,
+      "step": 734
+    },
+    {
+      "epoch": 1.614497528830313,
+      "grad_norm": 5868019.5,
+      "learning_rate": 2.5692182410423454e-05,
+      "loss": 1.0165,
+      "step": 735
+    },
+    {
+      "epoch": 1.6166941241076331,
+      "grad_norm": 2716921.5,
+      "learning_rate": 2.5651465798045604e-05,
+      "loss": 1.0318,
+      "step": 736
+    },
+    {
+      "epoch": 1.6188907193849533,
+      "grad_norm": 5350965.0,
+      "learning_rate": 2.5610749185667754e-05,
+      "loss": 0.9408,
+      "step": 737
+    },
+    {
+      "epoch": 1.6210873146622735,
+      "grad_norm": 3399871.5,
+      "learning_rate": 2.5570032573289905e-05,
+      "loss": 0.9137,
+      "step": 738
+    },
+    {
+      "epoch": 1.6232839099395937,
+      "grad_norm": 2762854.25,
+      "learning_rate": 2.5529315960912052e-05,
+      "loss": 0.9694,
+      "step": 739
+    },
+    {
+      "epoch": 1.6254805052169137,
+      "grad_norm": 2268797.0,
+      "learning_rate": 2.5488599348534202e-05,
+      "loss": 0.985,
+      "step": 740
+    },
+    {
+      "epoch": 1.627677100494234,
+      "grad_norm": 1372928.25,
+      "learning_rate": 2.5447882736156353e-05,
+      "loss": 1.1548,
+      "step": 741
+    },
+    {
+      "epoch": 1.629873695771554,
+      "grad_norm": 1926868.75,
+      "learning_rate": 2.5407166123778503e-05,
+      "loss": 0.8923,
+      "step": 742
+    },
+    {
+      "epoch": 1.6320702910488742,
+      "grad_norm": 3559582.25,
+      "learning_rate": 2.5366449511400654e-05,
+      "loss": 0.9491,
+      "step": 743
+    },
+    {
+      "epoch": 1.6342668863261944,
+      "grad_norm": 2918439.25,
+      "learning_rate": 2.5325732899022804e-05,
+      "loss": 0.7789,
+      "step": 744
+    },
+    {
+      "epoch": 1.6364634816035144,
+      "grad_norm": 4484149.0,
+      "learning_rate": 2.528501628664495e-05,
+      "loss": 0.8928,
+      "step": 745
+    },
+    {
+      "epoch": 1.6386600768808348,
+      "grad_norm": 1943964.625,
+      "learning_rate": 2.5244299674267102e-05,
+      "loss": 0.7128,
+      "step": 746
+    },
+    {
+      "epoch": 1.6408566721581548,
+      "grad_norm": 1141900.125,
+      "learning_rate": 2.5203583061889252e-05,
+      "loss": 0.9789,
+      "step": 747
+    },
+    {
+      "epoch": 1.643053267435475,
+      "grad_norm": 2497627.75,
+      "learning_rate": 2.5162866449511403e-05,
+      "loss": 0.8517,
+      "step": 748
+    },
+    {
+      "epoch": 1.6452498627127952,
+      "grad_norm": 3997427.25,
+      "learning_rate": 2.5122149837133553e-05,
+      "loss": 0.8005,
+      "step": 749
+    },
+    {
+      "epoch": 1.6474464579901154,
+      "grad_norm": 3010877.75,
+      "learning_rate": 2.50814332247557e-05,
+      "loss": 0.9369,
+      "step": 750
+    },
+    {
+      "epoch": 1.6496430532674355,
+      "grad_norm": 1498485.5,
+      "learning_rate": 2.504071661237785e-05,
+      "loss": 0.9907,
+      "step": 751
+    },
+    {
+      "epoch": 1.6518396485447555,
+      "grad_norm": 4192252.25,
+      "learning_rate": 2.5e-05,
+      "loss": 1.0085,
+      "step": 752
+    },
+    {
+      "epoch": 1.654036243822076,
+      "grad_norm": 14295537.0,
+      "learning_rate": 2.495928338762215e-05,
+      "loss": 1.2434,
+      "step": 753
+    },
+    {
+      "epoch": 1.656232839099396,
+      "grad_norm": 3102339.25,
+      "learning_rate": 2.4918566775244302e-05,
+      "loss": 0.7278,
+      "step": 754
+    },
+    {
+      "epoch": 1.658429434376716,
+      "grad_norm": 2099065.75,
+      "learning_rate": 2.487785016286645e-05,
+      "loss": 0.8411,
+      "step": 755
+    },
+    {
+      "epoch": 1.6606260296540363,
+      "grad_norm": 1186886.25,
+      "learning_rate": 2.48371335504886e-05,
+      "loss": 0.9813,
+      "step": 756
+    },
+    {
+      "epoch": 1.6628226249313562,
+      "grad_norm": 2484405.75,
+      "learning_rate": 2.479641693811075e-05,
+      "loss": 1.0318,
+      "step": 757
+    },
+    {
+      "epoch": 1.6650192202086767,
+      "grad_norm": 1520484.25,
+      "learning_rate": 2.47557003257329e-05,
+      "loss": 1.0638,
+      "step": 758
+    },
+    {
+      "epoch": 1.6672158154859966,
+      "grad_norm": 3790638.0,
+      "learning_rate": 2.471498371335505e-05,
+      "loss": 1.1124,
+      "step": 759
+    },
+    {
+      "epoch": 1.6694124107633168,
+      "grad_norm": 1392529.875,
+      "learning_rate": 2.4674267100977198e-05,
+      "loss": 1.0049,
+      "step": 760
+    },
+    {
+      "epoch": 1.671609006040637,
+      "grad_norm": 1831293.375,
+      "learning_rate": 2.463355048859935e-05,
+      "loss": 0.9615,
+      "step": 761
+    },
+    {
+      "epoch": 1.6738056013179572,
+      "grad_norm": 2202462.0,
+      "learning_rate": 2.45928338762215e-05,
+      "loss": 0.9262,
+      "step": 762
+    },
+    {
+      "epoch": 1.6760021965952774,
+      "grad_norm": 2974580.0,
+      "learning_rate": 2.455211726384365e-05,
+      "loss": 0.9581,
+      "step": 763
+    },
+    {
+      "epoch": 1.6781987918725974,
+      "grad_norm": 3410456.5,
+      "learning_rate": 2.45114006514658e-05,
+      "loss": 1.0848,
+      "step": 764
+    },
+    {
+      "epoch": 1.6803953871499178,
+      "grad_norm": 2628518.0,
+      "learning_rate": 2.447068403908795e-05,
+      "loss": 1.025,
+      "step": 765
+    },
+    {
+      "epoch": 1.6825919824272377,
+      "grad_norm": 3277198.5,
+      "learning_rate": 2.4429967426710097e-05,
+      "loss": 0.8477,
+      "step": 766
+    },
+    {
+      "epoch": 1.684788577704558,
+      "grad_norm": 1577086.125,
+      "learning_rate": 2.4389250814332248e-05,
+      "loss": 0.9896,
+      "step": 767
+    },
+    {
+      "epoch": 1.6869851729818781,
+      "grad_norm": 2099248.5,
+      "learning_rate": 2.43485342019544e-05,
+      "loss": 0.8798,
+      "step": 768
+    },
+    {
+      "epoch": 1.689181768259198,
+      "grad_norm": 2793630.25,
+      "learning_rate": 2.430781758957655e-05,
+      "loss": 0.8697,
+      "step": 769
+    },
+    {
+      "epoch": 1.6913783635365185,
+      "grad_norm": 1477864.625,
+      "learning_rate": 2.42671009771987e-05,
+      "loss": 1.0805,
+      "step": 770
+    },
+    {
+      "epoch": 1.6935749588138385,
+      "grad_norm": 1724719.5,
+      "learning_rate": 2.4226384364820846e-05,
+      "loss": 1.1061,
+      "step": 771
+    },
+    {
+      "epoch": 1.6957715540911587,
+      "grad_norm": 2306153.25,
+      "learning_rate": 2.4185667752442997e-05,
+      "loss": 1.1124,
+      "step": 772
+    },
+    {
+      "epoch": 1.6979681493684788,
+      "grad_norm": 1640113.375,
+      "learning_rate": 2.4144951140065147e-05,
+      "loss": 0.8557,
+      "step": 773
+    },
+    {
+      "epoch": 1.700164744645799,
+      "grad_norm": 1191419.25,
+      "learning_rate": 2.4104234527687298e-05,
+      "loss": 0.8333,
+      "step": 774
+    },
+    {
+      "epoch": 1.7023613399231192,
+      "grad_norm": 1614368.25,
+      "learning_rate": 2.4063517915309448e-05,
+      "loss": 1.0401,
+      "step": 775
+    },
+    {
+      "epoch": 1.7045579352004392,
+      "grad_norm": 5358047.5,
+      "learning_rate": 2.4022801302931595e-05,
+      "loss": 0.8094,
+      "step": 776
+    },
+    {
+      "epoch": 1.7067545304777596,
+      "grad_norm": 1548433.125,
+      "learning_rate": 2.3982084690553746e-05,
+      "loss": 0.9898,
+      "step": 777
+    },
+    {
+      "epoch": 1.7089511257550796,
+      "grad_norm": 2612244.75,
+      "learning_rate": 2.3941368078175896e-05,
+      "loss": 1.053,
+      "step": 778
+    },
+    {
+      "epoch": 1.7111477210323998,
+      "grad_norm": 1874481.0,
+      "learning_rate": 2.3900651465798047e-05,
+      "loss": 0.955,
+      "step": 779
+    },
+    {
+      "epoch": 1.71334431630972,
+      "grad_norm": 2458017.0,
+      "learning_rate": 2.3859934853420197e-05,
+      "loss": 1.018,
+      "step": 780
+    },
+    {
+      "epoch": 1.7155409115870401,
+      "grad_norm": 3250076.5,
+      "learning_rate": 2.3819218241042348e-05,
+      "loss": 0.8788,
+      "step": 781
+    },
+    {
+      "epoch": 1.7177375068643603,
+      "grad_norm": 2810063.5,
+      "learning_rate": 2.3778501628664495e-05,
+      "loss": 0.8829,
+      "step": 782
+    },
+    {
+      "epoch": 1.7199341021416803,
+      "grad_norm": 1110818.75,
+      "learning_rate": 2.3737785016286645e-05,
+      "loss": 1.0315,
+      "step": 783
+    },
+    {
+      "epoch": 1.7221306974190007,
+      "grad_norm": 4222312.5,
+      "learning_rate": 2.3697068403908796e-05,
+      "loss": 0.9115,
+      "step": 784
+    },
+    {
+      "epoch": 1.7243272926963207,
+      "grad_norm": 1993576.75,
+      "learning_rate": 2.3656351791530946e-05,
+      "loss": 1.1828,
+      "step": 785
+    },
+    {
+      "epoch": 1.7265238879736409,
+      "grad_norm": 2254578.75,
+      "learning_rate": 2.3615635179153096e-05,
+      "loss": 0.8008,
+      "step": 786
+    },
+    {
+      "epoch": 1.728720483250961,
+      "grad_norm": 1843266.875,
+      "learning_rate": 2.3574918566775244e-05,
+      "loss": 0.8577,
+      "step": 787
+    },
+    {
+      "epoch": 1.730917078528281,
+      "grad_norm": 10157919.0,
+      "learning_rate": 2.3534201954397394e-05,
+      "loss": 0.9596,
+      "step": 788
+    },
+    {
+      "epoch": 1.7331136738056014,
+      "grad_norm": 1058657.75,
+      "learning_rate": 2.3493485342019544e-05,
+      "loss": 1.0256,
+      "step": 789
+    },
+    {
+      "epoch": 1.7353102690829214,
+      "grad_norm": 2351505.5,
+      "learning_rate": 2.3452768729641695e-05,
+      "loss": 1.1603,
+      "step": 790
+    },
+    {
+      "epoch": 1.7375068643602416,
+      "grad_norm": 3747576.0,
+      "learning_rate": 2.3412052117263845e-05,
+      "loss": 0.942,
+      "step": 791
+    },
+    {
+      "epoch": 1.7397034596375618,
+      "grad_norm": 1952515.375,
+      "learning_rate": 2.3371335504885992e-05,
+      "loss": 0.8485,
+      "step": 792
+    },
+    {
+      "epoch": 1.741900054914882,
+      "grad_norm": 1214669.25,
+      "learning_rate": 2.3330618892508143e-05,
+      "loss": 0.9078,
+      "step": 793
+    },
+    {
+      "epoch": 1.7440966501922022,
+      "grad_norm": 1812717.75,
+      "learning_rate": 2.3289902280130293e-05,
+      "loss": 1.093,
+      "step": 794
+    },
+    {
+      "epoch": 1.7462932454695221,
+      "grad_norm": 2928997.0,
+      "learning_rate": 2.3249185667752444e-05,
+      "loss": 0.8756,
+      "step": 795
+    },
+    {
+      "epoch": 1.7484898407468425,
+      "grad_norm": 1408084.75,
+      "learning_rate": 2.3208469055374594e-05,
+      "loss": 0.9035,
+      "step": 796
+    },
+    {
+      "epoch": 1.7506864360241625,
+      "grad_norm": 947632.9375,
+      "learning_rate": 2.316775244299674e-05,
+      "loss": 1.0898,
+      "step": 797
+    },
+    {
+      "epoch": 1.7528830313014827,
+      "grad_norm": 632850.6875,
+      "learning_rate": 2.3127035830618892e-05,
+      "loss": 0.8856,
+      "step": 798
+    },
+    {
+      "epoch": 1.755079626578803,
+      "grad_norm": 1382651.625,
+      "learning_rate": 2.3086319218241042e-05,
+      "loss": 1.0601,
+      "step": 799
+    },
+    {
+      "epoch": 1.7572762218561229,
+      "grad_norm": 1378480.5,
+      "learning_rate": 2.3045602605863193e-05,
+      "loss": 1.1028,
+      "step": 800
+    },
+    {
+      "epoch": 1.7594728171334433,
+      "grad_norm": 2792537.25,
+      "learning_rate": 2.3004885993485343e-05,
+      "loss": 1.0955,
+      "step": 801
+    },
+    {
+      "epoch": 1.7616694124107632,
+      "grad_norm": 1353128.625,
+      "learning_rate": 2.2964169381107494e-05,
+      "loss": 1.2505,
+      "step": 802
+    },
+    {
+      "epoch": 1.7638660076880834,
+      "grad_norm": 2303027.25,
+      "learning_rate": 2.292345276872964e-05,
+      "loss": 1.345,
+      "step": 803
+    },
+    {
+      "epoch": 1.7660626029654036,
+      "grad_norm": 5693921.0,
+      "learning_rate": 2.288273615635179e-05,
+      "loss": 1.3822,
+      "step": 804
+    },
+    {
+      "epoch": 1.7682591982427238,
+      "grad_norm": 3738693.25,
+      "learning_rate": 2.284201954397394e-05,
+      "loss": 0.9503,
+      "step": 805
+    },
+    {
+      "epoch": 1.770455793520044,
+      "grad_norm": 2033062.875,
+      "learning_rate": 2.2801302931596092e-05,
+      "loss": 0.9286,
+      "step": 806
+    },
+    {
+      "epoch": 1.772652388797364,
+      "grad_norm": 3043184.5,
+      "learning_rate": 2.2760586319218243e-05,
+      "loss": 0.9552,
+      "step": 807
+    },
+    {
+      "epoch": 1.7748489840746844,
+      "grad_norm": 17073168.0,
+      "learning_rate": 2.271986970684039e-05,
+      "loss": 1.1412,
+      "step": 808
+    },
+    {
+      "epoch": 1.7770455793520044,
+      "grad_norm": 1847339.375,
+      "learning_rate": 2.267915309446254e-05,
+      "loss": 1.0545,
+      "step": 809
+    },
+    {
+      "epoch": 1.7792421746293245,
+      "grad_norm": 1654972.5,
+      "learning_rate": 2.263843648208469e-05,
+      "loss": 0.7561,
+      "step": 810
+    },
+    {
+      "epoch": 1.7814387699066447,
+      "grad_norm": 2546275.75,
+      "learning_rate": 2.259771986970684e-05,
+      "loss": 0.75,
+      "step": 811
+    },
+    {
+      "epoch": 1.7836353651839647,
+      "grad_norm": 1502084.375,
+      "learning_rate": 2.255700325732899e-05,
+      "loss": 0.9629,
+      "step": 812
+    },
+    {
+      "epoch": 1.7858319604612851,
+      "grad_norm": 3375901.25,
+      "learning_rate": 2.251628664495114e-05,
+      "loss": 1.0457,
+      "step": 813
+    },
+    {
+      "epoch": 1.788028555738605,
+      "grad_norm": 1292429.375,
+      "learning_rate": 2.247557003257329e-05,
+      "loss": 0.9544,
+      "step": 814
+    },
+    {
+      "epoch": 1.7902251510159253,
+      "grad_norm": 933108.0,
+      "learning_rate": 2.243485342019544e-05,
+      "loss": 1.0348,
+      "step": 815
+    },
+    {
+      "epoch": 1.7924217462932455,
+      "grad_norm": 2944729.75,
+      "learning_rate": 2.239413680781759e-05,
+      "loss": 1.0974,
+      "step": 816
+    },
+    {
+      "epoch": 1.7946183415705657,
+      "grad_norm": 1032196.125,
+      "learning_rate": 2.235342019543974e-05,
+      "loss": 0.798,
+      "step": 817
+    },
+    {
+      "epoch": 1.7968149368478858,
+      "grad_norm": 3021491.0,
+      "learning_rate": 2.231270358306189e-05,
+      "loss": 0.9869,
+      "step": 818
+    },
+    {
+      "epoch": 1.7990115321252058,
+      "grad_norm": 3395504.5,
+      "learning_rate": 2.2271986970684038e-05,
+      "loss": 1.0056,
+      "step": 819
+    },
+    {
+      "epoch": 1.8012081274025262,
+      "grad_norm": 5716750.0,
+      "learning_rate": 2.223127035830619e-05,
+      "loss": 1.0663,
+      "step": 820
+    },
+    {
+      "epoch": 1.8034047226798462,
+      "grad_norm": 3206699.0,
+      "learning_rate": 2.219055374592834e-05,
+      "loss": 0.9659,
+      "step": 821
+    },
+    {
+      "epoch": 1.8056013179571664,
+      "grad_norm": 1377826.25,
+      "learning_rate": 2.214983713355049e-05,
+      "loss": 1.1123,
+      "step": 822
+    },
+    {
+      "epoch": 1.8077979132344866,
+      "grad_norm": 4747385.5,
+      "learning_rate": 2.210912052117264e-05,
+      "loss": 1.0971,
+      "step": 823
+    },
+    {
+      "epoch": 1.8099945085118065,
+      "grad_norm": 1629752.25,
+      "learning_rate": 2.2068403908794787e-05,
+      "loss": 0.9088,
+      "step": 824
+    },
+    {
+      "epoch": 1.812191103789127,
+      "grad_norm": 4017747.5,
+      "learning_rate": 2.2027687296416937e-05,
+      "loss": 0.8199,
+      "step": 825
+    },
+    {
+      "epoch": 1.814387699066447,
+      "grad_norm": 1177915.75,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 0.7467,
+      "step": 826
+    },
+    {
+      "epoch": 1.8165842943437671,
+      "grad_norm": 2375430.0,
+      "learning_rate": 2.1946254071661238e-05,
+      "loss": 0.9912,
+      "step": 827
+    },
+    {
+      "epoch": 1.8187808896210873,
+      "grad_norm": 3976207.5,
+      "learning_rate": 2.190553745928339e-05,
+      "loss": 0.9033,
+      "step": 828
+    },
+    {
+      "epoch": 1.8209774848984075,
+      "grad_norm": 2068701.0,
+      "learning_rate": 2.186482084690554e-05,
+      "loss": 0.8418,
+      "step": 829
+    },
+    {
+      "epoch": 1.8231740801757277,
+      "grad_norm": 1827384.375,
+      "learning_rate": 2.182410423452769e-05,
+      "loss": 0.7273,
+      "step": 830
+    },
+    {
+      "epoch": 1.8253706754530477,
+      "grad_norm": 4664636.0,
+      "learning_rate": 2.178338762214984e-05,
+      "loss": 0.6621,
+      "step": 831
+    },
+    {
+      "epoch": 1.827567270730368,
+      "grad_norm": 1358408.0,
+      "learning_rate": 2.174267100977199e-05,
+      "loss": 0.9815,
+      "step": 832
+    },
+    {
+      "epoch": 1.829763866007688,
+      "grad_norm": 1412634.75,
+      "learning_rate": 2.1701954397394138e-05,
+      "loss": 0.9999,
+      "step": 833
+    },
+    {
+      "epoch": 1.8319604612850082,
+      "grad_norm": 1786990.5,
+      "learning_rate": 2.1661237785016288e-05,
+      "loss": 1.1193,
+      "step": 834
+    },
+    {
+      "epoch": 1.8341570565623284,
+      "grad_norm": 2609843.0,
+      "learning_rate": 2.162052117263844e-05,
+      "loss": 1.1148,
+      "step": 835
+    },
+    {
+      "epoch": 1.8363536518396484,
+      "grad_norm": 2844453.25,
+      "learning_rate": 2.157980456026059e-05,
+      "loss": 0.8861,
+      "step": 836
+    },
+    {
+      "epoch": 1.8385502471169688,
+      "grad_norm": 2739915.5,
+      "learning_rate": 2.153908794788274e-05,
+      "loss": 0.8702,
+      "step": 837
+    },
+    {
+      "epoch": 1.8407468423942888,
+      "grad_norm": 1631944.875,
+      "learning_rate": 2.149837133550489e-05,
+      "loss": 0.998,
+      "step": 838
+    },
+    {
+      "epoch": 1.842943437671609,
+      "grad_norm": 6121211.5,
+      "learning_rate": 2.1457654723127037e-05,
+      "loss": 0.8038,
+      "step": 839
+    },
+    {
+      "epoch": 1.8451400329489291,
+      "grad_norm": 4640500.5,
+      "learning_rate": 2.1416938110749187e-05,
+      "loss": 0.9678,
+      "step": 840
+    },
+    {
+      "epoch": 1.8473366282262493,
+      "grad_norm": 5313952.5,
+      "learning_rate": 2.1376221498371338e-05,
+      "loss": 0.878,
+      "step": 841
+    },
+    {
+      "epoch": 1.8495332235035695,
+      "grad_norm": 4175497.25,
+      "learning_rate": 2.1335504885993488e-05,
+      "loss": 0.926,
+      "step": 842
+    },
+    {
+      "epoch": 1.8517298187808895,
+      "grad_norm": 4675688.0,
+      "learning_rate": 2.129478827361564e-05,
+      "loss": 1.156,
+      "step": 843
+    },
+    {
+      "epoch": 1.85392641405821,
+      "grad_norm": 1082311.75,
+      "learning_rate": 2.1254071661237786e-05,
+      "loss": 0.9651,
+      "step": 844
+    },
+    {
+      "epoch": 1.8561230093355299,
+      "grad_norm": 2972914.25,
+      "learning_rate": 2.1213355048859936e-05,
+      "loss": 0.8766,
+      "step": 845
+    },
+    {
+      "epoch": 1.85831960461285,
+      "grad_norm": 2460822.75,
+      "learning_rate": 2.1172638436482087e-05,
+      "loss": 0.8205,
+      "step": 846
+    },
+    {
+      "epoch": 1.8605161998901703,
+      "grad_norm": 4384400.5,
+      "learning_rate": 2.1131921824104237e-05,
+      "loss": 1.1443,
+      "step": 847
+    },
+    {
+      "epoch": 1.8627127951674904,
+      "grad_norm": 1705934.875,
+      "learning_rate": 2.1091205211726388e-05,
+      "loss": 0.8186,
+      "step": 848
+    },
+    {
+      "epoch": 1.8649093904448106,
+      "grad_norm": 7292398.5,
+      "learning_rate": 2.1050488599348535e-05,
+      "loss": 0.8395,
+      "step": 849
+    },
+    {
+      "epoch": 1.8671059857221306,
+      "grad_norm": 2746214.75,
+      "learning_rate": 2.1009771986970685e-05,
+      "loss": 0.8542,
+      "step": 850
+    },
+    {
+      "epoch": 1.869302580999451,
+      "grad_norm": 7449180.0,
+      "learning_rate": 2.0969055374592836e-05,
+      "loss": 0.743,
+      "step": 851
+    },
+    {
+      "epoch": 1.871499176276771,
+      "grad_norm": 1611397.5,
+      "learning_rate": 2.0928338762214986e-05,
+      "loss": 0.9605,
+      "step": 852
+    },
+    {
+      "epoch": 1.8736957715540912,
+      "grad_norm": 4226208.0,
+      "learning_rate": 2.0887622149837137e-05,
+      "loss": 1.1856,
+      "step": 853
+    },
+    {
+      "epoch": 1.8758923668314114,
+      "grad_norm": 1047346.3125,
+      "learning_rate": 2.0846905537459287e-05,
+      "loss": 1.2013,
+      "step": 854
+    },
+    {
+      "epoch": 1.8780889621087313,
+      "grad_norm": 744690.6875,
+      "learning_rate": 2.0806188925081434e-05,
+      "loss": 0.8555,
+      "step": 855
+    },
+    {
+      "epoch": 1.8802855573860517,
+      "grad_norm": 2378983.75,
+      "learning_rate": 2.0765472312703585e-05,
+      "loss": 1.1242,
+      "step": 856
+    },
+    {
+      "epoch": 1.8824821526633717,
+      "grad_norm": 3008229.25,
+      "learning_rate": 2.0724755700325735e-05,
+      "loss": 0.9021,
+      "step": 857
+    },
+    {
+      "epoch": 1.884678747940692,
+      "grad_norm": 4206646.0,
+      "learning_rate": 2.0684039087947885e-05,
+      "loss": 1.0611,
+      "step": 858
+    },
+    {
+      "epoch": 1.886875343218012,
+      "grad_norm": 1867490.875,
+      "learning_rate": 2.0643322475570036e-05,
+      "loss": 0.7819,
+      "step": 859
+    },
+    {
+      "epoch": 1.8890719384953323,
+      "grad_norm": 2511457.75,
+      "learning_rate": 2.0602605863192183e-05,
+      "loss": 0.8572,
+      "step": 860
+    },
+    {
+      "epoch": 1.8912685337726525,
+      "grad_norm": 2866090.75,
+      "learning_rate": 2.0561889250814333e-05,
+      "loss": 1.0606,
+      "step": 861
+    },
+    {
+      "epoch": 1.8934651290499724,
+      "grad_norm": 1889533.125,
+      "learning_rate": 2.0521172638436484e-05,
+      "loss": 0.9014,
+      "step": 862
+    },
+    {
+      "epoch": 1.8956617243272929,
+      "grad_norm": 3488134.5,
+      "learning_rate": 2.0480456026058634e-05,
+      "loss": 1.3292,
+      "step": 863
+    },
+    {
+      "epoch": 1.8978583196046128,
+      "grad_norm": 4074890.25,
+      "learning_rate": 2.0439739413680785e-05,
+      "loss": 0.8382,
+      "step": 864
+    },
+    {
+      "epoch": 1.900054914881933,
+      "grad_norm": 935045.1875,
+      "learning_rate": 2.0399022801302932e-05,
+      "loss": 1.0265,
+      "step": 865
+    },
+    {
+      "epoch": 1.9022515101592532,
+      "grad_norm": 2350366.25,
+      "learning_rate": 2.0358306188925082e-05,
+      "loss": 0.9239,
+      "step": 866
+    },
+    {
+      "epoch": 1.9044481054365732,
+      "grad_norm": 2311624.25,
+      "learning_rate": 2.0317589576547233e-05,
+      "loss": 0.955,
+      "step": 867
+    },
+    {
+      "epoch": 1.9066447007138936,
+      "grad_norm": 3290345.5,
+      "learning_rate": 2.0276872964169383e-05,
+      "loss": 1.2169,
+      "step": 868
+    },
+    {
+      "epoch": 1.9088412959912135,
+      "grad_norm": 3364535.75,
+      "learning_rate": 2.0236156351791534e-05,
+      "loss": 1.2757,
+      "step": 869
+    },
+    {
+      "epoch": 1.9110378912685337,
+      "grad_norm": 1841278.5,
+      "learning_rate": 2.019543973941368e-05,
+      "loss": 0.9095,
+      "step": 870
+    },
+    {
+      "epoch": 1.913234486545854,
+      "grad_norm": 1730140.625,
+      "learning_rate": 2.015472312703583e-05,
+      "loss": 0.8983,
+      "step": 871
+    },
+    {
+      "epoch": 1.9154310818231741,
+      "grad_norm": 1985711.75,
+      "learning_rate": 2.0114006514657982e-05,
+      "loss": 1.0678,
+      "step": 872
+    },
+    {
+      "epoch": 1.9176276771004943,
+      "grad_norm": 2492912.5,
+      "learning_rate": 2.0073289902280132e-05,
+      "loss": 1.0632,
+      "step": 873
+    },
+    {
+      "epoch": 1.9198242723778143,
+      "grad_norm": 840524.5,
+      "learning_rate": 2.0032573289902283e-05,
+      "loss": 1.1989,
+      "step": 874
+    },
+    {
+      "epoch": 1.9220208676551347,
+      "grad_norm": 2586521.75,
+      "learning_rate": 1.9991856677524433e-05,
+      "loss": 1.0871,
+      "step": 875
+    },
+    {
+      "epoch": 1.9242174629324547,
+      "grad_norm": 3944735.5,
+      "learning_rate": 1.995114006514658e-05,
+      "loss": 1.076,
+      "step": 876
+    },
+    {
+      "epoch": 1.9264140582097748,
+      "grad_norm": 1728677.25,
+      "learning_rate": 1.991042345276873e-05,
+      "loss": 0.8351,
+      "step": 877
+    },
+    {
+      "epoch": 1.928610653487095,
+      "grad_norm": 1139669.375,
+      "learning_rate": 1.986970684039088e-05,
+      "loss": 1.1347,
+      "step": 878
+    },
+    {
+      "epoch": 1.930807248764415,
+      "grad_norm": 1689345.125,
+      "learning_rate": 1.982899022801303e-05,
+      "loss": 0.9774,
+      "step": 879
+    },
+    {
+      "epoch": 1.9330038440417354,
+      "grad_norm": 1877441.625,
+      "learning_rate": 1.9788273615635182e-05,
+      "loss": 0.9019,
+      "step": 880
+    },
+    {
+      "epoch": 1.9352004393190554,
+      "grad_norm": 4681564.5,
+      "learning_rate": 1.974755700325733e-05,
+      "loss": 0.9878,
+      "step": 881
+    },
+    {
+      "epoch": 1.9373970345963756,
+      "grad_norm": 1886807.625,
+      "learning_rate": 1.970684039087948e-05,
+      "loss": 0.8153,
+      "step": 882
+    },
+    {
+      "epoch": 1.9395936298736958,
+      "grad_norm": 1673156.875,
+      "learning_rate": 1.966612377850163e-05,
+      "loss": 1.1225,
+      "step": 883
+    },
+    {
+      "epoch": 1.941790225151016,
+      "grad_norm": 827876.875,
+      "learning_rate": 1.962540716612378e-05,
+      "loss": 1.0361,
+      "step": 884
+    },
+    {
+      "epoch": 1.9439868204283361,
+      "grad_norm": 2247420.0,
+      "learning_rate": 1.958469055374593e-05,
+      "loss": 1.3891,
+      "step": 885
+    },
+    {
+      "epoch": 1.9461834157056561,
+      "grad_norm": 7995183.5,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 0.96,
+      "step": 886
+    },
+    {
+      "epoch": 1.9483800109829765,
+      "grad_norm": 8394645.0,
+      "learning_rate": 1.950325732899023e-05,
+      "loss": 0.8842,
+      "step": 887
+    },
+    {
+      "epoch": 1.9505766062602965,
+      "grad_norm": 3071906.75,
+      "learning_rate": 1.946254071661238e-05,
+      "loss": 0.8371,
+      "step": 888
+    },
+    {
+      "epoch": 1.9527732015376167,
+      "grad_norm": 2119468.0,
+      "learning_rate": 1.942182410423453e-05,
+      "loss": 1.0326,
+      "step": 889
+    },
+    {
+      "epoch": 1.9549697968149369,
+      "grad_norm": 1544775.75,
+      "learning_rate": 1.938110749185668e-05,
+      "loss": 0.9704,
+      "step": 890
+    },
+    {
+      "epoch": 1.9571663920922568,
+      "grad_norm": 559410.1875,
+      "learning_rate": 1.9340390879478827e-05,
+      "loss": 1.0885,
+      "step": 891
+    },
+    {
+      "epoch": 1.9593629873695773,
+      "grad_norm": 1544833.625,
+      "learning_rate": 1.9299674267100977e-05,
+      "loss": 0.8355,
+      "step": 892
+    },
+    {
+      "epoch": 1.9615595826468972,
+      "grad_norm": 3066966.5,
+      "learning_rate": 1.9258957654723128e-05,
+      "loss": 0.9095,
+      "step": 893
+    },
+    {
+      "epoch": 1.9637561779242174,
+      "grad_norm": 1151919.5,
+      "learning_rate": 1.9218241042345278e-05,
+      "loss": 1.0314,
+      "step": 894
+    },
+    {
+      "epoch": 1.9659527732015376,
+      "grad_norm": 1155955.25,
+      "learning_rate": 1.917752442996743e-05,
+      "loss": 0.7379,
+      "step": 895
+    },
+    {
+      "epoch": 1.9681493684788578,
+      "grad_norm": 1341321.0,
+      "learning_rate": 1.913680781758958e-05,
+      "loss": 1.1137,
+      "step": 896
+    },
+    {
+      "epoch": 1.970345963756178,
+      "grad_norm": 2434594.5,
+      "learning_rate": 1.9096091205211726e-05,
+      "loss": 0.8449,
+      "step": 897
+    },
+    {
+      "epoch": 1.972542559033498,
+      "grad_norm": 3220293.0,
+      "learning_rate": 1.9055374592833877e-05,
+      "loss": 1.105,
+      "step": 898
+    },
+    {
+      "epoch": 1.9747391543108184,
+      "grad_norm": 1123356.875,
+      "learning_rate": 1.9014657980456027e-05,
+      "loss": 1.0869,
+      "step": 899
+    },
+    {
+      "epoch": 1.9769357495881383,
+      "grad_norm": 2139459.5,
+      "learning_rate": 1.8973941368078178e-05,
+      "loss": 0.8493,
+      "step": 900
+    },
+    {
+      "epoch": 1.9791323448654585,
+      "grad_norm": 1807368.5,
+      "learning_rate": 1.8933224755700328e-05,
+      "loss": 1.0569,
+      "step": 901
+    },
+    {
+      "epoch": 1.9813289401427787,
+      "grad_norm": 1604600.625,
+      "learning_rate": 1.8892508143322475e-05,
+      "loss": 1.0396,
+      "step": 902
+    },
+    {
+      "epoch": 1.9835255354200987,
+      "grad_norm": 973187.75,
+      "learning_rate": 1.8851791530944626e-05,
+      "loss": 0.9603,
+      "step": 903
+    },
+    {
+      "epoch": 1.985722130697419,
+      "grad_norm": 8093962.5,
+      "learning_rate": 1.8811074918566776e-05,
+      "loss": 0.8432,
+      "step": 904
+    },
+    {
+      "epoch": 1.987918725974739,
+      "grad_norm": 2768687.25,
+      "learning_rate": 1.8770358306188927e-05,
+      "loss": 0.8317,
+      "step": 905
+    },
+    {
+      "epoch": 1.9901153212520593,
+      "grad_norm": 3241238.5,
+      "learning_rate": 1.8729641693811077e-05,
+      "loss": 1.0601,
+      "step": 906
+    },
+    {
+      "epoch": 1.9923119165293794,
+      "grad_norm": 1686224.125,
+      "learning_rate": 1.8688925081433224e-05,
+      "loss": 0.7889,
+      "step": 907
+    },
+    {
+      "epoch": 1.9945085118066996,
+      "grad_norm": 1666758.0,
+      "learning_rate": 1.8648208469055375e-05,
+      "loss": 0.8242,
+      "step": 908
+    },
+    {
+      "epoch": 1.9967051070840198,
+      "grad_norm": 1214850.75,
+      "learning_rate": 1.8607491856677525e-05,
+      "loss": 0.7095,
+      "step": 909
+    },
+    {
+      "epoch": 1.9989017023613398,
+      "grad_norm": 2622711.5,
+      "learning_rate": 1.8566775244299675e-05,
+      "loss": 0.9368,
+      "step": 910
+    },
+    {
+      "epoch": 2.00109829763866,
+      "grad_norm": 1199788.25,
+      "learning_rate": 1.8526058631921826e-05,
+      "loss": 1.1,
+      "step": 911
+    },
+    {
+      "epoch": 2.00329489291598,
+      "grad_norm": 1947196.75,
+      "learning_rate": 1.8485342019543976e-05,
+      "loss": 0.917,
+      "step": 912
+    },
+    {
+      "epoch": 2.0054914881933006,
+      "grad_norm": 11854929.0,
+      "learning_rate": 1.8444625407166123e-05,
+      "loss": 0.7397,
+      "step": 913
+    },
+    {
+      "epoch": 2.0076880834706206,
+      "grad_norm": 8723692.0,
+      "learning_rate": 1.8403908794788274e-05,
+      "loss": 0.9708,
+      "step": 914
+    },
+    {
+      "epoch": 2.0098846787479405,
+      "grad_norm": 4998067.5,
+      "learning_rate": 1.8363192182410424e-05,
+      "loss": 0.9252,
+      "step": 915
+    },
+    {
+      "epoch": 2.012081274025261,
+      "grad_norm": 3107527.5,
+      "learning_rate": 1.8322475570032575e-05,
+      "loss": 1.1949,
+      "step": 916
+    },
+    {
+      "epoch": 2.014277869302581,
+      "grad_norm": 3669729.0,
+      "learning_rate": 1.8281758957654725e-05,
+      "loss": 1.1023,
+      "step": 917
+    },
+    {
+      "epoch": 2.0164744645799013,
+      "grad_norm": 3712110.75,
+      "learning_rate": 1.8241042345276872e-05,
+      "loss": 1.1003,
+      "step": 918
+    },
+    {
+      "epoch": 2.0186710598572213,
+      "grad_norm": 1057183.75,
+      "learning_rate": 1.8200325732899023e-05,
+      "loss": 0.8883,
+      "step": 919
+    },
+    {
+      "epoch": 2.0208676551345413,
+      "grad_norm": 1825676.0,
+      "learning_rate": 1.8159609120521173e-05,
+      "loss": 0.8535,
+      "step": 920
+    },
+    {
+      "epoch": 2.0230642504118617,
+      "grad_norm": 1339001.625,
+      "learning_rate": 1.8118892508143324e-05,
+      "loss": 0.8299,
+      "step": 921
+    },
+    {
+      "epoch": 2.0252608456891816,
+      "grad_norm": 1299639.5,
+      "learning_rate": 1.8078175895765474e-05,
+      "loss": 0.9494,
+      "step": 922
+    },
+    {
+      "epoch": 2.027457440966502,
+      "grad_norm": 1104215.125,
+      "learning_rate": 1.803745928338762e-05,
+      "loss": 1.1119,
+      "step": 923
+    },
+    {
+      "epoch": 2.029654036243822,
+      "grad_norm": 1521235.5,
+      "learning_rate": 1.799674267100977e-05,
+      "loss": 1.0934,
+      "step": 924
+    },
+    {
+      "epoch": 2.0318506315211424,
+      "grad_norm": 1450141.25,
+      "learning_rate": 1.7956026058631922e-05,
+      "loss": 0.9021,
+      "step": 925
+    },
+    {
+      "epoch": 2.0340472267984624,
+      "grad_norm": 2891009.25,
+      "learning_rate": 1.7915309446254073e-05,
+      "loss": 0.9486,
+      "step": 926
+    },
+    {
+      "epoch": 2.0362438220757824,
+      "grad_norm": 3796478.0,
+      "learning_rate": 1.7874592833876223e-05,
+      "loss": 1.0405,
+      "step": 927
+    },
+    {
+      "epoch": 2.0384404173531028,
+      "grad_norm": 1879201.0,
+      "learning_rate": 1.783387622149837e-05,
+      "loss": 0.8909,
+      "step": 928
+    },
+    {
+      "epoch": 2.0406370126304227,
+      "grad_norm": 2370561.5,
+      "learning_rate": 1.779315960912052e-05,
+      "loss": 0.808,
+      "step": 929
+    },
+    {
+      "epoch": 2.042833607907743,
+      "grad_norm": 4074646.5,
+      "learning_rate": 1.775244299674267e-05,
+      "loss": 0.7881,
+      "step": 930
+    },
+    {
+      "epoch": 2.045030203185063,
+      "grad_norm": 1228167.5,
+      "learning_rate": 1.771172638436482e-05,
+      "loss": 0.8824,
+      "step": 931
+    },
+    {
+      "epoch": 2.047226798462383,
+      "grad_norm": 2978557.5,
+      "learning_rate": 1.7671009771986972e-05,
+      "loss": 1.0255,
+      "step": 932
+    },
+    {
+      "epoch": 2.0494233937397035,
+      "grad_norm": 1542763.25,
+      "learning_rate": 1.7630293159609122e-05,
+      "loss": 0.8345,
+      "step": 933
+    },
+    {
+      "epoch": 2.0516199890170235,
+      "grad_norm": 7343809.5,
+      "learning_rate": 1.758957654723127e-05,
+      "loss": 0.8659,
+      "step": 934
+    },
+    {
+      "epoch": 2.053816584294344,
+      "grad_norm": 1709089.125,
+      "learning_rate": 1.754885993485342e-05,
+      "loss": 0.9165,
+      "step": 935
+    },
+    {
+      "epoch": 2.056013179571664,
+      "grad_norm": 1519979.0,
+      "learning_rate": 1.750814332247557e-05,
+      "loss": 1.076,
+      "step": 936
+    },
+    {
+      "epoch": 2.0582097748489843,
+      "grad_norm": 771057.125,
+      "learning_rate": 1.746742671009772e-05,
+      "loss": 0.9893,
+      "step": 937
+    },
+    {
+      "epoch": 2.0604063701263042,
+      "grad_norm": 2190010.25,
+      "learning_rate": 1.742671009771987e-05,
+      "loss": 0.7343,
+      "step": 938
+    },
+    {
+      "epoch": 2.062602965403624,
+      "grad_norm": 1742635.0,
+      "learning_rate": 1.738599348534202e-05,
+      "loss": 1.1407,
+      "step": 939
+    },
+    {
+      "epoch": 2.0647995606809446,
+      "grad_norm": 1713333.5,
+      "learning_rate": 1.734527687296417e-05,
+      "loss": 0.8241,
+      "step": 940
+    },
+    {
+      "epoch": 2.0669961559582646,
+      "grad_norm": 1141092.25,
+      "learning_rate": 1.730456026058632e-05,
+      "loss": 0.9051,
+      "step": 941
+    },
+    {
+      "epoch": 2.069192751235585,
+      "grad_norm": 529467.1875,
+      "learning_rate": 1.726384364820847e-05,
+      "loss": 0.7137,
+      "step": 942
+    },
+    {
+      "epoch": 2.071389346512905,
+      "grad_norm": 1868026.25,
+      "learning_rate": 1.722312703583062e-05,
+      "loss": 0.9524,
+      "step": 943
+    },
+    {
+      "epoch": 2.073585941790225,
+      "grad_norm": 1123170.5,
+      "learning_rate": 1.7182410423452767e-05,
+      "loss": 0.8669,
+      "step": 944
+    },
+    {
+      "epoch": 2.0757825370675453,
+      "grad_norm": 1219235.375,
+      "learning_rate": 1.7141693811074918e-05,
+      "loss": 0.8786,
+      "step": 945
+    },
+    {
+      "epoch": 2.0779791323448653,
+      "grad_norm": 2645246.5,
+      "learning_rate": 1.7100977198697068e-05,
+      "loss": 0.9934,
+      "step": 946
+    },
+    {
+      "epoch": 2.0801757276221857,
+      "grad_norm": 2224892.5,
+      "learning_rate": 1.706026058631922e-05,
+      "loss": 0.7758,
+      "step": 947
+    },
+    {
+      "epoch": 2.0823723228995057,
+      "grad_norm": 1505277.75,
+      "learning_rate": 1.701954397394137e-05,
+      "loss": 0.8261,
+      "step": 948
+    },
+    {
+      "epoch": 2.084568918176826,
+      "grad_norm": 2943298.5,
+      "learning_rate": 1.697882736156352e-05,
+      "loss": 1.1062,
+      "step": 949
+    },
+    {
+      "epoch": 2.086765513454146,
+      "grad_norm": 23076126.0,
+      "learning_rate": 1.6938110749185667e-05,
+      "loss": 0.9507,
+      "step": 950
+    },
+    {
+      "epoch": 2.088962108731466,
+      "grad_norm": 1525882.25,
+      "learning_rate": 1.6897394136807817e-05,
+      "loss": 0.9471,
+      "step": 951
+    },
+    {
+      "epoch": 2.0911587040087865,
+      "grad_norm": 5940633.0,
+      "learning_rate": 1.6856677524429968e-05,
+      "loss": 1.024,
+      "step": 952
+    },
+    {
+      "epoch": 2.0933552992861064,
+      "grad_norm": 3001057.5,
+      "learning_rate": 1.6815960912052118e-05,
+      "loss": 1.06,
+      "step": 953
+    },
+    {
+      "epoch": 2.095551894563427,
+      "grad_norm": 2065762.125,
+      "learning_rate": 1.677524429967427e-05,
+      "loss": 0.8582,
+      "step": 954
+    },
+    {
+      "epoch": 2.097748489840747,
+      "grad_norm": 1565316.5,
+      "learning_rate": 1.6734527687296416e-05,
+      "loss": 1.0399,
+      "step": 955
+    },
+    {
+      "epoch": 2.099945085118067,
+      "grad_norm": 2441446.0,
+      "learning_rate": 1.6693811074918566e-05,
+      "loss": 0.9849,
+      "step": 956
+    },
+    {
+      "epoch": 2.102141680395387,
+      "grad_norm": 1200210.375,
+      "learning_rate": 1.6653094462540716e-05,
+      "loss": 0.9446,
+      "step": 957
+    },
+    {
+      "epoch": 2.104338275672707,
+      "grad_norm": 5641191.5,
+      "learning_rate": 1.6612377850162867e-05,
+      "loss": 0.8163,
+      "step": 958
+    },
+    {
+      "epoch": 2.1065348709500276,
+      "grad_norm": 1119351.875,
+      "learning_rate": 1.6571661237785017e-05,
+      "loss": 0.9129,
+      "step": 959
+    },
+    {
+      "epoch": 2.1087314662273475,
+      "grad_norm": 698778.25,
+      "learning_rate": 1.6530944625407164e-05,
+      "loss": 0.8837,
+      "step": 960
+    },
+    {
+      "epoch": 2.110928061504668,
+      "grad_norm": 745734.3125,
+      "learning_rate": 1.6490228013029315e-05,
+      "loss": 1.0383,
+      "step": 961
+    },
+    {
+      "epoch": 2.113124656781988,
+      "grad_norm": 1575156.625,
+      "learning_rate": 1.6449511400651465e-05,
+      "loss": 0.923,
+      "step": 962
+    },
+    {
+      "epoch": 2.115321252059308,
+      "grad_norm": 2339435.75,
+      "learning_rate": 1.6408794788273616e-05,
+      "loss": 0.8075,
+      "step": 963
+    },
+    {
+      "epoch": 2.1175178473366283,
+      "grad_norm": 1824651.75,
+      "learning_rate": 1.6368078175895766e-05,
+      "loss": 0.8996,
+      "step": 964
+    },
+    {
+      "epoch": 2.1197144426139483,
+      "grad_norm": 2031222.0,
+      "learning_rate": 1.6327361563517913e-05,
+      "loss": 0.8833,
+      "step": 965
+    },
+    {
+      "epoch": 2.1219110378912687,
+      "grad_norm": 1933261.5,
+      "learning_rate": 1.6286644951140064e-05,
+      "loss": 0.826,
+      "step": 966
+    },
+    {
+      "epoch": 2.1241076331685886,
+      "grad_norm": 1464613.75,
+      "learning_rate": 1.6245928338762214e-05,
+      "loss": 0.8712,
+      "step": 967
+    },
+    {
+      "epoch": 2.1263042284459086,
+      "grad_norm": 1804412.75,
+      "learning_rate": 1.6205211726384365e-05,
+      "loss": 0.8866,
+      "step": 968
+    },
+    {
+      "epoch": 2.128500823723229,
+      "grad_norm": 2947552.5,
+      "learning_rate": 1.6164495114006515e-05,
+      "loss": 0.8547,
+      "step": 969
+    },
+    {
+      "epoch": 2.130697419000549,
+      "grad_norm": 1355221.75,
+      "learning_rate": 1.6123778501628666e-05,
+      "loss": 1.096,
+      "step": 970
+    },
+    {
+      "epoch": 2.1328940142778694,
+      "grad_norm": 4624274.0,
+      "learning_rate": 1.6083061889250813e-05,
+      "loss": 0.8897,
+      "step": 971
+    },
+    {
+      "epoch": 2.1350906095551894,
+      "grad_norm": 4975509.5,
+      "learning_rate": 1.6042345276872963e-05,
+      "loss": 1.2131,
+      "step": 972
+    },
+    {
+      "epoch": 2.13728720483251,
+      "grad_norm": 1781793.75,
+      "learning_rate": 1.6001628664495114e-05,
+      "loss": 0.9106,
+      "step": 973
+    },
+    {
+      "epoch": 2.1394838001098297,
+      "grad_norm": 3046921.5,
+      "learning_rate": 1.5960912052117264e-05,
+      "loss": 1.0682,
+      "step": 974
+    },
+    {
+      "epoch": 2.1416803953871497,
+      "grad_norm": 5085994.5,
+      "learning_rate": 1.5920195439739415e-05,
+      "loss": 1.0058,
+      "step": 975
+    },
+    {
+      "epoch": 2.14387699066447,
+      "grad_norm": 4169267.0,
+      "learning_rate": 1.587947882736156e-05,
+      "loss": 0.9488,
+      "step": 976
+    },
+    {
+      "epoch": 2.14607358594179,
+      "grad_norm": 1749402.125,
+      "learning_rate": 1.5838762214983712e-05,
+      "loss": 0.8982,
+      "step": 977
+    },
+    {
+      "epoch": 2.1482701812191105,
+      "grad_norm": 2908449.75,
+      "learning_rate": 1.5798045602605863e-05,
+      "loss": 0.8871,
+      "step": 978
+    },
+    {
+      "epoch": 2.1504667764964305,
+      "grad_norm": 8285533.5,
+      "learning_rate": 1.5757328990228013e-05,
+      "loss": 0.9313,
+      "step": 979
+    },
+    {
+      "epoch": 2.152663371773751,
+      "grad_norm": 7875153.5,
+      "learning_rate": 1.5716612377850163e-05,
+      "loss": 1.0804,
+      "step": 980
+    },
+    {
+      "epoch": 2.154859967051071,
+      "grad_norm": 6283799.5,
+      "learning_rate": 1.567589576547231e-05,
+      "loss": 0.7992,
+      "step": 981
+    },
+    {
+      "epoch": 2.157056562328391,
+      "grad_norm": 1566144.0,
+      "learning_rate": 1.563517915309446e-05,
+      "loss": 0.8221,
+      "step": 982
+    },
+    {
+      "epoch": 2.1592531576057112,
+      "grad_norm": 9990809.0,
+      "learning_rate": 1.5594462540716615e-05,
+      "loss": 1.0971,
+      "step": 983
+    },
+    {
+      "epoch": 2.161449752883031,
+      "grad_norm": 3094588.75,
+      "learning_rate": 1.5553745928338765e-05,
+      "loss": 0.8437,
+      "step": 984
+    },
+    {
+      "epoch": 2.1636463481603516,
+      "grad_norm": 996254.125,
+      "learning_rate": 1.5513029315960916e-05,
+      "loss": 0.8385,
+      "step": 985
+    },
+    {
+      "epoch": 2.1658429434376716,
+      "grad_norm": 2428077.25,
+      "learning_rate": 1.5472312703583063e-05,
+      "loss": 0.8488,
+      "step": 986
+    },
+    {
+      "epoch": 2.168039538714992,
+      "grad_norm": 4716001.5,
+      "learning_rate": 1.5431596091205213e-05,
+      "loss": 1.0623,
+      "step": 987
+    },
+    {
+      "epoch": 2.170236133992312,
+      "grad_norm": 11939434.0,
+      "learning_rate": 1.5390879478827364e-05,
+      "loss": 0.818,
+      "step": 988
+    },
+    {
+      "epoch": 2.172432729269632,
+      "grad_norm": 2631827.25,
+      "learning_rate": 1.5350162866449514e-05,
+      "loss": 0.8907,
+      "step": 989
+    },
+    {
+      "epoch": 2.1746293245469523,
+      "grad_norm": 3452244.75,
+      "learning_rate": 1.5309446254071665e-05,
+      "loss": 1.1876,
+      "step": 990
+    },
+    {
+      "epoch": 2.1768259198242723,
+      "grad_norm": 4227468.5,
+      "learning_rate": 1.5268729641693812e-05,
+      "loss": 0.8587,
+      "step": 991
+    },
+    {
+      "epoch": 2.1790225151015927,
+      "grad_norm": 2860340.5,
+      "learning_rate": 1.5228013029315962e-05,
+      "loss": 0.8116,
+      "step": 992
+    },
+    {
+      "epoch": 2.1812191103789127,
+      "grad_norm": 4269275.0,
+      "learning_rate": 1.5187296416938113e-05,
+      "loss": 0.983,
+      "step": 993
+    },
+    {
+      "epoch": 2.1834157056562327,
+      "grad_norm": 2099209.5,
+      "learning_rate": 1.5146579804560263e-05,
+      "loss": 0.8202,
+      "step": 994
+    },
+    {
+      "epoch": 2.185612300933553,
+      "grad_norm": 2718143.75,
+      "learning_rate": 1.5105863192182412e-05,
+      "loss": 0.8771,
+      "step": 995
+    },
+    {
+      "epoch": 2.187808896210873,
+      "grad_norm": 4178870.5,
+      "learning_rate": 1.5065146579804562e-05,
+      "loss": 1.0824,
+      "step": 996
+    },
+    {
+      "epoch": 2.1900054914881935,
+      "grad_norm": 6255556.5,
+      "learning_rate": 1.5024429967426711e-05,
+      "loss": 0.9757,
+      "step": 997
+    },
+    {
+      "epoch": 2.1922020867655134,
+      "grad_norm": 2294232.25,
+      "learning_rate": 1.4983713355048862e-05,
+      "loss": 0.9189,
+      "step": 998
+    },
+    {
+      "epoch": 2.1943986820428334,
+      "grad_norm": 5404711.0,
+      "learning_rate": 1.4942996742671012e-05,
+      "loss": 1.0752,
+      "step": 999
+    },
+    {
+      "epoch": 2.196595277320154,
+      "grad_norm": 3957902.5,
+      "learning_rate": 1.490228013029316e-05,
+      "loss": 0.8613,
+      "step": 1000
+    },
+    {
+      "epoch": 2.1987918725974738,
+      "grad_norm": 2663807.25,
+      "learning_rate": 1.4861563517915311e-05,
+      "loss": 0.6564,
+      "step": 1001
+    },
+    {
+      "epoch": 2.200988467874794,
+      "grad_norm": 1447628.5,
+      "learning_rate": 1.4820846905537462e-05,
+      "loss": 0.8891,
+      "step": 1002
+    },
+    {
+      "epoch": 2.203185063152114,
+      "grad_norm": 3801231.25,
+      "learning_rate": 1.478013029315961e-05,
+      "loss": 0.9416,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2053816584294346,
+      "grad_norm": 2295110.5,
+      "learning_rate": 1.4739413680781761e-05,
+      "loss": 0.7596,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2075782537067545,
+      "grad_norm": 1150046.125,
+      "learning_rate": 1.469869706840391e-05,
+      "loss": 0.9302,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2097748489840745,
+      "grad_norm": 4189615.0,
+      "learning_rate": 1.465798045602606e-05,
+      "loss": 0.9536,
+      "step": 1006
+    },
+    {
+      "epoch": 2.211971444261395,
+      "grad_norm": 2719024.25,
+      "learning_rate": 1.461726384364821e-05,
+      "loss": 0.7757,
+      "step": 1007
+    },
+    {
+      "epoch": 2.214168039538715,
+      "grad_norm": 2606171.75,
+      "learning_rate": 1.457654723127036e-05,
+      "loss": 1.1732,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2163646348160353,
+      "grad_norm": 2899244.75,
+      "learning_rate": 1.453583061889251e-05,
+      "loss": 1.0119,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2185612300933553,
+      "grad_norm": 4351610.0,
+      "learning_rate": 1.449511400651466e-05,
+      "loss": 0.8487,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2207578253706757,
+      "grad_norm": 1387850.625,
+      "learning_rate": 1.4454397394136809e-05,
+      "loss": 0.9458,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2229544206479956,
+      "grad_norm": 3179972.5,
+      "learning_rate": 1.441368078175896e-05,
+      "loss": 1.1096,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2251510159253156,
+      "grad_norm": 6551366.5,
+      "learning_rate": 1.4372964169381108e-05,
+      "loss": 0.8545,
+      "step": 1013
+    },
+    {
+      "epoch": 2.227347611202636,
+      "grad_norm": 2095185.125,
+      "learning_rate": 1.4332247557003259e-05,
+      "loss": 1.1,
+      "step": 1014
+    },
+    {
+      "epoch": 2.229544206479956,
+      "grad_norm": 4655149.0,
+      "learning_rate": 1.429153094462541e-05,
+      "loss": 1.1856,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2317408017572764,
+      "grad_norm": 4307353.0,
+      "learning_rate": 1.4250814332247558e-05,
+      "loss": 0.938,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2339373970345964,
+      "grad_norm": 3256821.0,
+      "learning_rate": 1.4210097719869708e-05,
+      "loss": 1.0719,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2361339923119163,
+      "grad_norm": 5372780.0,
+      "learning_rate": 1.4169381107491859e-05,
+      "loss": 1.0743,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2383305875892368,
+      "grad_norm": 4162222.75,
+      "learning_rate": 1.4128664495114008e-05,
+      "loss": 1.0769,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2405271828665567,
+      "grad_norm": 1537119.375,
+      "learning_rate": 1.4087947882736158e-05,
+      "loss": 0.9292,
+      "step": 1020
+    },
+    {
+      "epoch": 2.242723778143877,
+      "grad_norm": 1904608.875,
+      "learning_rate": 1.4047231270358307e-05,
+      "loss": 0.9351,
+      "step": 1021
+    },
+    {
+      "epoch": 2.244920373421197,
+      "grad_norm": 3816468.5,
+      "learning_rate": 1.4006514657980457e-05,
+      "loss": 0.6892,
+      "step": 1022
+    },
+    {
+      "epoch": 2.247116968698517,
+      "grad_norm": 1615283.5,
+      "learning_rate": 1.3965798045602608e-05,
+      "loss": 1.0105,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2493135639758375,
+      "grad_norm": 1716130.0,
+      "learning_rate": 1.3925081433224757e-05,
+      "loss": 0.931,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2515101592531575,
+      "grad_norm": 1263084.375,
+      "learning_rate": 1.3884364820846907e-05,
+      "loss": 0.8251,
+      "step": 1025
+    },
+    {
+      "epoch": 2.253706754530478,
+      "grad_norm": 1375688.25,
+      "learning_rate": 1.3843648208469056e-05,
+      "loss": 0.8337,
+      "step": 1026
+    },
+    {
+      "epoch": 2.255903349807798,
+      "grad_norm": 3485078.75,
+      "learning_rate": 1.3802931596091206e-05,
+      "loss": 0.8879,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2580999450851182,
+      "grad_norm": 5470211.5,
+      "learning_rate": 1.3762214983713357e-05,
+      "loss": 1.0105,
+      "step": 1028
+    },
+    {
+      "epoch": 2.260296540362438,
+      "grad_norm": 1462846.125,
+      "learning_rate": 1.3721498371335505e-05,
+      "loss": 1.0226,
+      "step": 1029
+    },
+    {
+      "epoch": 2.262493135639758,
+      "grad_norm": 1207471.0,
+      "learning_rate": 1.3680781758957656e-05,
+      "loss": 0.7893,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2646897309170786,
+      "grad_norm": 1491436.125,
+      "learning_rate": 1.3640065146579806e-05,
+      "loss": 0.7907,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2668863261943986,
+      "grad_norm": 2752501.25,
+      "learning_rate": 1.3599348534201955e-05,
+      "loss": 0.7851,
+      "step": 1032
+    },
+    {
+      "epoch": 2.269082921471719,
+      "grad_norm": 1462520.25,
+      "learning_rate": 1.3558631921824106e-05,
+      "loss": 1.1233,
+      "step": 1033
+    },
+    {
+      "epoch": 2.271279516749039,
+      "grad_norm": 882404.25,
+      "learning_rate": 1.3517915309446254e-05,
+      "loss": 0.8303,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2734761120263594,
+      "grad_norm": 2446822.5,
+      "learning_rate": 1.3477198697068405e-05,
+      "loss": 0.8208,
+      "step": 1035
+    },
+    {
+      "epoch": 2.2756727073036793,
+      "grad_norm": 1177807.25,
+      "learning_rate": 1.3436482084690555e-05,
+      "loss": 0.942,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2778693025809993,
+      "grad_norm": 837452.5625,
+      "learning_rate": 1.3395765472312704e-05,
+      "loss": 0.8626,
+      "step": 1037
+    },
+    {
+      "epoch": 2.2800658978583197,
+      "grad_norm": 4167568.0,
+      "learning_rate": 1.3355048859934855e-05,
+      "loss": 0.9156,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2822624931356397,
+      "grad_norm": 3274604.5,
+      "learning_rate": 1.3314332247557005e-05,
+      "loss": 0.8284,
+      "step": 1039
+    },
+    {
+      "epoch": 2.28445908841296,
+      "grad_norm": 6151152.0,
+      "learning_rate": 1.3273615635179154e-05,
+      "loss": 0.9516,
+      "step": 1040
+    },
+    {
+      "epoch": 2.28665568369028,
+      "grad_norm": 2699991.25,
+      "learning_rate": 1.3232899022801304e-05,
+      "loss": 0.8995,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2888522789676005,
+      "grad_norm": 1476659.5,
+      "learning_rate": 1.3192182410423453e-05,
+      "loss": 0.9802,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2910488742449204,
+      "grad_norm": 1324048.25,
+      "learning_rate": 1.3151465798045603e-05,
+      "loss": 0.7337,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2932454695222404,
+      "grad_norm": 2075367.0,
+      "learning_rate": 1.3110749185667754e-05,
+      "loss": 0.9263,
+      "step": 1044
+    },
+    {
+      "epoch": 2.295442064799561,
+      "grad_norm": 1149153.0,
+      "learning_rate": 1.3070032573289903e-05,
+      "loss": 0.9948,
+      "step": 1045
+    },
+    {
+      "epoch": 2.297638660076881,
+      "grad_norm": 1859404.0,
+      "learning_rate": 1.3029315960912053e-05,
+      "loss": 0.8585,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2998352553542007,
+      "grad_norm": 1141798.0,
+      "learning_rate": 1.2988599348534204e-05,
+      "loss": 0.7657,
+      "step": 1047
+    },
+    {
+      "epoch": 2.302031850631521,
+      "grad_norm": 4972899.0,
+      "learning_rate": 1.2947882736156352e-05,
+      "loss": 0.8913,
+      "step": 1048
+    },
+    {
+      "epoch": 2.304228445908841,
+      "grad_norm": 8308181.5,
+      "learning_rate": 1.2907166123778503e-05,
+      "loss": 0.9249,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3064250411861615,
+      "grad_norm": 1159885.75,
+      "learning_rate": 1.2866449511400652e-05,
+      "loss": 0.9553,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3086216364634815,
+      "grad_norm": 1752849.25,
+      "learning_rate": 1.2825732899022802e-05,
+      "loss": 0.678,
+      "step": 1051
+    },
+    {
+      "epoch": 2.310818231740802,
+      "grad_norm": 2692237.0,
+      "learning_rate": 1.2785016286644952e-05,
+      "loss": 0.861,
+      "step": 1052
+    },
+    {
+      "epoch": 2.313014827018122,
+      "grad_norm": 1686573.75,
+      "learning_rate": 1.2744299674267101e-05,
+      "loss": 0.9615,
+      "step": 1053
+    },
+    {
+      "epoch": 2.315211422295442,
+      "grad_norm": 2784386.0,
+      "learning_rate": 1.2703583061889252e-05,
+      "loss": 0.8789,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3174080175727623,
+      "grad_norm": 2572999.75,
+      "learning_rate": 1.2662866449511402e-05,
+      "loss": 1.0618,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3196046128500822,
+      "grad_norm": 1737379.75,
+      "learning_rate": 1.2622149837133551e-05,
+      "loss": 0.8946,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3218012081274026,
+      "grad_norm": 3722364.25,
+      "learning_rate": 1.2581433224755701e-05,
+      "loss": 1.1145,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3239978034047226,
+      "grad_norm": 2210988.5,
+      "learning_rate": 1.254071661237785e-05,
+      "loss": 0.6879,
+      "step": 1058
+    },
+    {
+      "epoch": 2.326194398682043,
+      "grad_norm": 3495376.5,
+      "learning_rate": 1.25e-05,
+      "loss": 0.8324,
+      "step": 1059
+    },
+    {
+      "epoch": 2.328390993959363,
+      "grad_norm": 3875286.75,
+      "learning_rate": 1.2459283387622151e-05,
+      "loss": 0.8942,
+      "step": 1060
+    },
+    {
+      "epoch": 2.330587589236683,
+      "grad_norm": 669766.5,
+      "learning_rate": 1.24185667752443e-05,
+      "loss": 0.9229,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3327841845140034,
+      "grad_norm": 1541036.125,
+      "learning_rate": 1.237785016286645e-05,
+      "loss": 0.9139,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3349807797913233,
+      "grad_norm": 2162414.25,
+      "learning_rate": 1.2337133550488599e-05,
+      "loss": 1.0492,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3371773750686438,
+      "grad_norm": 2440397.5,
+      "learning_rate": 1.229641693811075e-05,
+      "loss": 1.0252,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3393739703459637,
+      "grad_norm": 1311130.125,
+      "learning_rate": 1.22557003257329e-05,
+      "loss": 0.898,
+      "step": 1065
+    },
+    {
+      "epoch": 2.341570565623284,
+      "grad_norm": 3886045.5,
+      "learning_rate": 1.2214983713355049e-05,
+      "loss": 1.0149,
+      "step": 1066
+    },
+    {
+      "epoch": 2.343767160900604,
+      "grad_norm": 3510321.5,
+      "learning_rate": 1.21742671009772e-05,
+      "loss": 0.7841,
+      "step": 1067
+    },
+    {
+      "epoch": 2.345963756177924,
+      "grad_norm": 9004319.0,
+      "learning_rate": 1.213355048859935e-05,
+      "loss": 0.8999,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3481603514552445,
+      "grad_norm": 3036810.5,
+      "learning_rate": 1.2092833876221498e-05,
+      "loss": 1.0889,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3503569467325645,
+      "grad_norm": 5504600.5,
+      "learning_rate": 1.2052117263843649e-05,
+      "loss": 0.8415,
+      "step": 1070
+    },
+    {
+      "epoch": 2.352553542009885,
+      "grad_norm": 2847811.75,
+      "learning_rate": 1.2011400651465798e-05,
+      "loss": 0.912,
+      "step": 1071
+    },
+    {
+      "epoch": 2.354750137287205,
+      "grad_norm": 1708956.5,
+      "learning_rate": 1.1970684039087948e-05,
+      "loss": 0.7697,
+      "step": 1072
+    },
+    {
+      "epoch": 2.356946732564525,
+      "grad_norm": 2047011.125,
+      "learning_rate": 1.1929967426710099e-05,
+      "loss": 0.7525,
+      "step": 1073
+    },
+    {
+      "epoch": 2.359143327841845,
+      "grad_norm": 1327106.875,
+      "learning_rate": 1.1889250814332247e-05,
+      "loss": 0.8973,
+      "step": 1074
+    },
+    {
+      "epoch": 2.361339923119165,
+      "grad_norm": 2383668.0,
+      "learning_rate": 1.1848534201954398e-05,
+      "loss": 1.0114,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3635365183964856,
+      "grad_norm": 1537416.0,
+      "learning_rate": 1.1807817589576548e-05,
+      "loss": 0.8935,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3657331136738056,
+      "grad_norm": 1491642.5,
+      "learning_rate": 1.1767100977198697e-05,
+      "loss": 0.8742,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3679297089511255,
+      "grad_norm": 2734224.0,
+      "learning_rate": 1.1726384364820847e-05,
+      "loss": 0.9746,
+      "step": 1078
+    },
+    {
+      "epoch": 2.370126304228446,
+      "grad_norm": 1081284.5,
+      "learning_rate": 1.1685667752442996e-05,
+      "loss": 0.7082,
+      "step": 1079
+    },
+    {
+      "epoch": 2.372322899505766,
+      "grad_norm": 1304858.5,
+      "learning_rate": 1.1644951140065147e-05,
+      "loss": 0.8868,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3745194947830863,
+      "grad_norm": 10182949.0,
+      "learning_rate": 1.1604234527687297e-05,
+      "loss": 1.2137,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3767160900604063,
+      "grad_norm": 1948719.75,
+      "learning_rate": 1.1563517915309446e-05,
+      "loss": 0.8819,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3789126853377267,
+      "grad_norm": 4293130.0,
+      "learning_rate": 1.1522801302931596e-05,
+      "loss": 0.9482,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3811092806150467,
+      "grad_norm": 1263682.75,
+      "learning_rate": 1.1482084690553747e-05,
+      "loss": 1.0845,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3833058758923666,
+      "grad_norm": 745638.1875,
+      "learning_rate": 1.1441368078175896e-05,
+      "loss": 1.2173,
+      "step": 1085
+    },
+    {
+      "epoch": 2.385502471169687,
+      "grad_norm": 1798094.75,
+      "learning_rate": 1.1400651465798046e-05,
+      "loss": 0.9158,
+      "step": 1086
+    },
+    {
+      "epoch": 2.387699066447007,
+      "grad_norm": 4237468.0,
+      "learning_rate": 1.1359934853420195e-05,
+      "loss": 1.1924,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3898956617243274,
+      "grad_norm": 2902751.25,
+      "learning_rate": 1.1319218241042345e-05,
+      "loss": 0.8994,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3920922570016474,
+      "grad_norm": 1711488.125,
+      "learning_rate": 1.1278501628664496e-05,
+      "loss": 1.0191,
+      "step": 1089
+    },
+    {
+      "epoch": 2.394288852278968,
+      "grad_norm": 1366606.375,
+      "learning_rate": 1.1237785016286644e-05,
+      "loss": 0.8497,
+      "step": 1090
+    },
+    {
+      "epoch": 2.396485447556288,
+      "grad_norm": 993613.6875,
+      "learning_rate": 1.1197068403908795e-05,
+      "loss": 0.8729,
+      "step": 1091
+    },
+    {
+      "epoch": 2.3986820428336078,
+      "grad_norm": 2406019.0,
+      "learning_rate": 1.1156351791530945e-05,
+      "loss": 0.9458,
+      "step": 1092
+    },
+    {
+      "epoch": 2.400878638110928,
+      "grad_norm": 2713716.5,
+      "learning_rate": 1.1115635179153094e-05,
+      "loss": 0.9629,
+      "step": 1093
+    },
+    {
+      "epoch": 2.403075233388248,
+      "grad_norm": 2714554.0,
+      "learning_rate": 1.1074918566775245e-05,
+      "loss": 0.9843,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4052718286655685,
+      "grad_norm": 1991187.5,
+      "learning_rate": 1.1034201954397393e-05,
+      "loss": 0.8032,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4074684239428885,
+      "grad_norm": 1325641.875,
+      "learning_rate": 1.0993485342019544e-05,
+      "loss": 0.7473,
+      "step": 1096
+    },
+    {
+      "epoch": 2.409665019220209,
+      "grad_norm": 5467199.0,
+      "learning_rate": 1.0952768729641694e-05,
+      "loss": 1.0309,
+      "step": 1097
+    },
+    {
+      "epoch": 2.411861614497529,
+      "grad_norm": 1146516.5,
+      "learning_rate": 1.0912052117263845e-05,
+      "loss": 0.9754,
+      "step": 1098
+    },
+    {
+      "epoch": 2.414058209774849,
+      "grad_norm": 4022157.25,
+      "learning_rate": 1.0871335504885995e-05,
+      "loss": 0.9924,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4162548050521693,
+      "grad_norm": 6553877.0,
+      "learning_rate": 1.0830618892508144e-05,
+      "loss": 0.964,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4184514003294892,
+      "grad_norm": 5096666.0,
+      "learning_rate": 1.0789902280130294e-05,
+      "loss": 0.9239,
+      "step": 1101
+    },
+    {
+      "epoch": 2.420647995606809,
+      "grad_norm": 1214168.0,
+      "learning_rate": 1.0749185667752445e-05,
+      "loss": 1.0278,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4228445908841296,
+      "grad_norm": 1144715.625,
+      "learning_rate": 1.0708469055374594e-05,
+      "loss": 0.863,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4250411861614496,
+      "grad_norm": 1360645.75,
+      "learning_rate": 1.0667752442996744e-05,
+      "loss": 0.9893,
+      "step": 1104
+    },
+    {
+      "epoch": 2.42723778143877,
+      "grad_norm": 3539896.5,
+      "learning_rate": 1.0627035830618893e-05,
+      "loss": 1.004,
+      "step": 1105
+    },
+    {
+      "epoch": 2.42943437671609,
+      "grad_norm": 1051803.5,
+      "learning_rate": 1.0586319218241043e-05,
+      "loss": 0.8195,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4316309719934104,
+      "grad_norm": 6213618.0,
+      "learning_rate": 1.0545602605863194e-05,
+      "loss": 1.0477,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4338275672707304,
+      "grad_norm": 1593600.0,
+      "learning_rate": 1.0504885993485343e-05,
+      "loss": 1.0626,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4360241625480503,
+      "grad_norm": 4770628.5,
+      "learning_rate": 1.0464169381107493e-05,
+      "loss": 0.9243,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4382207578253707,
+      "grad_norm": 1934187.875,
+      "learning_rate": 1.0423452768729644e-05,
+      "loss": 0.9136,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4404173531026907,
+      "grad_norm": 2536422.75,
+      "learning_rate": 1.0382736156351792e-05,
+      "loss": 0.9452,
+      "step": 1111
+    },
+    {
+      "epoch": 2.442613948380011,
+      "grad_norm": 863557.375,
+      "learning_rate": 1.0342019543973943e-05,
+      "loss": 0.708,
+      "step": 1112
+    },
+    {
+      "epoch": 2.444810543657331,
+      "grad_norm": 2386128.0,
+      "learning_rate": 1.0301302931596091e-05,
+      "loss": 1.2572,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4470071389346515,
+      "grad_norm": 1260755.875,
+      "learning_rate": 1.0260586319218242e-05,
+      "loss": 1.0616,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4492037342119715,
+      "grad_norm": 1263228.25,
+      "learning_rate": 1.0219869706840392e-05,
+      "loss": 0.8843,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4514003294892914,
+      "grad_norm": 2332672.25,
+      "learning_rate": 1.0179153094462541e-05,
+      "loss": 0.7549,
+      "step": 1116
+    },
+    {
+      "epoch": 2.453596924766612,
+      "grad_norm": 2186742.0,
+      "learning_rate": 1.0138436482084692e-05,
+      "loss": 0.9752,
+      "step": 1117
+    },
+    {
+      "epoch": 2.455793520043932,
+      "grad_norm": 1302256.875,
+      "learning_rate": 1.009771986970684e-05,
+      "loss": 0.9125,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4579901153212522,
+      "grad_norm": 2253556.0,
+      "learning_rate": 1.0057003257328991e-05,
+      "loss": 0.9765,
+      "step": 1119
+    },
+    {
+      "epoch": 2.460186710598572,
+      "grad_norm": 2130001.75,
+      "learning_rate": 1.0016286644951141e-05,
+      "loss": 0.887,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4623833058758926,
+      "grad_norm": 3017812.0,
+      "learning_rate": 9.97557003257329e-06,
+      "loss": 0.9425,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4645799011532126,
+      "grad_norm": 2319200.25,
+      "learning_rate": 9.93485342019544e-06,
+      "loss": 1.3186,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4667764964305325,
+      "grad_norm": 1344412.5,
+      "learning_rate": 9.894136807817591e-06,
+      "loss": 0.9803,
+      "step": 1123
+    },
+    {
+      "epoch": 2.468973091707853,
+      "grad_norm": 2171796.0,
+      "learning_rate": 9.85342019543974e-06,
+      "loss": 0.9673,
+      "step": 1124
+    },
+    {
+      "epoch": 2.471169686985173,
+      "grad_norm": 2066248.5,
+      "learning_rate": 9.81270358306189e-06,
+      "loss": 0.8066,
+      "step": 1125
+    },
+    {
+      "epoch": 2.473366282262493,
+      "grad_norm": 942228.625,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 1.0971,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4755628775398133,
+      "grad_norm": 1388220.25,
+      "learning_rate": 9.73127035830619e-06,
+      "loss": 0.8418,
+      "step": 1127
+    },
+    {
+      "epoch": 2.4777594728171333,
+      "grad_norm": 1006753.125,
+      "learning_rate": 9.69055374592834e-06,
+      "loss": 0.9031,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4799560680944537,
+      "grad_norm": 2669757.5,
+      "learning_rate": 9.649837133550489e-06,
+      "loss": 1.152,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4821526633717736,
+      "grad_norm": 2930147.75,
+      "learning_rate": 9.609120521172639e-06,
+      "loss": 0.9873,
+      "step": 1130
+    },
+    {
+      "epoch": 2.484349258649094,
+      "grad_norm": 1066948.625,
+      "learning_rate": 9.56840390879479e-06,
+      "loss": 0.8597,
+      "step": 1131
+    },
+    {
+      "epoch": 2.486545853926414,
+      "grad_norm": 2585693.75,
+      "learning_rate": 9.527687296416938e-06,
+      "loss": 0.9589,
+      "step": 1132
+    },
+    {
+      "epoch": 2.488742449203734,
+      "grad_norm": 2357879.0,
+      "learning_rate": 9.486970684039089e-06,
+      "loss": 0.7995,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4909390444810544,
+      "grad_norm": 3151554.75,
+      "learning_rate": 9.446254071661238e-06,
+      "loss": 0.9351,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4931356397583744,
+      "grad_norm": 2803952.25,
+      "learning_rate": 9.405537459283388e-06,
+      "loss": 0.9485,
+      "step": 1135
+    },
+    {
+      "epoch": 2.495332235035695,
+      "grad_norm": 1258807.75,
+      "learning_rate": 9.364820846905538e-06,
+      "loss": 0.8499,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4975288303130148,
+      "grad_norm": 663777.3125,
+      "learning_rate": 9.324104234527687e-06,
+      "loss": 0.6709,
+      "step": 1137
+    },
+    {
+      "epoch": 2.499725425590335,
+      "grad_norm": 1318170.5,
+      "learning_rate": 9.283387622149838e-06,
+      "loss": 0.8936,
+      "step": 1138
+    },
+    {
+      "epoch": 2.501922020867655,
+      "grad_norm": 3373023.0,
+      "learning_rate": 9.242671009771988e-06,
+      "loss": 1.0656,
+      "step": 1139
+    },
+    {
+      "epoch": 2.504118616144975,
+      "grad_norm": 1677931.625,
+      "learning_rate": 9.201954397394137e-06,
+      "loss": 0.9142,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5063152114222955,
+      "grad_norm": 1333487.375,
+      "learning_rate": 9.161237785016287e-06,
+      "loss": 0.7976,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5085118066996155,
+      "grad_norm": 1737058.875,
+      "learning_rate": 9.120521172638436e-06,
+      "loss": 0.9696,
+      "step": 1142
+    },
+    {
+      "epoch": 2.510708401976936,
+      "grad_norm": 1557900.75,
+      "learning_rate": 9.079804560260587e-06,
+      "loss": 1.0436,
+      "step": 1143
+    },
+    {
+      "epoch": 2.512904997254256,
+      "grad_norm": 903309.4375,
+      "learning_rate": 9.039087947882737e-06,
+      "loss": 0.8953,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5151015925315763,
+      "grad_norm": 4292743.0,
+      "learning_rate": 8.998371335504886e-06,
+      "loss": 0.8931,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5172981878088962,
+      "grad_norm": 1909475.875,
+      "learning_rate": 8.957654723127036e-06,
+      "loss": 0.8903,
+      "step": 1146
+    },
+    {
+      "epoch": 2.519494783086216,
+      "grad_norm": 1347297.75,
+      "learning_rate": 8.916938110749185e-06,
+      "loss": 0.8689,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5216913783635366,
+      "grad_norm": 759535.5,
+      "learning_rate": 8.876221498371336e-06,
+      "loss": 0.7726,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5238879736408566,
+      "grad_norm": 3135357.75,
+      "learning_rate": 8.835504885993486e-06,
+      "loss": 0.9785,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5260845689181766,
+      "grad_norm": 7337167.5,
+      "learning_rate": 8.794788273615635e-06,
+      "loss": 1.1702,
+      "step": 1150
+    },
+    {
+      "epoch": 2.528281164195497,
+      "grad_norm": 1061835.125,
+      "learning_rate": 8.754071661237785e-06,
+      "loss": 0.978,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5304777594728174,
+      "grad_norm": 804567.125,
+      "learning_rate": 8.713355048859936e-06,
+      "loss": 0.9677,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5326743547501374,
+      "grad_norm": 695741.1875,
+      "learning_rate": 8.672638436482084e-06,
+      "loss": 0.9214,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5348709500274573,
+      "grad_norm": 1137866.5,
+      "learning_rate": 8.631921824104235e-06,
+      "loss": 0.8598,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5370675453047777,
+      "grad_norm": 1789339.25,
+      "learning_rate": 8.591205211726384e-06,
+      "loss": 0.8644,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5392641405820977,
+      "grad_norm": 546273.625,
+      "learning_rate": 8.550488599348534e-06,
+      "loss": 1.0591,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5414607358594177,
+      "grad_norm": 1518931.625,
+      "learning_rate": 8.509771986970685e-06,
+      "loss": 1.0593,
+      "step": 1157
+    },
+    {
+      "epoch": 2.543657331136738,
+      "grad_norm": 607515.0,
+      "learning_rate": 8.469055374592833e-06,
+      "loss": 0.8342,
+      "step": 1158
+    },
+    {
+      "epoch": 2.545853926414058,
+      "grad_norm": 1884551.25,
+      "learning_rate": 8.428338762214984e-06,
+      "loss": 0.8553,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5480505216913785,
+      "grad_norm": 893297.5625,
+      "learning_rate": 8.387622149837134e-06,
+      "loss": 0.9033,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5502471169686984,
+      "grad_norm": 1372398.125,
+      "learning_rate": 8.346905537459283e-06,
+      "loss": 0.7626,
+      "step": 1161
+    },
+    {
+      "epoch": 2.552443712246019,
+      "grad_norm": 1997930.875,
+      "learning_rate": 8.306188925081433e-06,
+      "loss": 0.9459,
+      "step": 1162
+    },
+    {
+      "epoch": 2.554640307523339,
+      "grad_norm": 899145.8125,
+      "learning_rate": 8.265472312703582e-06,
+      "loss": 0.7027,
+      "step": 1163
+    },
+    {
+      "epoch": 2.556836902800659,
+      "grad_norm": 5084283.0,
+      "learning_rate": 8.224755700325733e-06,
+      "loss": 1.0224,
+      "step": 1164
+    },
+    {
+      "epoch": 2.559033498077979,
+      "grad_norm": 2397002.25,
+      "learning_rate": 8.184039087947883e-06,
+      "loss": 1.0307,
+      "step": 1165
+    },
+    {
+      "epoch": 2.561230093355299,
+      "grad_norm": 638473.3125,
+      "learning_rate": 8.143322475570032e-06,
+      "loss": 0.8603,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5634266886326196,
+      "grad_norm": 978788.25,
+      "learning_rate": 8.102605863192182e-06,
+      "loss": 0.7585,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5656232839099395,
+      "grad_norm": 1170018.75,
+      "learning_rate": 8.061889250814333e-06,
+      "loss": 0.9557,
+      "step": 1168
+    },
+    {
+      "epoch": 2.56781987918726,
+      "grad_norm": 993072.1875,
+      "learning_rate": 8.021172638436482e-06,
+      "loss": 0.9192,
+      "step": 1169
+    },
+    {
+      "epoch": 2.57001647446458,
+      "grad_norm": 1483430.25,
+      "learning_rate": 7.980456026058632e-06,
+      "loss": 0.8372,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5722130697419,
+      "grad_norm": 1116923.0,
+      "learning_rate": 7.93973941368078e-06,
+      "loss": 0.8352,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5744096650192203,
+      "grad_norm": 3315607.75,
+      "learning_rate": 7.899022801302931e-06,
+      "loss": 0.8626,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5766062602965403,
+      "grad_norm": 915245.875,
+      "learning_rate": 7.858306188925082e-06,
+      "loss": 0.8104,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5788028555738602,
+      "grad_norm": 1556244.25,
+      "learning_rate": 7.81758957654723e-06,
+      "loss": 0.938,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5809994508511807,
+      "grad_norm": 1770848.0,
+      "learning_rate": 7.776872964169383e-06,
+      "loss": 0.9341,
+      "step": 1175
+    },
+    {
+      "epoch": 2.583196046128501,
+      "grad_norm": 1108543.0,
+      "learning_rate": 7.736156351791531e-06,
+      "loss": 0.8424,
+      "step": 1176
+    },
+    {
+      "epoch": 2.585392641405821,
+      "grad_norm": 2407417.75,
+      "learning_rate": 7.695439739413682e-06,
+      "loss": 1.1034,
+      "step": 1177
+    },
+    {
+      "epoch": 2.587589236683141,
+      "grad_norm": 1370503.5,
+      "learning_rate": 7.654723127035832e-06,
+      "loss": 0.8154,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5897858319604614,
+      "grad_norm": 4707321.5,
+      "learning_rate": 7.614006514657981e-06,
+      "loss": 1.0411,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5919824272377814,
+      "grad_norm": 594677.375,
+      "learning_rate": 7.5732899022801316e-06,
+      "loss": 0.784,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5941790225151014,
+      "grad_norm": 2094648.0,
+      "learning_rate": 7.532573289902281e-06,
+      "loss": 1.1086,
+      "step": 1181
+    },
+    {
+      "epoch": 2.5963756177924218,
+      "grad_norm": 720654.5,
+      "learning_rate": 7.491856677524431e-06,
+      "loss": 0.9586,
+      "step": 1182
+    },
+    {
+      "epoch": 2.598572213069742,
+      "grad_norm": 1930653.0,
+      "learning_rate": 7.45114006514658e-06,
+      "loss": 0.9136,
+      "step": 1183
+    },
+    {
+      "epoch": 2.600768808347062,
+      "grad_norm": 2009380.375,
+      "learning_rate": 7.410423452768731e-06,
+      "loss": 0.9653,
+      "step": 1184
+    },
+    {
+      "epoch": 2.602965403624382,
+      "grad_norm": 1149768.25,
+      "learning_rate": 7.3697068403908805e-06,
+      "loss": 1.0457,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6051619989017025,
+      "grad_norm": 588234.0625,
+      "learning_rate": 7.32899022801303e-06,
+      "loss": 0.9411,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6073585941790225,
+      "grad_norm": 1371944.375,
+      "learning_rate": 7.28827361563518e-06,
+      "loss": 0.7993,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6095551894563425,
+      "grad_norm": 1541022.5,
+      "learning_rate": 7.24755700325733e-06,
+      "loss": 0.8856,
+      "step": 1188
+    },
+    {
+      "epoch": 2.611751784733663,
+      "grad_norm": 1827986.5,
+      "learning_rate": 7.20684039087948e-06,
+      "loss": 0.7987,
+      "step": 1189
+    },
+    {
+      "epoch": 2.613948380010983,
+      "grad_norm": 2910553.5,
+      "learning_rate": 7.166123778501629e-06,
+      "loss": 0.9177,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6161449752883033,
+      "grad_norm": 2954863.0,
+      "learning_rate": 7.125407166123779e-06,
+      "loss": 0.8482,
+      "step": 1191
+    },
+    {
+      "epoch": 2.618341570565623,
+      "grad_norm": 1919593.875,
+      "learning_rate": 7.0846905537459294e-06,
+      "loss": 0.7339,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6205381658429436,
+      "grad_norm": 2761935.5,
+      "learning_rate": 7.043973941368079e-06,
+      "loss": 1.5727,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6227347611202636,
+      "grad_norm": 2628441.75,
+      "learning_rate": 7.003257328990229e-06,
+      "loss": 0.9985,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6249313563975836,
+      "grad_norm": 2100099.25,
+      "learning_rate": 6.962540716612378e-06,
+      "loss": 0.596,
+      "step": 1195
+    },
+    {
+      "epoch": 2.627127951674904,
+      "grad_norm": 1085557.875,
+      "learning_rate": 6.921824104234528e-06,
+      "loss": 0.8919,
+      "step": 1196
+    },
+    {
+      "epoch": 2.629324546952224,
+      "grad_norm": 1135327.625,
+      "learning_rate": 6.881107491856678e-06,
+      "loss": 0.8301,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6315211422295444,
+      "grad_norm": 1810503.25,
+      "learning_rate": 6.840390879478828e-06,
+      "loss": 0.9492,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6337177375068643,
+      "grad_norm": 2112446.5,
+      "learning_rate": 6.799674267100978e-06,
+      "loss": 0.7859,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6359143327841847,
+      "grad_norm": 1591997.75,
+      "learning_rate": 6.758957654723127e-06,
+      "loss": 0.8856,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6381109280615047,
+      "grad_norm": 1670775.625,
+      "learning_rate": 6.718241042345278e-06,
+      "loss": 0.7476,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6403075233388247,
+      "grad_norm": 1945860.125,
+      "learning_rate": 6.677524429967427e-06,
+      "loss": 0.7103,
+      "step": 1202
+    },
+    {
+      "epoch": 2.642504118616145,
+      "grad_norm": 1192853.0,
+      "learning_rate": 6.636807817589577e-06,
+      "loss": 1.1331,
+      "step": 1203
+    },
+    {
+      "epoch": 2.644700713893465,
+      "grad_norm": 3113124.0,
+      "learning_rate": 6.5960912052117265e-06,
+      "loss": 0.8415,
+      "step": 1204
+    },
+    {
+      "epoch": 2.646897309170785,
+      "grad_norm": 1809240.0,
+      "learning_rate": 6.555374592833877e-06,
+      "loss": 1.0128,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6490939044481054,
+      "grad_norm": 2133234.75,
+      "learning_rate": 6.5146579804560266e-06,
+      "loss": 1.0666,
+      "step": 1206
+    },
+    {
+      "epoch": 2.651290499725426,
+      "grad_norm": 1408382.75,
+      "learning_rate": 6.473941368078176e-06,
+      "loss": 0.9999,
+      "step": 1207
+    },
+    {
+      "epoch": 2.653487095002746,
+      "grad_norm": 1472510.25,
+      "learning_rate": 6.433224755700326e-06,
+      "loss": 0.9265,
+      "step": 1208
+    },
+    {
+      "epoch": 2.655683690280066,
+      "grad_norm": 1653248.125,
+      "learning_rate": 6.392508143322476e-06,
+      "loss": 0.763,
+      "step": 1209
+    },
+    {
+      "epoch": 2.657880285557386,
+      "grad_norm": 2138062.5,
+      "learning_rate": 6.351791530944626e-06,
+      "loss": 1.0772,
+      "step": 1210
+    },
+    {
+      "epoch": 2.660076880834706,
+      "grad_norm": 2783789.0,
+      "learning_rate": 6.3110749185667755e-06,
+      "loss": 0.9137,
+      "step": 1211
+    },
+    {
+      "epoch": 2.662273476112026,
+      "grad_norm": 1199928.75,
+      "learning_rate": 6.270358306188925e-06,
+      "loss": 0.7761,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6644700713893466,
+      "grad_norm": 1284395.5,
+      "learning_rate": 6.2296416938110755e-06,
+      "loss": 0.9672,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 1678511.125,
+      "learning_rate": 6.188925081433225e-06,
+      "loss": 0.7997,
+      "step": 1214
+    },
+    {
+      "epoch": 2.668863261943987,
+      "grad_norm": 1046202.9375,
+      "learning_rate": 6.148208469055375e-06,
+      "loss": 0.8468,
+      "step": 1215
+    },
+    {
+      "epoch": 2.671059857221307,
+      "grad_norm": 1519988.625,
+      "learning_rate": 6.107491856677524e-06,
+      "loss": 0.968,
+      "step": 1216
+    },
+    {
+      "epoch": 2.6732564524986273,
+      "grad_norm": 909830.3125,
+      "learning_rate": 6.066775244299675e-06,
+      "loss": 0.9578,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6754530477759473,
+      "grad_norm": 3272629.75,
+      "learning_rate": 6.0260586319218244e-06,
+      "loss": 0.8327,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6776496430532672,
+      "grad_norm": 3903972.25,
+      "learning_rate": 5.985342019543974e-06,
+      "loss": 0.8308,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6798462383305877,
+      "grad_norm": 40256360.0,
+      "learning_rate": 5.944625407166124e-06,
+      "loss": 0.914,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6820428336079076,
+      "grad_norm": 1642634.125,
+      "learning_rate": 5.903908794788274e-06,
+      "loss": 1.1433,
+      "step": 1221
+    },
+    {
+      "epoch": 2.684239428885228,
+      "grad_norm": 1448348.375,
+      "learning_rate": 5.863192182410424e-06,
+      "loss": 0.7058,
+      "step": 1222
+    },
+    {
+      "epoch": 2.686436024162548,
+      "grad_norm": 2249956.75,
+      "learning_rate": 5.822475570032573e-06,
+      "loss": 0.9459,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6886326194398684,
+      "grad_norm": 1917769.5,
+      "learning_rate": 5.781758957654723e-06,
+      "loss": 0.9408,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6908292147171884,
+      "grad_norm": 1487381.625,
+      "learning_rate": 5.741042345276873e-06,
+      "loss": 0.9913,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6930258099945084,
+      "grad_norm": 1656747.875,
+      "learning_rate": 5.700325732899023e-06,
+      "loss": 0.7321,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6952224052718288,
+      "grad_norm": 1595041.625,
+      "learning_rate": 5.659609120521173e-06,
+      "loss": 0.9881,
+      "step": 1227
+    },
+    {
+      "epoch": 2.6974190005491487,
+      "grad_norm": 3876064.25,
+      "learning_rate": 5.618892508143322e-06,
+      "loss": 0.8689,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6996155958264687,
+      "grad_norm": 2424973.25,
+      "learning_rate": 5.578175895765473e-06,
+      "loss": 1.0565,
+      "step": 1229
+    },
+    {
+      "epoch": 2.701812191103789,
+      "grad_norm": 12574552.0,
+      "learning_rate": 5.537459283387622e-06,
+      "loss": 0.7787,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7040087863811095,
+      "grad_norm": 1516350.75,
+      "learning_rate": 5.496742671009772e-06,
+      "loss": 0.7614,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7062053816584295,
+      "grad_norm": 1100186.25,
+      "learning_rate": 5.456026058631922e-06,
+      "loss": 0.9814,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7084019769357495,
+      "grad_norm": 2061419.375,
+      "learning_rate": 5.415309446254072e-06,
+      "loss": 0.8185,
+      "step": 1233
+    },
+    {
+      "epoch": 2.71059857221307,
+      "grad_norm": 1778106.0,
+      "learning_rate": 5.3745928338762225e-06,
+      "loss": 0.8616,
+      "step": 1234
+    },
+    {
+      "epoch": 2.71279516749039,
+      "grad_norm": 915714.75,
+      "learning_rate": 5.333876221498372e-06,
+      "loss": 1.0416,
+      "step": 1235
+    },
+    {
+      "epoch": 2.71499176276771,
+      "grad_norm": 1365092.25,
+      "learning_rate": 5.293159609120522e-06,
+      "loss": 1.1417,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7171883580450302,
+      "grad_norm": 2094327.75,
+      "learning_rate": 5.252442996742671e-06,
+      "loss": 0.95,
+      "step": 1237
+    },
+    {
+      "epoch": 2.71938495332235,
+      "grad_norm": 1992781.0,
+      "learning_rate": 5.211726384364822e-06,
+      "loss": 0.9561,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7215815485996706,
+      "grad_norm": 2370348.25,
+      "learning_rate": 5.171009771986971e-06,
+      "loss": 0.8522,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7237781438769906,
+      "grad_norm": 1446170.0,
+      "learning_rate": 5.130293159609121e-06,
+      "loss": 0.9364,
+      "step": 1240
+    },
+    {
+      "epoch": 2.725974739154311,
+      "grad_norm": 1551456.375,
+      "learning_rate": 5.089576547231271e-06,
+      "loss": 0.765,
+      "step": 1241
+    },
+    {
+      "epoch": 2.728171334431631,
+      "grad_norm": 1051578.25,
+      "learning_rate": 5.04885993485342e-06,
+      "loss": 0.8604,
+      "step": 1242
+    },
+    {
+      "epoch": 2.730367929708951,
+      "grad_norm": 2083879.625,
+      "learning_rate": 5.008143322475571e-06,
+      "loss": 0.9735,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7325645249862713,
+      "grad_norm": 833618.25,
+      "learning_rate": 4.96742671009772e-06,
+      "loss": 0.9976,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7347611202635913,
+      "grad_norm": 1959026.125,
+      "learning_rate": 4.92671009771987e-06,
+      "loss": 0.9973,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7369577155409117,
+      "grad_norm": 2064349.875,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 0.8947,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7391543108182317,
+      "grad_norm": 2012487.25,
+      "learning_rate": 4.84527687296417e-06,
+      "loss": 1.4479,
+      "step": 1247
+    },
+    {
+      "epoch": 2.741350906095552,
+      "grad_norm": 2349600.25,
+      "learning_rate": 4.8045602605863196e-06,
+      "loss": 0.8887,
+      "step": 1248
+    },
+    {
+      "epoch": 2.743547501372872,
+      "grad_norm": 1745776.625,
+      "learning_rate": 4.763843648208469e-06,
+      "loss": 1.0183,
+      "step": 1249
+    },
+    {
+      "epoch": 2.745744096650192,
+      "grad_norm": 1520103.625,
+      "learning_rate": 4.723127035830619e-06,
+      "loss": 0.9157,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7479406919275124,
+      "grad_norm": 3169934.25,
+      "learning_rate": 4.682410423452769e-06,
+      "loss": 1.0285,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7501372872048324,
+      "grad_norm": 1138326.375,
+      "learning_rate": 4.641693811074919e-06,
+      "loss": 1.0061,
+      "step": 1252
+    },
+    {
+      "epoch": 2.752333882482153,
+      "grad_norm": 2520258.5,
+      "learning_rate": 4.6009771986970685e-06,
+      "loss": 0.8141,
+      "step": 1253
+    },
+    {
+      "epoch": 2.754530477759473,
+      "grad_norm": 1981697.125,
+      "learning_rate": 4.560260586319218e-06,
+      "loss": 0.8471,
+      "step": 1254
+    },
+    {
+      "epoch": 2.756727073036793,
+      "grad_norm": 4664165.5,
+      "learning_rate": 4.5195439739413685e-06,
+      "loss": 1.0669,
+      "step": 1255
+    },
+    {
+      "epoch": 2.758923668314113,
+      "grad_norm": 5280658.0,
+      "learning_rate": 4.478827361563518e-06,
+      "loss": 0.7405,
+      "step": 1256
+    },
+    {
+      "epoch": 2.761120263591433,
+      "grad_norm": 977322.875,
+      "learning_rate": 4.438110749185668e-06,
+      "loss": 0.9411,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7633168588687536,
+      "grad_norm": 1431666.125,
+      "learning_rate": 4.397394136807817e-06,
+      "loss": 0.7526,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7655134541460735,
+      "grad_norm": 1787623.625,
+      "learning_rate": 4.356677524429968e-06,
+      "loss": 1.0046,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7677100494233935,
+      "grad_norm": 628591.625,
+      "learning_rate": 4.3159609120521174e-06,
+      "loss": 1.04,
+      "step": 1260
+    },
+    {
+      "epoch": 2.769906644700714,
+      "grad_norm": 6506424.0,
+      "learning_rate": 4.275244299674267e-06,
+      "loss": 0.9329,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7721032399780343,
+      "grad_norm": 2032906.25,
+      "learning_rate": 4.234527687296417e-06,
+      "loss": 0.9521,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7742998352553543,
+      "grad_norm": 877969.875,
+      "learning_rate": 4.193811074918567e-06,
+      "loss": 0.9607,
+      "step": 1263
+    },
+    {
+      "epoch": 2.7764964305326743,
+      "grad_norm": 636740.625,
+      "learning_rate": 4.153094462540717e-06,
+      "loss": 1.0672,
+      "step": 1264
+    },
+    {
+      "epoch": 2.7786930258099947,
+      "grad_norm": 910441.0,
+      "learning_rate": 4.112377850162866e-06,
+      "loss": 0.9637,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7808896210873146,
+      "grad_norm": 1192154.625,
+      "learning_rate": 4.071661237785016e-06,
+      "loss": 0.8037,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7830862163646346,
+      "grad_norm": 883359.125,
+      "learning_rate": 4.030944625407166e-06,
+      "loss": 0.779,
+      "step": 1267
+    },
+    {
+      "epoch": 2.785282811641955,
+      "grad_norm": 832570.8125,
+      "learning_rate": 3.990228013029316e-06,
+      "loss": 0.8469,
+      "step": 1268
+    },
+    {
+      "epoch": 2.787479406919275,
+      "grad_norm": 1072804.875,
+      "learning_rate": 3.949511400651466e-06,
+      "loss": 0.7789,
+      "step": 1269
+    },
+    {
+      "epoch": 2.7896760021965954,
+      "grad_norm": 1087654.125,
+      "learning_rate": 3.908794788273615e-06,
+      "loss": 0.8397,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7918725974739154,
+      "grad_norm": 1648599.25,
+      "learning_rate": 3.868078175895766e-06,
+      "loss": 0.8181,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7940691927512358,
+      "grad_norm": 2895951.5,
+      "learning_rate": 3.827361563517916e-06,
+      "loss": 0.8925,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7962657880285557,
+      "grad_norm": 2098716.75,
+      "learning_rate": 3.7866449511400658e-06,
+      "loss": 0.9499,
+      "step": 1273
+    },
+    {
+      "epoch": 2.7984623833058757,
+      "grad_norm": 1123537.5,
+      "learning_rate": 3.7459283387622154e-06,
+      "loss": 1.146,
+      "step": 1274
+    },
+    {
+      "epoch": 2.800658978583196,
+      "grad_norm": 3687213.75,
+      "learning_rate": 3.7052117263843654e-06,
+      "loss": 0.8588,
+      "step": 1275
+    },
+    {
+      "epoch": 2.802855573860516,
+      "grad_norm": 1790261.25,
+      "learning_rate": 3.664495114006515e-06,
+      "loss": 1.0259,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8050521691378365,
+      "grad_norm": 1815642.0,
+      "learning_rate": 3.623778501628665e-06,
+      "loss": 1.1725,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8072487644151565,
+      "grad_norm": 3437566.25,
+      "learning_rate": 3.5830618892508147e-06,
+      "loss": 0.9466,
+      "step": 1278
+    },
+    {
+      "epoch": 2.809445359692477,
+      "grad_norm": 4327000.0,
+      "learning_rate": 3.5423452768729647e-06,
+      "loss": 0.8548,
+      "step": 1279
+    },
+    {
+      "epoch": 2.811641954969797,
+      "grad_norm": 5509221.0,
+      "learning_rate": 3.5016286644951143e-06,
+      "loss": 0.9131,
+      "step": 1280
+    },
+    {
+      "epoch": 2.813838550247117,
+      "grad_norm": 2338416.75,
+      "learning_rate": 3.460912052117264e-06,
+      "loss": 1.003,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8160351455244372,
+      "grad_norm": 1033468.0625,
+      "learning_rate": 3.420195439739414e-06,
+      "loss": 0.88,
+      "step": 1282
+    },
+    {
+      "epoch": 2.818231740801757,
+      "grad_norm": 1177025.875,
+      "learning_rate": 3.3794788273615636e-06,
+      "loss": 1.102,
+      "step": 1283
+    },
+    {
+      "epoch": 2.820428336079077,
+      "grad_norm": 978581.9375,
+      "learning_rate": 3.3387622149837136e-06,
+      "loss": 1.031,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8226249313563976,
+      "grad_norm": 1016134.0,
+      "learning_rate": 3.2980456026058632e-06,
+      "loss": 0.9971,
+      "step": 1285
+    },
+    {
+      "epoch": 2.824821526633718,
+      "grad_norm": 2033483.375,
+      "learning_rate": 3.2573289902280133e-06,
+      "loss": 1.1467,
+      "step": 1286
+    },
+    {
+      "epoch": 2.827018121911038,
+      "grad_norm": 2450283.75,
+      "learning_rate": 3.216612377850163e-06,
+      "loss": 0.9003,
+      "step": 1287
+    },
+    {
+      "epoch": 2.829214717188358,
+      "grad_norm": 2191004.0,
+      "learning_rate": 3.175895765472313e-06,
+      "loss": 0.9573,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8314113124656783,
+      "grad_norm": 1361423.875,
+      "learning_rate": 3.1351791530944625e-06,
+      "loss": 0.943,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8336079077429983,
+      "grad_norm": 2481984.5,
+      "learning_rate": 3.0944625407166126e-06,
+      "loss": 1.0124,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8358045030203183,
+      "grad_norm": 1801705.125,
+      "learning_rate": 3.053745928338762e-06,
+      "loss": 0.7636,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8380010982976387,
+      "grad_norm": 1629957.375,
+      "learning_rate": 3.0130293159609122e-06,
+      "loss": 0.8272,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8401976935749587,
+      "grad_norm": 1378720.0,
+      "learning_rate": 2.972312703583062e-06,
+      "loss": 0.7251,
+      "step": 1293
+    },
+    {
+      "epoch": 2.842394288852279,
+      "grad_norm": 1243376.375,
+      "learning_rate": 2.931596091205212e-06,
+      "loss": 1.0439,
+      "step": 1294
+    },
+    {
+      "epoch": 2.844590884129599,
+      "grad_norm": 2272757.5,
+      "learning_rate": 2.8908794788273615e-06,
+      "loss": 0.9846,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8467874794069195,
+      "grad_norm": 2056938.75,
+      "learning_rate": 2.8501628664495115e-06,
+      "loss": 1.076,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8489840746842394,
+      "grad_norm": 1364959.5,
+      "learning_rate": 2.809446254071661e-06,
+      "loss": 0.8134,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8511806699615594,
+      "grad_norm": 2585693.75,
+      "learning_rate": 2.768729641693811e-06,
+      "loss": 1.1426,
+      "step": 1298
+    },
+    {
+      "epoch": 2.85337726523888,
+      "grad_norm": 1825060.75,
+      "learning_rate": 2.728013029315961e-06,
+      "loss": 0.943,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8555738605161998,
+      "grad_norm": 979609.375,
+      "learning_rate": 2.6872964169381112e-06,
+      "loss": 1.0538,
+      "step": 1300
+    },
+    {
+      "epoch": 2.85777045579352,
+      "grad_norm": 1030316.875,
+      "learning_rate": 2.646579804560261e-06,
+      "loss": 0.9976,
+      "step": 1301
+    },
+    {
+      "epoch": 2.85996705107084,
+      "grad_norm": 3794116.75,
+      "learning_rate": 2.605863192182411e-06,
+      "loss": 0.7937,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8621636463481606,
+      "grad_norm": 4988711.5,
+      "learning_rate": 2.5651465798045605e-06,
+      "loss": 1.1926,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8643602416254805,
+      "grad_norm": 2010447.875,
+      "learning_rate": 2.52442996742671e-06,
+      "loss": 0.7967,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8665568369028005,
+      "grad_norm": 2953106.25,
+      "learning_rate": 2.48371335504886e-06,
+      "loss": 0.8718,
+      "step": 1305
+    },
+    {
+      "epoch": 2.868753432180121,
+      "grad_norm": 3939449.75,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 0.9124,
+      "step": 1306
+    },
+    {
+      "epoch": 2.870950027457441,
+      "grad_norm": 1235599.875,
+      "learning_rate": 2.4022801302931598e-06,
+      "loss": 0.9391,
+      "step": 1307
+    },
+    {
+      "epoch": 2.873146622734761,
+      "grad_norm": 2248531.75,
+      "learning_rate": 2.3615635179153094e-06,
+      "loss": 0.7762,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8753432180120813,
+      "grad_norm": 8066388.5,
+      "learning_rate": 2.3208469055374594e-06,
+      "loss": 0.9556,
+      "step": 1309
+    },
+    {
+      "epoch": 2.8775398132894017,
+      "grad_norm": 6289998.0,
+      "learning_rate": 2.280130293159609e-06,
+      "loss": 1.1212,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8797364085667216,
+      "grad_norm": 4413165.0,
+      "learning_rate": 2.239413680781759e-06,
+      "loss": 0.9116,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8819330038440416,
+      "grad_norm": 2994462.75,
+      "learning_rate": 2.1986970684039087e-06,
+      "loss": 0.9691,
+      "step": 1312
+    },
+    {
+      "epoch": 2.884129599121362,
+      "grad_norm": 3440539.0,
+      "learning_rate": 2.1579804560260587e-06,
+      "loss": 0.9799,
+      "step": 1313
+    },
+    {
+      "epoch": 2.886326194398682,
+      "grad_norm": 1738647.125,
+      "learning_rate": 2.1172638436482083e-06,
+      "loss": 0.9699,
+      "step": 1314
+    },
+    {
+      "epoch": 2.888522789676002,
+      "grad_norm": 4423545.5,
+      "learning_rate": 2.0765472312703584e-06,
+      "loss": 0.8702,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8907193849533224,
+      "grad_norm": 1769714.25,
+      "learning_rate": 2.035830618892508e-06,
+      "loss": 0.8995,
+      "step": 1316
+    },
+    {
+      "epoch": 2.892915980230643,
+      "grad_norm": 3851606.5,
+      "learning_rate": 1.995114006514658e-06,
+      "loss": 0.8642,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8951125755079627,
+      "grad_norm": 2591223.25,
+      "learning_rate": 1.9543973941368076e-06,
+      "loss": 0.756,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8973091707852827,
+      "grad_norm": 4694927.5,
+      "learning_rate": 1.913680781758958e-06,
+      "loss": 0.7415,
+      "step": 1319
+    },
+    {
+      "epoch": 2.899505766062603,
+      "grad_norm": 4060603.0,
+      "learning_rate": 1.8729641693811077e-06,
+      "loss": 1.0408,
+      "step": 1320
+    },
+    {
+      "epoch": 2.901702361339923,
+      "grad_norm": 2173536.5,
+      "learning_rate": 1.8322475570032575e-06,
+      "loss": 0.7048,
+      "step": 1321
+    },
+    {
+      "epoch": 2.903898956617243,
+      "grad_norm": 1369903.875,
+      "learning_rate": 1.7915309446254073e-06,
+      "loss": 0.9383,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9060955518945635,
+      "grad_norm": 2814166.25,
+      "learning_rate": 1.7508143322475572e-06,
+      "loss": 0.8232,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9082921471718834,
+      "grad_norm": 2475005.5,
+      "learning_rate": 1.710097719869707e-06,
+      "loss": 0.9694,
+      "step": 1324
+    },
+    {
+      "epoch": 2.910488742449204,
+      "grad_norm": 1002695.4375,
+      "learning_rate": 1.6693811074918568e-06,
+      "loss": 0.8154,
+      "step": 1325
+    },
+    {
+      "epoch": 2.912685337726524,
+      "grad_norm": 2182041.75,
+      "learning_rate": 1.6286644951140066e-06,
+      "loss": 0.9101,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9148819330038442,
+      "grad_norm": 9407600.0,
+      "learning_rate": 1.5879478827361565e-06,
+      "loss": 0.9231,
+      "step": 1327
+    },
+    {
+      "epoch": 2.917078528281164,
+      "grad_norm": 2795357.25,
+      "learning_rate": 1.5472312703583063e-06,
+      "loss": 0.7344,
+      "step": 1328
+    },
+    {
+      "epoch": 2.919275123558484,
+      "grad_norm": 4107827.25,
+      "learning_rate": 1.5065146579804561e-06,
+      "loss": 0.7202,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9214717188358046,
+      "grad_norm": 4924422.5,
+      "learning_rate": 1.465798045602606e-06,
+      "loss": 0.8585,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9236683141131246,
+      "grad_norm": 5554205.0,
+      "learning_rate": 1.4250814332247558e-06,
+      "loss": 1.0892,
+      "step": 1331
+    },
+    {
+      "epoch": 2.925864909390445,
+      "grad_norm": 1239198.375,
+      "learning_rate": 1.3843648208469056e-06,
+      "loss": 0.5501,
+      "step": 1332
+    },
+    {
+      "epoch": 2.928061504667765,
+      "grad_norm": 3509285.75,
+      "learning_rate": 1.3436482084690556e-06,
+      "loss": 0.8339,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9302580999450853,
+      "grad_norm": 1792269.875,
+      "learning_rate": 1.3029315960912054e-06,
+      "loss": 0.8407,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9324546952224053,
+      "grad_norm": 2802601.5,
+      "learning_rate": 1.262214983713355e-06,
+      "loss": 0.7949,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9346512904997253,
+      "grad_norm": 1854249.75,
+      "learning_rate": 1.2214983713355049e-06,
+      "loss": 0.8019,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9368478857770457,
+      "grad_norm": 1740052.75,
+      "learning_rate": 1.1807817589576547e-06,
+      "loss": 0.7879,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9390444810543657,
+      "grad_norm": 2719202.75,
+      "learning_rate": 1.1400651465798045e-06,
+      "loss": 0.8041,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9412410763316856,
+      "grad_norm": 2500647.0,
+      "learning_rate": 1.0993485342019543e-06,
+      "loss": 0.839,
+      "step": 1339
+    },
+    {
+      "epoch": 2.943437671609006,
+      "grad_norm": 4806297.5,
+      "learning_rate": 1.0586319218241042e-06,
+      "loss": 0.6392,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9456342668863265,
+      "grad_norm": 1738768.25,
+      "learning_rate": 1.017915309446254e-06,
+      "loss": 0.8984,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9478308621636464,
+      "grad_norm": 4652460.5,
+      "learning_rate": 9.771986970684038e-07,
+      "loss": 0.9174,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9500274574409664,
+      "grad_norm": 2184450.0,
+      "learning_rate": 9.364820846905538e-07,
+      "loss": 0.9886,
+      "step": 1343
+    },
+    {
+      "epoch": 2.952224052718287,
+      "grad_norm": 1254930.125,
+      "learning_rate": 8.957654723127037e-07,
+      "loss": 1.0893,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9544206479956068,
+      "grad_norm": 2406315.75,
+      "learning_rate": 8.550488599348535e-07,
+      "loss": 0.9957,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9566172432729267,
+      "grad_norm": 12289882.0,
+      "learning_rate": 8.143322475570033e-07,
+      "loss": 0.8448,
+      "step": 1346
+    },
+    {
+      "epoch": 2.958813838550247,
+      "grad_norm": 3887086.0,
+      "learning_rate": 7.736156351791531e-07,
+      "loss": 0.8603,
+      "step": 1347
+    },
+    {
+      "epoch": 2.961010433827567,
+      "grad_norm": 5866402.5,
+      "learning_rate": 7.32899022801303e-07,
+      "loss": 0.7455,
+      "step": 1348
+    },
+    {
+      "epoch": 2.9632070291048875,
+      "grad_norm": 2806576.5,
+      "learning_rate": 6.921824104234528e-07,
+      "loss": 1.1782,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9654036243822075,
+      "grad_norm": 1751908.5,
+      "learning_rate": 6.514657980456027e-07,
+      "loss": 0.8152,
+      "step": 1350
+    },
+    {
+      "epoch": 2.967600219659528,
+      "grad_norm": 1759828.875,
+      "learning_rate": 6.107491856677524e-07,
+      "loss": 1.0757,
+      "step": 1351
+    },
+    {
+      "epoch": 2.969796814936848,
+      "grad_norm": 1662208.625,
+      "learning_rate": 5.700325732899023e-07,
+      "loss": 0.8954,
+      "step": 1352
+    },
+    {
+      "epoch": 2.971993410214168,
+      "grad_norm": 4824284.0,
+      "learning_rate": 5.293159609120521e-07,
+      "loss": 0.9703,
+      "step": 1353
+    },
+    {
+      "epoch": 2.9741900054914883,
+      "grad_norm": 1383399.5,
+      "learning_rate": 4.885993485342019e-07,
+      "loss": 1.0408,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9763866007688082,
+      "grad_norm": 2240621.0,
+      "learning_rate": 4.4788273615635184e-07,
+      "loss": 0.7816,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9785831960461286,
+      "grad_norm": 1203263.25,
+      "learning_rate": 4.0716612377850166e-07,
+      "loss": 1.0825,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9807797913234486,
+      "grad_norm": 1984830.5,
+      "learning_rate": 3.664495114006515e-07,
+      "loss": 0.9688,
+      "step": 1357
+    },
+    {
+      "epoch": 2.982976386600769,
+      "grad_norm": 3022934.25,
+      "learning_rate": 3.2573289902280136e-07,
+      "loss": 0.9687,
+      "step": 1358
+    },
+    {
+      "epoch": 2.985172981878089,
+      "grad_norm": 1848157.625,
+      "learning_rate": 2.8501628664495113e-07,
+      "loss": 1.0623,
+      "step": 1359
+    },
+    {
+      "epoch": 2.987369577155409,
+      "grad_norm": 2661720.25,
+      "learning_rate": 2.4429967426710095e-07,
+      "loss": 0.7379,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9895661724327294,
+      "grad_norm": 1405579.75,
+      "learning_rate": 2.0358306188925083e-07,
+      "loss": 0.9212,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9917627677100493,
+      "grad_norm": 1041576.3125,
+      "learning_rate": 1.6286644951140068e-07,
+      "loss": 1.178,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9939593629873693,
+      "grad_norm": 5677789.0,
+      "learning_rate": 1.2214983713355048e-07,
+      "loss": 0.9365,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9961559582646897,
+      "grad_norm": 1321839.75,
+      "learning_rate": 8.143322475570034e-08,
+      "loss": 0.9391,
+      "step": 1364
+    },
+    {
+      "epoch": 2.99835255354201,
+      "grad_norm": 3630052.5,
+      "learning_rate": 4.071661237785017e-08,
+      "loss": 0.9652,
+      "step": 1365
+    },
+    {
+      "epoch": 2.99835255354201,
+      "step": 1365,
+      "total_flos": 1.9058535824778854e+17,
+      "train_loss": 1.1155239005665203,
+      "train_runtime": 27996.883,
+      "train_samples_per_second": 0.78,
+      "train_steps_per_second": 0.049
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1365,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.9058535824778854e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}