diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28805 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.10004622758570351,
+  "eval_steps": 500,
+  "global_step": 4112,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.502673555476038,
+      "learning_rate": 8.103727714748784e-09,
+      "loss": 1.0313,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.658462234362208,
+      "learning_rate": 1.620745542949757e-08,
+      "loss": 0.8013,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.58082516844464,
+      "learning_rate": 2.4311183144246356e-08,
+      "loss": 0.8823,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.158369170130396,
+      "learning_rate": 3.241491085899514e-08,
+      "loss": 0.9733,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.635532754785615,
+      "learning_rate": 4.0518638573743925e-08,
+      "loss": 0.7794,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.000015513094192,
+      "learning_rate": 4.862236628849271e-08,
+      "loss": 1.1107,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.888280883207948,
+      "learning_rate": 5.6726094003241494e-08,
+      "loss": 0.9204,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.622181588960192,
+      "learning_rate": 6.482982171799027e-08,
+      "loss": 1.3566,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.237100847566673,
+      "learning_rate": 7.293354943273906e-08,
+      "loss": 0.9846,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.259269023701474,
+      "learning_rate": 8.103727714748785e-08,
+      "loss": 0.9483,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.113124125855178,
+      "learning_rate": 8.914100486223664e-08,
+      "loss": 1.0317,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.268904678242057,
+      "learning_rate": 9.724473257698543e-08,
+      "loss": 0.9106,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.837522583445786,
+      "learning_rate": 1.053484602917342e-07,
+      "loss": 0.8511,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.028370920550065,
+      "learning_rate": 1.1345218800648299e-07,
+      "loss": 0.534,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.86068263415377,
+      "learning_rate": 1.215559157212318e-07,
+      "loss": 0.7407,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.916825164009692,
+      "learning_rate": 1.2965964343598055e-07,
+      "loss": 0.5575,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.57943977607431,
+      "learning_rate": 1.3776337115072934e-07,
+      "loss": 0.748,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.501036217458406,
+      "learning_rate": 1.4586709886547813e-07,
+      "loss": 0.8239,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.899707917070728,
+      "learning_rate": 1.539708265802269e-07,
+      "loss": 0.76,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.19860333124517,
+      "learning_rate": 1.620745542949757e-07,
+      "loss": 0.7286,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.636099817175502,
+      "learning_rate": 1.7017828200972446e-07,
+      "loss": 0.8181,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.998551888492294,
+      "learning_rate": 1.7828200972447328e-07,
+      "loss": 0.8085,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.347819412793825,
+      "learning_rate": 1.8638573743922206e-07,
+      "loss": 0.7432,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.618766591270745,
+      "learning_rate": 1.9448946515397085e-07,
+      "loss": 0.6323,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.558446015785616,
+      "learning_rate": 2.0259319286871964e-07,
+      "loss": 0.7305,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.680499703254362,
+      "learning_rate": 2.106969205834684e-07,
+      "loss": 0.772,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.71132185081572,
+      "learning_rate": 2.1880064829821721e-07,
+      "loss": 0.8924,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.06338263943024,
+      "learning_rate": 2.2690437601296598e-07,
+      "loss": 0.9672,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.081346101626615,
+      "learning_rate": 2.3500810372771476e-07,
+      "loss": 1.038,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.137940071222836,
+      "learning_rate": 2.431118314424636e-07,
+      "loss": 0.5523,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.569755598051817,
+      "learning_rate": 2.512155591572123e-07,
+      "loss": 1.0804,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.685410984358743,
+      "learning_rate": 2.593192868719611e-07,
+      "loss": 0.7927,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.238654455644976,
+      "learning_rate": 2.6742301458670994e-07,
+      "loss": 1.2216,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.810990056838708,
+      "learning_rate": 2.755267423014587e-07,
+      "loss": 0.7208,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.013592106829662,
+      "learning_rate": 2.8363047001620746e-07,
+      "loss": 0.68,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.129656654524092,
+      "learning_rate": 2.9173419773095625e-07,
+      "loss": 0.8363,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.43723229429919,
+      "learning_rate": 2.9983792544570504e-07,
+      "loss": 0.8409,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.077648085782336,
+      "learning_rate": 3.079416531604538e-07,
+      "loss": 0.7023,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.841197997112697,
+      "learning_rate": 3.160453808752026e-07,
+      "loss": 0.4066,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.783803601024776,
+      "learning_rate": 3.241491085899514e-07,
+      "loss": 0.4545,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.188390765004614,
+      "learning_rate": 3.322528363047002e-07,
+      "loss": 0.8925,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.296339554138216,
+      "learning_rate": 3.403565640194489e-07,
+      "loss": 0.5936,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.493814526262367,
+      "learning_rate": 3.4846029173419776e-07,
+      "loss": 0.8092,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.028883498691142,
+      "learning_rate": 3.5656401944894655e-07,
+      "loss": 0.437,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.25887416856952,
+      "learning_rate": 3.646677471636953e-07,
+      "loss": 0.9594,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.374201066624943,
+      "learning_rate": 3.7277147487844413e-07,
+      "loss": 0.4671,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.3764789197598954,
+      "learning_rate": 3.808752025931929e-07,
+      "loss": 0.8344,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.803280275361416,
+      "learning_rate": 3.889789303079417e-07,
+      "loss": 0.49,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.627509953851554,
+      "learning_rate": 3.9708265802269044e-07,
+      "loss": 0.653,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.330121192258138,
+      "learning_rate": 4.051863857374393e-07,
+      "loss": 0.4291,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.95002241885734,
+      "learning_rate": 4.1329011345218807e-07,
+      "loss": 0.4645,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.906987091806532,
+      "learning_rate": 4.213938411669368e-07,
+      "loss": 0.746,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.32849657095027,
+      "learning_rate": 4.2949756888168564e-07,
+      "loss": 0.7291,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.296649970234043,
+      "learning_rate": 4.3760129659643443e-07,
+      "loss": 1.1122,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.978189174085387,
+      "learning_rate": 4.4570502431118316e-07,
+      "loss": 0.4716,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.4738910011865025,
+      "learning_rate": 4.5380875202593195e-07,
+      "loss": 0.4103,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.422614544611374,
+      "learning_rate": 4.619124797406808e-07,
+      "loss": 0.6453,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.23541553198637,
+      "learning_rate": 4.7001620745542953e-07,
+      "loss": 0.4255,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.016904854375009,
+      "learning_rate": 4.781199351701784e-07,
+      "loss": 0.9832,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.476761674769124,
+      "learning_rate": 4.862236628849272e-07,
+      "loss": 0.3079,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.42362589210662,
+      "learning_rate": 4.943273905996758e-07,
+      "loss": 0.3429,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.136564604908534,
+      "learning_rate": 5.024311183144246e-07,
+      "loss": 0.4754,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.747683586797518,
+      "learning_rate": 5.105348460291735e-07,
+      "loss": 0.72,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.352666066899136,
+      "learning_rate": 5.186385737439222e-07,
+      "loss": 0.2402,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.081652100103129,
+      "learning_rate": 5.26742301458671e-07,
+      "loss": 0.7386,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.04571270001964,
+      "learning_rate": 5.348460291734199e-07,
+      "loss": 0.4008,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.3279716947417,
+      "learning_rate": 5.429497568881686e-07,
+      "loss": 0.4565,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.361485616366148,
+      "learning_rate": 5.510534846029174e-07,
+      "loss": 0.6441,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.812062852069395,
+      "learning_rate": 5.591572123176661e-07,
+      "loss": 0.7238,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.140793776194712,
+      "learning_rate": 5.672609400324149e-07,
+      "loss": 0.2818,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.628240120516564,
+      "learning_rate": 5.753646677471637e-07,
+      "loss": 0.38,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.0859536025675744,
+      "learning_rate": 5.834683954619125e-07,
+      "loss": 0.5698,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.477575581184537,
+      "learning_rate": 5.915721231766613e-07,
+      "loss": 0.3507,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.063433478676999,
+      "learning_rate": 5.996758508914101e-07,
+      "loss": 0.68,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.119492020000943,
+      "learning_rate": 6.077795786061589e-07,
+      "loss": 0.6425,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.49147617297199,
+      "learning_rate": 6.158833063209077e-07,
+      "loss": 0.6417,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.358771198391835,
+      "learning_rate": 6.239870340356564e-07,
+      "loss": 0.4162,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.307108802627474,
+      "learning_rate": 6.320907617504052e-07,
+      "loss": 0.3458,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.306825331967706,
+      "learning_rate": 6.40194489465154e-07,
+      "loss": 0.3535,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.142057317837912,
+      "learning_rate": 6.482982171799028e-07,
+      "loss": 0.301,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.786648862392994,
+      "learning_rate": 6.564019448946515e-07,
+      "loss": 0.3363,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.94186459961697,
+      "learning_rate": 6.645056726094004e-07,
+      "loss": 0.2465,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9472268625023776,
+      "learning_rate": 6.726094003241492e-07,
+      "loss": 0.4245,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.324122360151815,
+      "learning_rate": 6.807131280388978e-07,
+      "loss": 0.5497,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.882138332365245,
+      "learning_rate": 6.888168557536467e-07,
+      "loss": 0.4684,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.11148478374892,
+      "learning_rate": 6.969205834683955e-07,
+      "loss": 0.4098,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8346945308496188,
+      "learning_rate": 7.050243111831442e-07,
+      "loss": 0.507,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.004451956428494,
+      "learning_rate": 7.131280388978931e-07,
+      "loss": 0.261,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.026187293805123,
+      "learning_rate": 7.212317666126419e-07,
+      "loss": 0.2963,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.258617220796378,
+      "learning_rate": 7.293354943273906e-07,
+      "loss": 0.7882,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.928678817447845,
+      "learning_rate": 7.374392220421395e-07,
+      "loss": 0.783,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.004713380918248,
+      "learning_rate": 7.455429497568883e-07,
+      "loss": 0.2381,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.252268982082319,
+      "learning_rate": 7.53646677471637e-07,
+      "loss": 0.4275,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.203327919608768,
+      "learning_rate": 7.617504051863858e-07,
+      "loss": 0.3739,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.29043837750598,
+      "learning_rate": 7.698541329011345e-07,
+      "loss": 0.2028,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.82349417722011,
+      "learning_rate": 7.779578606158834e-07,
+      "loss": 0.2092,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8843021260159696,
+      "learning_rate": 7.860615883306322e-07,
+      "loss": 0.2835,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.306515587943666,
+      "learning_rate": 7.941653160453809e-07,
+      "loss": 0.6496,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.029623159546317,
+      "learning_rate": 8.022690437601298e-07,
+      "loss": 0.454,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.809999658620271,
+      "learning_rate": 8.103727714748786e-07,
+      "loss": 0.6521,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.931486451411266,
+      "learning_rate": 8.184764991896272e-07,
+      "loss": 0.3613,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.352469659619574,
+      "learning_rate": 8.265802269043761e-07,
+      "loss": 0.5014,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1382361479318046,
+      "learning_rate": 8.346839546191249e-07,
+      "loss": 0.4852,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.476737810014492,
+      "learning_rate": 8.427876823338736e-07,
+      "loss": 0.625,
+      "step": 104
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.487849946909529,
+      "learning_rate": 8.508914100486225e-07,
+      "loss": 0.793,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.719559050668978,
+      "learning_rate": 8.589951377633713e-07,
+      "loss": 0.4177,
+      "step": 106
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.323206561682611,
+      "learning_rate": 8.6709886547812e-07,
+      "loss": 0.3544,
+      "step": 107
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.352650641606749,
+      "learning_rate": 8.752025931928689e-07,
+      "loss": 0.2939,
+      "step": 108
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6127756620782168,
+      "learning_rate": 8.833063209076175e-07,
+      "loss": 0.5211,
+      "step": 109
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6870935507013565,
+      "learning_rate": 8.914100486223663e-07,
+      "loss": 0.2495,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.87246320206598,
+      "learning_rate": 8.995137763371152e-07,
+      "loss": 0.4755,
+      "step": 111
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.854638588596419,
+      "learning_rate": 9.076175040518639e-07,
+      "loss": 0.2356,
+      "step": 112
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.341507156577239,
+      "learning_rate": 9.157212317666127e-07,
+      "loss": 0.809,
+      "step": 113
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.756047654022981,
+      "learning_rate": 9.238249594813616e-07,
+      "loss": 0.2074,
+      "step": 114
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.340176244268669,
+      "learning_rate": 9.319286871961103e-07,
+      "loss": 0.2054,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.076568858059618,
+      "learning_rate": 9.400324149108591e-07,
+      "loss": 0.5922,
+      "step": 116
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.626809020019448,
+      "learning_rate": 9.481361426256079e-07,
+      "loss": 0.3322,
+      "step": 117
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.387573974072711,
+      "learning_rate": 9.562398703403567e-07,
+      "loss": 0.385,
+      "step": 118
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.60002406231468,
+      "learning_rate": 9.643435980551054e-07,
+      "loss": 0.2094,
+      "step": 119
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.330109922630415,
+      "learning_rate": 9.724473257698543e-07,
+      "loss": 0.6843,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.820427570175162,
+      "learning_rate": 9.80551053484603e-07,
+      "loss": 0.7655,
+      "step": 121
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5467236356324716,
+      "learning_rate": 9.886547811993517e-07,
+      "loss": 0.6045,
+      "step": 122
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.850978699479514,
+      "learning_rate": 9.967585089141006e-07,
+      "loss": 0.1974,
+      "step": 123
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.734697254907031,
+      "learning_rate": 1.0048622366288492e-06,
+      "loss": 0.6607,
+      "step": 124
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.373379671591483,
+      "learning_rate": 1.0129659643435981e-06,
+      "loss": 0.3882,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.150705123654179,
+      "learning_rate": 1.021069692058347e-06,
+      "loss": 0.3118,
+      "step": 126
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.365116945603119,
+      "learning_rate": 1.0291734197730957e-06,
+      "loss": 0.6412,
+      "step": 127
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.15628438874197,
+      "learning_rate": 1.0372771474878444e-06,
+      "loss": 0.2086,
+      "step": 128
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.732704810612644,
+      "learning_rate": 1.0453808752025933e-06,
+      "loss": 0.5465,
+      "step": 129
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.2144878756916375,
+      "learning_rate": 1.053484602917342e-06,
+      "loss": 0.5853,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.154910710749087,
+      "learning_rate": 1.0615883306320909e-06,
+      "loss": 0.5573,
+      "step": 131
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.0364438833983,
+      "learning_rate": 1.0696920583468398e-06,
+      "loss": 0.2574,
+      "step": 132
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.893141227758161,
+      "learning_rate": 1.0777957860615884e-06,
+      "loss": 0.185,
+      "step": 133
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.245453012156529,
+      "learning_rate": 1.0858995137763371e-06,
+      "loss": 0.4147,
+      "step": 134
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.277018495499633,
+      "learning_rate": 1.094003241491086e-06,
+      "loss": 0.5133,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.936671972259915,
+      "learning_rate": 1.1021069692058347e-06,
+      "loss": 0.4147,
+      "step": 136
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.389531829119963,
+      "learning_rate": 1.1102106969205836e-06,
+      "loss": 0.3128,
+      "step": 137
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.691386023484968,
+      "learning_rate": 1.1183144246353323e-06,
+      "loss": 0.482,
+      "step": 138
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2041943695663204,
+      "learning_rate": 1.1264181523500812e-06,
+      "loss": 0.1927,
+      "step": 139
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.456553320707066,
+      "learning_rate": 1.1345218800648299e-06,
+      "loss": 0.6259,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.6973253020304435,
+      "learning_rate": 1.1426256077795787e-06,
+      "loss": 0.5953,
+      "step": 141
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.759767559100558,
+      "learning_rate": 1.1507293354943274e-06,
+      "loss": 0.2644,
+      "step": 142
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.605625632010758,
+      "learning_rate": 1.1588330632090761e-06,
+      "loss": 0.2718,
+      "step": 143
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.878767704482246,
+      "learning_rate": 1.166936790923825e-06,
+      "loss": 0.2419,
+      "step": 144
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.226429601707935,
+      "learning_rate": 1.175040518638574e-06,
+      "loss": 0.2443,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.603494157769408,
+      "learning_rate": 1.1831442463533226e-06,
+      "loss": 0.2774,
+      "step": 146
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.735976334724583,
+      "learning_rate": 1.1912479740680715e-06,
+      "loss": 0.423,
+      "step": 147
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.079886775736427,
+      "learning_rate": 1.1993517017828202e-06,
+      "loss": 0.7454,
+      "step": 148
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1694104057595673,
+      "learning_rate": 1.2074554294975688e-06,
+      "loss": 0.1765,
+      "step": 149
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0107910864982768,
+      "learning_rate": 1.2155591572123177e-06,
+      "loss": 0.358,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.436229842467225,
+      "learning_rate": 1.2236628849270666e-06,
+      "loss": 0.6371,
+      "step": 151
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3233766665896955,
+      "learning_rate": 1.2317666126418153e-06,
+      "loss": 0.2977,
+      "step": 152
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.465738201908871,
+      "learning_rate": 1.2398703403565642e-06,
+      "loss": 0.6907,
+      "step": 153
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.19253280303806,
+      "learning_rate": 1.2479740680713129e-06,
+      "loss": 0.7867,
+      "step": 154
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.332767043092093,
+      "learning_rate": 1.2560777957860618e-06,
+      "loss": 0.2089,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.204015711102247,
+      "learning_rate": 1.2641815235008105e-06,
+      "loss": 0.4326,
+      "step": 156
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.038319046415834,
+      "learning_rate": 1.2722852512155591e-06,
+      "loss": 0.5403,
+      "step": 157
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.479595090219046,
+      "learning_rate": 1.280388978930308e-06,
+      "loss": 0.3132,
+      "step": 158
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.649529303940417,
+      "learning_rate": 1.2884927066450567e-06,
+      "loss": 0.4483,
+      "step": 159
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2737266030482353,
+      "learning_rate": 1.2965964343598056e-06,
+      "loss": 0.3666,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5502218345049736,
+      "learning_rate": 1.3047001620745545e-06,
+      "loss": 0.246,
+      "step": 161
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.773885443658553,
+      "learning_rate": 1.312803889789303e-06,
+      "loss": 0.5446,
+      "step": 162
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3978748250350175,
+      "learning_rate": 1.3209076175040519e-06,
+      "loss": 0.6249,
+      "step": 163
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.62028143700258,
+      "learning_rate": 1.3290113452188008e-06,
+      "loss": 0.4355,
+      "step": 164
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.288610493618289,
+      "learning_rate": 1.3371150729335494e-06,
+      "loss": 0.31,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8918734466270286,
+      "learning_rate": 1.3452188006482983e-06,
+      "loss": 0.6963,
+      "step": 166
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.231699366654097,
+      "learning_rate": 1.3533225283630472e-06,
+      "loss": 0.298,
+      "step": 167
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0102906954235116,
+      "learning_rate": 1.3614262560777957e-06,
+      "loss": 0.4095,
+      "step": 168
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8455411877884678,
+      "learning_rate": 1.3695299837925446e-06,
+      "loss": 0.2894,
+      "step": 169
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7937848738398214,
+      "learning_rate": 1.3776337115072935e-06,
+      "loss": 0.6625,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.539378931568264,
+      "learning_rate": 1.3857374392220422e-06,
+      "loss": 0.8054,
+      "step": 171
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8506046055750716,
+      "learning_rate": 1.393841166936791e-06,
+      "loss": 0.3857,
+      "step": 172
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8756915059971173,
+      "learning_rate": 1.40194489465154e-06,
+      "loss": 0.3935,
+      "step": 173
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.106625722092892,
+      "learning_rate": 1.4100486223662884e-06,
+      "loss": 0.1481,
+      "step": 174
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.236474547751039,
+      "learning_rate": 1.4181523500810373e-06,
+      "loss": 0.4944,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2566930919702215,
+      "learning_rate": 1.4262560777957862e-06,
+      "loss": 0.2346,
+      "step": 176
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.905262256248179,
+      "learning_rate": 1.4343598055105349e-06,
+      "loss": 0.4087,
+      "step": 177
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0193487118066895,
+      "learning_rate": 1.4424635332252838e-06,
+      "loss": 0.1617,
+      "step": 178
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6391069913267864,
+      "learning_rate": 1.4505672609400327e-06,
+      "loss": 0.2556,
+      "step": 179
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.351834724059104,
+      "learning_rate": 1.4586709886547811e-06,
+      "loss": 0.2339,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.929447875569426,
+      "learning_rate": 1.46677471636953e-06,
+      "loss": 0.597,
+      "step": 181
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.2654799263096725,
+      "learning_rate": 1.474878444084279e-06,
+      "loss": 0.2093,
+      "step": 182
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7416838811103994,
+      "learning_rate": 1.4829821717990276e-06,
+      "loss": 0.2039,
+      "step": 183
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9492868781904953,
+      "learning_rate": 1.4910858995137765e-06,
+      "loss": 0.2009,
+      "step": 184
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.470336945388423,
+      "learning_rate": 1.4991896272285252e-06,
+      "loss": 0.7654,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8772697569262387,
+      "learning_rate": 1.507293354943274e-06,
+      "loss": 0.5529,
+      "step": 186
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4610008251174875,
+      "learning_rate": 1.5153970826580228e-06,
+      "loss": 0.5239,
+      "step": 187
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5155578945960992,
+      "learning_rate": 1.5235008103727717e-06,
+      "loss": 0.233,
+      "step": 188
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.828877510759205,
+      "learning_rate": 1.5316045380875206e-06,
+      "loss": 0.3879,
+      "step": 189
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5684939505127313,
+      "learning_rate": 1.539708265802269e-06,
+      "loss": 0.1762,
+      "step": 190
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.483264134214535,
+      "learning_rate": 1.547811993517018e-06,
+      "loss": 0.3862,
+      "step": 191
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.727200067709566,
+      "learning_rate": 1.5559157212317668e-06,
+      "loss": 0.6709,
+      "step": 192
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.806344607995004,
+      "learning_rate": 1.5640194489465155e-06,
+      "loss": 0.3782,
+      "step": 193
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6555578114568,
+      "learning_rate": 1.5721231766612644e-06,
+      "loss": 0.2949,
+      "step": 194
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.767784607382133,
+      "learning_rate": 1.5802269043760133e-06,
+      "loss": 0.5981,
+      "step": 195
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.776509995900383,
+      "learning_rate": 1.5883306320907617e-06,
+      "loss": 0.3773,
+      "step": 196
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.074686850908197,
+      "learning_rate": 1.5964343598055106e-06,
+      "loss": 0.3549,
+      "step": 197
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.292905992738229,
+      "learning_rate": 1.6045380875202595e-06,
+      "loss": 0.3812,
+      "step": 198
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.864009807765747,
+      "learning_rate": 1.6126418152350082e-06,
+      "loss": 0.4716,
+      "step": 199
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.415238603408013,
+      "learning_rate": 1.6207455429497571e-06,
+      "loss": 0.253,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.156777427307081,
+      "learning_rate": 1.628849270664506e-06,
+      "loss": 0.4182,
+      "step": 201
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.083089269752956,
+      "learning_rate": 1.6369529983792545e-06,
+      "loss": 0.6336,
+      "step": 202
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.025543298838364,
+      "learning_rate": 1.6450567260940034e-06,
+      "loss": 0.2047,
+      "step": 203
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.311621652057666,
+      "learning_rate": 1.6531604538087523e-06,
+      "loss": 0.572,
+      "step": 204
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.21888297463705,
+      "learning_rate": 1.661264181523501e-06,
+      "loss": 0.3791,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.815325662962679,
+      "learning_rate": 1.6693679092382498e-06,
+      "loss": 0.4181,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2068324135038955,
+      "learning_rate": 1.6774716369529987e-06,
+      "loss": 0.2186,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.43530669899021,
+      "learning_rate": 1.6855753646677472e-06,
+      "loss": 0.4646,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2778136112864615,
+      "learning_rate": 1.693679092382496e-06,
+      "loss": 0.5522,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.116159121262782,
+      "learning_rate": 1.701782820097245e-06,
+      "loss": 0.1847,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.17129871225742,
+      "learning_rate": 1.7098865478119937e-06,
+      "loss": 0.4018,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.050840819262028,
+      "learning_rate": 1.7179902755267426e-06,
+      "loss": 0.5433,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5915947463644944,
+      "learning_rate": 1.7260940032414912e-06,
+      "loss": 0.5618,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.317634964256702,
+      "learning_rate": 1.73419773095624e-06,
+      "loss": 0.2207,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8706388982548092,
+      "learning_rate": 1.7423014586709888e-06,
+      "loss": 0.2297,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9854575385743014,
+      "learning_rate": 1.7504051863857377e-06,
+      "loss": 0.587,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1265413107242166,
+      "learning_rate": 1.7585089141004864e-06,
+      "loss": 0.2581,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.318990577588553,
+      "learning_rate": 1.766612641815235e-06,
+      "loss": 0.7061,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.64603611336765,
+      "learning_rate": 1.774716369529984e-06,
+      "loss": 0.3323,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9092557601996254,
+      "learning_rate": 1.7828200972447327e-06,
+      "loss": 0.2298,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3725158591494933,
+      "learning_rate": 1.7909238249594815e-06,
+      "loss": 0.347,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3077569148709505,
+      "learning_rate": 1.7990275526742304e-06,
+      "loss": 0.2731,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1973357941530804,
+      "learning_rate": 1.807131280388979e-06,
+      "loss": 0.2505,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8046952860476435,
+      "learning_rate": 1.8152350081037278e-06,
+      "loss": 0.2252,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.166118436350418,
+      "learning_rate": 1.8233387358184767e-06,
+      "loss": 0.7477,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.522239464464581,
+      "learning_rate": 1.8314424635332254e-06,
+      "loss": 0.2818,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7472323695042973,
+      "learning_rate": 1.8395461912479743e-06,
+      "loss": 0.2527,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7539894037340664,
+      "learning_rate": 1.8476499189627232e-06,
+      "loss": 0.3069,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.63118367887673,
+      "learning_rate": 1.8557536466774716e-06,
+      "loss": 0.5543,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.151739689962817,
+      "learning_rate": 1.8638573743922205e-06,
+      "loss": 0.2094,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.83775980987546,
+      "learning_rate": 1.8719611021069694e-06,
+      "loss": 0.1642,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.396792941256002,
+      "learning_rate": 1.8800648298217181e-06,
+      "loss": 0.603,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.134280284854865,
+      "learning_rate": 1.888168557536467e-06,
+      "loss": 0.3608,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9190606827970678,
+      "learning_rate": 1.8962722852512159e-06,
+      "loss": 0.4579,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7777917771516467,
+      "learning_rate": 1.9043760129659644e-06,
+      "loss": 0.3955,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.86015602270858,
+      "learning_rate": 1.9124797406807135e-06,
+      "loss": 0.7013,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9159629699490797,
+      "learning_rate": 1.9205834683954624e-06,
+      "loss": 0.3592,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.01231764245415,
+      "learning_rate": 1.928687196110211e-06,
+      "loss": 0.3793,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.123762002761547,
+      "learning_rate": 1.9367909238249597e-06,
+      "loss": 0.4415,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1458867344313224,
+      "learning_rate": 1.9448946515397086e-06,
+      "loss": 0.3676,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6995468558086912,
+      "learning_rate": 1.952998379254457e-06,
+      "loss": 0.3977,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.214966836710618,
+      "learning_rate": 1.961102106969206e-06,
+      "loss": 0.3097,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7187159240187215,
+      "learning_rate": 1.969205834683955e-06,
+      "loss": 0.3423,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2374184007018,
+      "learning_rate": 1.9773095623987033e-06,
+      "loss": 0.4469,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8014616714947618,
+      "learning_rate": 1.9854132901134522e-06,
+      "loss": 0.3443,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.375545474030666,
+      "learning_rate": 1.993517017828201e-06,
+      "loss": 0.2596,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7439216306953234,
+      "learning_rate": 2.00162074554295e-06,
+      "loss": 0.4681,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3315995396207394,
+      "learning_rate": 2.0097244732576985e-06,
+      "loss": 0.2141,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.560852322264811,
+      "learning_rate": 2.0178282009724474e-06,
+      "loss": 0.2365,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.479874719688596,
+      "learning_rate": 2.0259319286871963e-06,
+      "loss": 0.2967,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.645879722697952,
+      "learning_rate": 2.034035656401945e-06,
+      "loss": 0.1187,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7161009192564993,
+      "learning_rate": 2.042139384116694e-06,
+      "loss": 0.6248,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.444635510804305,
+      "learning_rate": 2.0502431118314425e-06,
+      "loss": 0.3356,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.796305971290941,
+      "learning_rate": 2.0583468395461914e-06,
+      "loss": 0.3475,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0134207422490125,
+      "learning_rate": 2.0664505672609403e-06,
+      "loss": 0.4231,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.00616156351016,
+      "learning_rate": 2.074554294975689e-06,
+      "loss": 0.2695,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.01987624079023,
+      "learning_rate": 2.0826580226904377e-06,
+      "loss": 0.3253,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.74773314625552,
+      "learning_rate": 2.0907617504051866e-06,
+      "loss": 0.3756,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.725369337639468,
+      "learning_rate": 2.098865478119935e-06,
+      "loss": 0.2862,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2472083132681355,
+      "learning_rate": 2.106969205834684e-06,
+      "loss": 0.2528,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.136047712298029,
+      "learning_rate": 2.115072933549433e-06,
+      "loss": 0.5901,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.399599047295365,
+      "learning_rate": 2.1231766612641817e-06,
+      "loss": 0.4386,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.837237912650036,
+      "learning_rate": 2.1312803889789306e-06,
+      "loss": 0.3955,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.023467081009023,
+      "learning_rate": 2.1393841166936795e-06,
+      "loss": 0.3947,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8287699475092882,
+      "learning_rate": 2.147487844408428e-06,
+      "loss": 0.3811,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.515032129100183,
+      "learning_rate": 2.155591572123177e-06,
+      "loss": 0.538,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4771510965377512,
+      "learning_rate": 2.1636952998379258e-06,
+      "loss": 0.3436,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.24916814844127,
+      "learning_rate": 2.1717990275526743e-06,
+      "loss": 0.4186,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.705078012389009,
+      "learning_rate": 2.179902755267423e-06,
+      "loss": 0.3725,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.0279092067983395,
+      "learning_rate": 2.188006482982172e-06,
+      "loss": 0.211,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8220714251484176,
+      "learning_rate": 2.1961102106969205e-06,
+      "loss": 0.2113,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1091556471700343,
+      "learning_rate": 2.2042139384116694e-06,
+      "loss": 0.2247,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9836819812724045,
+      "learning_rate": 2.2123176661264183e-06,
+      "loss": 0.3268,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.206188039943177,
+      "learning_rate": 2.220421393841167e-06,
+      "loss": 0.3125,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.163719629618975,
+      "learning_rate": 2.228525121555916e-06,
+      "loss": 0.4495,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.546800507000177,
+      "learning_rate": 2.2366288492706646e-06,
+      "loss": 0.1593,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4685132057540145,
+      "learning_rate": 2.2447325769854134e-06,
+      "loss": 0.2596,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.503215760034495,
+      "learning_rate": 2.2528363047001623e-06,
+      "loss": 0.6969,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.406137482070885,
+      "learning_rate": 2.2609400324149112e-06,
+      "loss": 0.5497,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4927999825554927,
+      "learning_rate": 2.2690437601296597e-06,
+      "loss": 0.3005,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8751115569986276,
+      "learning_rate": 2.2771474878444086e-06,
+      "loss": 0.3937,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.186053533521565,
+      "learning_rate": 2.2852512155591575e-06,
+      "loss": 0.2728,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.117902912478666,
+      "learning_rate": 2.293354943273906e-06,
+      "loss": 0.4806,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.281885303574855,
+      "learning_rate": 2.301458670988655e-06,
+      "loss": 0.4677,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.34093117003832,
+      "learning_rate": 2.3095623987034037e-06,
+      "loss": 0.1987,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8679933065279126,
+      "learning_rate": 2.3176661264181522e-06,
+      "loss": 0.3156,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9004521688314893,
+      "learning_rate": 2.325769854132901e-06,
+      "loss": 0.2209,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.822880730039958,
+      "learning_rate": 2.33387358184765e-06,
+      "loss": 0.4889,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.657720764478753,
+      "learning_rate": 2.341977309562399e-06,
+      "loss": 0.222,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.388805095610131,
+      "learning_rate": 2.350081037277148e-06,
+      "loss": 0.2183,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9318500116178305,
+      "learning_rate": 2.3581847649918967e-06,
+      "loss": 0.2251,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4982263123939714,
+      "learning_rate": 2.366288492706645e-06,
+      "loss": 0.2403,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7314166237897086,
+      "learning_rate": 2.374392220421394e-06,
+      "loss": 0.201,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.218583037851503,
+      "learning_rate": 2.382495948136143e-06,
+      "loss": 0.5251,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.389115705199984,
+      "learning_rate": 2.3905996758508914e-06,
+      "loss": 0.3762,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2652680051321923,
+      "learning_rate": 2.3987034035656403e-06,
+      "loss": 0.3577,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.164171151073541,
+      "learning_rate": 2.406807131280389e-06,
+      "loss": 0.2132,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.762362131019772,
+      "learning_rate": 2.4149108589951377e-06,
+      "loss": 0.2762,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4974788031982125,
+      "learning_rate": 2.4230145867098866e-06,
+      "loss": 0.4058,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.8716413225334065,
+      "learning_rate": 2.4311183144246355e-06,
+      "loss": 0.5806,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0458740742764583,
+      "learning_rate": 2.4392220421393843e-06,
+      "loss": 0.196,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2146892765426287,
+      "learning_rate": 2.4473257698541332e-06,
+      "loss": 0.4029,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9251508064212532,
+      "learning_rate": 2.455429497568882e-06,
+      "loss": 0.3829,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.768450013679361,
+      "learning_rate": 2.4635332252836306e-06,
+      "loss": 0.5603,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3396391688815634,
+      "learning_rate": 2.4716369529983795e-06,
+      "loss": 0.502,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.162992621491299,
+      "learning_rate": 2.4797406807131284e-06,
+      "loss": 0.4213,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0345189371748504,
+      "learning_rate": 2.487844408427877e-06,
+      "loss": 0.2135,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.088796580421523,
+      "learning_rate": 2.4959481361426258e-06,
+      "loss": 0.3592,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.280284957661093,
+      "learning_rate": 2.5040518638573747e-06,
+      "loss": 0.3005,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4528291325233718,
+      "learning_rate": 2.5121555915721235e-06,
+      "loss": 0.5666,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.736721945316034,
+      "learning_rate": 2.5202593192868724e-06,
+      "loss": 0.6276,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.221111625642141,
+      "learning_rate": 2.528363047001621e-06,
+      "loss": 0.2796,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3039393358257683,
+      "learning_rate": 2.53646677471637e-06,
+      "loss": 0.2925,
+      "step": 313
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7851977030918147,
+      "learning_rate": 2.5445705024311183e-06,
+      "loss": 0.2315,
+      "step": 314
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.230434751502032,
+      "learning_rate": 2.552674230145867e-06,
+      "loss": 0.252,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9986105830366054,
+      "learning_rate": 2.560777957860616e-06,
+      "loss": 0.6968,
+      "step": 316
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1794709907142895,
+      "learning_rate": 2.568881685575365e-06,
+      "loss": 0.6751,
+      "step": 317
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2965269763814216,
+      "learning_rate": 2.5769854132901134e-06,
+      "loss": 0.358,
+      "step": 318
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.622536956919449,
+      "learning_rate": 2.5850891410048623e-06,
+      "loss": 0.2125,
+      "step": 319
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.138558234369301,
+      "learning_rate": 2.593192868719611e-06,
+      "loss": 0.537,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8100551575375383,
+      "learning_rate": 2.60129659643436e-06,
+      "loss": 0.3457,
+      "step": 321
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.834178463095216,
+      "learning_rate": 2.609400324149109e-06,
+      "loss": 0.3966,
+      "step": 322
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.558971883772141,
+      "learning_rate": 2.617504051863858e-06,
+      "loss": 0.4151,
+      "step": 323
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.380100470393718,
+      "learning_rate": 2.625607779578606e-06,
+      "loss": 0.3136,
+      "step": 324
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.687424977020075,
+      "learning_rate": 2.633711507293355e-06,
+      "loss": 0.4632,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.324577022821094,
+      "learning_rate": 2.6418152350081037e-06,
+      "loss": 0.183,
+      "step": 326
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0576279417679366,
+      "learning_rate": 2.6499189627228526e-06,
+      "loss": 0.2099,
+      "step": 327
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8976918621245558,
+      "learning_rate": 2.6580226904376015e-06,
+      "loss": 0.2853,
+      "step": 328
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.246465867771732,
+      "learning_rate": 2.6661264181523504e-06,
+      "loss": 0.2575,
+      "step": 329
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.753459748128563,
+      "learning_rate": 2.674230145867099e-06,
+      "loss": 0.3342,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9186808736687904,
+      "learning_rate": 2.6823338735818478e-06,
+      "loss": 0.4194,
+      "step": 331
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0223243663583648,
+      "learning_rate": 2.6904376012965967e-06,
+      "loss": 0.2214,
+      "step": 332
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3326429407094635,
+      "learning_rate": 2.6985413290113456e-06,
+      "loss": 0.2289,
+      "step": 333
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3931561639808443,
+      "learning_rate": 2.7066450567260944e-06,
+      "loss": 0.3163,
+      "step": 334
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0900970818104945,
+      "learning_rate": 2.7147487844408433e-06,
+      "loss": 0.2497,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.161266241768239,
+      "learning_rate": 2.7228525121555914e-06,
+      "loss": 0.2255,
+      "step": 336
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8943476239027746,
+      "learning_rate": 2.7309562398703403e-06,
+      "loss": 0.275,
+      "step": 337
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.489119683403253,
+      "learning_rate": 2.739059967585089e-06,
+      "loss": 0.4278,
+      "step": 338
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.033847403893962,
+      "learning_rate": 2.747163695299838e-06,
+      "loss": 0.3493,
+      "step": 339
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.806879032126168,
+      "learning_rate": 2.755267423014587e-06,
+      "loss": 0.3527,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.400048376628959,
+      "learning_rate": 2.763371150729336e-06,
+      "loss": 0.285,
+      "step": 341
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.125559413430035,
+      "learning_rate": 2.7714748784440843e-06,
+      "loss": 0.204,
+      "step": 342
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8499638952073334,
+      "learning_rate": 2.7795786061588332e-06,
+      "loss": 0.4274,
+      "step": 343
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1179596750026493,
+      "learning_rate": 2.787682333873582e-06,
+      "loss": 0.4245,
+      "step": 344
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9994288275140115,
+      "learning_rate": 2.795786061588331e-06,
+      "loss": 0.5032,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.832171690516967,
+      "learning_rate": 2.80388978930308e-06,
+      "loss": 0.4339,
+      "step": 346
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3242388812563695,
+      "learning_rate": 2.811993517017829e-06,
+      "loss": 0.2102,
+      "step": 347
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4286250891887944,
+      "learning_rate": 2.820097244732577e-06,
+      "loss": 0.1647,
+      "step": 348
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1111463526731735,
+      "learning_rate": 2.8282009724473257e-06,
+      "loss": 0.1962,
+      "step": 349
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2311965694473974,
+      "learning_rate": 2.8363047001620746e-06,
+      "loss": 0.5766,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4913919368893773,
+      "learning_rate": 2.8444084278768235e-06,
+      "loss": 0.2245,
+      "step": 351
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.388943835268522,
+      "learning_rate": 2.8525121555915724e-06,
+      "loss": 0.3877,
+      "step": 352
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.780160690610224,
+      "learning_rate": 2.8606158833063213e-06,
+      "loss": 0.4552,
+      "step": 353
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.423649033562793,
+      "learning_rate": 2.8687196110210698e-06,
+      "loss": 0.2083,
+      "step": 354
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.058066516515318,
+      "learning_rate": 2.8768233387358187e-06,
+      "loss": 0.4841,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3670673216152793,
+      "learning_rate": 2.8849270664505676e-06,
+      "loss": 0.4322,
+      "step": 356
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.318558412024612,
+      "learning_rate": 2.8930307941653165e-06,
+      "loss": 0.2765,
+      "step": 357
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2402997139659973,
+      "learning_rate": 2.9011345218800654e-06,
+      "loss": 0.4608,
+      "step": 358
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5190337170384596,
+      "learning_rate": 2.9092382495948142e-06,
+      "loss": 0.4505,
+      "step": 359
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.588428405096416,
+      "learning_rate": 2.9173419773095623e-06,
+      "loss": 0.5295,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7995881641729676,
+      "learning_rate": 2.925445705024311e-06,
+      "loss": 0.5465,
+      "step": 361
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8235204316094586,
+      "learning_rate": 2.93354943273906e-06,
+      "loss": 0.382,
+      "step": 362
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0105528201208713,
+      "learning_rate": 2.941653160453809e-06,
+      "loss": 0.3312,
+      "step": 363
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.728039338819667,
+      "learning_rate": 2.949756888168558e-06,
+      "loss": 0.4312,
+      "step": 364
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5219989963570337,
+      "learning_rate": 2.9578606158833068e-06,
+      "loss": 0.2183,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3985278830133177,
+      "learning_rate": 2.9659643435980552e-06,
+      "loss": 0.3815,
+      "step": 366
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2518459176385774,
+      "learning_rate": 2.974068071312804e-06,
+      "loss": 0.4626,
+      "step": 367
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.195252676779203,
+      "learning_rate": 2.982171799027553e-06,
+      "loss": 0.5451,
+      "step": 368
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.094311441667543,
+      "learning_rate": 2.990275526742302e-06,
+      "loss": 0.1666,
+      "step": 369
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.08564041252838,
+      "learning_rate": 2.9983792544570504e-06,
+      "loss": 0.4042,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.740344523189361,
+      "learning_rate": 3.0064829821717993e-06,
+      "loss": 0.3349,
+      "step": 371
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4595214160344296,
+      "learning_rate": 3.014586709886548e-06,
+      "loss": 0.2873,
+      "step": 372
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8288593076589525,
+      "learning_rate": 3.0226904376012966e-06,
+      "loss": 0.2327,
+      "step": 373
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1013152554837404,
+      "learning_rate": 3.0307941653160455e-06,
+      "loss": 0.3297,
+      "step": 374
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1573291859819688,
+      "learning_rate": 3.0388978930307944e-06,
+      "loss": 0.1694,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7736294778517205,
+      "learning_rate": 3.0470016207455433e-06,
+      "loss": 0.5156,
+      "step": 376
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.961375604190276,
+      "learning_rate": 3.0551053484602922e-06,
+      "loss": 0.5137,
+      "step": 377
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9441724032761756,
+      "learning_rate": 3.063209076175041e-06,
+      "loss": 0.3772,
+      "step": 378
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.390295355457491,
+      "learning_rate": 3.071312803889789e-06,
+      "loss": 0.3505,
+      "step": 379
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.703143188174364,
+      "learning_rate": 3.079416531604538e-06,
+      "loss": 0.6546,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7684254880223214,
+      "learning_rate": 3.087520259319287e-06,
+      "loss": 0.3858,
+      "step": 381
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7562713476453506,
+      "learning_rate": 3.095623987034036e-06,
+      "loss": 0.5216,
+      "step": 382
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.42321412995785,
+      "learning_rate": 3.1037277147487847e-06,
+      "loss": 0.2008,
+      "step": 383
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.975012726433493,
+      "learning_rate": 3.1118314424635336e-06,
+      "loss": 0.3343,
+      "step": 384
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9485163376400654,
+      "learning_rate": 3.119935170178282e-06,
+      "loss": 0.3031,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.067262993931404,
+      "learning_rate": 3.128038897893031e-06,
+      "loss": 0.2694,
+      "step": 386
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4202655489697626,
+      "learning_rate": 3.13614262560778e-06,
+      "loss": 0.2805,
+      "step": 387
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.73293226054898,
+      "learning_rate": 3.1442463533225288e-06,
+      "loss": 0.4378,
+      "step": 388
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.72730481118057,
+      "learning_rate": 3.1523500810372777e-06,
+      "loss": 0.3602,
+      "step": 389
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0459987456305986,
+      "learning_rate": 3.1604538087520266e-06,
+      "loss": 0.2796,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.72726153837423,
+      "learning_rate": 3.1685575364667746e-06,
+      "loss": 0.4599,
+      "step": 391
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.025793256247598,
+      "learning_rate": 3.1766612641815235e-06,
+      "loss": 0.2482,
+      "step": 392
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.549905245706064,
+      "learning_rate": 3.1847649918962724e-06,
+      "loss": 0.5942,
+      "step": 393
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4690565025388396,
+      "learning_rate": 3.1928687196110213e-06,
+      "loss": 0.2548,
+      "step": 394
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.272018288842226,
+      "learning_rate": 3.20097244732577e-06,
+      "loss": 0.3043,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7466640612750366,
+      "learning_rate": 3.209076175040519e-06,
+      "loss": 0.4959,
+      "step": 396
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9522605316958863,
+      "learning_rate": 3.2171799027552675e-06,
+      "loss": 0.3508,
+      "step": 397
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.260052576343187,
+      "learning_rate": 3.2252836304700164e-06,
+      "loss": 0.372,
+      "step": 398
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4769919591680005,
+      "learning_rate": 3.2333873581847653e-06,
+      "loss": 0.3513,
+      "step": 399
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0127291042219553,
+      "learning_rate": 3.2414910858995142e-06,
+      "loss": 0.439,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9218366704640766,
+      "learning_rate": 3.249594813614263e-06,
+      "loss": 0.3876,
+      "step": 401
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0213147382701284,
+      "learning_rate": 3.257698541329012e-06,
+      "loss": 0.4804,
+      "step": 402
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.646893764862711,
+      "learning_rate": 3.26580226904376e-06,
+      "loss": 0.4021,
+      "step": 403
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6226902859240315,
+      "learning_rate": 3.273905996758509e-06,
+      "loss": 0.2131,
+      "step": 404
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6562523336961257,
+      "learning_rate": 3.282009724473258e-06,
+      "loss": 0.3167,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5162965695562476,
+      "learning_rate": 3.2901134521880067e-06,
+      "loss": 0.3349,
+      "step": 406
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.817844593767384,
+      "learning_rate": 3.2982171799027556e-06,
+      "loss": 0.3582,
+      "step": 407
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8218197797101627,
+      "learning_rate": 3.3063209076175045e-06,
+      "loss": 0.399,
+      "step": 408
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.949392073320397,
+      "learning_rate": 3.314424635332253e-06,
+      "loss": 0.2369,
+      "step": 409
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.113674794670577,
+      "learning_rate": 3.322528363047002e-06,
+      "loss": 0.3102,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8760333899176596,
+      "learning_rate": 3.3306320907617508e-06,
+      "loss": 0.2544,
+      "step": 411
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.588507353026148,
+      "learning_rate": 3.3387358184764997e-06,
+      "loss": 0.3684,
+      "step": 412
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1742383204726954,
+      "learning_rate": 3.3468395461912486e-06,
+      "loss": 0.4798,
+      "step": 413
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8293543807306247,
+      "learning_rate": 3.3549432739059975e-06,
+      "loss": 0.3592,
+      "step": 414
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.65946738668728,
+      "learning_rate": 3.3630470016207455e-06,
+      "loss": 0.1867,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.27701588946181,
+      "learning_rate": 3.3711507293354944e-06,
+      "loss": 0.2821,
+      "step": 416
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9103486645408054,
+      "learning_rate": 3.3792544570502433e-06,
+      "loss": 0.5294,
+      "step": 417
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.008318609172654,
+      "learning_rate": 3.387358184764992e-06,
+      "loss": 0.4693,
+      "step": 418
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4073697283918385,
+      "learning_rate": 3.395461912479741e-06,
+      "loss": 0.403,
+      "step": 419
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.405028780480866,
+      "learning_rate": 3.40356564019449e-06,
+      "loss": 0.2485,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.872347043067743,
+      "learning_rate": 3.4116693679092384e-06,
+      "loss": 0.2437,
+      "step": 421
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.553908948211753,
+      "learning_rate": 3.4197730956239873e-06,
+      "loss": 0.697,
+      "step": 422
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.600400535104447,
+      "learning_rate": 3.4278768233387362e-06,
+      "loss": 0.3451,
+      "step": 423
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8134611606672864,
+      "learning_rate": 3.435980551053485e-06,
+      "loss": 0.1872,
+      "step": 424
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.024415313947084,
+      "learning_rate": 3.444084278768234e-06,
+      "loss": 0.363,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.487410927620263,
+      "learning_rate": 3.4521880064829825e-06,
+      "loss": 0.178,
+      "step": 426
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.186574372230303,
+      "learning_rate": 3.460291734197731e-06,
+      "loss": 0.1922,
+      "step": 427
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7921332903595744,
+      "learning_rate": 3.46839546191248e-06,
+      "loss": 0.4834,
+      "step": 428
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.311185553542691,
+      "learning_rate": 3.4764991896272287e-06,
+      "loss": 0.4904,
+      "step": 429
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.607005786702596,
+      "learning_rate": 3.4846029173419776e-06,
+      "loss": 0.2336,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4071699527532493,
+      "learning_rate": 3.4927066450567265e-06,
+      "loss": 0.4472,
+      "step": 431
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.710122109509036,
+      "learning_rate": 3.5008103727714754e-06,
+      "loss": 0.2993,
+      "step": 432
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2063866761718804,
+      "learning_rate": 3.508914100486224e-06,
+      "loss": 0.3627,
+      "step": 433
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2041683451660945,
+      "learning_rate": 3.517017828200973e-06,
+      "loss": 0.3592,
+      "step": 434
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8389713386763473,
+      "learning_rate": 3.5251215559157213e-06,
+      "loss": 0.271,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.638824132087506,
+      "learning_rate": 3.53322528363047e-06,
+      "loss": 0.4662,
+      "step": 436
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0127066391000867,
+      "learning_rate": 3.541329011345219e-06,
+      "loss": 0.1943,
+      "step": 437
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.372192844619086,
+      "learning_rate": 3.549432739059968e-06,
+      "loss": 0.2611,
+      "step": 438
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8830931963094804,
+      "learning_rate": 3.5575364667747164e-06,
+      "loss": 0.3677,
+      "step": 439
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.250761676446119,
+      "learning_rate": 3.5656401944894653e-06,
+      "loss": 0.5013,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9243081892401452,
+      "learning_rate": 3.573743922204214e-06,
+      "loss": 0.1874,
+      "step": 441
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2411001381321256,
+      "learning_rate": 3.581847649918963e-06,
+      "loss": 0.5127,
+      "step": 442
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.678853954981224,
+      "learning_rate": 3.589951377633712e-06,
+      "loss": 0.2608,
+      "step": 443
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.408348043930421,
+      "learning_rate": 3.598055105348461e-06,
+      "loss": 0.1868,
+      "step": 444
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.764139412632151,
+      "learning_rate": 3.606158833063209e-06,
+      "loss": 0.3675,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7760734591844756,
+      "learning_rate": 3.614262560777958e-06,
+      "loss": 0.7052,
+      "step": 446
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.359019433654961,
+      "learning_rate": 3.6223662884927067e-06,
+      "loss": 0.2483,
+      "step": 447
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0769398145495517,
+      "learning_rate": 3.6304700162074556e-06,
+      "loss": 0.4619,
+      "step": 448
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3543365753372596,
+      "learning_rate": 3.6385737439222045e-06,
+      "loss": 0.375,
+      "step": 449
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9543263867716574,
+      "learning_rate": 3.6466774716369534e-06,
+      "loss": 0.6276,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4877746145766015,
+      "learning_rate": 3.654781199351702e-06,
+      "loss": 0.6027,
+      "step": 451
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.639185563917071,
+      "learning_rate": 3.6628849270664508e-06,
+      "loss": 0.354,
+      "step": 452
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.952070279968592,
+      "learning_rate": 3.6709886547811997e-06,
+      "loss": 0.4734,
+      "step": 453
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.514207250910059,
+      "learning_rate": 3.6790923824959485e-06,
+      "loss": 0.4809,
+      "step": 454
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6544402240278338,
+      "learning_rate": 3.6871961102106974e-06,
+      "loss": 0.24,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.999255644324676,
+      "learning_rate": 3.6952998379254463e-06,
+      "loss": 0.2901,
+      "step": 456
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.504206515966085,
+      "learning_rate": 3.7034035656401944e-06,
+      "loss": 0.2267,
+      "step": 457
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.540587300960981,
+      "learning_rate": 3.7115072933549433e-06,
+      "loss": 0.5099,
+      "step": 458
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8901493402854155,
+      "learning_rate": 3.719611021069692e-06,
+      "loss": 0.4128,
+      "step": 459
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.215073021565053,
+      "learning_rate": 3.727714748784441e-06,
+      "loss": 0.2843,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1671747042713156,
+      "learning_rate": 3.73581847649919e-06,
+      "loss": 0.4137,
+      "step": 461
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8509567743915034,
+      "learning_rate": 3.743922204213939e-06,
+      "loss": 0.2527,
+      "step": 462
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4768283013734584,
+      "learning_rate": 3.7520259319286873e-06,
+      "loss": 0.3885,
+      "step": 463
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.112019228443134,
+      "learning_rate": 3.7601296596434362e-06,
+      "loss": 0.2349,
+      "step": 464
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.354416723388108,
+      "learning_rate": 3.768233387358185e-06,
+      "loss": 0.3929,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7492947757894286,
+      "learning_rate": 3.776337115072934e-06,
+      "loss": 0.3283,
+      "step": 466
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2400602968810044,
+      "learning_rate": 3.784440842787683e-06,
+      "loss": 0.3049,
+      "step": 467
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8167312780405687,
+      "learning_rate": 3.7925445705024318e-06,
+      "loss": 0.6593,
+      "step": 468
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.781515663839843,
+      "learning_rate": 3.80064829821718e-06,
+      "loss": 0.3884,
+      "step": 469
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2508265447378855,
+      "learning_rate": 3.8087520259319287e-06,
+      "loss": 0.1851,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8443941646539184,
+      "learning_rate": 3.816855753646678e-06,
+      "loss": 0.5586,
+      "step": 471
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9509582745185288,
+      "learning_rate": 3.824959481361427e-06,
+      "loss": 0.5023,
+      "step": 472
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9957027851641254,
+      "learning_rate": 3.833063209076176e-06,
+      "loss": 0.3774,
+      "step": 473
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.10368605014338,
+      "learning_rate": 3.841166936790925e-06,
+      "loss": 0.3423,
+      "step": 474
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.732927560537868,
+      "learning_rate": 3.849270664505673e-06,
+      "loss": 0.2924,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.287400901954352,
+      "learning_rate": 3.857374392220422e-06,
+      "loss": 0.5323,
+      "step": 476
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.410516090672485,
+      "learning_rate": 3.8654781199351706e-06,
+      "loss": 0.2308,
+      "step": 477
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7117805567344537,
+      "learning_rate": 3.8735818476499195e-06,
+      "loss": 0.523,
+      "step": 478
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7551075905528077,
+      "learning_rate": 3.881685575364668e-06,
+      "loss": 0.3131,
+      "step": 479
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2758155688128676,
+      "learning_rate": 3.889789303079417e-06,
+      "loss": 0.3656,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.88832142820442,
+      "learning_rate": 3.897893030794165e-06,
+      "loss": 0.2541,
+      "step": 481
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.123790270405033,
+      "learning_rate": 3.905996758508914e-06,
+      "loss": 0.463,
+      "step": 482
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7843761662439026,
+      "learning_rate": 3.914100486223663e-06,
+      "loss": 0.4105,
+      "step": 483
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.156926228730932,
+      "learning_rate": 3.922204213938412e-06,
+      "loss": 0.2373,
+      "step": 484
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1099756226243507,
+      "learning_rate": 3.930307941653161e-06,
+      "loss": 0.4951,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9666957795386413,
+      "learning_rate": 3.93841166936791e-06,
+      "loss": 0.7663,
+      "step": 486
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5229289666939985,
+      "learning_rate": 3.946515397082658e-06,
+      "loss": 0.4272,
+      "step": 487
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.628341251325886,
+      "learning_rate": 3.954619124797407e-06,
+      "loss": 0.3933,
+      "step": 488
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8201525106515692,
+      "learning_rate": 3.962722852512156e-06,
+      "loss": 0.4732,
+      "step": 489
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.903828065646083,
+      "learning_rate": 3.9708265802269045e-06,
+      "loss": 0.3674,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.210408272099592,
+      "learning_rate": 3.978930307941653e-06,
+      "loss": 0.2416,
+      "step": 491
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2813844510524808,
+      "learning_rate": 3.987034035656402e-06,
+      "loss": 0.185,
+      "step": 492
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6641404218742082,
+      "learning_rate": 3.995137763371151e-06,
+      "loss": 0.1681,
+      "step": 493
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5290864480577822,
+      "learning_rate": 4.0032414910859e-06,
+      "loss": 0.1743,
+      "step": 494
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4935724362518323,
+      "learning_rate": 4.011345218800649e-06,
+      "loss": 0.2287,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.264880248687236,
+      "learning_rate": 4.019448946515397e-06,
+      "loss": 0.3204,
+      "step": 496
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8557772489334883,
+      "learning_rate": 4.027552674230146e-06,
+      "loss": 0.4682,
+      "step": 497
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9703013726085852,
+      "learning_rate": 4.035656401944895e-06,
+      "loss": 0.1945,
+      "step": 498
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.199546163756787,
+      "learning_rate": 4.043760129659644e-06,
+      "loss": 0.3373,
+      "step": 499
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4719955622550183,
+      "learning_rate": 4.0518638573743926e-06,
+      "loss": 0.3677,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6474652151356763,
+      "learning_rate": 4.0599675850891415e-06,
+      "loss": 0.2232,
+      "step": 501
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4992516351213494,
+      "learning_rate": 4.06807131280389e-06,
+      "loss": 0.1977,
+      "step": 502
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9137438417871757,
+      "learning_rate": 4.076175040518639e-06,
+      "loss": 0.3626,
+      "step": 503
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.33170336801591,
+      "learning_rate": 4.084278768233388e-06,
+      "loss": 0.2327,
+      "step": 504
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.032018229872622,
+      "learning_rate": 4.092382495948137e-06,
+      "loss": 0.2592,
+      "step": 505
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8034857824086665,
+      "learning_rate": 4.100486223662885e-06,
+      "loss": 0.2995,
+      "step": 506
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1276861176569355,
+      "learning_rate": 4.108589951377634e-06,
+      "loss": 0.1706,
+      "step": 507
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.177877149791776,
+      "learning_rate": 4.116693679092383e-06,
+      "loss": 0.3582,
+      "step": 508
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2341323934238604,
+      "learning_rate": 4.124797406807132e-06,
+      "loss": 0.2337,
+      "step": 509
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3408180462790837,
+      "learning_rate": 4.132901134521881e-06,
+      "loss": 0.2413,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1327889267945417,
+      "learning_rate": 4.1410048622366296e-06,
+      "loss": 0.4498,
+      "step": 511
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2897961583262805,
+      "learning_rate": 4.149108589951378e-06,
+      "loss": 0.1479,
+      "step": 512
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4228371462533373,
+      "learning_rate": 4.1572123176661265e-06,
+      "loss": 0.3192,
+      "step": 513
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9985014671465144,
+      "learning_rate": 4.165316045380875e-06,
+      "loss": 0.3875,
+      "step": 514
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.41030521657357,
+      "learning_rate": 4.173419773095624e-06,
+      "loss": 0.2445,
+      "step": 515
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5037262018697266,
+      "learning_rate": 4.181523500810373e-06,
+      "loss": 0.5278,
+      "step": 516
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6454158949719369,
+      "learning_rate": 4.189627228525122e-06,
+      "loss": 0.1627,
+      "step": 517
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2736051056773117,
+      "learning_rate": 4.19773095623987e-06,
+      "loss": 0.3952,
+      "step": 518
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.465566344826595,
+      "learning_rate": 4.205834683954619e-06,
+      "loss": 0.3652,
+      "step": 519
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.648873228615363,
+      "learning_rate": 4.213938411669368e-06,
+      "loss": 0.2079,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7485572264720872,
+      "learning_rate": 4.222042139384117e-06,
+      "loss": 0.4876,
+      "step": 521
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9409370909424966,
+      "learning_rate": 4.230145867098866e-06,
+      "loss": 0.4361,
+      "step": 522
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.614372333558994,
+      "learning_rate": 4.238249594813615e-06,
+      "loss": 0.3246,
+      "step": 523
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4283733239729086,
+      "learning_rate": 4.2463533225283635e-06,
+      "loss": 0.5551,
+      "step": 524
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.960814138776678,
+      "learning_rate": 4.254457050243112e-06,
+      "loss": 0.3532,
+      "step": 525
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3250383561295447,
+      "learning_rate": 4.262560777957861e-06,
+      "loss": 0.4839,
+      "step": 526
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.984563731671893,
+      "learning_rate": 4.27066450567261e-06,
+      "loss": 0.2843,
+      "step": 527
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5111359313112387,
+      "learning_rate": 4.278768233387359e-06,
+      "loss": 0.3877,
+      "step": 528
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.504715331703049,
+      "learning_rate": 4.286871961102108e-06,
+      "loss": 0.4723,
+      "step": 529
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.375993176309867,
+      "learning_rate": 4.294975688816856e-06,
+      "loss": 0.6268,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8439301548997884,
+      "learning_rate": 4.303079416531605e-06,
+      "loss": 0.2911,
+      "step": 531
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.49832241277575,
+      "learning_rate": 4.311183144246354e-06,
+      "loss": 0.3482,
+      "step": 532
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.435487815306054,
+      "learning_rate": 4.319286871961103e-06,
+      "loss": 0.3593,
+      "step": 533
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7130223153697903,
+      "learning_rate": 4.3273905996758516e-06,
+      "loss": 0.4881,
+      "step": 534
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.65381226034885,
+      "learning_rate": 4.3354943273906005e-06,
+      "loss": 0.4552,
+      "step": 535
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.670848560041882,
+      "learning_rate": 4.3435980551053485e-06,
+      "loss": 0.3405,
+      "step": 536
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0827937566899886,
+      "learning_rate": 4.351701782820097e-06,
+      "loss": 0.5192,
+      "step": 537
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4268386496453216,
+      "learning_rate": 4.359805510534846e-06,
+      "loss": 0.3886,
+      "step": 538
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5222765750298093,
+      "learning_rate": 4.367909238249595e-06,
+      "loss": 0.6651,
+      "step": 539
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.719331536635033,
+      "learning_rate": 4.376012965964344e-06,
+      "loss": 0.3744,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.651415160923608,
+      "learning_rate": 4.384116693679093e-06,
+      "loss": 0.4273,
+      "step": 541
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4179866942888335,
+      "learning_rate": 4.392220421393841e-06,
+      "loss": 0.6255,
+      "step": 542
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7708502723539383,
+      "learning_rate": 4.40032414910859e-06,
+      "loss": 0.5403,
+      "step": 543
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9149006719662336,
+      "learning_rate": 4.408427876823339e-06,
+      "loss": 0.5139,
+      "step": 544
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5645172622458174,
+      "learning_rate": 4.416531604538088e-06,
+      "loss": 0.255,
+      "step": 545
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9538947520099104,
+      "learning_rate": 4.424635332252837e-06,
+      "loss": 0.5165,
+      "step": 546
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.416686003715565,
+      "learning_rate": 4.4327390599675855e-06,
+      "loss": 0.5741,
+      "step": 547
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3109987257248696,
+      "learning_rate": 4.440842787682334e-06,
+      "loss": 0.3904,
+      "step": 548
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1590888850987455,
+      "learning_rate": 4.448946515397083e-06,
+      "loss": 0.2584,
+      "step": 549
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6886727746099908,
+      "learning_rate": 4.457050243111832e-06,
+      "loss": 0.3423,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5672677276575784,
+      "learning_rate": 4.465153970826581e-06,
+      "loss": 0.4376,
+      "step": 551
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.093685248782853,
+      "learning_rate": 4.473257698541329e-06,
+      "loss": 0.2356,
+      "step": 552
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.706044070754052,
+      "learning_rate": 4.481361426256078e-06,
+      "loss": 0.3684,
+      "step": 553
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.291291775528102,
+      "learning_rate": 4.489465153970827e-06,
+      "loss": 0.2078,
+      "step": 554
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.648329425958528,
+      "learning_rate": 4.497568881685576e-06,
+      "loss": 0.4356,
+      "step": 555
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1033469462791743,
+      "learning_rate": 4.505672609400325e-06,
+      "loss": 0.2443,
+      "step": 556
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5158682877318426,
+      "learning_rate": 4.5137763371150736e-06,
+      "loss": 0.3257,
+      "step": 557
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5728149992747973,
+      "learning_rate": 4.5218800648298225e-06,
+      "loss": 0.5935,
+      "step": 558
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.718730270001918,
+      "learning_rate": 4.529983792544571e-06,
+      "loss": 0.3666,
+      "step": 559
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1262458627201934,
+      "learning_rate": 4.538087520259319e-06,
+      "loss": 0.3724,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1861010454752936,
+      "learning_rate": 4.546191247974068e-06,
+      "loss": 0.446,
+      "step": 561
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8164090012366287,
+      "learning_rate": 4.554294975688817e-06,
+      "loss": 0.3995,
+      "step": 562
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.018624587840169,
+      "learning_rate": 4.562398703403566e-06,
+      "loss": 0.4642,
+      "step": 563
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.371932170437454,
+      "learning_rate": 4.570502431118315e-06,
+      "loss": 0.4886,
+      "step": 564
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.360499934429888,
+      "learning_rate": 4.578606158833064e-06,
+      "loss": 0.5507,
+      "step": 565
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0872179972090916,
+      "learning_rate": 4.586709886547812e-06,
+      "loss": 0.5251,
+      "step": 566
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1943948225974714,
+      "learning_rate": 4.594813614262561e-06,
+      "loss": 0.2188,
+      "step": 567
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.437434195583917,
+      "learning_rate": 4.60291734197731e-06,
+      "loss": 0.4324,
+      "step": 568
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9314720923708797,
+      "learning_rate": 4.611021069692059e-06,
+      "loss": 0.1887,
+      "step": 569
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.229963168494108,
+      "learning_rate": 4.6191247974068075e-06,
+      "loss": 0.2647,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7494631481673544,
+      "learning_rate": 4.627228525121556e-06,
+      "loss": 0.3028,
+      "step": 571
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7428982783483535,
+      "learning_rate": 4.6353322528363044e-06,
+      "loss": 0.543,
+      "step": 572
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7656064868027186,
+      "learning_rate": 4.643435980551053e-06,
+      "loss": 0.3324,
+      "step": 573
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7498623078536755,
+      "learning_rate": 4.651539708265802e-06,
+      "loss": 0.1546,
+      "step": 574
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5547267549164467,
+      "learning_rate": 4.659643435980551e-06,
+      "loss": 0.3789,
+      "step": 575
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.268389409679173,
+      "learning_rate": 4.6677471636953e-06,
+      "loss": 0.1723,
+      "step": 576
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.251752206821709,
+      "learning_rate": 4.675850891410049e-06,
+      "loss": 0.5007,
+      "step": 577
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0963082839651834,
+      "learning_rate": 4.683954619124798e-06,
+      "loss": 0.2986,
+      "step": 578
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6108712071829245,
+      "learning_rate": 4.692058346839547e-06,
+      "loss": 0.34,
+      "step": 579
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0093630877864017,
+      "learning_rate": 4.700162074554296e-06,
+      "loss": 0.1717,
+      "step": 580
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.078758329503341,
+      "learning_rate": 4.7082658022690445e-06,
+      "loss": 0.2186,
+      "step": 581
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3173252984497252,
+      "learning_rate": 4.716369529983793e-06,
+      "loss": 0.5755,
+      "step": 582
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.739964583504613,
+      "learning_rate": 4.724473257698542e-06,
+      "loss": 0.2306,
+      "step": 583
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3680832933632923,
+      "learning_rate": 4.73257698541329e-06,
+      "loss": 0.4007,
+      "step": 584
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.146053800359869,
+      "learning_rate": 4.740680713128039e-06,
+      "loss": 0.8948,
+      "step": 585
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.356962103634448,
+      "learning_rate": 4.748784440842788e-06,
+      "loss": 0.2435,
+      "step": 586
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.33923980427795,
+      "learning_rate": 4.756888168557537e-06,
+      "loss": 0.5803,
+      "step": 587
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9918682066912592,
+      "learning_rate": 4.764991896272286e-06,
+      "loss": 0.221,
+      "step": 588
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.814522135116481,
+      "learning_rate": 4.773095623987035e-06,
+      "loss": 0.5565,
+      "step": 589
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.785774338929917,
+      "learning_rate": 4.781199351701783e-06,
+      "loss": 0.3839,
+      "step": 590
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6496875029873324,
+      "learning_rate": 4.789303079416532e-06,
+      "loss": 0.2543,
+      "step": 591
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5490133400923787,
+      "learning_rate": 4.797406807131281e-06,
+      "loss": 0.4136,
+      "step": 592
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0430922085320717,
+      "learning_rate": 4.8055105348460295e-06,
+      "loss": 0.4052,
+      "step": 593
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.121880984122331,
+      "learning_rate": 4.813614262560778e-06,
+      "loss": 0.168,
+      "step": 594
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.993886539953185,
+      "learning_rate": 4.821717990275527e-06,
+      "loss": 0.3982,
+      "step": 595
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3964259164608475,
+      "learning_rate": 4.829821717990275e-06,
+      "loss": 0.203,
+      "step": 596
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5950944643164202,
+      "learning_rate": 4.837925445705024e-06,
+      "loss": 0.2014,
+      "step": 597
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.021705486349326,
+      "learning_rate": 4.846029173419773e-06,
+      "loss": 0.1927,
+      "step": 598
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.151938883935647,
+      "learning_rate": 4.854132901134522e-06,
+      "loss": 0.2512,
+      "step": 599
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5442200730152904,
+      "learning_rate": 4.862236628849271e-06,
+      "loss": 0.271,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.152068115436321,
+      "learning_rate": 4.87034035656402e-06,
+      "loss": 0.3668,
+      "step": 601
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.343870738416362,
+      "learning_rate": 4.878444084278769e-06,
+      "loss": 0.2421,
+      "step": 602
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1594241317481413,
+      "learning_rate": 4.886547811993518e-06,
+      "loss": 0.2971,
+      "step": 603
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.76853326424161,
+      "learning_rate": 4.8946515397082665e-06,
+      "loss": 0.1949,
+      "step": 604
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.491456537465888,
+      "learning_rate": 4.902755267423015e-06,
+      "loss": 0.2933,
+      "step": 605
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0346324854393933,
+      "learning_rate": 4.910858995137764e-06,
+      "loss": 0.1705,
+      "step": 606
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3955095635081296,
+      "learning_rate": 4.918962722852513e-06,
+      "loss": 0.2517,
+      "step": 607
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4952104187793736,
+      "learning_rate": 4.927066450567261e-06,
+      "loss": 0.4059,
+      "step": 608
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.854497929768589,
+      "learning_rate": 4.93517017828201e-06,
+      "loss": 0.3992,
+      "step": 609
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3761296283729627,
+      "learning_rate": 4.943273905996759e-06,
+      "loss": 0.2492,
+      "step": 610
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.938562552574866,
+      "learning_rate": 4.951377633711508e-06,
+      "loss": 0.3773,
+      "step": 611
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6522450042456063,
+      "learning_rate": 4.959481361426257e-06,
+      "loss": 0.2225,
+      "step": 612
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4859124051115575,
+      "learning_rate": 4.967585089141006e-06,
+      "loss": 0.3239,
+      "step": 613
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.659087555343668,
+      "learning_rate": 4.975688816855754e-06,
+      "loss": 0.6588,
+      "step": 614
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4475545529758227,
+      "learning_rate": 4.983792544570503e-06,
+      "loss": 0.6981,
+      "step": 615
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.858428141677766,
+      "learning_rate": 4.9918962722852515e-06,
+      "loss": 0.4206,
+      "step": 616
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.560730264877578,
+      "learning_rate": 5e-06,
+      "loss": 0.4133,
+      "step": 617
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9596023269512581,
+      "learning_rate": 5.008103727714749e-06,
+      "loss": 0.1955,
+      "step": 618
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2239676764929976,
+      "learning_rate": 5.016207455429498e-06,
+      "loss": 0.1872,
+      "step": 619
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4176534222162447,
+      "learning_rate": 5.024311183144247e-06,
+      "loss": 0.2139,
+      "step": 620
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7600406046607744,
+      "learning_rate": 5.032414910858996e-06,
+      "loss": 0.1672,
+      "step": 621
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1999420819461455,
+      "learning_rate": 5.040518638573745e-06,
+      "loss": 0.3548,
+      "step": 622
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5526817508985165,
+      "learning_rate": 5.048622366288494e-06,
+      "loss": 0.245,
+      "step": 623
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.559514481373303,
+      "learning_rate": 5.056726094003242e-06,
+      "loss": 0.3866,
+      "step": 624
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6059088385337335,
+      "learning_rate": 5.064829821717991e-06,
+      "loss": 0.5229,
+      "step": 625
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3486053465190877,
+      "learning_rate": 5.07293354943274e-06,
+      "loss": 0.3343,
+      "step": 626
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.54006548511764,
+      "learning_rate": 5.081037277147488e-06,
+      "loss": 0.4424,
+      "step": 627
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0225130311159485,
+      "learning_rate": 5.0891410048622365e-06,
+      "loss": 0.246,
+      "step": 628
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5936613297530324,
+      "learning_rate": 5.0972447325769854e-06,
+      "loss": 0.3054,
+      "step": 629
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6071926573394055,
+      "learning_rate": 5.105348460291734e-06,
+      "loss": 0.3433,
+      "step": 630
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.347351143254067,
+      "learning_rate": 5.113452188006483e-06,
+      "loss": 0.2339,
+      "step": 631
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6442991461949634,
+      "learning_rate": 5.121555915721232e-06,
+      "loss": 0.2867,
+      "step": 632
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.574907688226903,
+      "learning_rate": 5.129659643435981e-06,
+      "loss": 0.44,
+      "step": 633
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0552871047325563,
+      "learning_rate": 5.13776337115073e-06,
+      "loss": 0.4243,
+      "step": 634
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.050627123172664,
+      "learning_rate": 5.145867098865479e-06,
+      "loss": 0.4193,
+      "step": 635
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1917673417955337,
+      "learning_rate": 5.153970826580227e-06,
+      "loss": 0.4859,
+      "step": 636
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.067365216525827,
+      "learning_rate": 5.162074554294976e-06,
+      "loss": 0.1407,
+      "step": 637
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.5381839200686835,
+      "learning_rate": 5.170178282009725e-06,
+      "loss": 0.2075,
+      "step": 638
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.464115113716842,
+      "learning_rate": 5.1782820097244735e-06,
+      "loss": 0.2148,
+      "step": 639
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4717443103382437,
+      "learning_rate": 5.186385737439222e-06,
+      "loss": 0.3969,
+      "step": 640
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.613780728035926,
+      "learning_rate": 5.194489465153971e-06,
+      "loss": 0.2749,
+      "step": 641
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7407810196047333,
+      "learning_rate": 5.20259319286872e-06,
+      "loss": 0.1555,
+      "step": 642
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1411571716060105,
+      "learning_rate": 5.210696920583469e-06,
+      "loss": 0.5545,
+      "step": 643
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2530066081113915,
+      "learning_rate": 5.218800648298218e-06,
+      "loss": 0.3401,
+      "step": 644
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.577285647308283,
+      "learning_rate": 5.226904376012967e-06,
+      "loss": 0.1789,
+      "step": 645
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.154883013590168,
+      "learning_rate": 5.235008103727716e-06,
+      "loss": 0.4004,
+      "step": 646
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2188414299622643,
+      "learning_rate": 5.243111831442465e-06,
+      "loss": 0.1699,
+      "step": 647
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3612041898376015,
+      "learning_rate": 5.251215559157212e-06,
+      "loss": 0.3946,
+      "step": 648
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8226102097233055,
+      "learning_rate": 5.259319286871961e-06,
+      "loss": 0.2851,
+      "step": 649
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.38207869466705,
+      "learning_rate": 5.26742301458671e-06,
+      "loss": 0.46,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0854426119939355,
+      "learning_rate": 5.2755267423014586e-06,
+      "loss": 0.4769,
+      "step": 651
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3811763832976967,
+      "learning_rate": 5.2836304700162074e-06,
+      "loss": 0.5724,
+      "step": 652
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.003808078228203,
+      "learning_rate": 5.291734197730956e-06,
+      "loss": 0.2956,
+      "step": 653
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8128048095835325,
+      "learning_rate": 5.299837925445705e-06,
+      "loss": 0.537,
+      "step": 654
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.847555443924769,
+      "learning_rate": 5.307941653160454e-06,
+      "loss": 0.2582,
+      "step": 655
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9108431831023576,
+      "learning_rate": 5.316045380875203e-06,
+      "loss": 0.5077,
+      "step": 656
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9517390865334994,
+      "learning_rate": 5.324149108589952e-06,
+      "loss": 0.2421,
+      "step": 657
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2431417508665867,
+      "learning_rate": 5.332252836304701e-06,
+      "loss": 0.4173,
+      "step": 658
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8777617996791793,
+      "learning_rate": 5.34035656401945e-06,
+      "loss": 0.3322,
+      "step": 659
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1742677886619433,
+      "learning_rate": 5.348460291734198e-06,
+      "loss": 0.2269,
+      "step": 660
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.035425373375686,
+      "learning_rate": 5.356564019448947e-06,
+      "loss": 0.2168,
+      "step": 661
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.986666952687346,
+      "learning_rate": 5.3646677471636955e-06,
+      "loss": 0.5926,
+      "step": 662
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6929300255089714,
+      "learning_rate": 5.3727714748784444e-06,
+      "loss": 0.3049,
+      "step": 663
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7362058100833975,
+      "learning_rate": 5.380875202593193e-06,
+      "loss": 0.1774,
+      "step": 664
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.433811098437835,
+      "learning_rate": 5.388978930307942e-06,
+      "loss": 0.1898,
+      "step": 665
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.762856040208506,
+      "learning_rate": 5.397082658022691e-06,
+      "loss": 0.2457,
+      "step": 666
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1600999738595488,
+      "learning_rate": 5.40518638573744e-06,
+      "loss": 0.3393,
+      "step": 667
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.282820897770217,
+      "learning_rate": 5.413290113452189e-06,
+      "loss": 0.4359,
+      "step": 668
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.57608744833594,
+      "learning_rate": 5.421393841166938e-06,
+      "loss": 0.1973,
+      "step": 669
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0028830423841693,
+      "learning_rate": 5.429497568881687e-06,
+      "loss": 0.3562,
+      "step": 670
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3497820662310227,
+      "learning_rate": 5.437601296596436e-06,
+      "loss": 0.76,
+      "step": 671
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.022068255772714,
+      "learning_rate": 5.445705024311183e-06,
+      "loss": 0.4124,
+      "step": 672
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8594897773511256,
+      "learning_rate": 5.453808752025932e-06,
+      "loss": 0.4131,
+      "step": 673
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.581507850496131,
+      "learning_rate": 5.4619124797406806e-06,
+      "loss": 0.4892,
+      "step": 674
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8844757244340338,
+      "learning_rate": 5.4700162074554295e-06,
+      "loss": 0.4981,
+      "step": 675
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.910539066814659,
+      "learning_rate": 5.478119935170178e-06,
+      "loss": 0.4077,
+      "step": 676
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2836799482637646,
+      "learning_rate": 5.486223662884927e-06,
+      "loss": 0.3491,
+      "step": 677
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.098822310436266,
+      "learning_rate": 5.494327390599676e-06,
+      "loss": 0.2274,
+      "step": 678
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.482241716307865,
+      "learning_rate": 5.502431118314425e-06,
+      "loss": 0.389,
+      "step": 679
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6557127184800207,
+      "learning_rate": 5.510534846029174e-06,
+      "loss": 0.3869,
+      "step": 680
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0592254595955124,
+      "learning_rate": 5.518638573743923e-06,
+      "loss": 0.2762,
+      "step": 681
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.42288606519937,
+      "learning_rate": 5.526742301458672e-06,
+      "loss": 0.3944,
+      "step": 682
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.251767391542139,
+      "learning_rate": 5.534846029173421e-06,
+      "loss": 0.2905,
+      "step": 683
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8832976677740036,
+      "learning_rate": 5.542949756888169e-06,
+      "loss": 0.2263,
+      "step": 684
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.909288509980412,
+      "learning_rate": 5.5510534846029175e-06,
+      "loss": 0.5254,
+      "step": 685
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.938835165265496,
+      "learning_rate": 5.5591572123176664e-06,
+      "loss": 0.4834,
+      "step": 686
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1777869384990867,
+      "learning_rate": 5.567260940032415e-06,
+      "loss": 0.1882,
+      "step": 687
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7935077422232548,
+      "learning_rate": 5.575364667747164e-06,
+      "loss": 0.2018,
+      "step": 688
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2869342635931624,
+      "learning_rate": 5.583468395461913e-06,
+      "loss": 0.1505,
+      "step": 689
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.026956330107726,
+      "learning_rate": 5.591572123176662e-06,
+      "loss": 0.2057,
+      "step": 690
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.267165956649647,
+      "learning_rate": 5.599675850891411e-06,
+      "loss": 0.6126,
+      "step": 691
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9685302502464395,
+      "learning_rate": 5.60777957860616e-06,
+      "loss": 0.3592,
+      "step": 692
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3148741391492056,
+      "learning_rate": 5.615883306320909e-06,
+      "loss": 0.2828,
+      "step": 693
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.930792962934235,
+      "learning_rate": 5.623987034035658e-06,
+      "loss": 0.1624,
+      "step": 694
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9746340137910334,
+      "learning_rate": 5.6320907617504065e-06,
+      "loss": 0.2176,
+      "step": 695
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.537779240206486,
+      "learning_rate": 5.640194489465154e-06,
+      "loss": 0.4024,
+      "step": 696
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2416004947700694,
+      "learning_rate": 5.6482982171799026e-06,
+      "loss": 0.3169,
+      "step": 697
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.199542616310568,
+      "learning_rate": 5.6564019448946515e-06,
+      "loss": 0.2797,
+      "step": 698
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0363542016540284,
+      "learning_rate": 5.6645056726094e-06,
+      "loss": 0.3177,
+      "step": 699
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2298908016477186,
+      "learning_rate": 5.672609400324149e-06,
+      "loss": 0.3667,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.357857259289626,
+      "learning_rate": 5.680713128038898e-06,
+      "loss": 0.415,
+      "step": 701
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.143269652443511,
+      "learning_rate": 5.688816855753647e-06,
+      "loss": 0.2883,
+      "step": 702
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.861263193689194,
+      "learning_rate": 5.696920583468396e-06,
+      "loss": 0.6982,
+      "step": 703
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3908957346782302,
+      "learning_rate": 5.705024311183145e-06,
+      "loss": 0.2555,
+      "step": 704
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8065294823977336,
+      "learning_rate": 5.713128038897894e-06,
+      "loss": 0.7304,
+      "step": 705
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5818408521827154,
+      "learning_rate": 5.721231766612643e-06,
+      "loss": 0.2248,
+      "step": 706
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.584269328419012,
+      "learning_rate": 5.7293354943273915e-06,
+      "loss": 0.2753,
+      "step": 707
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.512406601122236,
+      "learning_rate": 5.7374392220421396e-06,
+      "loss": 0.4526,
+      "step": 708
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.443841752939086,
+      "learning_rate": 5.7455429497568885e-06,
+      "loss": 0.201,
+      "step": 709
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2936359065313834,
+      "learning_rate": 5.753646677471637e-06,
+      "loss": 0.3701,
+      "step": 710
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.227727745724546,
+      "learning_rate": 5.761750405186386e-06,
+      "loss": 0.3791,
+      "step": 711
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2920147212464475,
+      "learning_rate": 5.769854132901135e-06,
+      "loss": 0.3949,
+      "step": 712
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6114961155105747,
+      "learning_rate": 5.777957860615884e-06,
+      "loss": 0.136,
+      "step": 713
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.100635870584026,
+      "learning_rate": 5.786061588330633e-06,
+      "loss": 0.6404,
+      "step": 714
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8831695518832234,
+      "learning_rate": 5.794165316045382e-06,
+      "loss": 0.2069,
+      "step": 715
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4547882394497744,
+      "learning_rate": 5.802269043760131e-06,
+      "loss": 0.3979,
+      "step": 716
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2288580664974518,
+      "learning_rate": 5.81037277147488e-06,
+      "loss": 0.6922,
+      "step": 717
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.30178043838276,
+      "learning_rate": 5.8184764991896285e-06,
+      "loss": 0.2361,
+      "step": 718
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8089913947556533,
+      "learning_rate": 5.826580226904377e-06,
+      "loss": 0.4019,
+      "step": 719
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.25550897031889,
+      "learning_rate": 5.834683954619125e-06,
+      "loss": 0.3837,
+      "step": 720
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2962003352040585,
+      "learning_rate": 5.8427876823338735e-06,
+      "loss": 0.4686,
+      "step": 721
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1195971441097066,
+      "learning_rate": 5.850891410048622e-06,
+      "loss": 0.2159,
+      "step": 722
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.651635548760485,
+      "learning_rate": 5.858995137763371e-06,
+      "loss": 0.382,
+      "step": 723
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8871529424262956,
+      "learning_rate": 5.86709886547812e-06,
+      "loss": 0.4015,
+      "step": 724
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.345284811386342,
+      "learning_rate": 5.875202593192869e-06,
+      "loss": 0.3111,
+      "step": 725
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.27518097398726,
+      "learning_rate": 5.883306320907618e-06,
+      "loss": 0.3227,
+      "step": 726
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.471795919240853,
+      "learning_rate": 5.891410048622367e-06,
+      "loss": 0.3755,
+      "step": 727
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.763799770887062,
+      "learning_rate": 5.899513776337116e-06,
+      "loss": 0.3024,
+      "step": 728
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.279044139745737,
+      "learning_rate": 5.907617504051865e-06,
+      "loss": 0.2352,
+      "step": 729
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2658373634519533,
+      "learning_rate": 5.9157212317666135e-06,
+      "loss": 0.2825,
+      "step": 730
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3514632904315027,
+      "learning_rate": 5.923824959481362e-06,
+      "loss": 0.4119,
+      "step": 731
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.329172462073205,
+      "learning_rate": 5.9319286871961105e-06,
+      "loss": 0.3401,
+      "step": 732
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.511138593909308,
+      "learning_rate": 5.940032414910859e-06,
+      "loss": 0.4517,
+      "step": 733
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4359516428450636,
+      "learning_rate": 5.948136142625608e-06,
+      "loss": 0.303,
+      "step": 734
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0290350061219793,
+      "learning_rate": 5.956239870340357e-06,
+      "loss": 0.3255,
+      "step": 735
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5436039739726402,
+      "learning_rate": 5.964343598055106e-06,
+      "loss": 0.2499,
+      "step": 736
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.043300125824239,
+      "learning_rate": 5.972447325769855e-06,
+      "loss": 0.5583,
+      "step": 737
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0333983705934844,
+      "learning_rate": 5.980551053484604e-06,
+      "loss": 0.2786,
+      "step": 738
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.260639025186152,
+      "learning_rate": 5.988654781199352e-06,
+      "loss": 0.3097,
+      "step": 739
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4868659121837937,
+      "learning_rate": 5.996758508914101e-06,
+      "loss": 0.3738,
+      "step": 740
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7968715056095146,
+      "learning_rate": 6.00486223662885e-06,
+      "loss": 0.2555,
+      "step": 741
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0483050119487043,
+      "learning_rate": 6.0129659643435986e-06,
+      "loss": 0.1917,
+      "step": 742
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.004068788179222,
+      "learning_rate": 6.0210696920583474e-06,
+      "loss": 0.2463,
+      "step": 743
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7534355843554335,
+      "learning_rate": 6.029173419773096e-06,
+      "loss": 0.2388,
+      "step": 744
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7452878407047696,
+      "learning_rate": 6.037277147487844e-06,
+      "loss": 0.4083,
+      "step": 745
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.137671149095801,
+      "learning_rate": 6.045380875202593e-06,
+      "loss": 0.3996,
+      "step": 746
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.747768648478345,
+      "learning_rate": 6.053484602917342e-06,
+      "loss": 0.2807,
+      "step": 747
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.659041979386983,
+      "learning_rate": 6.061588330632091e-06,
+      "loss": 0.4805,
+      "step": 748
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0662413145767284,
+      "learning_rate": 6.06969205834684e-06,
+      "loss": 0.3015,
+      "step": 749
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5207889740388256,
+      "learning_rate": 6.077795786061589e-06,
+      "loss": 0.2887,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6235415396887547,
+      "learning_rate": 6.085899513776338e-06,
+      "loss": 0.2072,
+      "step": 751
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.058097781419256,
+      "learning_rate": 6.094003241491087e-06,
+      "loss": 0.2534,
+      "step": 752
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.539926451035221,
+      "learning_rate": 6.1021069692058355e-06,
+      "loss": 0.2027,
+      "step": 753
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1404704887427073,
+      "learning_rate": 6.1102106969205844e-06,
+      "loss": 0.3638,
+      "step": 754
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.580231936103944,
+      "learning_rate": 6.118314424635333e-06,
+      "loss": 0.1974,
+      "step": 755
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.233135960229326,
+      "learning_rate": 6.126418152350082e-06,
+      "loss": 0.5416,
+      "step": 756
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.921339281714927,
+      "learning_rate": 6.13452188006483e-06,
+      "loss": 0.3844,
+      "step": 757
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5023001818000834,
+      "learning_rate": 6.142625607779578e-06,
+      "loss": 0.3532,
+      "step": 758
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6442674416184953,
+      "learning_rate": 6.150729335494327e-06,
+      "loss": 0.1799,
+      "step": 759
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0630710851540432,
+      "learning_rate": 6.158833063209076e-06,
+      "loss": 0.3067,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.479217586208125,
+      "learning_rate": 6.166936790923825e-06,
+      "loss": 0.465,
+      "step": 761
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4193978562010807,
+      "learning_rate": 6.175040518638574e-06,
+      "loss": 0.2556,
+      "step": 762
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2507584803395098,
+      "learning_rate": 6.183144246353323e-06,
+      "loss": 0.3753,
+      "step": 763
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7968789745281257,
+      "learning_rate": 6.191247974068072e-06,
+      "loss": 0.4201,
+      "step": 764
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4066751246989466,
+      "learning_rate": 6.1993517017828206e-06,
+      "loss": 0.3435,
+      "step": 765
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.313305933446064,
+      "learning_rate": 6.2074554294975695e-06,
+      "loss": 0.3945,
+      "step": 766
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4482489886030934,
+      "learning_rate": 6.215559157212318e-06,
+      "loss": 0.1964,
+      "step": 767
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5190737405026224,
+      "learning_rate": 6.223662884927067e-06,
+      "loss": 0.3512,
+      "step": 768
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.474008452593766,
+      "learning_rate": 6.231766612641815e-06,
+      "loss": 0.4829,
+      "step": 769
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3416105805555167,
+      "learning_rate": 6.239870340356564e-06,
+      "loss": 0.4228,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4515570774832884,
+      "learning_rate": 6.247974068071313e-06,
+      "loss": 0.2901,
+      "step": 771
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4698797369262424,
+      "learning_rate": 6.256077795786062e-06,
+      "loss": 0.4551,
+      "step": 772
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.875526525133319,
+      "learning_rate": 6.264181523500811e-06,
+      "loss": 0.4591,
+      "step": 773
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5343738635868482,
+      "learning_rate": 6.27228525121556e-06,
+      "loss": 0.5331,
+      "step": 774
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0927791035307863,
+      "learning_rate": 6.280388978930309e-06,
+      "loss": 0.3474,
+      "step": 775
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6987302231356713,
+      "learning_rate": 6.2884927066450575e-06,
+      "loss": 0.164,
+      "step": 776
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.040155264109265,
+      "learning_rate": 6.2965964343598064e-06,
+      "loss": 0.2289,
+      "step": 777
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7909145721914213,
+      "learning_rate": 6.304700162074555e-06,
+      "loss": 0.2154,
+      "step": 778
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9817744803371644,
+      "learning_rate": 6.312803889789304e-06,
+      "loss": 0.4226,
+      "step": 779
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7497704644895684,
+      "learning_rate": 6.320907617504053e-06,
+      "loss": 0.1529,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6347134079335872,
+      "learning_rate": 6.3290113452188e-06,
+      "loss": 0.1915,
+      "step": 781
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.82278708950005,
+      "learning_rate": 6.337115072933549e-06,
+      "loss": 0.4268,
+      "step": 782
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1950551421792497,
+      "learning_rate": 6.345218800648298e-06,
+      "loss": 0.4038,
+      "step": 783
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7312166938671427,
+      "learning_rate": 6.353322528363047e-06,
+      "loss": 0.3049,
+      "step": 784
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5865310379627564,
+      "learning_rate": 6.361426256077796e-06,
+      "loss": 0.5255,
+      "step": 785
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.755013880270662,
+      "learning_rate": 6.369529983792545e-06,
+      "loss": 0.2647,
+      "step": 786
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.004724733240294,
+      "learning_rate": 6.377633711507294e-06,
+      "loss": 0.2326,
+      "step": 787
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0259155842465004,
+      "learning_rate": 6.3857374392220426e-06,
+      "loss": 0.6588,
+      "step": 788
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5263337353178894,
+      "learning_rate": 6.3938411669367915e-06,
+      "loss": 0.3191,
+      "step": 789
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9176578192003944,
+      "learning_rate": 6.40194489465154e-06,
+      "loss": 0.2384,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.894961367631007,
+      "learning_rate": 6.410048622366289e-06,
+      "loss": 0.6247,
+      "step": 791
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4775209467567265,
+      "learning_rate": 6.418152350081038e-06,
+      "loss": 0.4561,
+      "step": 792
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7802359039485096,
+      "learning_rate": 6.426256077795786e-06,
+      "loss": 0.3332,
+      "step": 793
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.706149135057844,
+      "learning_rate": 6.434359805510535e-06,
+      "loss": 0.624,
+      "step": 794
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8066429823067147,
+      "learning_rate": 6.442463533225284e-06,
+      "loss": 0.301,
+      "step": 795
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7936077960730334,
+      "learning_rate": 6.450567260940033e-06,
+      "loss": 0.4281,
+      "step": 796
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2710793512109033,
+      "learning_rate": 6.458670988654782e-06,
+      "loss": 0.4301,
+      "step": 797
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0248968833995056,
+      "learning_rate": 6.466774716369531e-06,
+      "loss": 0.298,
+      "step": 798
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5991678343000557,
+      "learning_rate": 6.4748784440842796e-06,
+      "loss": 0.2879,
+      "step": 799
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4453903161485573,
+      "learning_rate": 6.4829821717990284e-06,
+      "loss": 0.2903,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.365828372573444,
+      "learning_rate": 6.491085899513777e-06,
+      "loss": 0.2778,
+      "step": 801
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3378006193575307,
+      "learning_rate": 6.499189627228526e-06,
+      "loss": 0.4345,
+      "step": 802
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4612205510179943,
+      "learning_rate": 6.507293354943275e-06,
+      "loss": 0.261,
+      "step": 803
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.827835273655188,
+      "learning_rate": 6.515397082658024e-06,
+      "loss": 0.284,
+      "step": 804
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.578166313996331,
+      "learning_rate": 6.523500810372771e-06,
+      "loss": 0.493,
+      "step": 805
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.476946467047892,
+      "learning_rate": 6.53160453808752e-06,
+      "loss": 0.4302,
+      "step": 806
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8715557412028507,
+      "learning_rate": 6.539708265802269e-06,
+      "loss": 0.3573,
+      "step": 807
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4083413473466724,
+      "learning_rate": 6.547811993517018e-06,
+      "loss": 0.3387,
+      "step": 808
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.380305419015888,
+      "learning_rate": 6.555915721231767e-06,
+      "loss": 0.2529,
+      "step": 809
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.271540712427097,
+      "learning_rate": 6.564019448946516e-06,
+      "loss": 0.3588,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3838927775023118,
+      "learning_rate": 6.572123176661265e-06,
+      "loss": 0.37,
+      "step": 811
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.243565332097389,
+      "learning_rate": 6.5802269043760135e-06,
+      "loss": 0.3168,
+      "step": 812
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0222146964641756,
+      "learning_rate": 6.588330632090762e-06,
+      "loss": 0.4367,
+      "step": 813
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.218784345441596,
+      "learning_rate": 6.596434359805511e-06,
+      "loss": 0.1964,
+      "step": 814
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.747302127334084,
+      "learning_rate": 6.60453808752026e-06,
+      "loss": 0.6813,
+      "step": 815
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8521965262607805,
+      "learning_rate": 6.612641815235009e-06,
+      "loss": 0.5183,
+      "step": 816
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.313335061727679,
+      "learning_rate": 6.620745542949757e-06,
+      "loss": 0.3505,
+      "step": 817
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8866852486076944,
+      "learning_rate": 6.628849270664506e-06,
+      "loss": 0.3114,
+      "step": 818
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.660296572262611,
+      "learning_rate": 6.636952998379255e-06,
+      "loss": 0.2861,
+      "step": 819
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2540177130946684,
+      "learning_rate": 6.645056726094004e-06,
+      "loss": 0.2711,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.129262589896115,
+      "learning_rate": 6.653160453808753e-06,
+      "loss": 0.2786,
+      "step": 821
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.584950347982462,
+      "learning_rate": 6.6612641815235016e-06,
+      "loss": 0.3551,
+      "step": 822
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5102267159580591,
+      "learning_rate": 6.6693679092382505e-06,
+      "loss": 0.1475,
+      "step": 823
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.018221356571632,
+      "learning_rate": 6.677471636952999e-06,
+      "loss": 0.2393,
+      "step": 824
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2054420218231496,
+      "learning_rate": 6.685575364667748e-06,
+      "loss": 0.2775,
+      "step": 825
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4607543756077326,
+      "learning_rate": 6.693679092382497e-06,
+      "loss": 0.4972,
+      "step": 826
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7177549711772704,
+      "learning_rate": 6.701782820097246e-06,
+      "loss": 0.3475,
+      "step": 827
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9341606984072826,
+      "learning_rate": 6.709886547811995e-06,
+      "loss": 0.2833,
+      "step": 828
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9451201776626512,
+      "learning_rate": 6.717990275526742e-06,
+      "loss": 0.2115,
+      "step": 829
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1304698976262535,
+      "learning_rate": 6.726094003241491e-06,
+      "loss": 0.229,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0887477389402727,
+      "learning_rate": 6.73419773095624e-06,
+      "loss": 0.2266,
+      "step": 831
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8094161822848767,
+      "learning_rate": 6.742301458670989e-06,
+      "loss": 0.2343,
+      "step": 832
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9350554561899664,
+      "learning_rate": 6.750405186385738e-06,
+      "loss": 0.5763,
+      "step": 833
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.605034753435821,
+      "learning_rate": 6.758508914100487e-06,
+      "loss": 0.3622,
+      "step": 834
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7267836519725792,
+      "learning_rate": 6.7666126418152355e-06,
+      "loss": 0.1722,
+      "step": 835
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.113711893374038,
+      "learning_rate": 6.774716369529984e-06,
+      "loss": 0.182,
+      "step": 836
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9145332069047625,
+      "learning_rate": 6.782820097244733e-06,
+      "loss": 0.3003,
+      "step": 837
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8934465508276948,
+      "learning_rate": 6.790923824959482e-06,
+      "loss": 0.2794,
+      "step": 838
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6653446233412317,
+      "learning_rate": 6.799027552674231e-06,
+      "loss": 0.3795,
+      "step": 839
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8371869603944788,
+      "learning_rate": 6.80713128038898e-06,
+      "loss": 0.28,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.360048678959445,
+      "learning_rate": 6.815235008103728e-06,
+      "loss": 0.5062,
+      "step": 841
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8124916182499073,
+      "learning_rate": 6.823338735818477e-06,
+      "loss": 0.4052,
+      "step": 842
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1905422908485184,
+      "learning_rate": 6.831442463533226e-06,
+      "loss": 0.123,
+      "step": 843
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4377775890059423,
+      "learning_rate": 6.839546191247975e-06,
+      "loss": 0.5716,
+      "step": 844
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.031581349303942,
+      "learning_rate": 6.8476499189627236e-06,
+      "loss": 0.5667,
+      "step": 845
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.238552537697188,
+      "learning_rate": 6.8557536466774725e-06,
+      "loss": 0.4277,
+      "step": 846
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5212662866243205,
+      "learning_rate": 6.863857374392221e-06,
+      "loss": 0.4999,
+      "step": 847
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7585324572029137,
+      "learning_rate": 6.87196110210697e-06,
+      "loss": 0.1443,
+      "step": 848
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0597666307255538,
+      "learning_rate": 6.880064829821719e-06,
+      "loss": 0.2001,
+      "step": 849
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.378211146936045,
+      "learning_rate": 6.888168557536468e-06,
+      "loss": 0.3557,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.512466005076358,
+      "learning_rate": 6.896272285251216e-06,
+      "loss": 0.3731,
+      "step": 851
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6997760909450257,
+      "learning_rate": 6.904376012965965e-06,
+      "loss": 0.5291,
+      "step": 852
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.343251906555452,
+      "learning_rate": 6.912479740680713e-06,
+      "loss": 0.2447,
+      "step": 853
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.035460626901196,
+      "learning_rate": 6.920583468395462e-06,
+      "loss": 0.1742,
+      "step": 854
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4814674469532223,
+      "learning_rate": 6.928687196110211e-06,
+      "loss": 0.3488,
+      "step": 855
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.464390567013009,
+      "learning_rate": 6.93679092382496e-06,
+      "loss": 0.4243,
+      "step": 856
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7633208329976653,
+      "learning_rate": 6.944894651539709e-06,
+      "loss": 0.1484,
+      "step": 857
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.253823898635695,
+      "learning_rate": 6.9529983792544575e-06,
+      "loss": 0.3792,
+      "step": 858
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.909001078890013,
+      "learning_rate": 6.961102106969206e-06,
+      "loss": 0.3685,
+      "step": 859
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2776039938646546,
+      "learning_rate": 6.969205834683955e-06,
+      "loss": 0.3622,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4696547142187857,
+      "learning_rate": 6.977309562398704e-06,
+      "loss": 0.3656,
+      "step": 861
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0106243564046173,
+      "learning_rate": 6.985413290113453e-06,
+      "loss": 0.2131,
+      "step": 862
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.788638400576771,
+      "learning_rate": 6.993517017828202e-06,
+      "loss": 0.2794,
+      "step": 863
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.47581913895901,
+      "learning_rate": 7.001620745542951e-06,
+      "loss": 0.265,
+      "step": 864
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1158904082410572,
+      "learning_rate": 7.0097244732577e-06,
+      "loss": 0.3237,
+      "step": 865
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.938362863891212,
+      "learning_rate": 7.017828200972448e-06,
+      "loss": 0.1878,
+      "step": 866
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9272134777869394,
+      "learning_rate": 7.025931928687197e-06,
+      "loss": 0.6449,
+      "step": 867
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9181504950098416,
+      "learning_rate": 7.034035656401946e-06,
+      "loss": 0.1444,
+      "step": 868
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5021323646825415,
+      "learning_rate": 7.0421393841166945e-06,
+      "loss": 0.3378,
+      "step": 869
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3280753548371416,
+      "learning_rate": 7.0502431118314425e-06,
+      "loss": 0.4191,
+      "step": 870
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4005184030937814,
+      "learning_rate": 7.058346839546191e-06,
+      "loss": 0.1957,
+      "step": 871
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3601043975801232,
+      "learning_rate": 7.06645056726094e-06,
+      "loss": 0.4233,
+      "step": 872
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2674255882727143,
+      "learning_rate": 7.074554294975689e-06,
+      "loss": 0.4909,
+      "step": 873
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6054630079306222,
+      "learning_rate": 7.082658022690438e-06,
+      "loss": 0.2554,
+      "step": 874
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.348058288475402,
+      "learning_rate": 7.090761750405187e-06,
+      "loss": 0.3663,
+      "step": 875
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.119560084507046,
+      "learning_rate": 7.098865478119936e-06,
+      "loss": 0.4762,
+      "step": 876
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8403758792728944,
+      "learning_rate": 7.106969205834685e-06,
+      "loss": 0.6885,
+      "step": 877
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6597130696167155,
+      "learning_rate": 7.115072933549433e-06,
+      "loss": 0.4788,
+      "step": 878
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7483566966277317,
+      "learning_rate": 7.123176661264182e-06,
+      "loss": 0.2655,
+      "step": 879
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8576646048532583,
+      "learning_rate": 7.131280388978931e-06,
+      "loss": 0.2244,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4084436895128354,
+      "learning_rate": 7.1393841166936795e-06,
+      "loss": 0.398,
+      "step": 881
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.894577025814168,
+      "learning_rate": 7.147487844408428e-06,
+      "loss": 0.2193,
+      "step": 882
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4802970646592652,
+      "learning_rate": 7.155591572123177e-06,
+      "loss": 0.3858,
+      "step": 883
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7485669894827667,
+      "learning_rate": 7.163695299837926e-06,
+      "loss": 0.5125,
+      "step": 884
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3896141533054465,
+      "learning_rate": 7.171799027552675e-06,
+      "loss": 0.2969,
+      "step": 885
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.782625237443837,
+      "learning_rate": 7.179902755267424e-06,
+      "loss": 0.2077,
+      "step": 886
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.543780121100909,
+      "learning_rate": 7.188006482982173e-06,
+      "loss": 0.2695,
+      "step": 887
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.490854199631467,
+      "learning_rate": 7.196110210696922e-06,
+      "loss": 0.5104,
+      "step": 888
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.464192759397532,
+      "learning_rate": 7.204213938411671e-06,
+      "loss": 0.4217,
+      "step": 889
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0043473831313663,
+      "learning_rate": 7.212317666126418e-06,
+      "loss": 0.3628,
+      "step": 890
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.255380026350978,
+      "learning_rate": 7.220421393841167e-06,
+      "loss": 0.2718,
+      "step": 891
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.915592811899007,
+      "learning_rate": 7.228525121555916e-06,
+      "loss": 0.2676,
+      "step": 892
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.081141548366811,
+      "learning_rate": 7.2366288492706645e-06,
+      "loss": 0.2858,
+      "step": 893
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.295271307132214,
+      "learning_rate": 7.2447325769854134e-06,
+      "loss": 0.4737,
+      "step": 894
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7008922319062463,
+      "learning_rate": 7.252836304700162e-06,
+      "loss": 0.5536,
+      "step": 895
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.914564145761027,
+      "learning_rate": 7.260940032414911e-06,
+      "loss": 0.3701,
+      "step": 896
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0192475922114297,
+      "learning_rate": 7.26904376012966e-06,
+      "loss": 0.3122,
+      "step": 897
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3018391675332643,
+      "learning_rate": 7.277147487844409e-06,
+      "loss": 0.3762,
+      "step": 898
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7883945151570373,
+      "learning_rate": 7.285251215559158e-06,
+      "loss": 0.2147,
+      "step": 899
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.309143426973912,
+      "learning_rate": 7.293354943273907e-06,
+      "loss": 0.2783,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.183621984718953,
+      "learning_rate": 7.301458670988656e-06,
+      "loss": 0.2691,
+      "step": 901
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.541118610942809,
+      "learning_rate": 7.309562398703404e-06,
+      "loss": 0.5362,
+      "step": 902
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.790523582102931,
+      "learning_rate": 7.317666126418153e-06,
+      "loss": 0.2319,
+      "step": 903
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.946573387236735,
+      "learning_rate": 7.3257698541329015e-06,
+      "loss": 0.5628,
+      "step": 904
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4309029651419523,
+      "learning_rate": 7.33387358184765e-06,
+      "loss": 0.2226,
+      "step": 905
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.204638425989733,
+      "learning_rate": 7.341977309562399e-06,
+      "loss": 0.4001,
+      "step": 906
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7287788944939615,
+      "learning_rate": 7.350081037277148e-06,
+      "loss": 0.2788,
+      "step": 907
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9131578365587636,
+      "learning_rate": 7.358184764991897e-06,
+      "loss": 0.5894,
+      "step": 908
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9353118619671412,
+      "learning_rate": 7.366288492706646e-06,
+      "loss": 0.4237,
+      "step": 909
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.079676751467975,
+      "learning_rate": 7.374392220421395e-06,
+      "loss": 0.2723,
+      "step": 910
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9681414934231425,
+      "learning_rate": 7.382495948136144e-06,
+      "loss": 0.5345,
+      "step": 911
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.860879792040135,
+      "learning_rate": 7.390599675850893e-06,
+      "loss": 0.2413,
+      "step": 912
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.308808873411622,
+      "learning_rate": 7.3987034035656416e-06,
+      "loss": 0.4648,
+      "step": 913
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6932408468437532,
+      "learning_rate": 7.406807131280389e-06,
+      "loss": 0.2498,
+      "step": 914
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6344822623209958,
+      "learning_rate": 7.414910858995138e-06,
+      "loss": 0.1871,
+      "step": 915
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9094777530673743,
+      "learning_rate": 7.4230145867098865e-06,
+      "loss": 0.3346,
+      "step": 916
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9175932615814237,
+      "learning_rate": 7.4311183144246354e-06,
+      "loss": 0.5678,
+      "step": 917
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4077277406226767,
+      "learning_rate": 7.439222042139384e-06,
+      "loss": 0.2216,
+      "step": 918
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.155508563228739,
+      "learning_rate": 7.447325769854133e-06,
+      "loss": 0.1553,
+      "step": 919
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4947519770407136,
+      "learning_rate": 7.455429497568882e-06,
+      "loss": 0.3898,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.325499603214447,
+      "learning_rate": 7.463533225283631e-06,
+      "loss": 0.1937,
+      "step": 921
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.073239404357678,
+      "learning_rate": 7.47163695299838e-06,
+      "loss": 0.6479,
+      "step": 922
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.443868649570717,
+      "learning_rate": 7.479740680713129e-06,
+      "loss": 0.3029,
+      "step": 923
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3392345992018897,
+      "learning_rate": 7.487844408427878e-06,
+      "loss": 0.3714,
+      "step": 924
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6994515910657766,
+      "learning_rate": 7.495948136142627e-06,
+      "loss": 0.3167,
+      "step": 925
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3218179051399135,
+      "learning_rate": 7.504051863857375e-06,
+      "loss": 0.3734,
+      "step": 926
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.564851403086989,
+      "learning_rate": 7.5121555915721235e-06,
+      "loss": 0.4653,
+      "step": 927
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9121452958714222,
+      "learning_rate": 7.5202593192868724e-06,
+      "loss": 0.4014,
+      "step": 928
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.511960855921232,
+      "learning_rate": 7.528363047001621e-06,
+      "loss": 0.3557,
+      "step": 929
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.372273574230356,
+      "learning_rate": 7.53646677471637e-06,
+      "loss": 0.4344,
+      "step": 930
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.863604289277082,
+      "learning_rate": 7.544570502431119e-06,
+      "loss": 0.2523,
+      "step": 931
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6564799410071551,
+      "learning_rate": 7.552674230145868e-06,
+      "loss": 0.1716,
+      "step": 932
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4661395419616907,
+      "learning_rate": 7.560777957860617e-06,
+      "loss": 0.1952,
+      "step": 933
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0342117578081718,
+      "learning_rate": 7.568881685575366e-06,
+      "loss": 0.344,
+      "step": 934
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.877258752239337,
+      "learning_rate": 7.576985413290115e-06,
+      "loss": 0.3479,
+      "step": 935
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.638264029471802,
+      "learning_rate": 7.5850891410048636e-06,
+      "loss": 0.2777,
+      "step": 936
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3542215388719456,
+      "learning_rate": 7.5931928687196125e-06,
+      "loss": 0.3716,
+      "step": 937
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5493389918465374,
+      "learning_rate": 7.60129659643436e-06,
+      "loss": 0.5592,
+      "step": 938
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.060680939075737,
+      "learning_rate": 7.6094003241491086e-06,
+      "loss": 0.2463,
+      "step": 939
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.182464895616671,
+      "learning_rate": 7.6175040518638575e-06,
+      "loss": 0.6658,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7490352433867553,
+      "learning_rate": 7.625607779578606e-06,
+      "loss": 0.2069,
+      "step": 941
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1936951989070006,
+      "learning_rate": 7.633711507293356e-06,
+      "loss": 0.3253,
+      "step": 942
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.680875559317055,
+      "learning_rate": 7.641815235008105e-06,
+      "loss": 0.5511,
+      "step": 943
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2212017869283343,
+      "learning_rate": 7.649918962722854e-06,
+      "loss": 0.4194,
+      "step": 944
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.534524142278459,
+      "learning_rate": 7.658022690437603e-06,
+      "loss": 0.1747,
+      "step": 945
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.204612665590637,
+      "learning_rate": 7.666126418152352e-06,
+      "loss": 0.4182,
+      "step": 946
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7274193278199688,
+      "learning_rate": 7.6742301458671e-06,
+      "loss": 0.3872,
+      "step": 947
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5269818172102587,
+      "learning_rate": 7.68233387358185e-06,
+      "loss": 0.3572,
+      "step": 948
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0663233850374207,
+      "learning_rate": 7.690437601296598e-06,
+      "loss": 0.2158,
+      "step": 949
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.230140429491099,
+      "learning_rate": 7.698541329011346e-06,
+      "loss": 0.546,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4030810023473084,
+      "learning_rate": 7.706645056726094e-06,
+      "loss": 0.2438,
+      "step": 951
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.206174795689911,
+      "learning_rate": 7.714748784440843e-06,
+      "loss": 0.3636,
+      "step": 952
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6066931459865725,
+      "learning_rate": 7.722852512155592e-06,
+      "loss": 0.2521,
+      "step": 953
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.119987606885945,
+      "learning_rate": 7.730956239870341e-06,
+      "loss": 0.6064,
+      "step": 954
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9854211264984167,
+      "learning_rate": 7.73905996758509e-06,
+      "loss": 0.2763,
+      "step": 955
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.436628008416,
+      "learning_rate": 7.747163695299839e-06,
+      "loss": 0.402,
+      "step": 956
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3434781739399368,
+      "learning_rate": 7.755267423014588e-06,
+      "loss": 0.1968,
+      "step": 957
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4258390164830184,
+      "learning_rate": 7.763371150729337e-06,
+      "loss": 0.504,
+      "step": 958
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7988657872287455,
+      "learning_rate": 7.771474878444086e-06,
+      "loss": 0.4546,
+      "step": 959
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6699235604984795,
+      "learning_rate": 7.779578606158834e-06,
+      "loss": 0.237,
+      "step": 960
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3983399125913127,
+      "learning_rate": 7.787682333873583e-06,
+      "loss": 0.4261,
+      "step": 961
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.962405542887829,
+      "learning_rate": 7.79578606158833e-06,
+      "loss": 0.479,
+      "step": 962
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.10905936881963,
+      "learning_rate": 7.80388978930308e-06,
+      "loss": 0.4278,
+      "step": 963
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5682859950880332,
+      "learning_rate": 7.811993517017828e-06,
+      "loss": 0.4258,
+      "step": 964
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.036695158440992,
+      "learning_rate": 7.820097244732577e-06,
+      "loss": 0.3851,
+      "step": 965
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5127053347705353,
+      "learning_rate": 7.828200972447326e-06,
+      "loss": 0.251,
+      "step": 966
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7542846582033844,
+      "learning_rate": 7.836304700162075e-06,
+      "loss": 0.4631,
+      "step": 967
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6893892573814526,
+      "learning_rate": 7.844408427876824e-06,
+      "loss": 0.119,
+      "step": 968
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.092866383747901,
+      "learning_rate": 7.852512155591573e-06,
+      "loss": 0.2578,
+      "step": 969
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9003753456201338,
+      "learning_rate": 7.860615883306322e-06,
+      "loss": 0.536,
+      "step": 970
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.680772789589768,
+      "learning_rate": 7.86871961102107e-06,
+      "loss": 0.4974,
+      "step": 971
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6303603867999263,
+      "learning_rate": 7.87682333873582e-06,
+      "loss": 0.4491,
+      "step": 972
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4093728299261037,
+      "learning_rate": 7.884927066450568e-06,
+      "loss": 0.4435,
+      "step": 973
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3334965662966733,
+      "learning_rate": 7.893030794165316e-06,
+      "loss": 0.4632,
+      "step": 974
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.558651080472022,
+      "learning_rate": 7.901134521880064e-06,
+      "loss": 0.3497,
+      "step": 975
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4060047198040806,
+      "learning_rate": 7.909238249594813e-06,
+      "loss": 0.3355,
+      "step": 976
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.947298402615608,
+      "learning_rate": 7.917341977309562e-06,
+      "loss": 0.1908,
+      "step": 977
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.848004512134337,
+      "learning_rate": 7.925445705024311e-06,
+      "loss": 0.4213,
+      "step": 978
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.258896755450016,
+      "learning_rate": 7.93354943273906e-06,
+      "loss": 0.32,
+      "step": 979
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0182702810774695,
+      "learning_rate": 7.941653160453809e-06,
+      "loss": 0.6476,
+      "step": 980
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9190024928353155,
+      "learning_rate": 7.949756888168558e-06,
+      "loss": 0.3392,
+      "step": 981
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.69195166322559,
+      "learning_rate": 7.957860615883307e-06,
+      "loss": 0.4677,
+      "step": 982
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9961259584050945,
+      "learning_rate": 7.965964343598056e-06,
+      "loss": 0.2395,
+      "step": 983
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7162207644627627,
+      "learning_rate": 7.974068071312805e-06,
+      "loss": 0.5829,
+      "step": 984
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3540266724684162,
+      "learning_rate": 7.982171799027553e-06,
+      "loss": 0.2531,
+      "step": 985
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1589888165260707,
+      "learning_rate": 7.990275526742302e-06,
+      "loss": 0.3397,
+      "step": 986
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8044779643499396,
+      "learning_rate": 7.998379254457051e-06,
+      "loss": 0.2096,
+      "step": 987
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.032911292540925,
+      "learning_rate": 8.0064829821718e-06,
+      "loss": 0.3175,
+      "step": 988
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.655264116097602,
+      "learning_rate": 8.014586709886549e-06,
+      "loss": 0.4094,
+      "step": 989
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0382239948885603,
+      "learning_rate": 8.022690437601298e-06,
+      "loss": 0.2394,
+      "step": 990
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.735587238686321,
+      "learning_rate": 8.030794165316047e-06,
+      "loss": 0.393,
+      "step": 991
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5086191965509603,
+      "learning_rate": 8.038897893030794e-06,
+      "loss": 0.5107,
+      "step": 992
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.493481243441639,
+      "learning_rate": 8.047001620745543e-06,
+      "loss": 0.417,
+      "step": 993
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.292500293715031,
+      "learning_rate": 8.055105348460292e-06,
+      "loss": 0.2417,
+      "step": 994
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.988219934968891,
+      "learning_rate": 8.06320907617504e-06,
+      "loss": 0.565,
+      "step": 995
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2049213823372606,
+      "learning_rate": 8.07131280388979e-06,
+      "loss": 0.404,
+      "step": 996
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2610028835457086,
+      "learning_rate": 8.079416531604538e-06,
+      "loss": 0.3252,
+      "step": 997
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2200345559368784,
+      "learning_rate": 8.087520259319287e-06,
+      "loss": 0.239,
+      "step": 998
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.668521945707589,
+      "learning_rate": 8.095623987034036e-06,
+      "loss": 0.1153,
+      "step": 999
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3300141983663805,
+      "learning_rate": 8.103727714748785e-06,
+      "loss": 0.4194,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0839613659358527,
+      "learning_rate": 8.111831442463534e-06,
+      "loss": 0.5825,
+      "step": 1001
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9635026347752458,
+      "learning_rate": 8.119935170178283e-06,
+      "loss": 0.3117,
+      "step": 1002
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.990713449306685,
+      "learning_rate": 8.128038897893032e-06,
+      "loss": 0.3316,
+      "step": 1003
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.051608421456767,
+      "learning_rate": 8.13614262560778e-06,
+      "loss": 0.6384,
+      "step": 1004
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.552177800836222,
+      "learning_rate": 8.14424635332253e-06,
+      "loss": 0.2838,
+      "step": 1005
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6469189614312425,
+      "learning_rate": 8.152350081037278e-06,
+      "loss": 0.258,
+      "step": 1006
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0931657567001225,
+      "learning_rate": 8.160453808752027e-06,
+      "loss": 0.4389,
+      "step": 1007
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8124012344680072,
+      "learning_rate": 8.168557536466776e-06,
+      "loss": 0.2715,
+      "step": 1008
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6570684910996225,
+      "learning_rate": 8.176661264181525e-06,
+      "loss": 0.6342,
+      "step": 1009
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.792185426784913,
+      "learning_rate": 8.184764991896274e-06,
+      "loss": 0.5323,
+      "step": 1010
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9284721052804479,
+      "learning_rate": 8.192868719611021e-06,
+      "loss": 0.295,
+      "step": 1011
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8957646081505732,
+      "learning_rate": 8.20097244732577e-06,
+      "loss": 0.272,
+      "step": 1012
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7254078559828327,
+      "learning_rate": 8.209076175040519e-06,
+      "loss": 0.5446,
+      "step": 1013
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0465076786701526,
+      "learning_rate": 8.217179902755268e-06,
+      "loss": 0.3318,
+      "step": 1014
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.11996385607633,
+      "learning_rate": 8.225283630470017e-06,
+      "loss": 0.3359,
+      "step": 1015
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1062526411385205,
+      "learning_rate": 8.233387358184766e-06,
+      "loss": 0.2757,
+      "step": 1016
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9308026987100608,
+      "learning_rate": 8.241491085899515e-06,
+      "loss": 0.2172,
+      "step": 1017
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8846366897518265,
+      "learning_rate": 8.249594813614264e-06,
+      "loss": 0.2473,
+      "step": 1018
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5164546901490077,
+      "learning_rate": 8.257698541329012e-06,
+      "loss": 0.2204,
+      "step": 1019
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.965946024832421,
+      "learning_rate": 8.265802269043761e-06,
+      "loss": 0.2682,
+      "step": 1020
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8770996025384163,
+      "learning_rate": 8.27390599675851e-06,
+      "loss": 0.4262,
+      "step": 1021
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7747551380415447,
+      "learning_rate": 8.282009724473259e-06,
+      "loss": 0.4501,
+      "step": 1022
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7646702573490158,
+      "learning_rate": 8.290113452188006e-06,
+      "loss": 0.2916,
+      "step": 1023
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.032020213747249,
+      "learning_rate": 8.298217179902755e-06,
+      "loss": 0.4006,
+      "step": 1024
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9632931237444256,
+      "learning_rate": 8.306320907617504e-06,
+      "loss": 0.5842,
+      "step": 1025
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.2655860982944525,
+      "learning_rate": 8.314424635332253e-06,
+      "loss": 0.2772,
+      "step": 1026
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5777340679281244,
+      "learning_rate": 8.322528363047002e-06,
+      "loss": 0.6111,
+      "step": 1027
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2272990096143674,
+      "learning_rate": 8.33063209076175e-06,
+      "loss": 0.2961,
+      "step": 1028
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8867529200585254,
+      "learning_rate": 8.3387358184765e-06,
+      "loss": 0.3721,
+      "step": 1029
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5396855220831567,
+      "learning_rate": 8.346839546191249e-06,
+      "loss": 0.5991,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7053120976699256,
+      "learning_rate": 8.354943273905997e-06,
+      "loss": 0.6307,
+      "step": 1031
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1322427659390026,
+      "learning_rate": 8.363047001620746e-06,
+      "loss": 0.3064,
+      "step": 1032
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5741821739169377,
+      "learning_rate": 8.371150729335495e-06,
+      "loss": 0.3448,
+      "step": 1033
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1238486452639105,
+      "learning_rate": 8.379254457050244e-06,
+      "loss": 0.3173,
+      "step": 1034
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.169036011429014,
+      "learning_rate": 8.387358184764991e-06,
+      "loss": 0.3799,
+      "step": 1035
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0350082988007436,
+      "learning_rate": 8.39546191247974e-06,
+      "loss": 0.3306,
+      "step": 1036
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.354525562304569,
+      "learning_rate": 8.403565640194489e-06,
+      "loss": 0.3223,
+      "step": 1037
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0986202021273646,
+      "learning_rate": 8.411669367909238e-06,
+      "loss": 0.307,
+      "step": 1038
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0546619189818123,
+      "learning_rate": 8.419773095623987e-06,
+      "loss": 0.3514,
+      "step": 1039
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0097824184950595,
+      "learning_rate": 8.427876823338736e-06,
+      "loss": 0.2622,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8730965512128366,
+      "learning_rate": 8.435980551053485e-06,
+      "loss": 0.6751,
+      "step": 1041
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6796045318304387,
+      "learning_rate": 8.444084278768234e-06,
+      "loss": 0.228,
+      "step": 1042
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9648058074260562,
+      "learning_rate": 8.452188006482982e-06,
+      "loss": 0.1877,
+      "step": 1043
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0178936345101848,
+      "learning_rate": 8.460291734197731e-06,
+      "loss": 0.3236,
+      "step": 1044
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.499365887092295,
+      "learning_rate": 8.46839546191248e-06,
+      "loss": 0.5304,
+      "step": 1045
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.750043223864979,
+      "learning_rate": 8.47649918962723e-06,
+      "loss": 0.3324,
+      "step": 1046
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.390432533918871,
+      "learning_rate": 8.484602917341978e-06,
+      "loss": 0.3128,
+      "step": 1047
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.361342462412532,
+      "learning_rate": 8.492706645056727e-06,
+      "loss": 0.4278,
+      "step": 1048
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9970842505530577,
+      "learning_rate": 8.500810372771476e-06,
+      "loss": 0.2969,
+      "step": 1049
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3394517352880442,
+      "learning_rate": 8.508914100486225e-06,
+      "loss": 0.2509,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1868432621292886,
+      "learning_rate": 8.517017828200974e-06,
+      "loss": 0.3775,
+      "step": 1051
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4366667313571386,
+      "learning_rate": 8.525121555915723e-06,
+      "loss": 0.4347,
+      "step": 1052
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6829115830918275,
+      "learning_rate": 8.533225283630471e-06,
+      "loss": 0.2121,
+      "step": 1053
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.247445054949991,
+      "learning_rate": 8.54132901134522e-06,
+      "loss": 0.2186,
+      "step": 1054
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.270861762565885,
+      "learning_rate": 8.54943273905997e-06,
+      "loss": 0.5022,
+      "step": 1055
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7503713369211311,
+      "learning_rate": 8.557536466774718e-06,
+      "loss": 0.2279,
+      "step": 1056
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0164677517810854,
+      "learning_rate": 8.565640194489467e-06,
+      "loss": 0.1833,
+      "step": 1057
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.446995025312077,
+      "learning_rate": 8.573743922204216e-06,
+      "loss": 0.2738,
+      "step": 1058
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0007975243618166,
+      "learning_rate": 8.581847649918963e-06,
+      "loss": 0.1865,
+      "step": 1059
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0637859323794205,
+      "learning_rate": 8.589951377633712e-06,
+      "loss": 0.4058,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.782686066365747,
+      "learning_rate": 8.598055105348461e-06,
+      "loss": 0.3157,
+      "step": 1061
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.739087713879544,
+      "learning_rate": 8.60615883306321e-06,
+      "loss": 0.5574,
+      "step": 1062
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0718574907612974,
+      "learning_rate": 8.614262560777959e-06,
+      "loss": 0.3182,
+      "step": 1063
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9407690065429373,
+      "learning_rate": 8.622366288492708e-06,
+      "loss": 0.4389,
+      "step": 1064
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.784290869455431,
+      "learning_rate": 8.630470016207456e-06,
+      "loss": 0.4286,
+      "step": 1065
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5775046120589797,
+      "learning_rate": 8.638573743922205e-06,
+      "loss": 0.2037,
+      "step": 1066
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.437714139384116,
+      "learning_rate": 8.646677471636954e-06,
+      "loss": 0.5343,
+      "step": 1067
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.483349156462633,
+      "learning_rate": 8.654781199351703e-06,
+      "loss": 0.3788,
+      "step": 1068
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.709821738101605,
+      "learning_rate": 8.662884927066452e-06,
+      "loss": 0.5581,
+      "step": 1069
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.400755790081389,
+      "learning_rate": 8.670988654781201e-06,
+      "loss": 0.3647,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8934228623978604,
+      "learning_rate": 8.679092382495948e-06,
+      "loss": 0.2116,
+      "step": 1071
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3004669985187696,
+      "learning_rate": 8.687196110210697e-06,
+      "loss": 0.3732,
+      "step": 1072
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1032243489780567,
+      "learning_rate": 8.695299837925446e-06,
+      "loss": 0.3316,
+      "step": 1073
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.521279478132474,
+      "learning_rate": 8.703403565640195e-06,
+      "loss": 0.5415,
+      "step": 1074
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.707217222488944,
+      "learning_rate": 8.711507293354944e-06,
+      "loss": 0.5892,
+      "step": 1075
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5265250435000723,
+      "learning_rate": 8.719611021069693e-06,
+      "loss": 0.5907,
+      "step": 1076
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9974593364231343,
+      "learning_rate": 8.727714748784441e-06,
+      "loss": 0.2184,
+      "step": 1077
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0892230101317923,
+      "learning_rate": 8.73581847649919e-06,
+      "loss": 0.2349,
+      "step": 1078
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.385666360793876,
+      "learning_rate": 8.74392220421394e-06,
+      "loss": 0.3804,
+      "step": 1079
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.160349500955361,
+      "learning_rate": 8.752025931928688e-06,
+      "loss": 0.178,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.689755795218436,
+      "learning_rate": 8.760129659643437e-06,
+      "loss": 0.1572,
+      "step": 1081
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.294056506038017,
+      "learning_rate": 8.768233387358186e-06,
+      "loss": 0.4049,
+      "step": 1082
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.781935628754395,
+      "learning_rate": 8.776337115072933e-06,
+      "loss": 0.2775,
+      "step": 1083
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1276263037797993,
+      "learning_rate": 8.784440842787682e-06,
+      "loss": 0.413,
+      "step": 1084
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0913658018479344,
+      "learning_rate": 8.792544570502431e-06,
+      "loss": 0.3203,
+      "step": 1085
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0957566506364347,
+      "learning_rate": 8.80064829821718e-06,
+      "loss": 0.4394,
+      "step": 1086
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.006125960695718,
+      "learning_rate": 8.808752025931929e-06,
+      "loss": 0.4378,
+      "step": 1087
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.910310554253882,
+      "learning_rate": 8.816855753646678e-06,
+      "loss": 0.4747,
+      "step": 1088
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0500485398776305,
+      "learning_rate": 8.824959481361426e-06,
+      "loss": 0.5352,
+      "step": 1089
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.406907443855871,
+      "learning_rate": 8.833063209076175e-06,
+      "loss": 0.2424,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5540340532804793,
+      "learning_rate": 8.841166936790924e-06,
+      "loss": 0.4409,
+      "step": 1091
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7845565075266903,
+      "learning_rate": 8.849270664505673e-06,
+      "loss": 0.618,
+      "step": 1092
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0972709870265454,
+      "learning_rate": 8.857374392220422e-06,
+      "loss": 0.1855,
+      "step": 1093
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.207830314807331,
+      "learning_rate": 8.865478119935171e-06,
+      "loss": 0.4526,
+      "step": 1094
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8186664649780988,
+      "learning_rate": 8.87358184764992e-06,
+      "loss": 0.1782,
+      "step": 1095
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.086442205063557,
+      "learning_rate": 8.881685575364669e-06,
+      "loss": 0.7286,
+      "step": 1096
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.369013925232263,
+      "learning_rate": 8.889789303079418e-06,
+      "loss": 0.3849,
+      "step": 1097
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.337818230676887,
+      "learning_rate": 8.897893030794167e-06,
+      "loss": 0.4231,
+      "step": 1098
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3275928273112925,
+      "learning_rate": 8.905996758508915e-06,
+      "loss": 0.2443,
+      "step": 1099
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.92485488369392,
+      "learning_rate": 8.914100486223664e-06,
+      "loss": 0.2581,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.212567762910474,
+      "learning_rate": 8.922204213938413e-06,
+      "loss": 0.435,
+      "step": 1101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6340983389207246,
+      "learning_rate": 8.930307941653162e-06,
+      "loss": 0.5674,
+      "step": 1102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8792845727791028,
+      "learning_rate": 8.93841166936791e-06,
+      "loss": 0.3916,
+      "step": 1103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8346828658062146,
+      "learning_rate": 8.946515397082658e-06,
+      "loss": 0.1879,
+      "step": 1104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8883772327301065,
+      "learning_rate": 8.954619124797407e-06,
+      "loss": 0.2627,
+      "step": 1105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3098134854392,
+      "learning_rate": 8.962722852512156e-06,
+      "loss": 0.3936,
+      "step": 1106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.873029519837041,
+      "learning_rate": 8.970826580226905e-06,
+      "loss": 0.2136,
+      "step": 1107
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5415235828293907,
+      "learning_rate": 8.978930307941654e-06,
+      "loss": 0.3727,
+      "step": 1108
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6098679375202454,
+      "learning_rate": 8.987034035656403e-06,
+      "loss": 0.3905,
+      "step": 1109
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.303265905793004,
+      "learning_rate": 8.995137763371152e-06,
+      "loss": 0.5198,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.158183147907476,
+      "learning_rate": 9.0032414910859e-06,
+      "loss": 0.6048,
+      "step": 1111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.974654479758728,
+      "learning_rate": 9.01134521880065e-06,
+      "loss": 0.7033,
+      "step": 1112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1155563982909067,
+      "learning_rate": 9.019448946515398e-06,
+      "loss": 0.3928,
+      "step": 1113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9258232954031986,
+      "learning_rate": 9.027552674230147e-06,
+      "loss": 0.3591,
+      "step": 1114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4972054836908346,
+      "learning_rate": 9.035656401944896e-06,
+      "loss": 0.4009,
+      "step": 1115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.447497266323679,
+      "learning_rate": 9.043760129659645e-06,
+      "loss": 0.2289,
+      "step": 1116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.302481503955969,
+      "learning_rate": 9.051863857374394e-06,
+      "loss": 0.484,
+      "step": 1117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9837637733409186,
+      "learning_rate": 9.059967585089143e-06,
+      "loss": 0.4079,
+      "step": 1118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5428953848714007,
+      "learning_rate": 9.068071312803892e-06,
+      "loss": 0.508,
+      "step": 1119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3310487297247255,
+      "learning_rate": 9.076175040518639e-06,
+      "loss": 0.2705,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5827794570793854,
+      "learning_rate": 9.084278768233388e-06,
+      "loss": 0.271,
+      "step": 1121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9916189691453106,
+      "learning_rate": 9.092382495948137e-06,
+      "loss": 0.3741,
+      "step": 1122
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6029654835234486,
+      "learning_rate": 9.100486223662885e-06,
+      "loss": 0.2491,
+      "step": 1123
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6843323780801662,
+      "learning_rate": 9.108589951377634e-06,
+      "loss": 0.4605,
+      "step": 1124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8937327403046844,
+      "learning_rate": 9.116693679092383e-06,
+      "loss": 0.2121,
+      "step": 1125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5748500031616681,
+      "learning_rate": 9.124797406807132e-06,
+      "loss": 0.1992,
+      "step": 1126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.584456097254127,
+      "learning_rate": 9.132901134521881e-06,
+      "loss": 0.6839,
+      "step": 1127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.002472951006482,
+      "learning_rate": 9.14100486223663e-06,
+      "loss": 0.3942,
+      "step": 1128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.290926073800616,
+      "learning_rate": 9.149108589951379e-06,
+      "loss": 0.5362,
+      "step": 1129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1523777704613742,
+      "learning_rate": 9.157212317666128e-06,
+      "loss": 0.361,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9861406971173527,
+      "learning_rate": 9.165316045380877e-06,
+      "loss": 0.4203,
+      "step": 1131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.732132591656581,
+      "learning_rate": 9.173419773095624e-06,
+      "loss": 0.5565,
+      "step": 1132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5061909433194374,
+      "learning_rate": 9.181523500810373e-06,
+      "loss": 0.5791,
+      "step": 1133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.152541826483666,
+      "learning_rate": 9.189627228525122e-06,
+      "loss": 0.2494,
+      "step": 1134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.361574537169059,
+      "learning_rate": 9.19773095623987e-06,
+      "loss": 0.3892,
+      "step": 1135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6356223127935579,
+      "learning_rate": 9.20583468395462e-06,
+      "loss": 0.3579,
+      "step": 1136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9122696245353248,
+      "learning_rate": 9.213938411669368e-06,
+      "loss": 0.3245,
+      "step": 1137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.592290214724769,
+      "learning_rate": 9.222042139384117e-06,
+      "loss": 0.4281,
+      "step": 1138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.283611629806301,
+      "learning_rate": 9.230145867098866e-06,
+      "loss": 0.3588,
+      "step": 1139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3546012938577308,
+      "learning_rate": 9.238249594813615e-06,
+      "loss": 0.4201,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.570659295198111,
+      "learning_rate": 9.246353322528364e-06,
+      "loss": 0.5381,
+      "step": 1141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2796721423274344,
+      "learning_rate": 9.254457050243113e-06,
+      "loss": 0.4437,
+      "step": 1142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.781949943082391,
+      "learning_rate": 9.262560777957862e-06,
+      "loss": 0.1712,
+      "step": 1143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.221000611200382,
+      "learning_rate": 9.270664505672609e-06,
+      "loss": 0.3394,
+      "step": 1144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3654410309307337,
+      "learning_rate": 9.278768233387358e-06,
+      "loss": 0.6641,
+      "step": 1145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2343345385002387,
+      "learning_rate": 9.286871961102107e-06,
+      "loss": 0.5177,
+      "step": 1146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4288478639394393,
+      "learning_rate": 9.294975688816856e-06,
+      "loss": 0.3886,
+      "step": 1147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1534648128820675,
+      "learning_rate": 9.303079416531604e-06,
+      "loss": 0.4025,
+      "step": 1148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5633151312363531,
+      "learning_rate": 9.311183144246353e-06,
+      "loss": 0.2489,
+      "step": 1149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8768516457521371,
+      "learning_rate": 9.319286871961102e-06,
+      "loss": 0.2715,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3102412146591247,
+      "learning_rate": 9.327390599675851e-06,
+      "loss": 0.422,
+      "step": 1151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.117606378481867,
+      "learning_rate": 9.3354943273906e-06,
+      "loss": 0.5086,
+      "step": 1152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3068418326224625,
+      "learning_rate": 9.343598055105349e-06,
+      "loss": 0.3282,
+      "step": 1153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5690326808938497,
+      "learning_rate": 9.351701782820098e-06,
+      "loss": 0.4448,
+      "step": 1154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4023837944518864,
+      "learning_rate": 9.359805510534847e-06,
+      "loss": 0.1758,
+      "step": 1155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.06395559224549,
+      "learning_rate": 9.367909238249596e-06,
+      "loss": 0.2894,
+      "step": 1156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0284483588643107,
+      "learning_rate": 9.376012965964344e-06,
+      "loss": 0.2281,
+      "step": 1157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8466007371120743,
+      "learning_rate": 9.384116693679093e-06,
+      "loss": 0.2625,
+      "step": 1158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.073249831255183,
+      "learning_rate": 9.392220421393842e-06,
+      "loss": 0.1806,
+      "step": 1159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0604521996320924,
+      "learning_rate": 9.400324149108591e-06,
+      "loss": 0.2566,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7743048794399432,
+      "learning_rate": 9.40842787682334e-06,
+      "loss": 0.2897,
+      "step": 1161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0941673546074093,
+      "learning_rate": 9.416531604538089e-06,
+      "loss": 0.4121,
+      "step": 1162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8438867683923401,
+      "learning_rate": 9.424635332252838e-06,
+      "loss": 0.2835,
+      "step": 1163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.260141898369755,
+      "learning_rate": 9.432739059967587e-06,
+      "loss": 0.7149,
+      "step": 1164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.417710612481021,
+      "learning_rate": 9.440842787682336e-06,
+      "loss": 0.5424,
+      "step": 1165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4420254585851078,
+      "learning_rate": 9.448946515397085e-06,
+      "loss": 0.602,
+      "step": 1166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.521025352153083,
+      "learning_rate": 9.457050243111833e-06,
+      "loss": 0.4274,
+      "step": 1167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2965914038659796,
+      "learning_rate": 9.46515397082658e-06,
+      "loss": 0.3162,
+      "step": 1168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1434557081584558,
+      "learning_rate": 9.47325769854133e-06,
+      "loss": 0.3215,
+      "step": 1169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.710430394655179,
+      "learning_rate": 9.481361426256078e-06,
+      "loss": 0.4365,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.969973081312627,
+      "learning_rate": 9.489465153970827e-06,
+      "loss": 0.7312,
+      "step": 1171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6483006601350636,
+      "learning_rate": 9.497568881685576e-06,
+      "loss": 0.2341,
+      "step": 1172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2489303854518714,
+      "learning_rate": 9.505672609400325e-06,
+      "loss": 0.2919,
+      "step": 1173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.471839673528933,
+      "learning_rate": 9.513776337115074e-06,
+      "loss": 0.3787,
+      "step": 1174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5581616335253563,
+      "learning_rate": 9.521880064829823e-06,
+      "loss": 0.4093,
+      "step": 1175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1737918765052533,
+      "learning_rate": 9.529983792544572e-06,
+      "loss": 0.4631,
+      "step": 1176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2481791360428085,
+      "learning_rate": 9.53808752025932e-06,
+      "loss": 0.3476,
+      "step": 1177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1651818517005577,
+      "learning_rate": 9.54619124797407e-06,
+      "loss": 0.3632,
+      "step": 1178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8903275878526102,
+      "learning_rate": 9.554294975688818e-06,
+      "loss": 0.2973,
+      "step": 1179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1934616289226057,
+      "learning_rate": 9.562398703403566e-06,
+      "loss": 0.3305,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.279622189564768,
+      "learning_rate": 9.570502431118315e-06,
+      "loss": 0.2162,
+      "step": 1181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.271580950716776,
+      "learning_rate": 9.578606158833063e-06,
+      "loss": 0.3093,
+      "step": 1182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8171795274195364,
+      "learning_rate": 9.586709886547812e-06,
+      "loss": 0.4487,
+      "step": 1183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.619931187673807,
+      "learning_rate": 9.594813614262561e-06,
+      "loss": 0.2849,
+      "step": 1184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6214271367497153,
+      "learning_rate": 9.60291734197731e-06,
+      "loss": 0.6501,
+      "step": 1185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6359199085765381,
+      "learning_rate": 9.611021069692059e-06,
+      "loss": 0.3297,
+      "step": 1186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6411329943245123,
+      "learning_rate": 9.619124797406808e-06,
+      "loss": 0.3808,
+      "step": 1187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3265245812415074,
+      "learning_rate": 9.627228525121557e-06,
+      "loss": 0.4688,
+      "step": 1188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9236239562158506,
+      "learning_rate": 9.635332252836306e-06,
+      "loss": 0.3809,
+      "step": 1189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2305234912361285,
+      "learning_rate": 9.643435980551055e-06,
+      "loss": 0.378,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.223409483765205,
+      "learning_rate": 9.651539708265803e-06,
+      "loss": 0.23,
+      "step": 1191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.664790213410784,
+      "learning_rate": 9.65964343598055e-06,
+      "loss": 0.2293,
+      "step": 1192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.822316458300743,
+      "learning_rate": 9.6677471636953e-06,
+      "loss": 0.35,
+      "step": 1193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2219674529787303,
+      "learning_rate": 9.675850891410048e-06,
+      "loss": 0.4275,
+      "step": 1194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6294195592213816,
+      "learning_rate": 9.683954619124797e-06,
+      "loss": 0.2521,
+      "step": 1195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.614518640991949,
+      "learning_rate": 9.692058346839546e-06,
+      "loss": 0.2258,
+      "step": 1196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.34959181177634,
+      "learning_rate": 9.700162074554295e-06,
+      "loss": 0.2422,
+      "step": 1197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4947640209321977,
+      "learning_rate": 9.708265802269044e-06,
+      "loss": 0.204,
+      "step": 1198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5026407887465723,
+      "learning_rate": 9.716369529983793e-06,
+      "loss": 0.322,
+      "step": 1199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4295713390781977,
+      "learning_rate": 9.724473257698542e-06,
+      "loss": 0.3953,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1943300702942254,
+      "learning_rate": 9.73257698541329e-06,
+      "loss": 0.2129,
+      "step": 1201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6069829249854433,
+      "learning_rate": 9.74068071312804e-06,
+      "loss": 0.3327,
+      "step": 1202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7979319325001946,
+      "learning_rate": 9.748784440842789e-06,
+      "loss": 0.209,
+      "step": 1203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4363453465511022,
+      "learning_rate": 9.756888168557537e-06,
+      "loss": 0.244,
+      "step": 1204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.396626655564512,
+      "learning_rate": 9.764991896272286e-06,
+      "loss": 0.4459,
+      "step": 1205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1032069384618035,
+      "learning_rate": 9.773095623987035e-06,
+      "loss": 0.2811,
+      "step": 1206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.818709049960038,
+      "learning_rate": 9.781199351701784e-06,
+      "loss": 0.2779,
+      "step": 1207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9869287196130316,
+      "learning_rate": 9.789303079416533e-06,
+      "loss": 0.3301,
+      "step": 1208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9283405653335193,
+      "learning_rate": 9.797406807131282e-06,
+      "loss": 0.3047,
+      "step": 1209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.15457778164452,
+      "learning_rate": 9.80551053484603e-06,
+      "loss": 0.2626,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7561992893330345,
+      "learning_rate": 9.81361426256078e-06,
+      "loss": 0.3394,
+      "step": 1211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8645936994069112,
+      "learning_rate": 9.821717990275529e-06,
+      "loss": 0.1985,
+      "step": 1212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2956354014085467,
+      "learning_rate": 9.829821717990277e-06,
+      "loss": 0.194,
+      "step": 1213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7835259169382427,
+      "learning_rate": 9.837925445705026e-06,
+      "loss": 0.3692,
+      "step": 1214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9701594570651717,
+      "learning_rate": 9.846029173419774e-06,
+      "loss": 0.3978,
+      "step": 1215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2384912415640494,
+      "learning_rate": 9.854132901134522e-06,
+      "loss": 0.3364,
+      "step": 1216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.659194829466432,
+      "learning_rate": 9.862236628849271e-06,
+      "loss": 0.3029,
+      "step": 1217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.950740570840422,
+      "learning_rate": 9.87034035656402e-06,
+      "loss": 0.1626,
+      "step": 1218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8087368797365904,
+      "learning_rate": 9.878444084278769e-06,
+      "loss": 0.3146,
+      "step": 1219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9149109165261406,
+      "learning_rate": 9.886547811993518e-06,
+      "loss": 0.2789,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.676804089485406,
+      "learning_rate": 9.894651539708267e-06,
+      "loss": 0.7441,
+      "step": 1221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0251399922095366,
+      "learning_rate": 9.902755267423016e-06,
+      "loss": 0.5809,
+      "step": 1222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7823018860185944,
+      "learning_rate": 9.910858995137765e-06,
+      "loss": 0.3594,
+      "step": 1223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3984286155908903,
+      "learning_rate": 9.918962722852514e-06,
+      "loss": 0.562,
+      "step": 1224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2053665364464177,
+      "learning_rate": 9.927066450567262e-06,
+      "loss": 0.4011,
+      "step": 1225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1406280350489575,
+      "learning_rate": 9.935170178282011e-06,
+      "loss": 0.3252,
+      "step": 1226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2704139640855256,
+      "learning_rate": 9.94327390599676e-06,
+      "loss": 0.3393,
+      "step": 1227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6554788030943057,
+      "learning_rate": 9.951377633711507e-06,
+      "loss": 0.6732,
+      "step": 1228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8548889037144014,
+      "learning_rate": 9.959481361426256e-06,
+      "loss": 0.2144,
+      "step": 1229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3588771042087946,
+      "learning_rate": 9.967585089141005e-06,
+      "loss": 0.4642,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.263588917802719,
+      "learning_rate": 9.975688816855754e-06,
+      "loss": 0.4061,
+      "step": 1231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9410124749741353,
+      "learning_rate": 9.983792544570503e-06,
+      "loss": 0.2857,
+      "step": 1232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0345075821926257,
+      "learning_rate": 9.991896272285252e-06,
+      "loss": 0.2277,
+      "step": 1233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.569341876299674,
+      "learning_rate": 1e-05,
+      "loss": 0.4141,
+      "step": 1234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.694073433720551,
+      "learning_rate": 9.99999998447568e-06,
+      "loss": 0.435,
+      "step": 1235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1761356660935496,
+      "learning_rate": 9.999999937902713e-06,
+      "loss": 0.4686,
+      "step": 1236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0837918651153284,
+      "learning_rate": 9.999999860281105e-06,
+      "loss": 0.3658,
+      "step": 1237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.035419121310948,
+      "learning_rate": 9.999999751610852e-06,
+      "loss": 0.2821,
+      "step": 1238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.081368900968512,
+      "learning_rate": 9.999999611891956e-06,
+      "loss": 0.47,
+      "step": 1239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.432534613446358,
+      "learning_rate": 9.99999944112442e-06,
+      "loss": 0.5572,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.227260118919454,
+      "learning_rate": 9.999999239308242e-06,
+      "loss": 0.4274,
+      "step": 1241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8554747330418546,
+      "learning_rate": 9.999999006443427e-06,
+      "loss": 0.2017,
+      "step": 1242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0890570042037857,
+      "learning_rate": 9.999998742529973e-06,
+      "loss": 0.3677,
+      "step": 1243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.388685583541598,
+      "learning_rate": 9.999998447567883e-06,
+      "loss": 0.3138,
+      "step": 1244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9598959098029918,
+      "learning_rate": 9.999998121557156e-06,
+      "loss": 0.2605,
+      "step": 1245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8843219215040872,
+      "learning_rate": 9.999997764497802e-06,
+      "loss": 0.3755,
+      "step": 1246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.129530318246904,
+      "learning_rate": 9.999997376389814e-06,
+      "loss": 0.2561,
+      "step": 1247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0360232097299584,
+      "learning_rate": 9.9999969572332e-06,
+      "loss": 0.2763,
+      "step": 1248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.900860660369104,
+      "learning_rate": 9.99999650702796e-06,
+      "loss": 0.3827,
+      "step": 1249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8563723281408049,
+      "learning_rate": 9.9999960257741e-06,
+      "loss": 0.2636,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9003101478997055,
+      "learning_rate": 9.999995513471617e-06,
+      "loss": 0.4108,
+      "step": 1251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9647302215570617,
+      "learning_rate": 9.999994970120522e-06,
+      "loss": 0.237,
+      "step": 1252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.620508176882178,
+      "learning_rate": 9.99999439572081e-06,
+      "loss": 0.6288,
+      "step": 1253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9843955264193798,
+      "learning_rate": 9.99999379027249e-06,
+      "loss": 0.3364,
+      "step": 1254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.271788628391138,
+      "learning_rate": 9.999993153775567e-06,
+      "loss": 0.3854,
+      "step": 1255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2669735019466235,
+      "learning_rate": 9.999992486230041e-06,
+      "loss": 0.4857,
+      "step": 1256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6477589455344397,
+      "learning_rate": 9.999991787635918e-06,
+      "loss": 0.3463,
+      "step": 1257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.45320692651794,
+      "learning_rate": 9.999991057993201e-06,
+      "loss": 0.1731,
+      "step": 1258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8509201029923288,
+      "learning_rate": 9.999990297301897e-06,
+      "loss": 0.3241,
+      "step": 1259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2009512715448407,
+      "learning_rate": 9.999989505562008e-06,
+      "loss": 0.2067,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.362001649580629,
+      "learning_rate": 9.999988682773541e-06,
+      "loss": 0.317,
+      "step": 1261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.631922545980802,
+      "learning_rate": 9.999987828936502e-06,
+      "loss": 0.4987,
+      "step": 1262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9508335652667639,
+      "learning_rate": 9.999986944050892e-06,
+      "loss": 0.2317,
+      "step": 1263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.936516313001885,
+      "learning_rate": 9.999986028116719e-06,
+      "loss": 0.7129,
+      "step": 1264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5541815883849868,
+      "learning_rate": 9.99998508113399e-06,
+      "loss": 0.5558,
+      "step": 1265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7196775469491075,
+      "learning_rate": 9.99998410310271e-06,
+      "loss": 0.2777,
+      "step": 1266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9391060986408841,
+      "learning_rate": 9.999983094022884e-06,
+      "loss": 0.2453,
+      "step": 1267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1103764794996973,
+      "learning_rate": 9.99998205389452e-06,
+      "loss": 0.231,
+      "step": 1268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.773477043622143,
+      "learning_rate": 9.999980982717621e-06,
+      "loss": 0.3201,
+      "step": 1269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3458749039615845,
+      "learning_rate": 9.999979880492199e-06,
+      "loss": 0.3342,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9057042009304825,
+      "learning_rate": 9.999978747218257e-06,
+      "loss": 0.5203,
+      "step": 1271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.159158541792798,
+      "learning_rate": 9.999977582895804e-06,
+      "loss": 0.5068,
+      "step": 1272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.664970567605134,
+      "learning_rate": 9.999976387524843e-06,
+      "loss": 0.594,
+      "step": 1273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6822012629906427,
+      "learning_rate": 9.999975161105388e-06,
+      "loss": 0.7875,
+      "step": 1274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8064769678575865,
+      "learning_rate": 9.99997390363744e-06,
+      "loss": 0.3053,
+      "step": 1275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6327682174444997,
+      "learning_rate": 9.999972615121014e-06,
+      "loss": 0.4892,
+      "step": 1276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.792579244916774,
+      "learning_rate": 9.999971295556112e-06,
+      "loss": 0.3489,
+      "step": 1277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0351754332240546,
+      "learning_rate": 9.999969944942745e-06,
+      "loss": 0.442,
+      "step": 1278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1101150627311256,
+      "learning_rate": 9.99996856328092e-06,
+      "loss": 0.3765,
+      "step": 1279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5514570337049634,
+      "learning_rate": 9.999967150570646e-06,
+      "loss": 0.3102,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6471780128385554,
+      "learning_rate": 9.999965706811933e-06,
+      "loss": 0.2303,
+      "step": 1281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7135987260546286,
+      "learning_rate": 9.999964232004788e-06,
+      "loss": 0.4245,
+      "step": 1282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1794456703273775,
+      "learning_rate": 9.999962726149221e-06,
+      "loss": 0.4632,
+      "step": 1283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.394420941722879,
+      "learning_rate": 9.999961189245244e-06,
+      "loss": 0.4007,
+      "step": 1284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.318684890120838,
+      "learning_rate": 9.99995962129286e-06,
+      "loss": 0.4112,
+      "step": 1285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.502921007117672,
+      "learning_rate": 9.999958022292087e-06,
+      "loss": 0.1756,
+      "step": 1286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.397025399677403,
+      "learning_rate": 9.999956392242928e-06,
+      "loss": 0.2066,
+      "step": 1287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9734261134129487,
+      "learning_rate": 9.999954731145396e-06,
+      "loss": 0.3154,
+      "step": 1288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7855897289500497,
+      "learning_rate": 9.999953038999502e-06,
+      "loss": 0.2899,
+      "step": 1289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3373591584598503,
+      "learning_rate": 9.999951315805255e-06,
+      "loss": 0.3529,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.289757974450396,
+      "learning_rate": 9.999949561562667e-06,
+      "loss": 0.3074,
+      "step": 1291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.340149148518494,
+      "learning_rate": 9.999947776271748e-06,
+      "loss": 0.6423,
+      "step": 1292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4712587116379265,
+      "learning_rate": 9.999945959932509e-06,
+      "loss": 0.5544,
+      "step": 1293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0617473413841823,
+      "learning_rate": 9.999944112544962e-06,
+      "loss": 0.3361,
+      "step": 1294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.610604310706024,
+      "learning_rate": 9.999942234109118e-06,
+      "loss": 0.5632,
+      "step": 1295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5887974485417184,
+      "learning_rate": 9.999940324624987e-06,
+      "loss": 0.3976,
+      "step": 1296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2338062509521004,
+      "learning_rate": 9.999938384092587e-06,
+      "loss": 0.5098,
+      "step": 1297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.620036740655126,
+      "learning_rate": 9.99993641251192e-06,
+      "loss": 0.4891,
+      "step": 1298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.204514433964072,
+      "learning_rate": 9.99993440988301e-06,
+      "loss": 0.3208,
+      "step": 1299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.418813708530229,
+      "learning_rate": 9.999932376205858e-06,
+      "loss": 0.2286,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.010388654965022,
+      "learning_rate": 9.999930311480485e-06,
+      "loss": 0.3676,
+      "step": 1301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8783257354946814,
+      "learning_rate": 9.9999282157069e-06,
+      "loss": 0.3395,
+      "step": 1302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7052195873523384,
+      "learning_rate": 9.999926088885117e-06,
+      "loss": 0.37,
+      "step": 1303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4356346083978413,
+      "learning_rate": 9.99992393101515e-06,
+      "loss": 0.5735,
+      "step": 1304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7483548613329343,
+      "learning_rate": 9.99992174209701e-06,
+      "loss": 0.3151,
+      "step": 1305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0837899558018655,
+      "learning_rate": 9.999919522130712e-06,
+      "loss": 0.3112,
+      "step": 1306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7713073675682205,
+      "learning_rate": 9.99991727111627e-06,
+      "loss": 0.3749,
+      "step": 1307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8670631909988316,
+      "learning_rate": 9.999914989053697e-06,
+      "loss": 0.1907,
+      "step": 1308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3326143711177176,
+      "learning_rate": 9.999912675943009e-06,
+      "loss": 0.2172,
+      "step": 1309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5718377342603485,
+      "learning_rate": 9.99991033178422e-06,
+      "loss": 0.1897,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.343518802047933,
+      "learning_rate": 9.999907956577341e-06,
+      "loss": 0.3662,
+      "step": 1311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1405058883328922,
+      "learning_rate": 9.999905550322392e-06,
+      "loss": 0.3923,
+      "step": 1312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.540066095228122,
+      "learning_rate": 9.999903113019386e-06,
+      "loss": 0.2942,
+      "step": 1313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4617109748332564,
+      "learning_rate": 9.999900644668337e-06,
+      "loss": 0.4357,
+      "step": 1314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.710985632659204,
+      "learning_rate": 9.999898145269261e-06,
+      "loss": 0.1944,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.226185701511042,
+      "learning_rate": 9.999895614822172e-06,
+      "loss": 0.2609,
+      "step": 1316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.152547410974486,
+      "learning_rate": 9.999893053327088e-06,
+      "loss": 0.193,
+      "step": 1317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.16878734633421,
+      "learning_rate": 9.999890460784026e-06,
+      "loss": 0.3721,
+      "step": 1318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.657495318935287,
+      "learning_rate": 9.999887837193e-06,
+      "loss": 0.4849,
+      "step": 1319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4463425468643702,
+      "learning_rate": 9.999885182554023e-06,
+      "loss": 0.3568,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8309926200239603,
+      "learning_rate": 9.999882496867118e-06,
+      "loss": 0.72,
+      "step": 1321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1052018670560524,
+      "learning_rate": 9.999879780132298e-06,
+      "loss": 0.3412,
+      "step": 1322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1245605000230716,
+      "learning_rate": 9.99987703234958e-06,
+      "loss": 0.2554,
+      "step": 1323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.375946101170959,
+      "learning_rate": 9.999874253518983e-06,
+      "loss": 0.4104,
+      "step": 1324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.012641569157623,
+      "learning_rate": 9.999871443640523e-06,
+      "loss": 0.3807,
+      "step": 1325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.05579373999473,
+      "learning_rate": 9.999868602714216e-06,
+      "loss": 0.37,
+      "step": 1326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6490884372506835,
+      "learning_rate": 9.999865730740082e-06,
+      "loss": 0.2643,
+      "step": 1327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.562294876215932,
+      "learning_rate": 9.999862827718138e-06,
+      "loss": 0.391,
+      "step": 1328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0123819091458377,
+      "learning_rate": 9.999859893648401e-06,
+      "loss": 0.2714,
+      "step": 1329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.378556549978178,
+      "learning_rate": 9.99985692853089e-06,
+      "loss": 0.5183,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2206402638042566,
+      "learning_rate": 9.999853932365623e-06,
+      "loss": 0.2923,
+      "step": 1331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7227143431402716,
+      "learning_rate": 9.999850905152622e-06,
+      "loss": 0.187,
+      "step": 1332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.146875074248327,
+      "learning_rate": 9.9998478468919e-06,
+      "loss": 0.3546,
+      "step": 1333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0260426125773465,
+      "learning_rate": 9.999844757583479e-06,
+      "loss": 0.2793,
+      "step": 1334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2160534120347473,
+      "learning_rate": 9.999841637227378e-06,
+      "loss": 0.337,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.324955331465567,
+      "learning_rate": 9.999838485823619e-06,
+      "loss": 0.5096,
+      "step": 1336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6873219564146256,
+      "learning_rate": 9.999835303372217e-06,
+      "loss": 0.2964,
+      "step": 1337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.009727680776595,
+      "learning_rate": 9.999832089873194e-06,
+      "loss": 0.4556,
+      "step": 1338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.783733052449225,
+      "learning_rate": 9.99982884532657e-06,
+      "loss": 0.5448,
+      "step": 1339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.393902451812398,
+      "learning_rate": 9.999825569732364e-06,
+      "loss": 0.381,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4997122111016483,
+      "learning_rate": 9.999822263090597e-06,
+      "loss": 0.3057,
+      "step": 1341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4255745779751803,
+      "learning_rate": 9.99981892540129e-06,
+      "loss": 0.5455,
+      "step": 1342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0103261330994853,
+      "learning_rate": 9.999815556664466e-06,
+      "loss": 0.2818,
+      "step": 1343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9267976826245445,
+      "learning_rate": 9.99981215688014e-06,
+      "loss": 0.3626,
+      "step": 1344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2287611796825586,
+      "learning_rate": 9.999808726048339e-06,
+      "loss": 0.208,
+      "step": 1345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.897631361180933,
+      "learning_rate": 9.999805264169079e-06,
+      "loss": 0.2219,
+      "step": 1346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.308097379350296,
+      "learning_rate": 9.999801771242385e-06,
+      "loss": 0.2643,
+      "step": 1347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8144554220213176,
+      "learning_rate": 9.99979824726828e-06,
+      "loss": 0.3011,
+      "step": 1348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5360352472069296,
+      "learning_rate": 9.999794692246782e-06,
+      "loss": 0.5592,
+      "step": 1349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7107295085699246,
+      "learning_rate": 9.999791106177917e-06,
+      "loss": 0.6647,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2610152176507694,
+      "learning_rate": 9.999787489061702e-06,
+      "loss": 0.3024,
+      "step": 1351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.517826434972325,
+      "learning_rate": 9.999783840898166e-06,
+      "loss": 0.1584,
+      "step": 1352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.732187947478687,
+      "learning_rate": 9.999780161687326e-06,
+      "loss": 0.5173,
+      "step": 1353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.467650277219839,
+      "learning_rate": 9.999776451429209e-06,
+      "loss": 0.4099,
+      "step": 1354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.910734592804878,
+      "learning_rate": 9.999772710123836e-06,
+      "loss": 0.8473,
+      "step": 1355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7392172363322325,
+      "learning_rate": 9.99976893777123e-06,
+      "loss": 0.367,
+      "step": 1356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5137541015739497,
+      "learning_rate": 9.999765134371412e-06,
+      "loss": 0.3025,
+      "step": 1357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6701586460664135,
+      "learning_rate": 9.999761299924413e-06,
+      "loss": 0.3904,
+      "step": 1358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7713662290026886,
+      "learning_rate": 9.99975743443025e-06,
+      "loss": 0.1539,
+      "step": 1359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1407858586259993,
+      "learning_rate": 9.999753537888949e-06,
+      "loss": 0.4371,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7945859321958295,
+      "learning_rate": 9.999749610300536e-06,
+      "loss": 0.2114,
+      "step": 1361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.108927205910015,
+      "learning_rate": 9.999745651665033e-06,
+      "loss": 0.2386,
+      "step": 1362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8595011972829703,
+      "learning_rate": 9.999741661982466e-06,
+      "loss": 0.3893,
+      "step": 1363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9792368231473416,
+      "learning_rate": 9.999737641252859e-06,
+      "loss": 0.979,
+      "step": 1364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6464109296923168,
+      "learning_rate": 9.999733589476237e-06,
+      "loss": 0.2222,
+      "step": 1365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.214016515998653,
+      "learning_rate": 9.999729506652626e-06,
+      "loss": 0.3077,
+      "step": 1366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9033763988798125,
+      "learning_rate": 9.99972539278205e-06,
+      "loss": 0.2291,
+      "step": 1367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.529394093416209,
+      "learning_rate": 9.999721247864535e-06,
+      "loss": 0.5258,
+      "step": 1368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2879903743258403,
+      "learning_rate": 9.999717071900109e-06,
+      "loss": 0.3917,
+      "step": 1369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.909291347526172,
+      "learning_rate": 9.999712864888793e-06,
+      "loss": 0.2629,
+      "step": 1370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.285265210806212,
+      "learning_rate": 9.999708626830617e-06,
+      "loss": 0.5047,
+      "step": 1371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.690673883505778,
+      "learning_rate": 9.999704357725608e-06,
+      "loss": 0.3496,
+      "step": 1372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0807875019119066,
+      "learning_rate": 9.999700057573788e-06,
+      "loss": 0.4371,
+      "step": 1373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.022966529832893,
+      "learning_rate": 9.99969572637519e-06,
+      "loss": 0.2269,
+      "step": 1374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4764058999210503,
+      "learning_rate": 9.999691364129836e-06,
+      "loss": 0.5404,
+      "step": 1375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2739013906476897,
+      "learning_rate": 9.999686970837755e-06,
+      "loss": 0.5644,
+      "step": 1376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6061120751459321,
+      "learning_rate": 9.999682546498974e-06,
+      "loss": 0.3374,
+      "step": 1377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7026520956191262,
+      "learning_rate": 9.999678091113518e-06,
+      "loss": 0.3017,
+      "step": 1378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.584168088351041,
+      "learning_rate": 9.99967360468142e-06,
+      "loss": 0.2739,
+      "step": 1379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8965652637144135,
+      "learning_rate": 9.999669087202703e-06,
+      "loss": 0.3328,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.579462785135159,
+      "learning_rate": 9.999664538677398e-06,
+      "loss": 0.6169,
+      "step": 1381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6680022386987607,
+      "learning_rate": 9.999659959105531e-06,
+      "loss": 0.7266,
+      "step": 1382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.838094936296466,
+      "learning_rate": 9.999655348487132e-06,
+      "loss": 0.3356,
+      "step": 1383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6325704906296257,
+      "learning_rate": 9.99965070682223e-06,
+      "loss": 0.2688,
+      "step": 1384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.397483946773337,
+      "learning_rate": 9.999646034110853e-06,
+      "loss": 0.4261,
+      "step": 1385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5142759413885325,
+      "learning_rate": 9.99964133035303e-06,
+      "loss": 0.4347,
+      "step": 1386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.416600266865305,
+      "learning_rate": 9.999636595548789e-06,
+      "loss": 0.3079,
+      "step": 1387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.128582599081169,
+      "learning_rate": 9.999631829698162e-06,
+      "loss": 0.2142,
+      "step": 1388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.782037305842524,
+      "learning_rate": 9.999627032801177e-06,
+      "loss": 0.2364,
+      "step": 1389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3854157719895017,
+      "learning_rate": 9.999622204857864e-06,
+      "loss": 0.232,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1155865200887636,
+      "learning_rate": 9.999617345868253e-06,
+      "loss": 0.3944,
+      "step": 1391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6349638913085527,
+      "learning_rate": 9.999612455832374e-06,
+      "loss": 0.4311,
+      "step": 1392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.627660606441004,
+      "learning_rate": 9.999607534750257e-06,
+      "loss": 0.4226,
+      "step": 1393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1178228132429626,
+      "learning_rate": 9.999602582621935e-06,
+      "loss": 0.1692,
+      "step": 1394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7108186701254797,
+      "learning_rate": 9.999597599447436e-06,
+      "loss": 0.5013,
+      "step": 1395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.228668434995432,
+      "learning_rate": 9.999592585226791e-06,
+      "loss": 0.3458,
+      "step": 1396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.579171898723055,
+      "learning_rate": 9.999587539960034e-06,
+      "loss": 0.5255,
+      "step": 1397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.315303958357559,
+      "learning_rate": 9.999582463647193e-06,
+      "loss": 0.378,
+      "step": 1398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2970363145090875,
+      "learning_rate": 9.9995773562883e-06,
+      "loss": 0.5867,
+      "step": 1399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6346323307618584,
+      "learning_rate": 9.99957221788339e-06,
+      "loss": 0.1862,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0127845237243465,
+      "learning_rate": 9.99956704843249e-06,
+      "loss": 0.3819,
+      "step": 1401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6638520246473758,
+      "learning_rate": 9.999561847935635e-06,
+      "loss": 0.1776,
+      "step": 1402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9835640684583704,
+      "learning_rate": 9.999556616392858e-06,
+      "loss": 0.2611,
+      "step": 1403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1009270866734022,
+      "learning_rate": 9.999551353804189e-06,
+      "loss": 0.3511,
+      "step": 1404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1174175939772466,
+      "learning_rate": 9.999546060169662e-06,
+      "loss": 0.2631,
+      "step": 1405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.653673795528741,
+      "learning_rate": 9.99954073548931e-06,
+      "loss": 0.289,
+      "step": 1406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5850542409765884,
+      "learning_rate": 9.999535379763168e-06,
+      "loss": 0.3428,
+      "step": 1407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6080216161496141,
+      "learning_rate": 9.999529992991265e-06,
+      "loss": 0.1767,
+      "step": 1408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0267203416215933,
+      "learning_rate": 9.999524575173639e-06,
+      "loss": 0.2793,
+      "step": 1409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.986613816940732,
+      "learning_rate": 9.999519126310317e-06,
+      "loss": 0.6728,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3193582826843007,
+      "learning_rate": 9.99951364640134e-06,
+      "loss": 0.4834,
+      "step": 1411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.849839831962927,
+      "learning_rate": 9.99950813544674e-06,
+      "loss": 0.3523,
+      "step": 1412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9784354657789496,
+      "learning_rate": 9.999502593446548e-06,
+      "loss": 0.4551,
+      "step": 1413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.394489648902538,
+      "learning_rate": 9.999497020400804e-06,
+      "loss": 0.6288,
+      "step": 1414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3652418188238702,
+      "learning_rate": 9.999491416309536e-06,
+      "loss": 0.1891,
+      "step": 1415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5615210208074919,
+      "learning_rate": 9.999485781172783e-06,
+      "loss": 0.3495,
+      "step": 1416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6759760549998184,
+      "learning_rate": 9.99948011499058e-06,
+      "loss": 0.3012,
+      "step": 1417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2010807536371026,
+      "learning_rate": 9.999474417762962e-06,
+      "loss": 0.4865,
+      "step": 1418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.23641526498059,
+      "learning_rate": 9.999468689489964e-06,
+      "loss": 0.5067,
+      "step": 1419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9964456639933956,
+      "learning_rate": 9.999462930171619e-06,
+      "loss": 0.3433,
+      "step": 1420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.498877311241541,
+      "learning_rate": 9.999457139807967e-06,
+      "loss": 0.2343,
+      "step": 1421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0594688596739763,
+      "learning_rate": 9.999451318399042e-06,
+      "loss": 0.4891,
+      "step": 1422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0355781659649588,
+      "learning_rate": 9.999445465944878e-06,
+      "loss": 0.3694,
+      "step": 1423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.431810242891925,
+      "learning_rate": 9.999439582445517e-06,
+      "loss": 0.5243,
+      "step": 1424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.000206985044732,
+      "learning_rate": 9.999433667900991e-06,
+      "loss": 0.3291,
+      "step": 1425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.944521921441219,
+      "learning_rate": 9.999427722311337e-06,
+      "loss": 0.3318,
+      "step": 1426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.087731779102821,
+      "learning_rate": 9.999421745676592e-06,
+      "loss": 0.467,
+      "step": 1427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9190459882617232,
+      "learning_rate": 9.999415737996795e-06,
+      "loss": 0.2345,
+      "step": 1428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.019707939490743,
+      "learning_rate": 9.999409699271982e-06,
+      "loss": 0.3493,
+      "step": 1429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.059436632776768,
+      "learning_rate": 9.99940362950219e-06,
+      "loss": 0.3783,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.172041097642954,
+      "learning_rate": 9.999397528687459e-06,
+      "loss": 0.2363,
+      "step": 1431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4555754253385715,
+      "learning_rate": 9.999391396827824e-06,
+      "loss": 0.1864,
+      "step": 1432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.229225087551542,
+      "learning_rate": 9.999385233923324e-06,
+      "loss": 0.3791,
+      "step": 1433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2493144553924687,
+      "learning_rate": 9.999379039973997e-06,
+      "loss": 0.332,
+      "step": 1434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.738578230959231,
+      "learning_rate": 9.999372814979883e-06,
+      "loss": 0.3844,
+      "step": 1435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.494736709998176,
+      "learning_rate": 9.999366558941021e-06,
+      "loss": 0.5854,
+      "step": 1436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1177568831338123,
+      "learning_rate": 9.999360271857447e-06,
+      "loss": 0.4817,
+      "step": 1437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8262541767688798,
+      "learning_rate": 9.999353953729203e-06,
+      "loss": 0.2646,
+      "step": 1438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.22822429146836,
+      "learning_rate": 9.999347604556325e-06,
+      "loss": 0.3855,
+      "step": 1439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.729173322965547,
+      "learning_rate": 9.999341224338854e-06,
+      "loss": 0.2725,
+      "step": 1440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.329746225933963,
+      "learning_rate": 9.999334813076832e-06,
+      "loss": 0.2071,
+      "step": 1441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0479221177214537,
+      "learning_rate": 9.999328370770296e-06,
+      "loss": 0.3916,
+      "step": 1442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.157840173503807,
+      "learning_rate": 9.999321897419285e-06,
+      "loss": 0.4066,
+      "step": 1443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0286704336559302,
+      "learning_rate": 9.999315393023844e-06,
+      "loss": 0.3525,
+      "step": 1444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9298222452293154,
+      "learning_rate": 9.999308857584008e-06,
+      "loss": 0.3598,
+      "step": 1445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4669532394082974,
+      "learning_rate": 9.99930229109982e-06,
+      "loss": 0.2876,
+      "step": 1446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4322418266325303,
+      "learning_rate": 9.99929569357132e-06,
+      "loss": 0.5316,
+      "step": 1447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.094370835339109,
+      "learning_rate": 9.99928906499855e-06,
+      "loss": 0.2395,
+      "step": 1448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0107638157352477,
+      "learning_rate": 9.999282405381552e-06,
+      "loss": 0.284,
+      "step": 1449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3191692284350465,
+      "learning_rate": 9.999275714720364e-06,
+      "loss": 0.5675,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.359518681502751,
+      "learning_rate": 9.99926899301503e-06,
+      "loss": 0.3855,
+      "step": 1451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8820476798879977,
+      "learning_rate": 9.999262240265592e-06,
+      "loss": 0.4327,
+      "step": 1452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2787916772581114,
+      "learning_rate": 9.999255456472092e-06,
+      "loss": 0.3213,
+      "step": 1453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0949448016105534,
+      "learning_rate": 9.999248641634571e-06,
+      "loss": 0.3507,
+      "step": 1454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.99824077750447,
+      "learning_rate": 9.999241795753072e-06,
+      "loss": 0.3218,
+      "step": 1455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.174139225060894,
+      "learning_rate": 9.999234918827638e-06,
+      "loss": 0.455,
+      "step": 1456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2472524225147876,
+      "learning_rate": 9.999228010858311e-06,
+      "loss": 0.5748,
+      "step": 1457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.121382193158183,
+      "learning_rate": 9.999221071845133e-06,
+      "loss": 0.3388,
+      "step": 1458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.754148421224697,
+      "learning_rate": 9.999214101788147e-06,
+      "loss": 0.4872,
+      "step": 1459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4993513587361538,
+      "learning_rate": 9.999207100687398e-06,
+      "loss": 0.3014,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4842834996582013,
+      "learning_rate": 9.999200068542928e-06,
+      "loss": 0.2399,
+      "step": 1461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8339931946888715,
+      "learning_rate": 9.999193005354783e-06,
+      "loss": 0.2087,
+      "step": 1462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.588391565746357,
+      "learning_rate": 9.999185911123005e-06,
+      "loss": 0.4598,
+      "step": 1463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9193156071948023,
+      "learning_rate": 9.999178785847638e-06,
+      "loss": 0.2425,
+      "step": 1464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5522373435609895,
+      "learning_rate": 9.999171629528727e-06,
+      "loss": 0.3403,
+      "step": 1465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.252641898410673,
+      "learning_rate": 9.999164442166315e-06,
+      "loss": 0.2651,
+      "step": 1466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5458288911217757,
+      "learning_rate": 9.999157223760448e-06,
+      "loss": 0.555,
+      "step": 1467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.901268747345127,
+      "learning_rate": 9.99914997431117e-06,
+      "loss": 0.302,
+      "step": 1468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5991103310259234,
+      "learning_rate": 9.999142693818528e-06,
+      "loss": 0.2433,
+      "step": 1469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9176123903732873,
+      "learning_rate": 9.999135382282565e-06,
+      "loss": 0.2678,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0867193307022434,
+      "learning_rate": 9.999128039703326e-06,
+      "loss": 0.4795,
+      "step": 1471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.62338082012044,
+      "learning_rate": 9.999120666080859e-06,
+      "loss": 0.6636,
+      "step": 1472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.074389822419331,
+      "learning_rate": 9.999113261415207e-06,
+      "loss": 0.1901,
+      "step": 1473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3792000731618663,
+      "learning_rate": 9.999105825706419e-06,
+      "loss": 0.4219,
+      "step": 1474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.201547109710211,
+      "learning_rate": 9.99909835895454e-06,
+      "loss": 0.2293,
+      "step": 1475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0467694670283874,
+      "learning_rate": 9.999090861159615e-06,
+      "loss": 0.4251,
+      "step": 1476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0023409881751353,
+      "learning_rate": 9.99908333232169e-06,
+      "loss": 0.4829,
+      "step": 1477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7385219450194493,
+      "learning_rate": 9.999075772440816e-06,
+      "loss": 0.3144,
+      "step": 1478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.18993067233752,
+      "learning_rate": 9.999068181517036e-06,
+      "loss": 0.4186,
+      "step": 1479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.250643995629138,
+      "learning_rate": 9.999060559550398e-06,
+      "loss": 0.4559,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6639182647410866,
+      "learning_rate": 9.99905290654095e-06,
+      "loss": 0.4472,
+      "step": 1481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.237558617703114,
+      "learning_rate": 9.99904522248874e-06,
+      "loss": 0.5821,
+      "step": 1482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1835192358852513,
+      "learning_rate": 9.999037507393815e-06,
+      "loss": 0.5655,
+      "step": 1483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9178343296533549,
+      "learning_rate": 9.999029761256221e-06,
+      "loss": 0.3428,
+      "step": 1484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.044283358217653,
+      "learning_rate": 9.99902198407601e-06,
+      "loss": 0.3492,
+      "step": 1485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6504880977942216,
+      "learning_rate": 9.999014175853227e-06,
+      "loss": 0.3103,
+      "step": 1486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.189110750976256,
+      "learning_rate": 9.999006336587922e-06,
+      "loss": 0.6815,
+      "step": 1487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1441533758083504,
+      "learning_rate": 9.998998466280145e-06,
+      "loss": 0.2947,
+      "step": 1488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9441262912432622,
+      "learning_rate": 9.998990564929942e-06,
+      "loss": 0.2375,
+      "step": 1489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.106818861506256,
+      "learning_rate": 9.998982632537362e-06,
+      "loss": 0.1753,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9121836568959296,
+      "learning_rate": 9.998974669102458e-06,
+      "loss": 0.118,
+      "step": 1491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.517890170981269,
+      "learning_rate": 9.998966674625275e-06,
+      "loss": 0.6586,
+      "step": 1492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3566237317679106,
+      "learning_rate": 9.998958649105866e-06,
+      "loss": 0.2261,
+      "step": 1493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8194701410942826,
+      "learning_rate": 9.998950592544279e-06,
+      "loss": 0.3243,
+      "step": 1494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8746515249445235,
+      "learning_rate": 9.998942504940565e-06,
+      "loss": 0.9093,
+      "step": 1495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.264172086752697,
+      "learning_rate": 9.998934386294774e-06,
+      "loss": 0.4658,
+      "step": 1496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6592886990306877,
+      "learning_rate": 9.998926236606956e-06,
+      "loss": 0.4813,
+      "step": 1497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0493339081553796,
+      "learning_rate": 9.99891805587716e-06,
+      "loss": 0.3769,
+      "step": 1498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.406549088356879,
+      "learning_rate": 9.998909844105442e-06,
+      "loss": 0.5571,
+      "step": 1499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.440142715024762,
+      "learning_rate": 9.998901601291846e-06,
+      "loss": 0.3005,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.408057896055102,
+      "learning_rate": 9.99889332743643e-06,
+      "loss": 0.2724,
+      "step": 1501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6873683878119603,
+      "learning_rate": 9.99888502253924e-06,
+      "loss": 0.3303,
+      "step": 1502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8808840579252446,
+      "learning_rate": 9.998876686600333e-06,
+      "loss": 0.3289,
+      "step": 1503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5082340985219784,
+      "learning_rate": 9.998868319619754e-06,
+      "loss": 0.1628,
+      "step": 1504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.529962859974287,
+      "learning_rate": 9.998859921597559e-06,
+      "loss": 0.3879,
+      "step": 1505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4374358828988294,
+      "learning_rate": 9.998851492533802e-06,
+      "loss": 0.2078,
+      "step": 1506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.501525044205487,
+      "learning_rate": 9.998843032428528e-06,
+      "loss": 0.6175,
+      "step": 1507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.861603543378376,
+      "learning_rate": 9.998834541281798e-06,
+      "loss": 0.3324,
+      "step": 1508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1590982022344805,
+      "learning_rate": 9.998826019093662e-06,
+      "loss": 0.2856,
+      "step": 1509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3288634228471583,
+      "learning_rate": 9.998817465864172e-06,
+      "loss": 0.2161,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0006858126673217,
+      "learning_rate": 9.99880888159338e-06,
+      "loss": 0.3398,
+      "step": 1511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5259763622442484,
+      "learning_rate": 9.998800266281341e-06,
+      "loss": 0.1578,
+      "step": 1512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.660037257912359,
+      "learning_rate": 9.998791619928108e-06,
+      "loss": 0.7484,
+      "step": 1513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6396239541857391,
+      "learning_rate": 9.998782942533734e-06,
+      "loss": 0.3108,
+      "step": 1514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5079674123292635,
+      "learning_rate": 9.998774234098274e-06,
+      "loss": 0.2516,
+      "step": 1515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.034025929061132,
+      "learning_rate": 9.998765494621783e-06,
+      "loss": 0.3956,
+      "step": 1516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0229146136344824,
+      "learning_rate": 9.998756724104312e-06,
+      "loss": 0.2544,
+      "step": 1517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7118636975884773,
+      "learning_rate": 9.99874792254592e-06,
+      "loss": 0.5099,
+      "step": 1518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.036518358645166,
+      "learning_rate": 9.998739089946657e-06,
+      "loss": 0.4024,
+      "step": 1519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.499801592818062,
+      "learning_rate": 9.99873022630658e-06,
+      "loss": 0.1773,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1675172971521475,
+      "learning_rate": 9.998721331625744e-06,
+      "loss": 0.5132,
+      "step": 1521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4598496308525208,
+      "learning_rate": 9.998712405904205e-06,
+      "loss": 0.2103,
+      "step": 1522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.498653716941641,
+      "learning_rate": 9.998703449142018e-06,
+      "loss": 0.7119,
+      "step": 1523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3259531832398412,
+      "learning_rate": 9.998694461339237e-06,
+      "loss": 0.2362,
+      "step": 1524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.09073499677438,
+      "learning_rate": 9.998685442495921e-06,
+      "loss": 0.1946,
+      "step": 1525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8398142688762287,
+      "learning_rate": 9.998676392612121e-06,
+      "loss": 0.2145,
+      "step": 1526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0720236903988263,
+      "learning_rate": 9.9986673116879e-06,
+      "loss": 0.2088,
+      "step": 1527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0869534443703777,
+      "learning_rate": 9.998658199723308e-06,
+      "loss": 0.3318,
+      "step": 1528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7376593401882565,
+      "learning_rate": 9.998649056718405e-06,
+      "loss": 0.4605,
+      "step": 1529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.474923852269309,
+      "learning_rate": 9.998639882673247e-06,
+      "loss": 0.5239,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5637137544888526,
+      "learning_rate": 9.998630677587892e-06,
+      "loss": 0.452,
+      "step": 1531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5710919714301097,
+      "learning_rate": 9.998621441462395e-06,
+      "loss": 0.4579,
+      "step": 1532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.139438704183835,
+      "learning_rate": 9.998612174296815e-06,
+      "loss": 0.4193,
+      "step": 1533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3290154719086544,
+      "learning_rate": 9.998602876091209e-06,
+      "loss": 0.3619,
+      "step": 1534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8510443248060637,
+      "learning_rate": 9.998593546845635e-06,
+      "loss": 0.4839,
+      "step": 1535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4021502719285626,
+      "learning_rate": 9.998584186560149e-06,
+      "loss": 0.3981,
+      "step": 1536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6435263230232775,
+      "learning_rate": 9.998574795234812e-06,
+      "loss": 0.2639,
+      "step": 1537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1073154850391695,
+      "learning_rate": 9.998565372869682e-06,
+      "loss": 0.501,
+      "step": 1538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8539162495275505,
+      "learning_rate": 9.998555919464816e-06,
+      "loss": 0.3216,
+      "step": 1539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.407786745581009,
+      "learning_rate": 9.998546435020274e-06,
+      "loss": 0.4046,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9281271096748176,
+      "learning_rate": 9.998536919536114e-06,
+      "loss": 0.365,
+      "step": 1541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6503028507705424,
+      "learning_rate": 9.998527373012395e-06,
+      "loss": 0.2689,
+      "step": 1542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.396069645457234,
+      "learning_rate": 9.998517795449177e-06,
+      "loss": 0.2942,
+      "step": 1543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0435057179770313,
+      "learning_rate": 9.998508186846518e-06,
+      "loss": 0.9898,
+      "step": 1544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9471173410293074,
+      "learning_rate": 9.99849854720448e-06,
+      "loss": 0.5652,
+      "step": 1545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.172804564921422,
+      "learning_rate": 9.998488876523122e-06,
+      "loss": 0.544,
+      "step": 1546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.069064113015389,
+      "learning_rate": 9.998479174802503e-06,
+      "loss": 0.3834,
+      "step": 1547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3706753155744,
+      "learning_rate": 9.998469442042685e-06,
+      "loss": 0.5206,
+      "step": 1548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8567535289013333,
+      "learning_rate": 9.998459678243727e-06,
+      "loss": 0.3079,
+      "step": 1549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2972686261613293,
+      "learning_rate": 9.998449883405687e-06,
+      "loss": 0.5078,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.226060449315398,
+      "learning_rate": 9.998440057528633e-06,
+      "loss": 0.3775,
+      "step": 1551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2113680066567443,
+      "learning_rate": 9.99843020061262e-06,
+      "loss": 0.4163,
+      "step": 1552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.044523453542722,
+      "learning_rate": 9.99842031265771e-06,
+      "loss": 0.2043,
+      "step": 1553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0473268279586154,
+      "learning_rate": 9.998410393663968e-06,
+      "loss": 0.3883,
+      "step": 1554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.699779899332694,
+      "learning_rate": 9.99840044363145e-06,
+      "loss": 0.4614,
+      "step": 1555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1993524771597044,
+      "learning_rate": 9.998390462560224e-06,
+      "loss": 0.3929,
+      "step": 1556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9073908784384033,
+      "learning_rate": 9.998380450450347e-06,
+      "loss": 0.4242,
+      "step": 1557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.25163054988827,
+      "learning_rate": 9.998370407301883e-06,
+      "loss": 0.4419,
+      "step": 1558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5272378031090164,
+      "learning_rate": 9.998360333114896e-06,
+      "loss": 0.2511,
+      "step": 1559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.094960083649503,
+      "learning_rate": 9.998350227889446e-06,
+      "loss": 0.2641,
+      "step": 1560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8456338859367447,
+      "learning_rate": 9.998340091625595e-06,
+      "loss": 0.3253,
+      "step": 1561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.453979489151378,
+      "learning_rate": 9.99832992432341e-06,
+      "loss": 0.5957,
+      "step": 1562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8703048811066598,
+      "learning_rate": 9.998319725982951e-06,
+      "loss": 0.3314,
+      "step": 1563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3010498045579837,
+      "learning_rate": 9.998309496604282e-06,
+      "loss": 0.2786,
+      "step": 1564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4265743746592845,
+      "learning_rate": 9.998299236187466e-06,
+      "loss": 0.4284,
+      "step": 1565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7673978513393356,
+      "learning_rate": 9.998288944732568e-06,
+      "loss": 0.3646,
+      "step": 1566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9831563406462602,
+      "learning_rate": 9.998278622239653e-06,
+      "loss": 0.2828,
+      "step": 1567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6951863092490667,
+      "learning_rate": 9.99826826870878e-06,
+      "loss": 0.2811,
+      "step": 1568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8748755175276148,
+      "learning_rate": 9.998257884140019e-06,
+      "loss": 0.4528,
+      "step": 1569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3947425890510443,
+      "learning_rate": 9.998247468533431e-06,
+      "loss": 0.1831,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.528837568604896,
+      "learning_rate": 9.998237021889081e-06,
+      "loss": 0.1633,
+      "step": 1571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2219781897253017,
+      "learning_rate": 9.998226544207035e-06,
+      "loss": 0.265,
+      "step": 1572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1308192206497103,
+      "learning_rate": 9.998216035487359e-06,
+      "loss": 0.4895,
+      "step": 1573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.332901771406108,
+      "learning_rate": 9.998205495730117e-06,
+      "loss": 0.4023,
+      "step": 1574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7906193143944074,
+      "learning_rate": 9.998194924935374e-06,
+      "loss": 0.2858,
+      "step": 1575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3664709020014927,
+      "learning_rate": 9.998184323103194e-06,
+      "loss": 0.2607,
+      "step": 1576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4584497987239056,
+      "learning_rate": 9.998173690233648e-06,
+      "loss": 0.466,
+      "step": 1577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6365646204228343,
+      "learning_rate": 9.998163026326798e-06,
+      "loss": 0.4631,
+      "step": 1578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.374151831308621,
+      "learning_rate": 9.998152331382712e-06,
+      "loss": 0.3135,
+      "step": 1579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4637708040537674,
+      "learning_rate": 9.998141605401453e-06,
+      "loss": 0.1787,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2858427271879975,
+      "learning_rate": 9.998130848383091e-06,
+      "loss": 0.4321,
+      "step": 1581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0962608142862726,
+      "learning_rate": 9.998120060327694e-06,
+      "loss": 0.6065,
+      "step": 1582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.979875829259674,
+      "learning_rate": 9.998109241235324e-06,
+      "loss": 0.3181,
+      "step": 1583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4084051254509367,
+      "learning_rate": 9.998098391106052e-06,
+      "loss": 0.2147,
+      "step": 1584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6374684621962488,
+      "learning_rate": 9.998087509939946e-06,
+      "loss": 0.1611,
+      "step": 1585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3821988175388658,
+      "learning_rate": 9.998076597737073e-06,
+      "loss": 0.2141,
+      "step": 1586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.322873356049032,
+      "learning_rate": 9.998065654497496e-06,
+      "loss": 0.3393,
+      "step": 1587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0471238065153483,
+      "learning_rate": 9.99805468022129e-06,
+      "loss": 0.4499,
+      "step": 1588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.967222840019733,
+      "learning_rate": 9.998043674908518e-06,
+      "loss": 0.3618,
+      "step": 1589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2380156147198154,
+      "learning_rate": 9.998032638559252e-06,
+      "loss": 0.2875,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.345581280214218,
+      "learning_rate": 9.998021571173558e-06,
+      "loss": 0.3298,
+      "step": 1591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5306670612618625,
+      "learning_rate": 9.998010472751504e-06,
+      "loss": 0.4166,
+      "step": 1592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4073616655363606,
+      "learning_rate": 9.997999343293164e-06,
+      "loss": 0.2866,
+      "step": 1593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7574704155256515,
+      "learning_rate": 9.9979881827986e-06,
+      "loss": 0.3393,
+      "step": 1594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.303183876537632,
+      "learning_rate": 9.997976991267885e-06,
+      "loss": 0.2422,
+      "step": 1595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6079084417784049,
+      "learning_rate": 9.997965768701091e-06,
+      "loss": 0.3108,
+      "step": 1596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.55952887736246,
+      "learning_rate": 9.997954515098285e-06,
+      "loss": 0.3033,
+      "step": 1597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5067689394050405,
+      "learning_rate": 9.997943230459534e-06,
+      "loss": 0.74,
+      "step": 1598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0494293243587385,
+      "learning_rate": 9.997931914784913e-06,
+      "loss": 0.2423,
+      "step": 1599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.062011462208657,
+      "learning_rate": 9.99792056807449e-06,
+      "loss": 0.2329,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.09776584331852,
+      "learning_rate": 9.997909190328336e-06,
+      "loss": 0.4041,
+      "step": 1601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6031488215106184,
+      "learning_rate": 9.997897781546523e-06,
+      "loss": 0.2114,
+      "step": 1602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.211532219607657,
+      "learning_rate": 9.997886341729117e-06,
+      "loss": 0.4158,
+      "step": 1603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.711530450181109,
+      "learning_rate": 9.997874870876195e-06,
+      "loss": 0.7069,
+      "step": 1604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.557489206229782,
+      "learning_rate": 9.997863368987823e-06,
+      "loss": 0.6961,
+      "step": 1605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7662728058030412,
+      "learning_rate": 9.997851836064076e-06,
+      "loss": 0.4201,
+      "step": 1606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.797713681082383,
+      "learning_rate": 9.997840272105027e-06,
+      "loss": 0.311,
+      "step": 1607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7945546333855467,
+      "learning_rate": 9.997828677110741e-06,
+      "loss": 0.2333,
+      "step": 1608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.153287552564895,
+      "learning_rate": 9.997817051081299e-06,
+      "loss": 0.3592,
+      "step": 1609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.709811634928875,
+      "learning_rate": 9.997805394016764e-06,
+      "loss": 0.3744,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.042514517124371,
+      "learning_rate": 9.997793705917216e-06,
+      "loss": 0.2427,
+      "step": 1611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.455241722792465,
+      "learning_rate": 9.997781986782724e-06,
+      "loss": 0.3157,
+      "step": 1612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7583636680740788,
+      "learning_rate": 9.99777023661336e-06,
+      "loss": 0.2327,
+      "step": 1613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9477823044510998,
+      "learning_rate": 9.997758455409198e-06,
+      "loss": 0.2276,
+      "step": 1614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0748705297891106,
+      "learning_rate": 9.997746643170315e-06,
+      "loss": 0.2588,
+      "step": 1615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.772090717670352,
+      "learning_rate": 9.997734799896777e-06,
+      "loss": 0.3512,
+      "step": 1616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.535313610330402,
+      "learning_rate": 9.997722925588663e-06,
+      "loss": 0.197,
+      "step": 1617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9597618218564288,
+      "learning_rate": 9.997711020246045e-06,
+      "loss": 0.254,
+      "step": 1618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6183601564464851,
+      "learning_rate": 9.997699083868998e-06,
+      "loss": 0.208,
+      "step": 1619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.037827881660084,
+      "learning_rate": 9.997687116457593e-06,
+      "loss": 0.4192,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6478415862473246,
+      "learning_rate": 9.997675118011907e-06,
+      "loss": 0.7951,
+      "step": 1621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9496147044098855,
+      "learning_rate": 9.997663088532015e-06,
+      "loss": 0.2727,
+      "step": 1622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7702545536149967,
+      "learning_rate": 9.99765102801799e-06,
+      "loss": 0.2379,
+      "step": 1623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2824667925367113,
+      "learning_rate": 9.997638936469907e-06,
+      "loss": 0.4922,
+      "step": 1624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.10481781531319,
+      "learning_rate": 9.997626813887842e-06,
+      "loss": 0.3261,
+      "step": 1625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.60719320401925,
+      "learning_rate": 9.997614660271868e-06,
+      "loss": 0.3873,
+      "step": 1626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.049195021296708,
+      "learning_rate": 9.997602475622066e-06,
+      "loss": 0.2591,
+      "step": 1627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7432502981701543,
+      "learning_rate": 9.997590259938506e-06,
+      "loss": 0.2714,
+      "step": 1628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8809562452174122,
+      "learning_rate": 9.997578013221265e-06,
+      "loss": 0.2482,
+      "step": 1629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.242019135271419,
+      "learning_rate": 9.99756573547042e-06,
+      "loss": 0.3711,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9794203411627018,
+      "learning_rate": 9.99755342668605e-06,
+      "loss": 0.3418,
+      "step": 1631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.147014054469053,
+      "learning_rate": 9.997541086868227e-06,
+      "loss": 0.3289,
+      "step": 1632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.280349109382446,
+      "learning_rate": 9.997528716017029e-06,
+      "loss": 0.4795,
+      "step": 1633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7245866702735577,
+      "learning_rate": 9.997516314132532e-06,
+      "loss": 0.2975,
+      "step": 1634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2135362242984569,
+      "learning_rate": 9.997503881214816e-06,
+      "loss": 0.1717,
+      "step": 1635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2588182802862837,
+      "learning_rate": 9.997491417263955e-06,
+      "loss": 0.4747,
+      "step": 1636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.876902365511883,
+      "learning_rate": 9.997478922280029e-06,
+      "loss": 0.3001,
+      "step": 1637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.386222356942886,
+      "learning_rate": 9.997466396263113e-06,
+      "loss": 0.4638,
+      "step": 1638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3490940965235525,
+      "learning_rate": 9.997453839213287e-06,
+      "loss": 0.1955,
+      "step": 1639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.684098758763749,
+      "learning_rate": 9.997441251130626e-06,
+      "loss": 0.3404,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.333555531707633,
+      "learning_rate": 9.997428632015211e-06,
+      "loss": 0.3806,
+      "step": 1641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1364853363587337,
+      "learning_rate": 9.99741598186712e-06,
+      "loss": 0.2105,
+      "step": 1642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8786457301486073,
+      "learning_rate": 9.997403300686432e-06,
+      "loss": 0.3644,
+      "step": 1643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5967411924884716,
+      "learning_rate": 9.997390588473224e-06,
+      "loss": 0.8759,
+      "step": 1644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3374071091283986,
+      "learning_rate": 9.997377845227577e-06,
+      "loss": 0.2221,
+      "step": 1645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1436959813154655,
+      "learning_rate": 9.997365070949566e-06,
+      "loss": 0.3808,
+      "step": 1646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9335478035448173,
+      "learning_rate": 9.997352265639275e-06,
+      "loss": 0.4564,
+      "step": 1647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6495785402509535,
+      "learning_rate": 9.997339429296782e-06,
+      "loss": 0.3752,
+      "step": 1648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7679373866058377,
+      "learning_rate": 9.997326561922167e-06,
+      "loss": 0.3068,
+      "step": 1649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.465335014101197,
+      "learning_rate": 9.997313663515506e-06,
+      "loss": 0.2423,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6821590647879856,
+      "learning_rate": 9.997300734076887e-06,
+      "loss": 0.3102,
+      "step": 1651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.831343521274611,
+      "learning_rate": 9.997287773606383e-06,
+      "loss": 0.2611,
+      "step": 1652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7180103531154698,
+      "learning_rate": 9.997274782104078e-06,
+      "loss": 0.1941,
+      "step": 1653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4864695678461042,
+      "learning_rate": 9.997261759570051e-06,
+      "loss": 0.3478,
+      "step": 1654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.233240546486307,
+      "learning_rate": 9.997248706004385e-06,
+      "loss": 0.3545,
+      "step": 1655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8086125844547363,
+      "learning_rate": 9.99723562140716e-06,
+      "loss": 0.3654,
+      "step": 1656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3510403734664727,
+      "learning_rate": 9.997222505778457e-06,
+      "loss": 0.4636,
+      "step": 1657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2311449979870805,
+      "learning_rate": 9.997209359118355e-06,
+      "loss": 0.3178,
+      "step": 1658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.195696020055465,
+      "learning_rate": 9.997196181426943e-06,
+      "loss": 0.4039,
+      "step": 1659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.096313483670769,
+      "learning_rate": 9.997182972704294e-06,
+      "loss": 0.4018,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6577401407865064,
+      "learning_rate": 9.997169732950496e-06,
+      "loss": 0.2257,
+      "step": 1661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.019592892943761,
+      "learning_rate": 9.997156462165628e-06,
+      "loss": 0.3843,
+      "step": 1662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7226029515401018,
+      "learning_rate": 9.997143160349773e-06,
+      "loss": 0.1916,
+      "step": 1663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5979655203601606,
+      "learning_rate": 9.997129827503016e-06,
+      "loss": 0.6668,
+      "step": 1664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5404803440160957,
+      "learning_rate": 9.997116463625439e-06,
+      "loss": 0.6779,
+      "step": 1665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.300397015719468,
+      "learning_rate": 9.997103068717121e-06,
+      "loss": 0.1726,
+      "step": 1666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9778875199361476,
+      "learning_rate": 9.99708964277815e-06,
+      "loss": 0.4551,
+      "step": 1667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.208462121345029,
+      "learning_rate": 9.997076185808607e-06,
+      "loss": 0.473,
+      "step": 1668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.105757872170148,
+      "learning_rate": 9.997062697808577e-06,
+      "loss": 0.2731,
+      "step": 1669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6528222013474294,
+      "learning_rate": 9.997049178778142e-06,
+      "loss": 0.2679,
+      "step": 1670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.91043430822324,
+      "learning_rate": 9.997035628717386e-06,
+      "loss": 0.2209,
+      "step": 1671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3833902752905405,
+      "learning_rate": 9.997022047626394e-06,
+      "loss": 0.4857,
+      "step": 1672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3615868791869263,
+      "learning_rate": 9.997008435505254e-06,
+      "loss": 0.5165,
+      "step": 1673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2056404285353404,
+      "learning_rate": 9.996994792354043e-06,
+      "loss": 0.2378,
+      "step": 1674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4751432084112241,
+      "learning_rate": 9.99698111817285e-06,
+      "loss": 0.2669,
+      "step": 1675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.950755971394177,
+      "learning_rate": 9.996967412961761e-06,
+      "loss": 0.3704,
+      "step": 1676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0491636199748826,
+      "learning_rate": 9.99695367672086e-06,
+      "loss": 0.3696,
+      "step": 1677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7067758761744267,
+      "learning_rate": 9.99693990945023e-06,
+      "loss": 0.2583,
+      "step": 1678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.629391578665575,
+      "learning_rate": 9.996926111149959e-06,
+      "loss": 0.3991,
+      "step": 1679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3693247548226926,
+      "learning_rate": 9.996912281820132e-06,
+      "loss": 0.5957,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.043438849146086,
+      "learning_rate": 9.996898421460835e-06,
+      "loss": 0.2074,
+      "step": 1681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7665071688897167,
+      "learning_rate": 9.996884530072153e-06,
+      "loss": 0.3248,
+      "step": 1682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.610927665207955,
+      "learning_rate": 9.996870607654176e-06,
+      "loss": 0.3264,
+      "step": 1683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.621198433208274,
+      "learning_rate": 9.996856654206985e-06,
+      "loss": 0.5433,
+      "step": 1684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.225232546955284,
+      "learning_rate": 9.996842669730671e-06,
+      "loss": 0.4727,
+      "step": 1685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5632535245233634,
+      "learning_rate": 9.99682865422532e-06,
+      "loss": 0.6108,
+      "step": 1686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.582022212441191,
+      "learning_rate": 9.996814607691015e-06,
+      "loss": 0.5751,
+      "step": 1687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5729368589591453,
+      "learning_rate": 9.996800530127848e-06,
+      "loss": 0.2414,
+      "step": 1688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3856763924913191,
+      "learning_rate": 9.996786421535906e-06,
+      "loss": 0.1653,
+      "step": 1689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.69311203626592,
+      "learning_rate": 9.996772281915274e-06,
+      "loss": 0.4276,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1923501197059574,
+      "learning_rate": 9.996758111266042e-06,
+      "loss": 0.2671,
+      "step": 1691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4672646460330836,
+      "learning_rate": 9.996743909588299e-06,
+      "loss": 0.6446,
+      "step": 1692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4365939632750857,
+      "learning_rate": 9.996729676882129e-06,
+      "loss": 0.3386,
+      "step": 1693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2720553745646725,
+      "learning_rate": 9.996715413147625e-06,
+      "loss": 0.2454,
+      "step": 1694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8078361514628907,
+      "learning_rate": 9.99670111838487e-06,
+      "loss": 0.3006,
+      "step": 1695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1305376608950315,
+      "learning_rate": 9.996686792593959e-06,
+      "loss": 0.2129,
+      "step": 1696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7836729268542988,
+      "learning_rate": 9.996672435774977e-06,
+      "loss": 0.3379,
+      "step": 1697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.313017793627772,
+      "learning_rate": 9.996658047928016e-06,
+      "loss": 0.2665,
+      "step": 1698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.21922088213354,
+      "learning_rate": 9.996643629053161e-06,
+      "loss": 0.2466,
+      "step": 1699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9994727609493204,
+      "learning_rate": 9.996629179150507e-06,
+      "loss": 0.2375,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7309536674262371,
+      "learning_rate": 9.99661469822014e-06,
+      "loss": 0.2356,
+      "step": 1701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5408424671779026,
+      "learning_rate": 9.996600186262149e-06,
+      "loss": 0.2115,
+      "step": 1702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7879734201827826,
+      "learning_rate": 9.996585643276628e-06,
+      "loss": 0.2364,
+      "step": 1703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.200167971126635,
+      "learning_rate": 9.996571069263665e-06,
+      "loss": 0.5728,
+      "step": 1704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9768047485870166,
+      "learning_rate": 9.99655646422335e-06,
+      "loss": 0.3611,
+      "step": 1705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.054560610258007,
+      "learning_rate": 9.996541828155775e-06,
+      "loss": 0.5293,
+      "step": 1706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.154462266731005,
+      "learning_rate": 9.99652716106103e-06,
+      "loss": 0.461,
+      "step": 1707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1112723612236364,
+      "learning_rate": 9.996512462939208e-06,
+      "loss": 0.5856,
+      "step": 1708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0647279742283513,
+      "learning_rate": 9.996497733790395e-06,
+      "loss": 0.4074,
+      "step": 1709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8050511476820092,
+      "learning_rate": 9.99648297361469e-06,
+      "loss": 0.3826,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3489478124071366,
+      "learning_rate": 9.99646818241218e-06,
+      "loss": 0.28,
+      "step": 1711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.330238105098598,
+      "learning_rate": 9.996453360182956e-06,
+      "loss": 0.2378,
+      "step": 1712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.124336770284101,
+      "learning_rate": 9.996438506927113e-06,
+      "loss": 0.3794,
+      "step": 1713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.15182089988826,
+      "learning_rate": 9.996423622644742e-06,
+      "loss": 0.2498,
+      "step": 1714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2004167160105947,
+      "learning_rate": 9.996408707335933e-06,
+      "loss": 0.1643,
+      "step": 1715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9967551432713229,
+      "learning_rate": 9.996393761000784e-06,
+      "loss": 0.5689,
+      "step": 1716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6382488391237005,
+      "learning_rate": 9.996378783639385e-06,
+      "loss": 0.3626,
+      "step": 1717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1124613682188427,
+      "learning_rate": 9.996363775251826e-06,
+      "loss": 0.3466,
+      "step": 1718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5382543659331138,
+      "learning_rate": 9.996348735838207e-06,
+      "loss": 0.3984,
+      "step": 1719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7789384792928156,
+      "learning_rate": 9.996333665398614e-06,
+      "loss": 0.3352,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.617197299319446,
+      "learning_rate": 9.996318563933146e-06,
+      "loss": 0.3808,
+      "step": 1721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9148432248759553,
+      "learning_rate": 9.996303431441893e-06,
+      "loss": 0.3799,
+      "step": 1722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.279668050451631,
+      "learning_rate": 9.996288267924953e-06,
+      "loss": 0.4677,
+      "step": 1723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.531958245926336,
+      "learning_rate": 9.996273073382414e-06,
+      "loss": 0.2022,
+      "step": 1724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0391554318001495,
+      "learning_rate": 9.996257847814378e-06,
+      "loss": 0.3496,
+      "step": 1725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7615988134424496,
+      "learning_rate": 9.996242591220934e-06,
+      "loss": 0.1796,
+      "step": 1726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9244412214971605,
+      "learning_rate": 9.99622730360218e-06,
+      "loss": 0.2741,
+      "step": 1727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.631220440913731,
+      "learning_rate": 9.996211984958208e-06,
+      "loss": 0.4366,
+      "step": 1728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6273534403643848,
+      "learning_rate": 9.996196635289115e-06,
+      "loss": 0.1819,
+      "step": 1729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.694415089577331,
+      "learning_rate": 9.996181254594996e-06,
+      "loss": 0.6271,
+      "step": 1730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4371652128704877,
+      "learning_rate": 9.996165842875947e-06,
+      "loss": 0.5072,
+      "step": 1731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8475799811972378,
+      "learning_rate": 9.996150400132062e-06,
+      "loss": 0.3291,
+      "step": 1732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.224199955669571,
+      "learning_rate": 9.99613492636344e-06,
+      "loss": 0.3678,
+      "step": 1733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.914287349573077,
+      "learning_rate": 9.996119421570173e-06,
+      "loss": 0.2433,
+      "step": 1734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.014669531944294,
+      "learning_rate": 9.99610388575236e-06,
+      "loss": 0.2493,
+      "step": 1735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2045666529723107,
+      "learning_rate": 9.996088318910095e-06,
+      "loss": 0.6067,
+      "step": 1736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.655695723509572,
+      "learning_rate": 9.996072721043479e-06,
+      "loss": 0.2,
+      "step": 1737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.032166699651038,
+      "learning_rate": 9.996057092152607e-06,
+      "loss": 0.4049,
+      "step": 1738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9119680214088317,
+      "learning_rate": 9.996041432237573e-06,
+      "loss": 0.2664,
+      "step": 1739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9649535920096246,
+      "learning_rate": 9.996025741298479e-06,
+      "loss": 0.3807,
+      "step": 1740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2631652000419145,
+      "learning_rate": 9.996010019335419e-06,
+      "loss": 0.4596,
+      "step": 1741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3275603979460815,
+      "learning_rate": 9.995994266348491e-06,
+      "loss": 0.366,
+      "step": 1742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.667420052981176,
+      "learning_rate": 9.995978482337794e-06,
+      "loss": 0.5643,
+      "step": 1743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.920521810250357,
+      "learning_rate": 9.995962667303427e-06,
+      "loss": 0.3586,
+      "step": 1744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4276232487005032,
+      "learning_rate": 9.995946821245484e-06,
+      "loss": 0.176,
+      "step": 1745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.862284251652549,
+      "learning_rate": 9.99593094416407e-06,
+      "loss": 0.1997,
+      "step": 1746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3806107651777415,
+      "learning_rate": 9.995915036059276e-06,
+      "loss": 0.3962,
+      "step": 1747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.32670553189169,
+      "learning_rate": 9.995899096931207e-06,
+      "loss": 0.3611,
+      "step": 1748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.946870629855252,
+      "learning_rate": 9.995883126779958e-06,
+      "loss": 0.3888,
+      "step": 1749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8060135118970735,
+      "learning_rate": 9.99586712560563e-06,
+      "loss": 0.2047,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7695501928599746,
+      "learning_rate": 9.995851093408323e-06,
+      "loss": 0.3507,
+      "step": 1751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3364671215198793,
+      "learning_rate": 9.995835030188135e-06,
+      "loss": 0.4443,
+      "step": 1752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1982721896943684,
+      "learning_rate": 9.995818935945167e-06,
+      "loss": 0.3114,
+      "step": 1753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5879113192900474,
+      "learning_rate": 9.995802810679519e-06,
+      "loss": 0.3822,
+      "step": 1754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8082867685232302,
+      "learning_rate": 9.995786654391289e-06,
+      "loss": 0.329,
+      "step": 1755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.073688044068036,
+      "learning_rate": 9.995770467080577e-06,
+      "loss": 0.2111,
+      "step": 1756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9970240916380595,
+      "learning_rate": 9.995754248747489e-06,
+      "loss": 0.3256,
+      "step": 1757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.118278109908388,
+      "learning_rate": 9.995737999392119e-06,
+      "loss": 0.3701,
+      "step": 1758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8532374085786303,
+      "learning_rate": 9.995721719014574e-06,
+      "loss": 0.365,
+      "step": 1759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3287400642895633,
+      "learning_rate": 9.995705407614948e-06,
+      "loss": 0.2502,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7294306407555462,
+      "learning_rate": 9.99568906519335e-06,
+      "loss": 0.2156,
+      "step": 1761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.53127579278048,
+      "learning_rate": 9.995672691749877e-06,
+      "loss": 0.1862,
+      "step": 1762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5451566611802579,
+      "learning_rate": 9.99565628728463e-06,
+      "loss": 0.178,
+      "step": 1763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7892215426963505,
+      "learning_rate": 9.995639851797714e-06,
+      "loss": 0.7502,
+      "step": 1764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1662688533509296,
+      "learning_rate": 9.995623385289229e-06,
+      "loss": 0.4292,
+      "step": 1765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8585800908000862,
+      "learning_rate": 9.995606887759275e-06,
+      "loss": 0.2694,
+      "step": 1766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2654176502054337,
+      "learning_rate": 9.995590359207961e-06,
+      "loss": 0.2116,
+      "step": 1767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.004750614852611,
+      "learning_rate": 9.995573799635383e-06,
+      "loss": 0.1796,
+      "step": 1768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9009279508850974,
+      "learning_rate": 9.995557209041647e-06,
+      "loss": 0.2881,
+      "step": 1769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7749752066452213,
+      "learning_rate": 9.995540587426857e-06,
+      "loss": 0.3067,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8001905370241063,
+      "learning_rate": 9.995523934791114e-06,
+      "loss": 0.3073,
+      "step": 1771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4097175384339162,
+      "learning_rate": 9.995507251134522e-06,
+      "loss": 0.591,
+      "step": 1772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1789464750574212,
+      "learning_rate": 9.995490536457184e-06,
+      "loss": 0.4808,
+      "step": 1773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.282879299828798,
+      "learning_rate": 9.995473790759204e-06,
+      "loss": 0.3581,
+      "step": 1774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4712407217486736,
+      "learning_rate": 9.99545701404069e-06,
+      "loss": 0.183,
+      "step": 1775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0056156652987593,
+      "learning_rate": 9.99544020630174e-06,
+      "loss": 0.3096,
+      "step": 1776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3936784666344093,
+      "learning_rate": 9.995423367542462e-06,
+      "loss": 0.3672,
+      "step": 1777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.29441603259932,
+      "learning_rate": 9.99540649776296e-06,
+      "loss": 0.38,
+      "step": 1778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.216377070668171,
+      "learning_rate": 9.995389596963338e-06,
+      "loss": 0.2983,
+      "step": 1779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2323299666047487,
+      "learning_rate": 9.9953726651437e-06,
+      "loss": 0.2494,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.702447485246879,
+      "learning_rate": 9.995355702304154e-06,
+      "loss": 0.1652,
+      "step": 1781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5259675844216267,
+      "learning_rate": 9.995338708444804e-06,
+      "loss": 0.4693,
+      "step": 1782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9939735036197228,
+      "learning_rate": 9.995321683565756e-06,
+      "loss": 0.402,
+      "step": 1783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.87231493860473,
+      "learning_rate": 9.995304627667114e-06,
+      "loss": 0.1539,
+      "step": 1784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.421012121734305,
+      "learning_rate": 9.995287540748986e-06,
+      "loss": 0.3754,
+      "step": 1785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.554989071451894,
+      "learning_rate": 9.995270422811477e-06,
+      "loss": 0.1751,
+      "step": 1786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7824937176025235,
+      "learning_rate": 9.995253273854692e-06,
+      "loss": 0.2079,
+      "step": 1787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0599835335347687,
+      "learning_rate": 9.995236093878739e-06,
+      "loss": 0.2019,
+      "step": 1788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2774130921980233,
+      "learning_rate": 9.995218882883727e-06,
+      "loss": 0.2202,
+      "step": 1789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1983302406764555,
+      "learning_rate": 9.995201640869758e-06,
+      "loss": 0.508,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3096462180828126,
+      "learning_rate": 9.995184367836942e-06,
+      "loss": 0.5551,
+      "step": 1791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0663178069832675,
+      "learning_rate": 9.995167063785388e-06,
+      "loss": 0.4013,
+      "step": 1792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0265004530138384,
+      "learning_rate": 9.995149728715198e-06,
+      "loss": 0.3925,
+      "step": 1793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4585533909866102,
+      "learning_rate": 9.995132362626485e-06,
+      "loss": 0.1891,
+      "step": 1794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.733219371189641,
+      "learning_rate": 9.995114965519355e-06,
+      "loss": 0.7867,
+      "step": 1795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1528832647327434,
+      "learning_rate": 9.995097537393914e-06,
+      "loss": 0.4151,
+      "step": 1796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.220736192173553,
+      "learning_rate": 9.995080078250271e-06,
+      "loss": 0.4266,
+      "step": 1797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8736155127103735,
+      "learning_rate": 9.995062588088539e-06,
+      "loss": 0.3034,
+      "step": 1798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9347589738755762,
+      "learning_rate": 9.995045066908821e-06,
+      "loss": 0.4013,
+      "step": 1799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7060022751720365,
+      "learning_rate": 9.995027514711228e-06,
+      "loss": 0.7692,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.630655463457621,
+      "learning_rate": 9.99500993149587e-06,
+      "loss": 0.3251,
+      "step": 1801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0114059130928315,
+      "learning_rate": 9.994992317262855e-06,
+      "loss": 0.3524,
+      "step": 1802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1791337584848693,
+      "learning_rate": 9.99497467201229e-06,
+      "loss": 0.4262,
+      "step": 1803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1030730989041968,
+      "learning_rate": 9.994956995744289e-06,
+      "loss": 0.2212,
+      "step": 1804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3245644596374075,
+      "learning_rate": 9.99493928845896e-06,
+      "loss": 0.1848,
+      "step": 1805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.224743866643779,
+      "learning_rate": 9.99492155015641e-06,
+      "loss": 0.3689,
+      "step": 1806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.077333041275042,
+      "learning_rate": 9.994903780836756e-06,
+      "loss": 0.4852,
+      "step": 1807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.746094270925893,
+      "learning_rate": 9.9948859805001e-06,
+      "loss": 0.2898,
+      "step": 1808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5352071347447134,
+      "learning_rate": 9.99486814914656e-06,
+      "loss": 0.2601,
+      "step": 1809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4730718011196247,
+      "learning_rate": 9.994850286776241e-06,
+      "loss": 0.2802,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.960661832650745,
+      "learning_rate": 9.994832393389257e-06,
+      "loss": 0.2502,
+      "step": 1811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7901845853084146,
+      "learning_rate": 9.99481446898572e-06,
+      "loss": 0.3422,
+      "step": 1812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5491561722632963,
+      "learning_rate": 9.99479651356574e-06,
+      "loss": 0.2247,
+      "step": 1813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.319868982999637,
+      "learning_rate": 9.994778527129426e-06,
+      "loss": 0.3672,
+      "step": 1814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5907458457120662,
+      "learning_rate": 9.994760509676892e-06,
+      "loss": 0.2284,
+      "step": 1815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.432422401963583,
+      "learning_rate": 9.994742461208251e-06,
+      "loss": 0.2991,
+      "step": 1816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0199638669249964,
+      "learning_rate": 9.994724381723616e-06,
+      "loss": 0.1104,
+      "step": 1817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5693159171483995,
+      "learning_rate": 9.994706271223093e-06,
+      "loss": 0.5009,
+      "step": 1818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9964055304948822,
+      "learning_rate": 9.9946881297068e-06,
+      "loss": 0.2534,
+      "step": 1819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.289361888588641,
+      "learning_rate": 9.99466995717485e-06,
+      "loss": 0.4118,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0228241226443453,
+      "learning_rate": 9.994651753627353e-06,
+      "loss": 0.2191,
+      "step": 1821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1888598166496784,
+      "learning_rate": 9.994633519064422e-06,
+      "loss": 0.3841,
+      "step": 1822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9547870125813924,
+      "learning_rate": 9.994615253486173e-06,
+      "loss": 0.3668,
+      "step": 1823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.126316882656444,
+      "learning_rate": 9.994596956892719e-06,
+      "loss": 0.2581,
+      "step": 1824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6151561996023265,
+      "learning_rate": 9.99457862928417e-06,
+      "loss": 0.3273,
+      "step": 1825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.174451572272321,
+      "learning_rate": 9.994560270660645e-06,
+      "loss": 0.5952,
+      "step": 1826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6246918101766255,
+      "learning_rate": 9.994541881022253e-06,
+      "loss": 0.1434,
+      "step": 1827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7596668634845722,
+      "learning_rate": 9.99452346036911e-06,
+      "loss": 0.3678,
+      "step": 1828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0604728432323447,
+      "learning_rate": 9.994505008701333e-06,
+      "loss": 0.4405,
+      "step": 1829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.580093297112418,
+      "learning_rate": 9.994486526019033e-06,
+      "loss": 0.6359,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6835699190369353,
+      "learning_rate": 9.994468012322327e-06,
+      "loss": 0.357,
+      "step": 1831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3910685296103655,
+      "learning_rate": 9.99444946761133e-06,
+      "loss": 0.4117,
+      "step": 1832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.343796430763644,
+      "learning_rate": 9.994430891886155e-06,
+      "loss": 0.5187,
+      "step": 1833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.433888278240232,
+      "learning_rate": 9.994412285146918e-06,
+      "loss": 0.3318,
+      "step": 1834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.496900150616417,
+      "learning_rate": 9.994393647393737e-06,
+      "loss": 0.4564,
+      "step": 1835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8168805169762203,
+      "learning_rate": 9.994374978626725e-06,
+      "loss": 0.255,
+      "step": 1836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 43.61924727583175,
+      "learning_rate": 9.994356278845997e-06,
+      "loss": 0.4799,
+      "step": 1837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6090547002179911,
+      "learning_rate": 9.994337548051674e-06,
+      "loss": 0.1697,
+      "step": 1838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1215824983199014,
+      "learning_rate": 9.994318786243869e-06,
+      "loss": 0.3714,
+      "step": 1839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.840540097429281,
+      "learning_rate": 9.994299993422698e-06,
+      "loss": 0.2582,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4732626468391385,
+      "learning_rate": 9.994281169588278e-06,
+      "loss": 0.3506,
+      "step": 1841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.02723528915489,
+      "learning_rate": 9.994262314740727e-06,
+      "loss": 0.4119,
+      "step": 1842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9068908044278272,
+      "learning_rate": 9.994243428880162e-06,
+      "loss": 0.2859,
+      "step": 1843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4646593499465426,
+      "learning_rate": 9.994224512006699e-06,
+      "loss": 0.7167,
+      "step": 1844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.38861208674792,
+      "learning_rate": 9.994205564120457e-06,
+      "loss": 0.4636,
+      "step": 1845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.516539555821553,
+      "learning_rate": 9.994186585221553e-06,
+      "loss": 0.213,
+      "step": 1846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5058289461735181,
+      "learning_rate": 9.994167575310104e-06,
+      "loss": 0.3242,
+      "step": 1847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4335208154466188,
+      "learning_rate": 9.99414853438623e-06,
+      "loss": 0.208,
+      "step": 1848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.142258369002531,
+      "learning_rate": 9.994129462450046e-06,
+      "loss": 0.2266,
+      "step": 1849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1734345072020185,
+      "learning_rate": 9.994110359501674e-06,
+      "loss": 0.3919,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1531893528150823,
+      "learning_rate": 9.994091225541232e-06,
+      "loss": 0.2648,
+      "step": 1851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.160554566315358,
+      "learning_rate": 9.994072060568836e-06,
+      "loss": 0.2867,
+      "step": 1852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.47250190129637,
+      "learning_rate": 9.994052864584608e-06,
+      "loss": 0.2264,
+      "step": 1853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2950744593576147,
+      "learning_rate": 9.994033637588665e-06,
+      "loss": 0.8703,
+      "step": 1854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6238693175007857,
+      "learning_rate": 9.99401437958113e-06,
+      "loss": 0.5931,
+      "step": 1855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7860089024965855,
+      "learning_rate": 9.993995090562119e-06,
+      "loss": 0.3126,
+      "step": 1856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8492036110532681,
+      "learning_rate": 9.993975770531752e-06,
+      "loss": 0.3068,
+      "step": 1857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5074576024440207,
+      "learning_rate": 9.993956419490149e-06,
+      "loss": 0.1995,
+      "step": 1858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5759403384862205,
+      "learning_rate": 9.993937037437434e-06,
+      "loss": 0.1687,
+      "step": 1859
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1452550517792806,
+      "learning_rate": 9.99391762437372e-06,
+      "loss": 0.5159,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.532931108068555,
+      "learning_rate": 9.993898180299137e-06,
+      "loss": 0.2186,
+      "step": 1861
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.137897257112392,
+      "learning_rate": 9.993878705213798e-06,
+      "loss": 0.532,
+      "step": 1862
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.854602796360591,
+      "learning_rate": 9.993859199117827e-06,
+      "loss": 0.3385,
+      "step": 1863
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.690971749145029,
+      "learning_rate": 9.993839662011343e-06,
+      "loss": 0.1989,
+      "step": 1864
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4645817716300442,
+      "learning_rate": 9.993820093894473e-06,
+      "loss": 0.1992,
+      "step": 1865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.886538550259056,
+      "learning_rate": 9.993800494767332e-06,
+      "loss": 0.233,
+      "step": 1866
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.864130250574101,
+      "learning_rate": 9.993780864630044e-06,
+      "loss": 0.3899,
+      "step": 1867
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1303029579529746,
+      "learning_rate": 9.993761203482732e-06,
+      "loss": 0.2633,
+      "step": 1868
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.189278164067253,
+      "learning_rate": 9.993741511325517e-06,
+      "loss": 0.2381,
+      "step": 1869
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.363143352552296,
+      "learning_rate": 9.993721788158521e-06,
+      "loss": 0.4707,
+      "step": 1870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6341607491303247,
+      "learning_rate": 9.993702033981869e-06,
+      "loss": 0.4741,
+      "step": 1871
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3050127737390564,
+      "learning_rate": 9.99368224879568e-06,
+      "loss": 0.5023,
+      "step": 1872
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.322964339780993,
+      "learning_rate": 9.99366243260008e-06,
+      "loss": 0.2366,
+      "step": 1873
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1110017843644377,
+      "learning_rate": 9.993642585395188e-06,
+      "loss": 0.2537,
+      "step": 1874
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4192728275965645,
+      "learning_rate": 9.993622707181132e-06,
+      "loss": 0.2666,
+      "step": 1875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.261384032349233,
+      "learning_rate": 9.993602797958033e-06,
+      "loss": 0.1461,
+      "step": 1876
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.396256203581213,
+      "learning_rate": 9.993582857726016e-06,
+      "loss": 0.4181,
+      "step": 1877
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6862419878724773,
+      "learning_rate": 9.993562886485202e-06,
+      "loss": 0.2301,
+      "step": 1878
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0373218019644503,
+      "learning_rate": 9.993542884235717e-06,
+      "loss": 0.3539,
+      "step": 1879
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9596779794390253,
+      "learning_rate": 9.993522850977685e-06,
+      "loss": 0.4004,
+      "step": 1880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5676764299361299,
+      "learning_rate": 9.99350278671123e-06,
+      "loss": 0.2175,
+      "step": 1881
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3440787752341943,
+      "learning_rate": 9.993482691436479e-06,
+      "loss": 0.5618,
+      "step": 1882
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.429217454111678,
+      "learning_rate": 9.993462565153553e-06,
+      "loss": 0.1873,
+      "step": 1883
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0944077744769474,
+      "learning_rate": 9.99344240786258e-06,
+      "loss": 0.1722,
+      "step": 1884
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.405175154765388,
+      "learning_rate": 9.993422219563683e-06,
+      "loss": 0.4344,
+      "step": 1885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0521689566795622,
+      "learning_rate": 9.993402000256987e-06,
+      "loss": 0.3335,
+      "step": 1886
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.915112888405196,
+      "learning_rate": 9.99338174994262e-06,
+      "loss": 0.329,
+      "step": 1887
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3643245695475095,
+      "learning_rate": 9.993361468620705e-06,
+      "loss": 0.5769,
+      "step": 1888
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0984917864449026,
+      "learning_rate": 9.993341156291374e-06,
+      "loss": 0.4109,
+      "step": 1889
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3078420643672202,
+      "learning_rate": 9.993320812954746e-06,
+      "loss": 0.216,
+      "step": 1890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9344011408028645,
+      "learning_rate": 9.99330043861095e-06,
+      "loss": 0.2015,
+      "step": 1891
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0162309615495095,
+      "learning_rate": 9.99328003326011e-06,
+      "loss": 0.241,
+      "step": 1892
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0097391670742506,
+      "learning_rate": 9.993259596902359e-06,
+      "loss": 0.3108,
+      "step": 1893
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8068441955602066,
+      "learning_rate": 9.993239129537819e-06,
+      "loss": 0.7482,
+      "step": 1894
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.931291562160918,
+      "learning_rate": 9.993218631166616e-06,
+      "loss": 0.324,
+      "step": 1895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.386532590508089,
+      "learning_rate": 9.993198101788882e-06,
+      "loss": 0.2376,
+      "step": 1896
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9801991804049517,
+      "learning_rate": 9.993177541404742e-06,
+      "loss": 0.2688,
+      "step": 1897
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.106847186211924,
+      "learning_rate": 9.993156950014323e-06,
+      "loss": 0.3045,
+      "step": 1898
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2368302071543449,
+      "learning_rate": 9.993136327617754e-06,
+      "loss": 0.1617,
+      "step": 1899
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6129429996012545,
+      "learning_rate": 9.993115674215162e-06,
+      "loss": 0.1952,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.247625853406234,
+      "learning_rate": 9.993094989806676e-06,
+      "loss": 0.2238,
+      "step": 1901
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.106219499321894,
+      "learning_rate": 9.993074274392425e-06,
+      "loss": 0.3193,
+      "step": 1902
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.831070688813547,
+      "learning_rate": 9.993053527972535e-06,
+      "loss": 0.1912,
+      "step": 1903
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7093303352043945,
+      "learning_rate": 9.993032750547138e-06,
+      "loss": 0.3475,
+      "step": 1904
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0166483906636645,
+      "learning_rate": 9.993011942116362e-06,
+      "loss": 0.2078,
+      "step": 1905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0862146900610457,
+      "learning_rate": 9.992991102680337e-06,
+      "loss": 0.5976,
+      "step": 1906
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8934574997401752,
+      "learning_rate": 9.992970232239189e-06,
+      "loss": 0.3342,
+      "step": 1907
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.198280032496014,
+      "learning_rate": 9.992949330793052e-06,
+      "loss": 0.5127,
+      "step": 1908
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2284164206032493,
+      "learning_rate": 9.992928398342052e-06,
+      "loss": 0.5492,
+      "step": 1909
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9496259244943657,
+      "learning_rate": 9.99290743488632e-06,
+      "loss": 0.3062,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2043207589450406,
+      "learning_rate": 9.992886440425991e-06,
+      "loss": 0.3815,
+      "step": 1911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.195501224374571,
+      "learning_rate": 9.992865414961188e-06,
+      "loss": 0.374,
+      "step": 1912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9260698744425997,
+      "learning_rate": 9.992844358492045e-06,
+      "loss": 0.2317,
+      "step": 1913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2969098510174204,
+      "learning_rate": 9.992823271018695e-06,
+      "loss": 0.5302,
+      "step": 1914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9352314612165054,
+      "learning_rate": 9.992802152541263e-06,
+      "loss": 0.3982,
+      "step": 1915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3161984259807062,
+      "learning_rate": 9.992781003059888e-06,
+      "loss": 0.1545,
+      "step": 1916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0640398620588445,
+      "learning_rate": 9.992759822574693e-06,
+      "loss": 0.4095,
+      "step": 1917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0267866615079537,
+      "learning_rate": 9.992738611085817e-06,
+      "loss": 0.306,
+      "step": 1918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8135163893920092,
+      "learning_rate": 9.992717368593385e-06,
+      "loss": 0.3298,
+      "step": 1919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.277753462668123,
+      "learning_rate": 9.992696095097533e-06,
+      "loss": 0.5242,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.30947275742037,
+      "learning_rate": 9.992674790598395e-06,
+      "loss": 0.5515,
+      "step": 1921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8179721112582907,
+      "learning_rate": 9.992653455096098e-06,
+      "loss": 0.2191,
+      "step": 1922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.759453736301902,
+      "learning_rate": 9.992632088590779e-06,
+      "loss": 0.7382,
+      "step": 1923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0608490707686316,
+      "learning_rate": 9.992610691082568e-06,
+      "loss": 0.4148,
+      "step": 1924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7556637409942804,
+      "learning_rate": 9.992589262571597e-06,
+      "loss": 0.3787,
+      "step": 1925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.322287531333119,
+      "learning_rate": 9.992567803058001e-06,
+      "loss": 0.3592,
+      "step": 1926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.844988164734549,
+      "learning_rate": 9.992546312541915e-06,
+      "loss": 0.4891,
+      "step": 1927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8578434996177617,
+      "learning_rate": 9.99252479102347e-06,
+      "loss": 0.4944,
+      "step": 1928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.282047060631267,
+      "learning_rate": 9.9925032385028e-06,
+      "loss": 0.5029,
+      "step": 1929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1766726518678188,
+      "learning_rate": 9.992481654980037e-06,
+      "loss": 0.2595,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7309110304473,
+      "learning_rate": 9.992460040455319e-06,
+      "loss": 0.2919,
+      "step": 1931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.697708888994451,
+      "learning_rate": 9.992438394928778e-06,
+      "loss": 0.2765,
+      "step": 1932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3479235922971073,
+      "learning_rate": 9.992416718400548e-06,
+      "loss": 0.1693,
+      "step": 1933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7218244235370075,
+      "learning_rate": 9.992395010870764e-06,
+      "loss": 0.2704,
+      "step": 1934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9592318697036126,
+      "learning_rate": 9.992373272339562e-06,
+      "loss": 0.3199,
+      "step": 1935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7077008206497484,
+      "learning_rate": 9.992351502807074e-06,
+      "loss": 0.3584,
+      "step": 1936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2300483220524967,
+      "learning_rate": 9.992329702273438e-06,
+      "loss": 0.4778,
+      "step": 1937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.378838592970308,
+      "learning_rate": 9.992307870738789e-06,
+      "loss": 0.4766,
+      "step": 1938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9839473886904684,
+      "learning_rate": 9.992286008203263e-06,
+      "loss": 0.3986,
+      "step": 1939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.951036526832253,
+      "learning_rate": 9.992264114666991e-06,
+      "loss": 0.4263,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5903057251986636,
+      "learning_rate": 9.992242190130117e-06,
+      "loss": 0.1599,
+      "step": 1941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9213105279922231,
+      "learning_rate": 9.992220234592772e-06,
+      "loss": 0.4938,
+      "step": 1942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.572737707026805,
+      "learning_rate": 9.992198248055091e-06,
+      "loss": 0.2122,
+      "step": 1943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6634778464270825,
+      "learning_rate": 9.992176230517215e-06,
+      "loss": 0.2231,
+      "step": 1944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8317817931905276,
+      "learning_rate": 9.992154181979279e-06,
+      "loss": 0.1547,
+      "step": 1945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1092261014590243,
+      "learning_rate": 9.992132102441418e-06,
+      "loss": 0.2196,
+      "step": 1946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0814486340919953,
+      "learning_rate": 9.99210999190377e-06,
+      "loss": 0.3773,
+      "step": 1947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.206705376527072,
+      "learning_rate": 9.992087850366474e-06,
+      "loss": 0.5555,
+      "step": 1948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.900726912811723,
+      "learning_rate": 9.992065677829667e-06,
+      "loss": 0.3817,
+      "step": 1949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6291642650486478,
+      "learning_rate": 9.992043474293484e-06,
+      "loss": 0.5882,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3698964542720315,
+      "learning_rate": 9.992021239758067e-06,
+      "loss": 0.2221,
+      "step": 1951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1128307257166488,
+      "learning_rate": 9.991998974223552e-06,
+      "loss": 0.4346,
+      "step": 1952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3128990524955384,
+      "learning_rate": 9.991976677690076e-06,
+      "loss": 0.3848,
+      "step": 1953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2220654621793148,
+      "learning_rate": 9.991954350157779e-06,
+      "loss": 0.2025,
+      "step": 1954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8700411788593463,
+      "learning_rate": 9.9919319916268e-06,
+      "loss": 0.5097,
+      "step": 1955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6596610890375243,
+      "learning_rate": 9.991909602097277e-06,
+      "loss": 0.2132,
+      "step": 1956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8632467874502865,
+      "learning_rate": 9.991887181569347e-06,
+      "loss": 0.3318,
+      "step": 1957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.462201674781625,
+      "learning_rate": 9.991864730043154e-06,
+      "loss": 0.2833,
+      "step": 1958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.541505096464584,
+      "learning_rate": 9.991842247518834e-06,
+      "loss": 0.4971,
+      "step": 1959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6303808172575207,
+      "learning_rate": 9.991819733996528e-06,
+      "loss": 0.2502,
+      "step": 1960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1307913771985407,
+      "learning_rate": 9.991797189476374e-06,
+      "loss": 0.355,
+      "step": 1961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.800454401735212,
+      "learning_rate": 9.991774613958513e-06,
+      "loss": 0.3977,
+      "step": 1962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6171323965666495,
+      "learning_rate": 9.991752007443087e-06,
+      "loss": 0.3646,
+      "step": 1963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.337737522195154,
+      "learning_rate": 9.991729369930236e-06,
+      "loss": 0.2509,
+      "step": 1964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0793798838589415,
+      "learning_rate": 9.991706701420097e-06,
+      "loss": 0.2231,
+      "step": 1965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.351366001243859,
+      "learning_rate": 9.991684001912813e-06,
+      "loss": 0.1917,
+      "step": 1966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9148404593733204,
+      "learning_rate": 9.991661271408528e-06,
+      "loss": 0.2432,
+      "step": 1967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0989016169480785,
+      "learning_rate": 9.991638509907378e-06,
+      "loss": 0.3984,
+      "step": 1968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.362565271041051,
+      "learning_rate": 9.991615717409507e-06,
+      "loss": 0.5539,
+      "step": 1969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.851516600333654,
+      "learning_rate": 9.991592893915056e-06,
+      "loss": 0.3676,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9220495183529447,
+      "learning_rate": 9.991570039424168e-06,
+      "loss": 0.3699,
+      "step": 1971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.109062263476321,
+      "learning_rate": 9.991547153936982e-06,
+      "loss": 0.5248,
+      "step": 1972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.949348105226423,
+      "learning_rate": 9.991524237453642e-06,
+      "loss": 0.4604,
+      "step": 1973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.477118450784252,
+      "learning_rate": 9.991501289974293e-06,
+      "loss": 0.4358,
+      "step": 1974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7605699654895905,
+      "learning_rate": 9.991478311499072e-06,
+      "loss": 0.196,
+      "step": 1975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.046215540817737,
+      "learning_rate": 9.991455302028126e-06,
+      "loss": 0.303,
+      "step": 1976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4414212489586062,
+      "learning_rate": 9.991432261561595e-06,
+      "loss": 0.3082,
+      "step": 1977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.435814907740232,
+      "learning_rate": 9.991409190099623e-06,
+      "loss": 0.56,
+      "step": 1978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.70610012480873,
+      "learning_rate": 9.991386087642356e-06,
+      "loss": 0.2046,
+      "step": 1979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8720708935596917,
+      "learning_rate": 9.991362954189933e-06,
+      "loss": 0.2104,
+      "step": 1980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8686101114740343,
+      "learning_rate": 9.9913397897425e-06,
+      "loss": 0.1783,
+      "step": 1981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2535878688786033,
+      "learning_rate": 9.9913165943002e-06,
+      "loss": 0.5394,
+      "step": 1982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4708954747530294,
+      "learning_rate": 9.99129336786318e-06,
+      "loss": 0.5378,
+      "step": 1983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.272798094350415,
+      "learning_rate": 9.99127011043158e-06,
+      "loss": 0.5905,
+      "step": 1984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9809814857814536,
+      "learning_rate": 9.991246822005547e-06,
+      "loss": 0.401,
+      "step": 1985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6609431375924295,
+      "learning_rate": 9.991223502585225e-06,
+      "loss": 0.7112,
+      "step": 1986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.150268384904771,
+      "learning_rate": 9.991200152170758e-06,
+      "loss": 0.4005,
+      "step": 1987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6388246285462658,
+      "learning_rate": 9.991176770762292e-06,
+      "loss": 0.2116,
+      "step": 1988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.137217350168572,
+      "learning_rate": 9.991153358359973e-06,
+      "loss": 0.3849,
+      "step": 1989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4640163598732814,
+      "learning_rate": 9.991129914963944e-06,
+      "loss": 0.4047,
+      "step": 1990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4334482942947755,
+      "learning_rate": 9.991106440574351e-06,
+      "loss": 0.2331,
+      "step": 1991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2605174865447095,
+      "learning_rate": 9.991082935191343e-06,
+      "loss": 0.4588,
+      "step": 1992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9911777794644396,
+      "learning_rate": 9.991059398815061e-06,
+      "loss": 0.3832,
+      "step": 1993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1492550629488147,
+      "learning_rate": 9.991035831445656e-06,
+      "loss": 0.5628,
+      "step": 1994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.726033496511212,
+      "learning_rate": 9.991012233083271e-06,
+      "loss": 0.3422,
+      "step": 1995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4388165525274323,
+      "learning_rate": 9.990988603728054e-06,
+      "loss": 0.5172,
+      "step": 1996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2180206752287304,
+      "learning_rate": 9.990964943380151e-06,
+      "loss": 0.536,
+      "step": 1997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1193140530267107,
+      "learning_rate": 9.99094125203971e-06,
+      "loss": 0.26,
+      "step": 1998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2285399874426215,
+      "learning_rate": 9.990917529706878e-06,
+      "loss": 0.3871,
+      "step": 1999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.588730243297102,
+      "learning_rate": 9.9908937763818e-06,
+      "loss": 0.5636,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.179111582374085,
+      "learning_rate": 9.990869992064625e-06,
+      "loss": 0.4066,
+      "step": 2001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2976646590508625,
+      "learning_rate": 9.990846176755502e-06,
+      "loss": 0.3839,
+      "step": 2002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.296510505300505,
+      "learning_rate": 9.990822330454578e-06,
+      "loss": 0.6738,
+      "step": 2003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.46459361165844,
+      "learning_rate": 9.990798453162e-06,
+      "loss": 0.486,
+      "step": 2004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.069945947118741,
+      "learning_rate": 9.99077454487792e-06,
+      "loss": 0.3861,
+      "step": 2005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5943624731414374,
+      "learning_rate": 9.99075060560248e-06,
+      "loss": 0.1953,
+      "step": 2006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.354465637943378,
+      "learning_rate": 9.990726635335834e-06,
+      "loss": 0.5094,
+      "step": 2007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.269658372375154,
+      "learning_rate": 9.990702634078127e-06,
+      "loss": 0.5917,
+      "step": 2008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3810205256411696,
+      "learning_rate": 9.990678601829513e-06,
+      "loss": 0.5362,
+      "step": 2009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6357586418576848,
+      "learning_rate": 9.990654538590136e-06,
+      "loss": 0.2583,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6208555281012145,
+      "learning_rate": 9.99063044436015e-06,
+      "loss": 0.7463,
+      "step": 2011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6703608956387013,
+      "learning_rate": 9.990606319139702e-06,
+      "loss": 0.3961,
+      "step": 2012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6793208498329253,
+      "learning_rate": 9.990582162928942e-06,
+      "loss": 0.3116,
+      "step": 2013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.980855097133235,
+      "learning_rate": 9.990557975728019e-06,
+      "loss": 0.322,
+      "step": 2014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0802525550167354,
+      "learning_rate": 9.990533757537086e-06,
+      "loss": 0.3459,
+      "step": 2015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6062207830973207,
+      "learning_rate": 9.990509508356291e-06,
+      "loss": 0.6563,
+      "step": 2016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2812532725376014,
+      "learning_rate": 9.990485228185786e-06,
+      "loss": 0.4644,
+      "step": 2017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4403318675819161,
+      "learning_rate": 9.99046091702572e-06,
+      "loss": 0.2556,
+      "step": 2018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.199160891867781,
+      "learning_rate": 9.990436574876247e-06,
+      "loss": 0.4789,
+      "step": 2019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.974671586412195,
+      "learning_rate": 9.990412201737515e-06,
+      "loss": 0.1763,
+      "step": 2020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5296553706768412,
+      "learning_rate": 9.990387797609677e-06,
+      "loss": 0.2066,
+      "step": 2021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5685690980690326,
+      "learning_rate": 9.990363362492885e-06,
+      "loss": 0.2204,
+      "step": 2022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0593949266317693,
+      "learning_rate": 9.990338896387288e-06,
+      "loss": 0.5461,
+      "step": 2023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3161840670013167,
+      "learning_rate": 9.990314399293042e-06,
+      "loss": 0.4029,
+      "step": 2024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3659108751376214,
+      "learning_rate": 9.990289871210296e-06,
+      "loss": 0.6258,
+      "step": 2025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6202668068603743,
+      "learning_rate": 9.990265312139204e-06,
+      "loss": 0.3928,
+      "step": 2026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9983668141096282,
+      "learning_rate": 9.990240722079916e-06,
+      "loss": 0.4363,
+      "step": 2027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7111331444626656,
+      "learning_rate": 9.990216101032587e-06,
+      "loss": 0.5384,
+      "step": 2028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9857370478013032,
+      "learning_rate": 9.990191448997371e-06,
+      "loss": 0.3588,
+      "step": 2029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8871922188580004,
+      "learning_rate": 9.990166765974418e-06,
+      "loss": 0.6382,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6070842017433598,
+      "learning_rate": 9.990142051963885e-06,
+      "loss": 0.2002,
+      "step": 2031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.972330531055075,
+      "learning_rate": 9.99011730696592e-06,
+      "loss": 0.3049,
+      "step": 2032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2824011014860988,
+      "learning_rate": 9.990092530980681e-06,
+      "loss": 0.4417,
+      "step": 2033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1507654131975906,
+      "learning_rate": 9.990067724008323e-06,
+      "loss": 0.4553,
+      "step": 2034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6322152309337064,
+      "learning_rate": 9.990042886048997e-06,
+      "loss": 0.2493,
+      "step": 2035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9437962295026914,
+      "learning_rate": 9.990018017102856e-06,
+      "loss": 0.3447,
+      "step": 2036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3803473201684036,
+      "learning_rate": 9.989993117170058e-06,
+      "loss": 0.2443,
+      "step": 2037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.054617681151076,
+      "learning_rate": 9.989968186250754e-06,
+      "loss": 0.4462,
+      "step": 2038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.966051306989094,
+      "learning_rate": 9.989943224345104e-06,
+      "loss": 0.3312,
+      "step": 2039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2911500492969137,
+      "learning_rate": 9.989918231453257e-06,
+      "loss": 0.4155,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.906819403140148,
+      "learning_rate": 9.989893207575373e-06,
+      "loss": 0.2202,
+      "step": 2041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7253476033882924,
+      "learning_rate": 9.989868152711605e-06,
+      "loss": 0.2656,
+      "step": 2042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.333525434566184,
+      "learning_rate": 9.989843066862109e-06,
+      "loss": 0.2719,
+      "step": 2043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9324770428933566,
+      "learning_rate": 9.989817950027041e-06,
+      "loss": 0.3475,
+      "step": 2044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3877368292614758,
+      "learning_rate": 9.989792802206556e-06,
+      "loss": 0.2167,
+      "step": 2045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.351324601670619,
+      "learning_rate": 9.989767623400813e-06,
+      "loss": 0.2629,
+      "step": 2046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.12957967203933,
+      "learning_rate": 9.989742413609964e-06,
+      "loss": 0.6212,
+      "step": 2047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0968578584513304,
+      "learning_rate": 9.989717172834168e-06,
+      "loss": 0.3531,
+      "step": 2048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.793486231673412,
+      "learning_rate": 9.989691901073582e-06,
+      "loss": 0.2118,
+      "step": 2049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5089362321439963,
+      "learning_rate": 9.989666598328364e-06,
+      "loss": 0.4311,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7570284133148701,
+      "learning_rate": 9.989641264598668e-06,
+      "loss": 0.3932,
+      "step": 2051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5470760580851215,
+      "learning_rate": 9.989615899884653e-06,
+      "loss": 0.3556,
+      "step": 2052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5544689015200777,
+      "learning_rate": 9.989590504186478e-06,
+      "loss": 0.3744,
+      "step": 2053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0934316122607473,
+      "learning_rate": 9.989565077504298e-06,
+      "loss": 0.505,
+      "step": 2054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.078762200549686,
+      "learning_rate": 9.989539619838273e-06,
+      "loss": 0.3633,
+      "step": 2055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5605704505104023,
+      "learning_rate": 9.98951413118856e-06,
+      "loss": 0.2182,
+      "step": 2056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.630370594491851,
+      "learning_rate": 9.989488611555317e-06,
+      "loss": 0.1836,
+      "step": 2057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7728918093056798,
+      "learning_rate": 9.989463060938701e-06,
+      "loss": 0.2999,
+      "step": 2058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7832912452542644,
+      "learning_rate": 9.989437479338875e-06,
+      "loss": 0.3349,
+      "step": 2059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6462957278655386,
+      "learning_rate": 9.989411866755996e-06,
+      "loss": 0.2435,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7925264149479239,
+      "learning_rate": 9.98938622319022e-06,
+      "loss": 0.3609,
+      "step": 2061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8252095874536347,
+      "learning_rate": 9.98936054864171e-06,
+      "loss": 0.388,
+      "step": 2062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3406351104657412,
+      "learning_rate": 9.989334843110625e-06,
+      "loss": 0.3618,
+      "step": 2063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3493349649268835,
+      "learning_rate": 9.989309106597122e-06,
+      "loss": 0.2012,
+      "step": 2064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0472056835172343,
+      "learning_rate": 9.989283339101364e-06,
+      "loss": 0.399,
+      "step": 2065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0193725833506457,
+      "learning_rate": 9.989257540623509e-06,
+      "loss": 0.3361,
+      "step": 2066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6744652623623333,
+      "learning_rate": 9.989231711163718e-06,
+      "loss": 0.4002,
+      "step": 2067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1604410228686732,
+      "learning_rate": 9.989205850722153e-06,
+      "loss": 0.4098,
+      "step": 2068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7094200559846495,
+      "learning_rate": 9.989179959298969e-06,
+      "loss": 0.2475,
+      "step": 2069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1313554097282545,
+      "learning_rate": 9.989154036894333e-06,
+      "loss": 0.2565,
+      "step": 2070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0148407571096625,
+      "learning_rate": 9.989128083508403e-06,
+      "loss": 0.356,
+      "step": 2071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.384755602416426,
+      "learning_rate": 9.989102099141342e-06,
+      "loss": 0.5226,
+      "step": 2072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.941861582211428,
+      "learning_rate": 9.989076083793307e-06,
+      "loss": 0.6693,
+      "step": 2073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6611146561461816,
+      "learning_rate": 9.989050037464464e-06,
+      "loss": 0.3001,
+      "step": 2074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6712349932288575,
+      "learning_rate": 9.989023960154975e-06,
+      "loss": 0.3917,
+      "step": 2075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7232593979326216,
+      "learning_rate": 9.988997851864998e-06,
+      "loss": 0.4166,
+      "step": 2076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9300640372044158,
+      "learning_rate": 9.9889717125947e-06,
+      "loss": 0.4168,
+      "step": 2077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6415664242462071,
+      "learning_rate": 9.988945542344239e-06,
+      "loss": 0.3599,
+      "step": 2078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9394828221716416,
+      "learning_rate": 9.988919341113778e-06,
+      "loss": 0.3194,
+      "step": 2079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6708977597278178,
+      "learning_rate": 9.988893108903483e-06,
+      "loss": 0.2135,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.313621152559792,
+      "learning_rate": 9.988866845713514e-06,
+      "loss": 0.5201,
+      "step": 2081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3268108205147198,
+      "learning_rate": 9.988840551544034e-06,
+      "loss": 0.1774,
+      "step": 2082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.961354716625363,
+      "learning_rate": 9.98881422639521e-06,
+      "loss": 0.2804,
+      "step": 2083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.910898751772887,
+      "learning_rate": 9.988787870267201e-06,
+      "loss": 0.1993,
+      "step": 2084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.929730312552028,
+      "learning_rate": 9.988761483160172e-06,
+      "loss": 0.2841,
+      "step": 2085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3793145504290085,
+      "learning_rate": 9.988735065074286e-06,
+      "loss": 0.5895,
+      "step": 2086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1616383963961545,
+      "learning_rate": 9.98870861600971e-06,
+      "loss": 0.1738,
+      "step": 2087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7081597065985101,
+      "learning_rate": 9.988682135966607e-06,
+      "loss": 0.2799,
+      "step": 2088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8373032258543858,
+      "learning_rate": 9.988655624945139e-06,
+      "loss": 0.3332,
+      "step": 2089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.417005328518372,
+      "learning_rate": 9.988629082945473e-06,
+      "loss": 0.4414,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7532534048682358,
+      "learning_rate": 9.988602509967775e-06,
+      "loss": 0.3665,
+      "step": 2091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.234633050575824,
+      "learning_rate": 9.988575906012208e-06,
+      "loss": 0.2708,
+      "step": 2092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.400846632002923,
+      "learning_rate": 9.988549271078936e-06,
+      "loss": 0.5518,
+      "step": 2093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5613989790691836,
+      "learning_rate": 9.988522605168127e-06,
+      "loss": 0.307,
+      "step": 2094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.264497190148389,
+      "learning_rate": 9.988495908279945e-06,
+      "loss": 0.5354,
+      "step": 2095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1442076364179306,
+      "learning_rate": 9.988469180414558e-06,
+      "loss": 0.4132,
+      "step": 2096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.543548102848484,
+      "learning_rate": 9.988442421572129e-06,
+      "loss": 0.3772,
+      "step": 2097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.147581276637871,
+      "learning_rate": 9.988415631752824e-06,
+      "loss": 0.2786,
+      "step": 2098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.558553896536106,
+      "learning_rate": 9.988388810956812e-06,
+      "loss": 0.6973,
+      "step": 2099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.115458334833686,
+      "learning_rate": 9.988361959184259e-06,
+      "loss": 0.2192,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.504738179989282,
+      "learning_rate": 9.988335076435331e-06,
+      "loss": 0.873,
+      "step": 2101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7517821910062228,
+      "learning_rate": 9.988308162710193e-06,
+      "loss": 0.3593,
+      "step": 2102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9837258771572635,
+      "learning_rate": 9.988281218009017e-06,
+      "loss": 0.4124,
+      "step": 2103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3829238615434885,
+      "learning_rate": 9.988254242331967e-06,
+      "loss": 0.5375,
+      "step": 2104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6967217514481567,
+      "learning_rate": 9.98822723567921e-06,
+      "loss": 0.2122,
+      "step": 2105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.960508086768422,
+      "learning_rate": 9.988200198050916e-06,
+      "loss": 0.381,
+      "step": 2106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.191419757794587,
+      "learning_rate": 9.988173129447251e-06,
+      "loss": 0.348,
+      "step": 2107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8520939462641588,
+      "learning_rate": 9.988146029868383e-06,
+      "loss": 0.4266,
+      "step": 2108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.070566903084276,
+      "learning_rate": 9.988118899314481e-06,
+      "loss": 0.5085,
+      "step": 2109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3661042054366077,
+      "learning_rate": 9.988091737785715e-06,
+      "loss": 0.5127,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.998031041892702,
+      "learning_rate": 9.98806454528225e-06,
+      "loss": 0.4014,
+      "step": 2111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3475963703618612,
+      "learning_rate": 9.988037321804259e-06,
+      "loss": 0.1598,
+      "step": 2112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9449959271383421,
+      "learning_rate": 9.988010067351908e-06,
+      "loss": 0.2669,
+      "step": 2113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8600048513054392,
+      "learning_rate": 9.987982781925367e-06,
+      "loss": 0.1986,
+      "step": 2114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.025876129582457,
+      "learning_rate": 9.987955465524807e-06,
+      "loss": 0.27,
+      "step": 2115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.696181795179872,
+      "learning_rate": 9.987928118150394e-06,
+      "loss": 0.3841,
+      "step": 2116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9318442126486306,
+      "learning_rate": 9.987900739802303e-06,
+      "loss": 0.2212,
+      "step": 2117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.518924744093847,
+      "learning_rate": 9.9878733304807e-06,
+      "loss": 0.2462,
+      "step": 2118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5936118742748757,
+      "learning_rate": 9.987845890185755e-06,
+      "loss": 0.1962,
+      "step": 2119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6246562603373753,
+      "learning_rate": 9.98781841891764e-06,
+      "loss": 0.2571,
+      "step": 2120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.296380632097604,
+      "learning_rate": 9.987790916676527e-06,
+      "loss": 0.4637,
+      "step": 2121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3108402946203155,
+      "learning_rate": 9.987763383462586e-06,
+      "loss": 0.5754,
+      "step": 2122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6486952024309054,
+      "learning_rate": 9.987735819275984e-06,
+      "loss": 0.2948,
+      "step": 2123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7130046054059445,
+      "learning_rate": 9.987708224116896e-06,
+      "loss": 0.4917,
+      "step": 2124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8394450385368983,
+      "learning_rate": 9.987680597985494e-06,
+      "loss": 0.2804,
+      "step": 2125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9844343662768256,
+      "learning_rate": 9.98765294088195e-06,
+      "loss": 0.2176,
+      "step": 2126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.243456995538705,
+      "learning_rate": 9.98762525280643e-06,
+      "loss": 0.1697,
+      "step": 2127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.753674991034061,
+      "learning_rate": 9.987597533759112e-06,
+      "loss": 0.218,
+      "step": 2128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8264754180773433,
+      "learning_rate": 9.987569783740165e-06,
+      "loss": 0.3486,
+      "step": 2129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.348405969018486,
+      "learning_rate": 9.987542002749762e-06,
+      "loss": 0.4348,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.308176955085865,
+      "learning_rate": 9.987514190788077e-06,
+      "loss": 0.6446,
+      "step": 2131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8270601876179982,
+      "learning_rate": 9.987486347855282e-06,
+      "loss": 0.3028,
+      "step": 2132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5879031530112995,
+      "learning_rate": 9.987458473951548e-06,
+      "loss": 0.272,
+      "step": 2133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4944179040767853,
+      "learning_rate": 9.98743056907705e-06,
+      "loss": 0.2651,
+      "step": 2134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.459845928333925,
+      "learning_rate": 9.987402633231959e-06,
+      "loss": 0.3021,
+      "step": 2135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4624049095731975,
+      "learning_rate": 9.987374666416454e-06,
+      "loss": 0.653,
+      "step": 2136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0955562771189933,
+      "learning_rate": 9.987346668630701e-06,
+      "loss": 0.3959,
+      "step": 2137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9917649435617619,
+      "learning_rate": 9.987318639874881e-06,
+      "loss": 0.1914,
+      "step": 2138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9171492851819893,
+      "learning_rate": 9.987290580149164e-06,
+      "loss": 0.3872,
+      "step": 2139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0709128103285432,
+      "learning_rate": 9.987262489453724e-06,
+      "loss": 0.437,
+      "step": 2140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.82559070825285,
+      "learning_rate": 9.987234367788739e-06,
+      "loss": 0.3475,
+      "step": 2141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.208323034826293,
+      "learning_rate": 9.987206215154378e-06,
+      "loss": 0.4963,
+      "step": 2142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5374062753086866,
+      "learning_rate": 9.98717803155082e-06,
+      "loss": 0.2382,
+      "step": 2143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5315873038513983,
+      "learning_rate": 9.98714981697824e-06,
+      "loss": 0.2145,
+      "step": 2144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4618529301884993,
+      "learning_rate": 9.987121571436812e-06,
+      "loss": 0.1799,
+      "step": 2145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.311698929475138,
+      "learning_rate": 9.98709329492671e-06,
+      "loss": 0.6766,
+      "step": 2146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7770671482530518,
+      "learning_rate": 9.987064987448114e-06,
+      "loss": 0.2079,
+      "step": 2147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.462269498232157,
+      "learning_rate": 9.987036649001194e-06,
+      "loss": 0.6479,
+      "step": 2148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9683426859504523,
+      "learning_rate": 9.987008279586131e-06,
+      "loss": 0.3937,
+      "step": 2149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9575439778925958,
+      "learning_rate": 9.986979879203098e-06,
+      "loss": 0.2699,
+      "step": 2150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8445820628126566,
+      "learning_rate": 9.986951447852273e-06,
+      "loss": 0.3286,
+      "step": 2151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0335554391166415,
+      "learning_rate": 9.986922985533831e-06,
+      "loss": 0.3056,
+      "step": 2152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2040296465930584,
+      "learning_rate": 9.986894492247952e-06,
+      "loss": 0.2246,
+      "step": 2153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6597615524532914,
+      "learning_rate": 9.986865967994809e-06,
+      "loss": 0.2277,
+      "step": 2154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.362870443910324,
+      "learning_rate": 9.986837412774582e-06,
+      "loss": 0.3927,
+      "step": 2155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5094077314062062,
+      "learning_rate": 9.986808826587445e-06,
+      "loss": 0.4181,
+      "step": 2156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9126168533230479,
+      "learning_rate": 9.98678020943358e-06,
+      "loss": 0.4813,
+      "step": 2157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4223408761111207,
+      "learning_rate": 9.98675156131316e-06,
+      "loss": 0.2415,
+      "step": 2158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.102919463080274,
+      "learning_rate": 9.986722882226367e-06,
+      "loss": 0.4495,
+      "step": 2159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7501819890624846,
+      "learning_rate": 9.986694172173378e-06,
+      "loss": 0.2583,
+      "step": 2160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5805006134321042,
+      "learning_rate": 9.986665431154368e-06,
+      "loss": 0.3094,
+      "step": 2161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2774592479908065,
+      "learning_rate": 9.98663665916952e-06,
+      "loss": 0.5321,
+      "step": 2162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.267160336090504,
+      "learning_rate": 9.98660785621901e-06,
+      "loss": 0.4857,
+      "step": 2163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.548623616305164,
+      "learning_rate": 9.986579022303017e-06,
+      "loss": 0.2202,
+      "step": 2164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.665350885362241,
+      "learning_rate": 9.986550157421722e-06,
+      "loss": 0.4751,
+      "step": 2165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7577769463600457,
+      "learning_rate": 9.986521261575303e-06,
+      "loss": 0.6926,
+      "step": 2166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6137439373716531,
+      "learning_rate": 9.98649233476394e-06,
+      "loss": 0.1924,
+      "step": 2167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5011320486550455,
+      "learning_rate": 9.986463376987809e-06,
+      "loss": 0.3499,
+      "step": 2168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.426527122109478,
+      "learning_rate": 9.986434388247095e-06,
+      "loss": 0.2264,
+      "step": 2169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6600210460748075,
+      "learning_rate": 9.986405368541976e-06,
+      "loss": 0.3982,
+      "step": 2170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6634429754467719,
+      "learning_rate": 9.986376317872633e-06,
+      "loss": 0.2305,
+      "step": 2171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1731040962142014,
+      "learning_rate": 9.986347236239242e-06,
+      "loss": 0.4121,
+      "step": 2172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0320411205235347,
+      "learning_rate": 9.98631812364199e-06,
+      "loss": 0.2133,
+      "step": 2173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0567917365252204,
+      "learning_rate": 9.986288980081054e-06,
+      "loss": 0.2956,
+      "step": 2174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7731461306614802,
+      "learning_rate": 9.986259805556614e-06,
+      "loss": 0.3484,
+      "step": 2175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.164075720571079,
+      "learning_rate": 9.986230600068856e-06,
+      "loss": 0.2411,
+      "step": 2176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3777286107704882,
+      "learning_rate": 9.986201363617955e-06,
+      "loss": 0.398,
+      "step": 2177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.511532623846168,
+      "learning_rate": 9.986172096204099e-06,
+      "loss": 0.2464,
+      "step": 2178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0097663812663713,
+      "learning_rate": 9.986142797827466e-06,
+      "loss": 0.3295,
+      "step": 2179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9113425528119947,
+      "learning_rate": 9.986113468488236e-06,
+      "loss": 0.3425,
+      "step": 2180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4163761841278384,
+      "learning_rate": 9.986084108186596e-06,
+      "loss": 0.6465,
+      "step": 2181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0284613613993097,
+      "learning_rate": 9.986054716922723e-06,
+      "loss": 0.3059,
+      "step": 2182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.180032796325274,
+      "learning_rate": 9.986025294696807e-06,
+      "loss": 0.2641,
+      "step": 2183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.116093532401973,
+      "learning_rate": 9.985995841509024e-06,
+      "loss": 0.5797,
+      "step": 2184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5188743107748395,
+      "learning_rate": 9.985966357359557e-06,
+      "loss": 0.2366,
+      "step": 2185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.779634743026586,
+      "learning_rate": 9.985936842248594e-06,
+      "loss": 0.4523,
+      "step": 2186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2366222345135587,
+      "learning_rate": 9.985907296176315e-06,
+      "loss": 0.2215,
+      "step": 2187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5789400950755903,
+      "learning_rate": 9.985877719142903e-06,
+      "loss": 0.8049,
+      "step": 2188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6353706279478388,
+      "learning_rate": 9.985848111148542e-06,
+      "loss": 0.2222,
+      "step": 2189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7346091499014054,
+      "learning_rate": 9.985818472193419e-06,
+      "loss": 0.3938,
+      "step": 2190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9350569285081995,
+      "learning_rate": 9.985788802277714e-06,
+      "loss": 0.3562,
+      "step": 2191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3458030482472148,
+      "learning_rate": 9.985759101401612e-06,
+      "loss": 0.5767,
+      "step": 2192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3551733591061226,
+      "learning_rate": 9.985729369565299e-06,
+      "loss": 0.2479,
+      "step": 2193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2078659504699205,
+      "learning_rate": 9.985699606768959e-06,
+      "loss": 0.545,
+      "step": 2194
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1577653518569737,
+      "learning_rate": 9.985669813012775e-06,
+      "loss": 0.3872,
+      "step": 2195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8890801869077918,
+      "learning_rate": 9.985639988296936e-06,
+      "loss": 0.2559,
+      "step": 2196
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5050263234304568,
+      "learning_rate": 9.985610132621624e-06,
+      "loss": 0.2139,
+      "step": 2197
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1626296136854384,
+      "learning_rate": 9.985580245987023e-06,
+      "loss": 0.3611,
+      "step": 2198
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6693075946433857,
+      "learning_rate": 9.985550328393323e-06,
+      "loss": 0.2209,
+      "step": 2199
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9693125477394664,
+      "learning_rate": 9.985520379840708e-06,
+      "loss": 0.2157,
+      "step": 2200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5551989682058736,
+      "learning_rate": 9.985490400329362e-06,
+      "loss": 0.464,
+      "step": 2201
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.809281949465528,
+      "learning_rate": 9.985460389859474e-06,
+      "loss": 0.2994,
+      "step": 2202
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2329495558108285,
+      "learning_rate": 9.98543034843123e-06,
+      "loss": 0.5632,
+      "step": 2203
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0588660883411585,
+      "learning_rate": 9.985400276044811e-06,
+      "loss": 0.3521,
+      "step": 2204
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0725699002711955,
+      "learning_rate": 9.985370172700413e-06,
+      "loss": 0.2056,
+      "step": 2205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7430799063235047,
+      "learning_rate": 9.985340038398216e-06,
+      "loss": 0.4211,
+      "step": 2206
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4024749365095344,
+      "learning_rate": 9.985309873138412e-06,
+      "loss": 0.1678,
+      "step": 2207
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2243248194026144,
+      "learning_rate": 9.985279676921184e-06,
+      "loss": 0.1968,
+      "step": 2208
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.543489498676701,
+      "learning_rate": 9.98524944974672e-06,
+      "loss": 0.3792,
+      "step": 2209
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0705657696103823,
+      "learning_rate": 9.985219191615211e-06,
+      "loss": 0.3825,
+      "step": 2210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4009887210834737,
+      "learning_rate": 9.985188902526842e-06,
+      "loss": 0.2101,
+      "step": 2211
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.777295633964251,
+      "learning_rate": 9.985158582481802e-06,
+      "loss": 0.3039,
+      "step": 2212
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.418951970483186,
+      "learning_rate": 9.98512823148028e-06,
+      "loss": 0.1692,
+      "step": 2213
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0984300148859543,
+      "learning_rate": 9.985097849522462e-06,
+      "loss": 0.3622,
+      "step": 2214
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.505890520617321,
+      "learning_rate": 9.98506743660854e-06,
+      "loss": 0.1612,
+      "step": 2215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0733306783225878,
+      "learning_rate": 9.9850369927387e-06,
+      "loss": 0.2868,
+      "step": 2216
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9056197577016096,
+      "learning_rate": 9.985006517913134e-06,
+      "loss": 0.1668,
+      "step": 2217
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.312876889729703,
+      "learning_rate": 9.984976012132029e-06,
+      "loss": 0.134,
+      "step": 2218
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4132106438623975,
+      "learning_rate": 9.984945475395573e-06,
+      "loss": 0.2879,
+      "step": 2219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.029967251470039,
+      "learning_rate": 9.984914907703959e-06,
+      "loss": 0.523,
+      "step": 2220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4633853878242205,
+      "learning_rate": 9.984884309057375e-06,
+      "loss": 0.5399,
+      "step": 2221
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8315877804706153,
+      "learning_rate": 9.984853679456011e-06,
+      "loss": 0.3993,
+      "step": 2222
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.595796320888971,
+      "learning_rate": 9.984823018900059e-06,
+      "loss": 0.1856,
+      "step": 2223
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.243210868031592,
+      "learning_rate": 9.984792327389707e-06,
+      "loss": 0.3551,
+      "step": 2224
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.691495104563612,
+      "learning_rate": 9.984761604925147e-06,
+      "loss": 0.3196,
+      "step": 2225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.469192570841636,
+      "learning_rate": 9.984730851506569e-06,
+      "loss": 0.4671,
+      "step": 2226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.386237681586086,
+      "learning_rate": 9.984700067134164e-06,
+      "loss": 0.4034,
+      "step": 2227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.069418105779924,
+      "learning_rate": 9.984669251808123e-06,
+      "loss": 0.5291,
+      "step": 2228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0088769080315756,
+      "learning_rate": 9.98463840552864e-06,
+      "loss": 0.5553,
+      "step": 2229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.03675744883621,
+      "learning_rate": 9.984607528295903e-06,
+      "loss": 0.2377,
+      "step": 2230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0883183554147493,
+      "learning_rate": 9.984576620110105e-06,
+      "loss": 0.5175,
+      "step": 2231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.065968151560459,
+      "learning_rate": 9.984545680971439e-06,
+      "loss": 0.4551,
+      "step": 2232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4238791626677332,
+      "learning_rate": 9.984514710880093e-06,
+      "loss": 0.1601,
+      "step": 2233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.836070465792075,
+      "learning_rate": 9.984483709836265e-06,
+      "loss": 0.5045,
+      "step": 2234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.822884637460434,
+      "learning_rate": 9.984452677840147e-06,
+      "loss": 0.466,
+      "step": 2235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3683157142768554,
+      "learning_rate": 9.984421614891926e-06,
+      "loss": 0.5166,
+      "step": 2236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.015829331627998,
+      "learning_rate": 9.9843905209918e-06,
+      "loss": 0.4201,
+      "step": 2237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.556424833151749,
+      "learning_rate": 9.98435939613996e-06,
+      "loss": 0.3215,
+      "step": 2238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6755997953009725,
+      "learning_rate": 9.984328240336602e-06,
+      "loss": 0.2584,
+      "step": 2239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.051592933243118,
+      "learning_rate": 9.984297053581916e-06,
+      "loss": 0.5271,
+      "step": 2240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.10324461173117,
+      "learning_rate": 9.984265835876096e-06,
+      "loss": 0.4324,
+      "step": 2241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7953848423339123,
+      "learning_rate": 9.98423458721934e-06,
+      "loss": 0.1987,
+      "step": 2242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9087836945150984,
+      "learning_rate": 9.984203307611838e-06,
+      "loss": 0.2977,
+      "step": 2243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9497078356513111,
+      "learning_rate": 9.984171997053782e-06,
+      "loss": 0.2144,
+      "step": 2244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1461796253521674,
+      "learning_rate": 9.984140655545374e-06,
+      "loss": 0.5969,
+      "step": 2245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.763261225514421,
+      "learning_rate": 9.984109283086801e-06,
+      "loss": 0.2249,
+      "step": 2246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.583190432692848,
+      "learning_rate": 9.984077879678263e-06,
+      "loss": 0.4079,
+      "step": 2247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.53587727200618,
+      "learning_rate": 9.984046445319953e-06,
+      "loss": 0.2296,
+      "step": 2248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9719234918535273,
+      "learning_rate": 9.984014980012064e-06,
+      "loss": 0.2143,
+      "step": 2249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2089524405925896,
+      "learning_rate": 9.983983483754795e-06,
+      "loss": 0.5975,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6642791580140344,
+      "learning_rate": 9.983951956548341e-06,
+      "loss": 0.1672,
+      "step": 2251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.827014039406393,
+      "learning_rate": 9.983920398392898e-06,
+      "loss": 0.2328,
+      "step": 2252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0396948743628145,
+      "learning_rate": 9.98388880928866e-06,
+      "loss": 0.4668,
+      "step": 2253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7807400877056936,
+      "learning_rate": 9.983857189235824e-06,
+      "loss": 0.2215,
+      "step": 2254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2577822856354317,
+      "learning_rate": 9.983825538234587e-06,
+      "loss": 0.1517,
+      "step": 2255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2317754921478925,
+      "learning_rate": 9.983793856285144e-06,
+      "loss": 0.2393,
+      "step": 2256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7211675958812687,
+      "learning_rate": 9.983762143387693e-06,
+      "loss": 0.1991,
+      "step": 2257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.529191534139017,
+      "learning_rate": 9.98373039954243e-06,
+      "loss": 0.3409,
+      "step": 2258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.911447786584458,
+      "learning_rate": 9.983698624749555e-06,
+      "loss": 0.3101,
+      "step": 2259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0574385679084566,
+      "learning_rate": 9.983666819009261e-06,
+      "loss": 0.3696,
+      "step": 2260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5974476288664234,
+      "learning_rate": 9.98363498232175e-06,
+      "loss": 0.2504,
+      "step": 2261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.057385584298721,
+      "learning_rate": 9.983603114687214e-06,
+      "loss": 0.1986,
+      "step": 2262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4738379542861604,
+      "learning_rate": 9.983571216105858e-06,
+      "loss": 0.2715,
+      "step": 2263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8763476853098218,
+      "learning_rate": 9.983539286577875e-06,
+      "loss": 0.3516,
+      "step": 2264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.406166112354036,
+      "learning_rate": 9.983507326103462e-06,
+      "loss": 0.4527,
+      "step": 2265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.006856078257975,
+      "learning_rate": 9.983475334682824e-06,
+      "loss": 0.3063,
+      "step": 2266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1198370918511835,
+      "learning_rate": 9.983443312316155e-06,
+      "loss": 0.4258,
+      "step": 2267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0079345876513295,
+      "learning_rate": 9.983411259003653e-06,
+      "loss": 1.0289,
+      "step": 2268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8955273123617125,
+      "learning_rate": 9.98337917474552e-06,
+      "loss": 0.3381,
+      "step": 2269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0554776547837914,
+      "learning_rate": 9.983347059541952e-06,
+      "loss": 0.1989,
+      "step": 2270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7693787902769875,
+      "learning_rate": 9.983314913393152e-06,
+      "loss": 0.3159,
+      "step": 2271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0968314330468485,
+      "learning_rate": 9.983282736299318e-06,
+      "loss": 0.2799,
+      "step": 2272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0520821541493963,
+      "learning_rate": 9.983250528260649e-06,
+      "loss": 0.2858,
+      "step": 2273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0222764089885192,
+      "learning_rate": 9.983218289277346e-06,
+      "loss": 0.228,
+      "step": 2274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0073809948581736,
+      "learning_rate": 9.983186019349608e-06,
+      "loss": 0.2594,
+      "step": 2275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.453234882246251,
+      "learning_rate": 9.983153718477636e-06,
+      "loss": 0.4602,
+      "step": 2276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.98501267135984,
+      "learning_rate": 9.983121386661633e-06,
+      "loss": 0.3481,
+      "step": 2277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.708243622595295,
+      "learning_rate": 9.983089023901795e-06,
+      "loss": 0.3431,
+      "step": 2278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.657988977600275,
+      "learning_rate": 9.983056630198329e-06,
+      "loss": 0.5065,
+      "step": 2279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.18850427462775,
+      "learning_rate": 9.98302420555143e-06,
+      "loss": 0.2416,
+      "step": 2280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3887262561735758,
+      "learning_rate": 9.982991749961304e-06,
+      "loss": 0.2934,
+      "step": 2281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.429334081232062,
+      "learning_rate": 9.98295926342815e-06,
+      "loss": 0.4643,
+      "step": 2282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.803934154131335,
+      "learning_rate": 9.982926745952169e-06,
+      "loss": 0.2424,
+      "step": 2283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.15584135714984,
+      "learning_rate": 9.982894197533566e-06,
+      "loss": 0.2498,
+      "step": 2284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5191700225453975,
+      "learning_rate": 9.98286161817254e-06,
+      "loss": 0.345,
+      "step": 2285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.118436311326718,
+      "learning_rate": 9.982829007869297e-06,
+      "loss": 0.2834,
+      "step": 2286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9491056340637194,
+      "learning_rate": 9.982796366624035e-06,
+      "loss": 0.3359,
+      "step": 2287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.792688921365772,
+      "learning_rate": 9.982763694436959e-06,
+      "loss": 0.2849,
+      "step": 2288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.285338831397362,
+      "learning_rate": 9.982730991308273e-06,
+      "loss": 0.2506,
+      "step": 2289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.323750860738364,
+      "learning_rate": 9.98269825723818e-06,
+      "loss": 0.5436,
+      "step": 2290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.04529664655447,
+      "learning_rate": 9.98266549222688e-06,
+      "loss": 0.5814,
+      "step": 2291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4721797430222365,
+      "learning_rate": 9.982632696274578e-06,
+      "loss": 0.2502,
+      "step": 2292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.433488476861182,
+      "learning_rate": 9.98259986938148e-06,
+      "loss": 0.1829,
+      "step": 2293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5476158347670685,
+      "learning_rate": 9.982567011547788e-06,
+      "loss": 0.1858,
+      "step": 2294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.885539908798307,
+      "learning_rate": 9.982534122773707e-06,
+      "loss": 0.3759,
+      "step": 2295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.036779342206595,
+      "learning_rate": 9.982501203059438e-06,
+      "loss": 0.3821,
+      "step": 2296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.198584653190189,
+      "learning_rate": 9.982468252405189e-06,
+      "loss": 0.4875,
+      "step": 2297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1985502294121044,
+      "learning_rate": 9.982435270811163e-06,
+      "loss": 0.5087,
+      "step": 2298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1747605619506243,
+      "learning_rate": 9.982402258277567e-06,
+      "loss": 0.4117,
+      "step": 2299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8596732237019007,
+      "learning_rate": 9.982369214804604e-06,
+      "loss": 0.2037,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4459013427970557,
+      "learning_rate": 9.982336140392476e-06,
+      "loss": 0.5655,
+      "step": 2301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.550757649007274,
+      "learning_rate": 9.982303035041396e-06,
+      "loss": 0.2255,
+      "step": 2302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.713911574344924,
+      "learning_rate": 9.982269898751564e-06,
+      "loss": 0.5157,
+      "step": 2303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5707575845752322,
+      "learning_rate": 9.982236731523187e-06,
+      "loss": 0.2931,
+      "step": 2304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.461041826731311,
+      "learning_rate": 9.982203533356471e-06,
+      "loss": 0.2698,
+      "step": 2305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2371557145325824,
+      "learning_rate": 9.982170304251621e-06,
+      "loss": 0.5115,
+      "step": 2306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.004606871793467,
+      "learning_rate": 9.982137044208847e-06,
+      "loss": 0.38,
+      "step": 2307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7496614820603826,
+      "learning_rate": 9.982103753228352e-06,
+      "loss": 0.2667,
+      "step": 2308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.932569722196642,
+      "learning_rate": 9.982070431310344e-06,
+      "loss": 0.2092,
+      "step": 2309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.374909452979246,
+      "learning_rate": 9.982037078455029e-06,
+      "loss": 0.1744,
+      "step": 2310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.144572975016821,
+      "learning_rate": 9.982003694662616e-06,
+      "loss": 0.5561,
+      "step": 2311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8433734986220291,
+      "learning_rate": 9.981970279933311e-06,
+      "loss": 0.2382,
+      "step": 2312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1259218348203945,
+      "learning_rate": 9.981936834267322e-06,
+      "loss": 0.3326,
+      "step": 2313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0951810472687917,
+      "learning_rate": 9.981903357664854e-06,
+      "loss": 0.2856,
+      "step": 2314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7089730271457215,
+      "learning_rate": 9.98186985012612e-06,
+      "loss": 0.396,
+      "step": 2315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0138021106951,
+      "learning_rate": 9.981836311651324e-06,
+      "loss": 0.3795,
+      "step": 2316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7534490426947646,
+      "learning_rate": 9.981802742240678e-06,
+      "loss": 0.2573,
+      "step": 2317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3528156553387367,
+      "learning_rate": 9.981769141894387e-06,
+      "loss": 0.4308,
+      "step": 2318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9004875699414328,
+      "learning_rate": 9.981735510612659e-06,
+      "loss": 0.4032,
+      "step": 2319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9762809311619538,
+      "learning_rate": 9.981701848395705e-06,
+      "loss": 0.4802,
+      "step": 2320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.136775397852174,
+      "learning_rate": 9.981668155243735e-06,
+      "loss": 0.5247,
+      "step": 2321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1452526657917907,
+      "learning_rate": 9.981634431156956e-06,
+      "loss": 0.3922,
+      "step": 2322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0709433692030417,
+      "learning_rate": 9.98160067613558e-06,
+      "loss": 0.6012,
+      "step": 2323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5183836935739259,
+      "learning_rate": 9.981566890179811e-06,
+      "loss": 0.1986,
+      "step": 2324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2044719339229335,
+      "learning_rate": 9.981533073289866e-06,
+      "loss": 0.2347,
+      "step": 2325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.175062790326055,
+      "learning_rate": 9.981499225465949e-06,
+      "loss": 0.635,
+      "step": 2326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9355857144644129,
+      "learning_rate": 9.981465346708275e-06,
+      "loss": 0.3286,
+      "step": 2327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9777391648013216,
+      "learning_rate": 9.981431437017051e-06,
+      "loss": 0.3819,
+      "step": 2328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3717948467799217,
+      "learning_rate": 9.98139749639249e-06,
+      "loss": 0.2071,
+      "step": 2329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8490449722513214,
+      "learning_rate": 9.9813635248348e-06,
+      "loss": 0.1812,
+      "step": 2330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2373255075579093,
+      "learning_rate": 9.981329522344195e-06,
+      "loss": 0.477,
+      "step": 2331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6644987823877784,
+      "learning_rate": 9.981295488920883e-06,
+      "loss": 0.2475,
+      "step": 2332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.45390887870018,
+      "learning_rate": 9.981261424565079e-06,
+      "loss": 0.1835,
+      "step": 2333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2770386322976095,
+      "learning_rate": 9.981227329276991e-06,
+      "loss": 0.5698,
+      "step": 2334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.492708797404311,
+      "learning_rate": 9.981193203056834e-06,
+      "loss": 0.6228,
+      "step": 2335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2660461560503813,
+      "learning_rate": 9.981159045904817e-06,
+      "loss": 0.3952,
+      "step": 2336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7211208096358381,
+      "learning_rate": 9.981124857821151e-06,
+      "loss": 0.3471,
+      "step": 2337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8588668705948015,
+      "learning_rate": 9.981090638806053e-06,
+      "loss": 0.3449,
+      "step": 2338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.591754720866897,
+      "learning_rate": 9.981056388859734e-06,
+      "loss": 0.2899,
+      "step": 2339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5700913173851896,
+      "learning_rate": 9.981022107982403e-06,
+      "loss": 0.2385,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.756302545311238,
+      "learning_rate": 9.980987796174278e-06,
+      "loss": 0.3277,
+      "step": 2341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.27138340628025,
+      "learning_rate": 9.98095345343557e-06,
+      "loss": 0.6876,
+      "step": 2342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.090361739519814,
+      "learning_rate": 9.980919079766488e-06,
+      "loss": 0.4051,
+      "step": 2343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9763148684173346,
+      "learning_rate": 9.980884675167253e-06,
+      "loss": 0.3472,
+      "step": 2344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8592511223952113,
+      "learning_rate": 9.980850239638073e-06,
+      "loss": 0.2741,
+      "step": 2345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.205191137156398,
+      "learning_rate": 9.980815773179166e-06,
+      "loss": 0.4075,
+      "step": 2346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1344973822285214,
+      "learning_rate": 9.980781275790744e-06,
+      "loss": 0.2456,
+      "step": 2347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.886359857895421,
+      "learning_rate": 9.98074674747302e-06,
+      "loss": 0.2386,
+      "step": 2348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9642783312689758,
+      "learning_rate": 9.98071218822621e-06,
+      "loss": 0.2164,
+      "step": 2349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.546981582076433,
+      "learning_rate": 9.980677598050526e-06,
+      "loss": 0.1336,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6286556053346262,
+      "learning_rate": 9.980642976946187e-06,
+      "loss": 0.2677,
+      "step": 2351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.643676304071862,
+      "learning_rate": 9.980608324913406e-06,
+      "loss": 0.4442,
+      "step": 2352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0075042224479773,
+      "learning_rate": 9.980573641952396e-06,
+      "loss": 0.3472,
+      "step": 2353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9323348831389124,
+      "learning_rate": 9.980538928063378e-06,
+      "loss": 0.4689,
+      "step": 2354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3521645304436523,
+      "learning_rate": 9.98050418324656e-06,
+      "loss": 0.5178,
+      "step": 2355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8528544952232329,
+      "learning_rate": 9.980469407502164e-06,
+      "loss": 0.3815,
+      "step": 2356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9355502046773931,
+      "learning_rate": 9.980434600830403e-06,
+      "loss": 0.4362,
+      "step": 2357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5322715990925184,
+      "learning_rate": 9.980399763231494e-06,
+      "loss": 0.1869,
+      "step": 2358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7940676657287205,
+      "learning_rate": 9.980364894705654e-06,
+      "loss": 0.2571,
+      "step": 2359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.605157364861877,
+      "learning_rate": 9.980329995253099e-06,
+      "loss": 0.2502,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7848226540217176,
+      "learning_rate": 9.980295064874043e-06,
+      "loss": 0.309,
+      "step": 2361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0328613345755246,
+      "learning_rate": 9.980260103568708e-06,
+      "loss": 0.315,
+      "step": 2362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9199239744405823,
+      "learning_rate": 9.980225111337307e-06,
+      "loss": 0.3694,
+      "step": 2363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8196053993902646,
+      "learning_rate": 9.980190088180059e-06,
+      "loss": 0.3691,
+      "step": 2364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5815030240428885,
+      "learning_rate": 9.980155034097183e-06,
+      "loss": 0.375,
+      "step": 2365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1546414774958293,
+      "learning_rate": 9.980119949088894e-06,
+      "loss": 0.2582,
+      "step": 2366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.317384017140691,
+      "learning_rate": 9.980084833155408e-06,
+      "loss": 0.22,
+      "step": 2367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5321915776311106,
+      "learning_rate": 9.980049686296948e-06,
+      "loss": 0.1801,
+      "step": 2368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9604197156688505,
+      "learning_rate": 9.98001450851373e-06,
+      "loss": 0.3756,
+      "step": 2369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9812246080329796,
+      "learning_rate": 9.979979299805974e-06,
+      "loss": 0.2276,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1589669338879447,
+      "learning_rate": 9.979944060173894e-06,
+      "loss": 0.2449,
+      "step": 2371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.340931403662524,
+      "learning_rate": 9.979908789617714e-06,
+      "loss": 0.1582,
+      "step": 2372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2486401329428314,
+      "learning_rate": 9.97987348813765e-06,
+      "loss": 0.2226,
+      "step": 2373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9000812760963723,
+      "learning_rate": 9.979838155733924e-06,
+      "loss": 0.4032,
+      "step": 2374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.106633517046254,
+      "learning_rate": 9.979802792406752e-06,
+      "loss": 0.5234,
+      "step": 2375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4319081284226032,
+      "learning_rate": 9.979767398156357e-06,
+      "loss": 0.2212,
+      "step": 2376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.380644872326581,
+      "learning_rate": 9.979731972982955e-06,
+      "loss": 0.2632,
+      "step": 2377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6501656916839087,
+      "learning_rate": 9.979696516886768e-06,
+      "loss": 0.2683,
+      "step": 2378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4680074275566872,
+      "learning_rate": 9.979661029868017e-06,
+      "loss": 0.2467,
+      "step": 2379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2971969690232803,
+      "learning_rate": 9.979625511926922e-06,
+      "loss": 0.4667,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8313003649242403,
+      "learning_rate": 9.979589963063703e-06,
+      "loss": 0.5934,
+      "step": 2381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7185088486990254,
+      "learning_rate": 9.97955438327858e-06,
+      "loss": 0.1582,
+      "step": 2382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0558796894128784,
+      "learning_rate": 9.979518772571774e-06,
+      "loss": 0.5233,
+      "step": 2383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.688213187223298,
+      "learning_rate": 9.97948313094351e-06,
+      "loss": 0.6289,
+      "step": 2384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6387296627799195,
+      "learning_rate": 9.979447458394004e-06,
+      "loss": 0.2653,
+      "step": 2385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8303916198527173,
+      "learning_rate": 9.97941175492348e-06,
+      "loss": 0.433,
+      "step": 2386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6216777841013998,
+      "learning_rate": 9.979376020532159e-06,
+      "loss": 0.2085,
+      "step": 2387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.870635596848988,
+      "learning_rate": 9.979340255220265e-06,
+      "loss": 0.2971,
+      "step": 2388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1009786815939377,
+      "learning_rate": 9.979304458988017e-06,
+      "loss": 0.2106,
+      "step": 2389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7048892324018445,
+      "learning_rate": 9.979268631835638e-06,
+      "loss": 0.3054,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4594097224066203,
+      "learning_rate": 9.979232773763352e-06,
+      "loss": 0.2249,
+      "step": 2391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9490206891117048,
+      "learning_rate": 9.979196884771382e-06,
+      "loss": 0.2354,
+      "step": 2392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5887934709269778,
+      "learning_rate": 9.979160964859948e-06,
+      "loss": 0.1905,
+      "step": 2393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.18226696989192,
+      "learning_rate": 9.979125014029275e-06,
+      "loss": 0.2607,
+      "step": 2394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4759890015655186,
+      "learning_rate": 9.97908903227959e-06,
+      "loss": 0.3778,
+      "step": 2395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7385229971298286,
+      "learning_rate": 9.979053019611107e-06,
+      "loss": 0.3625,
+      "step": 2396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.725711457415747,
+      "learning_rate": 9.979016976024059e-06,
+      "loss": 0.3313,
+      "step": 2397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.381444645130067,
+      "learning_rate": 9.978980901518663e-06,
+      "loss": 0.2129,
+      "step": 2398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4706492783781533,
+      "learning_rate": 9.978944796095146e-06,
+      "loss": 0.2447,
+      "step": 2399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4443084476890669,
+      "learning_rate": 9.978908659753734e-06,
+      "loss": 0.1903,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5193401746081494,
+      "learning_rate": 9.97887249249465e-06,
+      "loss": 0.1807,
+      "step": 2401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6854777139595252,
+      "learning_rate": 9.978836294318117e-06,
+      "loss": 0.2338,
+      "step": 2402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.038430040023618,
+      "learning_rate": 9.978800065224361e-06,
+      "loss": 0.2358,
+      "step": 2403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.287044931443664,
+      "learning_rate": 9.978763805213606e-06,
+      "loss": 0.4109,
+      "step": 2404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5904226373916528,
+      "learning_rate": 9.97872751428608e-06,
+      "loss": 0.1747,
+      "step": 2405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6235877098813305,
+      "learning_rate": 9.978691192442005e-06,
+      "loss": 0.2174,
+      "step": 2406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2653605076537255,
+      "learning_rate": 9.978654839681607e-06,
+      "loss": 0.2699,
+      "step": 2407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.847453015463012,
+      "learning_rate": 9.978618456005116e-06,
+      "loss": 0.4737,
+      "step": 2408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9331336890172441,
+      "learning_rate": 9.978582041412752e-06,
+      "loss": 0.2076,
+      "step": 2409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3176586760155455,
+      "learning_rate": 9.978545595904744e-06,
+      "loss": 0.2232,
+      "step": 2410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.214586502017903,
+      "learning_rate": 9.97850911948132e-06,
+      "loss": 0.4614,
+      "step": 2411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9392389984094187,
+      "learning_rate": 9.978472612142703e-06,
+      "loss": 0.3256,
+      "step": 2412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8357435408058702,
+      "learning_rate": 9.97843607388912e-06,
+      "loss": 0.3321,
+      "step": 2413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6359605730513347,
+      "learning_rate": 9.9783995047208e-06,
+      "loss": 0.5077,
+      "step": 2414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6901748548888862,
+      "learning_rate": 9.978362904637971e-06,
+      "loss": 0.303,
+      "step": 2415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7871670447878354,
+      "learning_rate": 9.978326273640856e-06,
+      "loss": 0.3754,
+      "step": 2416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4915146711699534,
+      "learning_rate": 9.978289611729687e-06,
+      "loss": 0.5687,
+      "step": 2417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5689974753979186,
+      "learning_rate": 9.978252918904688e-06,
+      "loss": 0.3571,
+      "step": 2418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3118610079182824,
+      "learning_rate": 9.97821619516609e-06,
+      "loss": 0.4551,
+      "step": 2419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6636939065893637,
+      "learning_rate": 9.978179440514119e-06,
+      "loss": 0.3385,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7452911648859286,
+      "learning_rate": 9.978142654949004e-06,
+      "loss": 0.3961,
+      "step": 2421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7427980567657868,
+      "learning_rate": 9.978105838470973e-06,
+      "loss": 0.2234,
+      "step": 2422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.211192712780311,
+      "learning_rate": 9.978068991080254e-06,
+      "loss": 0.5656,
+      "step": 2423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7680203985693055,
+      "learning_rate": 9.978032112777078e-06,
+      "loss": 0.1551,
+      "step": 2424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.022650576716512,
+      "learning_rate": 9.977995203561671e-06,
+      "loss": 0.4149,
+      "step": 2425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.077205508317472,
+      "learning_rate": 9.977958263434267e-06,
+      "loss": 0.4217,
+      "step": 2426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6700531757334707,
+      "learning_rate": 9.97792129239509e-06,
+      "loss": 0.2307,
+      "step": 2427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.942300653805785,
+      "learning_rate": 9.977884290444373e-06,
+      "loss": 0.358,
+      "step": 2428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.612399023982663,
+      "learning_rate": 9.977847257582343e-06,
+      "loss": 0.387,
+      "step": 2429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9703124644924588,
+      "learning_rate": 9.977810193809231e-06,
+      "loss": 0.2157,
+      "step": 2430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7519569864730211,
+      "learning_rate": 9.97777309912527e-06,
+      "loss": 0.1748,
+      "step": 2431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0647222150468196,
+      "learning_rate": 9.977735973530687e-06,
+      "loss": 0.3265,
+      "step": 2432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5641313988794718,
+      "learning_rate": 9.977698817025715e-06,
+      "loss": 0.1855,
+      "step": 2433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5033693643353143,
+      "learning_rate": 9.977661629610582e-06,
+      "loss": 0.1883,
+      "step": 2434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4926258706112883,
+      "learning_rate": 9.97762441128552e-06,
+      "loss": 0.7336,
+      "step": 2435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7306476256100467,
+      "learning_rate": 9.97758716205076e-06,
+      "loss": 0.3781,
+      "step": 2436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.88716724348875,
+      "learning_rate": 9.977549881906533e-06,
+      "loss": 0.3628,
+      "step": 2437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2209244771264003,
+      "learning_rate": 9.977512570853074e-06,
+      "loss": 0.2693,
+      "step": 2438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9485503141878227,
+      "learning_rate": 9.97747522889061e-06,
+      "loss": 0.3757,
+      "step": 2439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8537275108067459,
+      "learning_rate": 9.977437856019376e-06,
+      "loss": 0.1924,
+      "step": 2440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.094358213534428,
+      "learning_rate": 9.977400452239602e-06,
+      "loss": 0.5826,
+      "step": 2441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.379189760950063,
+      "learning_rate": 9.977363017551521e-06,
+      "loss": 0.5988,
+      "step": 2442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9624233126083594,
+      "learning_rate": 9.977325551955366e-06,
+      "loss": 0.3453,
+      "step": 2443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9401580666515654,
+      "learning_rate": 9.977288055451368e-06,
+      "loss": 0.4052,
+      "step": 2444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7667466162601542,
+      "learning_rate": 9.977250528039762e-06,
+      "loss": 0.3927,
+      "step": 2445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9423042922070095,
+      "learning_rate": 9.977212969720782e-06,
+      "loss": 0.4433,
+      "step": 2446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6194497536924226,
+      "learning_rate": 9.977175380494657e-06,
+      "loss": 0.416,
+      "step": 2447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2864448354247184,
+      "learning_rate": 9.977137760361623e-06,
+      "loss": 0.1798,
+      "step": 2448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1761856251222125,
+      "learning_rate": 9.977100109321914e-06,
+      "loss": 0.3942,
+      "step": 2449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1320678385596517,
+      "learning_rate": 9.977062427375762e-06,
+      "loss": 0.2448,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.726071874894813,
+      "learning_rate": 9.977024714523403e-06,
+      "loss": 0.3337,
+      "step": 2451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5568766899551476,
+      "learning_rate": 9.97698697076507e-06,
+      "loss": 0.2242,
+      "step": 2452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7367496633159372,
+      "learning_rate": 9.976949196100997e-06,
+      "loss": 0.2904,
+      "step": 2453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4970206030834743,
+      "learning_rate": 9.976911390531421e-06,
+      "loss": 0.3801,
+      "step": 2454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9778621682777418,
+      "learning_rate": 9.976873554056574e-06,
+      "loss": 0.4903,
+      "step": 2455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.503260035209016,
+      "learning_rate": 9.976835686676691e-06,
+      "loss": 0.1763,
+      "step": 2456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7396485469685743,
+      "learning_rate": 9.97679778839201e-06,
+      "loss": 0.3158,
+      "step": 2457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8146512710734746,
+      "learning_rate": 9.976759859202763e-06,
+      "loss": 0.2182,
+      "step": 2458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8555234238449747,
+      "learning_rate": 9.976721899109188e-06,
+      "loss": 0.3645,
+      "step": 2459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3104255496084503,
+      "learning_rate": 9.97668390811152e-06,
+      "loss": 0.6005,
+      "step": 2460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4784118232924377,
+      "learning_rate": 9.976645886209995e-06,
+      "loss": 0.2957,
+      "step": 2461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9104490621295958,
+      "learning_rate": 9.976607833404847e-06,
+      "loss": 0.4216,
+      "step": 2462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2720257728085125,
+      "learning_rate": 9.976569749696316e-06,
+      "loss": 0.3072,
+      "step": 2463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5821674853405183,
+      "learning_rate": 9.976531635084634e-06,
+      "loss": 0.1633,
+      "step": 2464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9153626677717273,
+      "learning_rate": 9.976493489570044e-06,
+      "loss": 0.3516,
+      "step": 2465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2989932432809628,
+      "learning_rate": 9.976455313152775e-06,
+      "loss": 0.3735,
+      "step": 2466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8601509046097253,
+      "learning_rate": 9.97641710583307e-06,
+      "loss": 0.2871,
+      "step": 2467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.085409197538289,
+      "learning_rate": 9.976378867611165e-06,
+      "loss": 0.2753,
+      "step": 2468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7701659843102422,
+      "learning_rate": 9.976340598487296e-06,
+      "loss": 0.2968,
+      "step": 2469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4129751724878448,
+      "learning_rate": 9.976302298461702e-06,
+      "loss": 0.1567,
+      "step": 2470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.273355885643302,
+      "learning_rate": 9.976263967534621e-06,
+      "loss": 0.4217,
+      "step": 2471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8420316156206895,
+      "learning_rate": 9.976225605706289e-06,
+      "loss": 0.1414,
+      "step": 2472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5157657833193068,
+      "learning_rate": 9.976187212976947e-06,
+      "loss": 0.249,
+      "step": 2473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4558212927767027,
+      "learning_rate": 9.976148789346831e-06,
+      "loss": 0.5459,
+      "step": 2474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3680972264876883,
+      "learning_rate": 9.976110334816181e-06,
+      "loss": 0.3325,
+      "step": 2475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4449536971328008,
+      "learning_rate": 9.976071849385234e-06,
+      "loss": 0.4391,
+      "step": 2476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8683538225854306,
+      "learning_rate": 9.976033333054232e-06,
+      "loss": 0.229,
+      "step": 2477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1847354142712345,
+      "learning_rate": 9.975994785823412e-06,
+      "loss": 0.519,
+      "step": 2478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7935139381941467,
+      "learning_rate": 9.975956207693014e-06,
+      "loss": 0.2502,
+      "step": 2479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9281349268430517,
+      "learning_rate": 9.975917598663277e-06,
+      "loss": 0.2331,
+      "step": 2480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.083965730449643,
+      "learning_rate": 9.975878958734442e-06,
+      "loss": 0.4526,
+      "step": 2481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8454659728427487,
+      "learning_rate": 9.975840287906747e-06,
+      "loss": 0.3848,
+      "step": 2482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9453538331079656,
+      "learning_rate": 9.975801586180435e-06,
+      "loss": 0.2691,
+      "step": 2483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.511024598686311,
+      "learning_rate": 9.975762853555744e-06,
+      "loss": 0.6234,
+      "step": 2484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9312200107607076,
+      "learning_rate": 9.975724090032915e-06,
+      "loss": 0.2226,
+      "step": 2485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0755106823209095,
+      "learning_rate": 9.975685295612187e-06,
+      "loss": 0.473,
+      "step": 2486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.851911395233814,
+      "learning_rate": 9.975646470293804e-06,
+      "loss": 0.2004,
+      "step": 2487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2097996186340234,
+      "learning_rate": 9.975607614078008e-06,
+      "loss": 0.2254,
+      "step": 2488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3971809970522646,
+      "learning_rate": 9.975568726965037e-06,
+      "loss": 0.1959,
+      "step": 2489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0294719983960636,
+      "learning_rate": 9.975529808955132e-06,
+      "loss": 0.399,
+      "step": 2490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4712538290138855,
+      "learning_rate": 9.975490860048537e-06,
+      "loss": 0.3924,
+      "step": 2491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5448451957529967,
+      "learning_rate": 9.975451880245494e-06,
+      "loss": 0.211,
+      "step": 2492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1151997606539643,
+      "learning_rate": 9.975412869546243e-06,
+      "loss": 0.3259,
+      "step": 2493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.008375122415761,
+      "learning_rate": 9.975373827951029e-06,
+      "loss": 0.3786,
+      "step": 2494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1539867624061584,
+      "learning_rate": 9.975334755460092e-06,
+      "loss": 0.3155,
+      "step": 2495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.236223185046902,
+      "learning_rate": 9.975295652073675e-06,
+      "loss": 0.2777,
+      "step": 2496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8065410380366294,
+      "learning_rate": 9.975256517792022e-06,
+      "loss": 0.2356,
+      "step": 2497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1686797685822037,
+      "learning_rate": 9.975217352615376e-06,
+      "loss": 0.5682,
+      "step": 2498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9109923282346535,
+      "learning_rate": 9.975178156543978e-06,
+      "loss": 0.4033,
+      "step": 2499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7571886439067843,
+      "learning_rate": 9.975138929578074e-06,
+      "loss": 0.2945,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4248736614076316,
+      "learning_rate": 9.975099671717908e-06,
+      "loss": 0.4736,
+      "step": 2501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6375045767658278,
+      "learning_rate": 9.975060382963719e-06,
+      "loss": 0.1767,
+      "step": 2502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.095845692718157,
+      "learning_rate": 9.975021063315758e-06,
+      "loss": 0.4487,
+      "step": 2503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2023896823192084,
+      "learning_rate": 9.974981712774263e-06,
+      "loss": 0.6318,
+      "step": 2504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5930548383659557,
+      "learning_rate": 9.974942331339482e-06,
+      "loss": 0.165,
+      "step": 2505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.925159028841324,
+      "learning_rate": 9.974902919011656e-06,
+      "loss": 0.2693,
+      "step": 2506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.486383663734859,
+      "learning_rate": 9.974863475791034e-06,
+      "loss": 0.2416,
+      "step": 2507
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.04284989282944,
+      "learning_rate": 9.97482400167786e-06,
+      "loss": 0.4525,
+      "step": 2508
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9456263599213315,
+      "learning_rate": 9.974784496672376e-06,
+      "loss": 0.2413,
+      "step": 2509
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8777718559509715,
+      "learning_rate": 9.974744960774832e-06,
+      "loss": 0.4285,
+      "step": 2510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3516213609292866,
+      "learning_rate": 9.97470539398547e-06,
+      "loss": 0.3348,
+      "step": 2511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3960595551057327,
+      "learning_rate": 9.974665796304534e-06,
+      "loss": 0.263,
+      "step": 2512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6377518700402802,
+      "learning_rate": 9.974626167732275e-06,
+      "loss": 0.3381,
+      "step": 2513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.643842553937773,
+      "learning_rate": 9.974586508268936e-06,
+      "loss": 0.3011,
+      "step": 2514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8112740892581451,
+      "learning_rate": 9.974546817914765e-06,
+      "loss": 0.3523,
+      "step": 2515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4878046018410453,
+      "learning_rate": 9.974507096670007e-06,
+      "loss": 0.204,
+      "step": 2516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2824857774335254,
+      "learning_rate": 9.974467344534909e-06,
+      "loss": 0.5423,
+      "step": 2517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9760582246691616,
+      "learning_rate": 9.974427561509716e-06,
+      "loss": 0.4488,
+      "step": 2518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.204591675479842,
+      "learning_rate": 9.97438774759468e-06,
+      "loss": 0.4867,
+      "step": 2519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.166493142191309,
+      "learning_rate": 9.974347902790042e-06,
+      "loss": 0.7144,
+      "step": 2520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8874556798342033,
+      "learning_rate": 9.974308027096056e-06,
+      "loss": 0.1809,
+      "step": 2521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9154117733146936,
+      "learning_rate": 9.974268120512964e-06,
+      "loss": 0.2259,
+      "step": 2522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5645574184901017,
+      "learning_rate": 9.974228183041016e-06,
+      "loss": 0.3654,
+      "step": 2523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9962707859072648,
+      "learning_rate": 9.974188214680462e-06,
+      "loss": 0.2977,
+      "step": 2524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8173023638581178,
+      "learning_rate": 9.974148215431547e-06,
+      "loss": 0.2425,
+      "step": 2525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.461268922880687,
+      "learning_rate": 9.97410818529452e-06,
+      "loss": 0.2285,
+      "step": 2526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8050325732581023,
+      "learning_rate": 9.97406812426963e-06,
+      "loss": 0.3406,
+      "step": 2527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.816077017941295,
+      "learning_rate": 9.974028032357127e-06,
+      "loss": 0.3387,
+      "step": 2528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7383555664292338,
+      "learning_rate": 9.97398790955726e-06,
+      "loss": 0.1576,
+      "step": 2529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.788082308524688,
+      "learning_rate": 9.973947755870274e-06,
+      "loss": 0.406,
+      "step": 2530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.00173261591053,
+      "learning_rate": 9.973907571296423e-06,
+      "loss": 0.2401,
+      "step": 2531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2137033991898618,
+      "learning_rate": 9.973867355835955e-06,
+      "loss": 0.1742,
+      "step": 2532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.441931553594118,
+      "learning_rate": 9.97382710948912e-06,
+      "loss": 0.7089,
+      "step": 2533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4510080468385496,
+      "learning_rate": 9.973786832256169e-06,
+      "loss": 0.1994,
+      "step": 2534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.23294791417634,
+      "learning_rate": 9.973746524137349e-06,
+      "loss": 0.5461,
+      "step": 2535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0191451354741163,
+      "learning_rate": 9.973706185132912e-06,
+      "loss": 0.5215,
+      "step": 2536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.088571835218341,
+      "learning_rate": 9.973665815243108e-06,
+      "loss": 0.371,
+      "step": 2537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.748190779063834,
+      "learning_rate": 9.973625414468189e-06,
+      "loss": 0.2717,
+      "step": 2538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.528962040245427,
+      "learning_rate": 9.973584982808404e-06,
+      "loss": 0.4242,
+      "step": 2539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6045935628375554,
+      "learning_rate": 9.973544520264007e-06,
+      "loss": 0.2473,
+      "step": 2540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.04213891763259,
+      "learning_rate": 9.973504026835245e-06,
+      "loss": 0.5411,
+      "step": 2541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2120616278923821,
+      "learning_rate": 9.973463502522373e-06,
+      "loss": 0.1591,
+      "step": 2542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2988414715401344,
+      "learning_rate": 9.973422947325643e-06,
+      "loss": 0.3109,
+      "step": 2543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.997096319088373,
+      "learning_rate": 9.973382361245302e-06,
+      "loss": 0.2362,
+      "step": 2544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8091436625290132,
+      "learning_rate": 9.973341744281608e-06,
+      "loss": 0.3812,
+      "step": 2545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5967713029341357,
+      "learning_rate": 9.97330109643481e-06,
+      "loss": 0.3244,
+      "step": 2546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.473832288157415,
+      "learning_rate": 9.973260417705162e-06,
+      "loss": 0.6471,
+      "step": 2547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3398608810485473,
+      "learning_rate": 9.973219708092913e-06,
+      "loss": 0.2773,
+      "step": 2548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.667477381243996,
+      "learning_rate": 9.973178967598322e-06,
+      "loss": 0.2005,
+      "step": 2549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.159365241860894,
+      "learning_rate": 9.973138196221635e-06,
+      "loss": 0.3906,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6125267610621068,
+      "learning_rate": 9.97309739396311e-06,
+      "loss": 0.2336,
+      "step": 2551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6270065115679357,
+      "learning_rate": 9.973056560822999e-06,
+      "loss": 0.176,
+      "step": 2552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 46.76553270284337,
+      "learning_rate": 9.973015696801556e-06,
+      "loss": 0.6608,
+      "step": 2553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 29.946936280941095,
+      "learning_rate": 9.972974801899034e-06,
+      "loss": 0.3715,
+      "step": 2554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.105131044357961,
+      "learning_rate": 9.972933876115686e-06,
+      "loss": 0.4575,
+      "step": 2555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.233096503901494,
+      "learning_rate": 9.972892919451768e-06,
+      "loss": 0.6701,
+      "step": 2556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0438862649450735,
+      "learning_rate": 9.972851931907534e-06,
+      "loss": 0.3344,
+      "step": 2557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6601234131521467,
+      "learning_rate": 9.972810913483237e-06,
+      "loss": 0.7947,
+      "step": 2558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5915598312526327,
+      "learning_rate": 9.972769864179134e-06,
+      "loss": 0.2027,
+      "step": 2559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.306191553971187,
+      "learning_rate": 9.972728783995478e-06,
+      "loss": 0.1928,
+      "step": 2560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.303276263566012,
+      "learning_rate": 9.972687672932526e-06,
+      "loss": 0.3288,
+      "step": 2561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8062322577756789,
+      "learning_rate": 9.972646530990532e-06,
+      "loss": 0.4282,
+      "step": 2562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4974686204621013,
+      "learning_rate": 9.97260535816975e-06,
+      "loss": 0.5153,
+      "step": 2563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0815745384190047,
+      "learning_rate": 9.97256415447044e-06,
+      "loss": 0.3397,
+      "step": 2564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4898874221425997,
+      "learning_rate": 9.972522919892855e-06,
+      "loss": 0.1727,
+      "step": 2565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.874192314231832,
+      "learning_rate": 9.97248165443725e-06,
+      "loss": 0.334,
+      "step": 2566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.200402502670886,
+      "learning_rate": 9.97244035810388e-06,
+      "loss": 0.1929,
+      "step": 2567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.501545219909724,
+      "learning_rate": 9.972399030893008e-06,
+      "loss": 0.3812,
+      "step": 2568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1599467883579306,
+      "learning_rate": 9.972357672804886e-06,
+      "loss": 0.3487,
+      "step": 2569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5172592934848814,
+      "learning_rate": 9.97231628383977e-06,
+      "loss": 0.3062,
+      "step": 2570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.810813599549353,
+      "learning_rate": 9.97227486399792e-06,
+      "loss": 0.5845,
+      "step": 2571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.506444290353463,
+      "learning_rate": 9.97223341327959e-06,
+      "loss": 0.4225,
+      "step": 2572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.717571397701489,
+      "learning_rate": 9.972191931685038e-06,
+      "loss": 0.4892,
+      "step": 2573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8024759572339595,
+      "learning_rate": 9.972150419214526e-06,
+      "loss": 0.5062,
+      "step": 2574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8457897630167606,
+      "learning_rate": 9.972108875868306e-06,
+      "loss": 0.4825,
+      "step": 2575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9554837479665843,
+      "learning_rate": 9.972067301646639e-06,
+      "loss": 0.5081,
+      "step": 2576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.804541255755723,
+      "learning_rate": 9.972025696549782e-06,
+      "loss": 0.383,
+      "step": 2577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.309687037673716,
+      "learning_rate": 9.971984060577996e-06,
+      "loss": 0.5556,
+      "step": 2578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.012094109895181,
+      "learning_rate": 9.971942393731536e-06,
+      "loss": 0.2232,
+      "step": 2579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9507400600009488,
+      "learning_rate": 9.971900696010661e-06,
+      "loss": 0.4647,
+      "step": 2580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.529694086621748,
+      "learning_rate": 9.971858967415632e-06,
+      "loss": 0.2016,
+      "step": 2581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5023875309796995,
+      "learning_rate": 9.971817207946707e-06,
+      "loss": 0.4009,
+      "step": 2582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.638195282259088,
+      "learning_rate": 9.971775417604146e-06,
+      "loss": 0.3654,
+      "step": 2583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.387264036372206,
+      "learning_rate": 9.971733596388207e-06,
+      "loss": 0.1983,
+      "step": 2584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5792725894484518,
+      "learning_rate": 9.971691744299152e-06,
+      "loss": 0.3497,
+      "step": 2585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9872338762719184,
+      "learning_rate": 9.971649861337238e-06,
+      "loss": 0.2683,
+      "step": 2586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.269452129395675,
+      "learning_rate": 9.971607947502729e-06,
+      "loss": 0.4726,
+      "step": 2587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1667896122773755,
+      "learning_rate": 9.971566002795883e-06,
+      "loss": 0.3445,
+      "step": 2588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1376516898637123,
+      "learning_rate": 9.97152402721696e-06,
+      "loss": 0.2523,
+      "step": 2589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.680970514786702,
+      "learning_rate": 9.971482020766219e-06,
+      "loss": 0.2205,
+      "step": 2590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.571053864718254,
+      "learning_rate": 9.971439983443925e-06,
+      "loss": 0.437,
+      "step": 2591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.488377077890734,
+      "learning_rate": 9.971397915250336e-06,
+      "loss": 0.8096,
+      "step": 2592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.818623192719725,
+      "learning_rate": 9.971355816185714e-06,
+      "loss": 0.6495,
+      "step": 2593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7895089970767803,
+      "learning_rate": 9.971313686250322e-06,
+      "loss": 0.2836,
+      "step": 2594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7269399109623689,
+      "learning_rate": 9.97127152544442e-06,
+      "loss": 0.4468,
+      "step": 2595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.319028492678077,
+      "learning_rate": 9.971229333768269e-06,
+      "loss": 0.4739,
+      "step": 2596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8394171428745765,
+      "learning_rate": 9.971187111222134e-06,
+      "loss": 0.2626,
+      "step": 2597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4462426800859944,
+      "learning_rate": 9.971144857806273e-06,
+      "loss": 0.2306,
+      "step": 2598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.195626341626338,
+      "learning_rate": 9.971102573520952e-06,
+      "loss": 0.4231,
+      "step": 2599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1195540662100587,
+      "learning_rate": 9.971060258366432e-06,
+      "loss": 0.2162,
+      "step": 2600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.841382707243828,
+      "learning_rate": 9.971017912342977e-06,
+      "loss": 0.2462,
+      "step": 2601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.345114755649089,
+      "learning_rate": 9.970975535450846e-06,
+      "loss": 0.6783,
+      "step": 2602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6451780628370896,
+      "learning_rate": 9.970933127690308e-06,
+      "loss": 0.1761,
+      "step": 2603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.49747415310138,
+      "learning_rate": 9.970890689061622e-06,
+      "loss": 0.2145,
+      "step": 2604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3933720864766737,
+      "learning_rate": 9.970848219565054e-06,
+      "loss": 0.1912,
+      "step": 2605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.930275240670234,
+      "learning_rate": 9.970805719200866e-06,
+      "loss": 0.4753,
+      "step": 2606
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6588368665908884,
+      "learning_rate": 9.970763187969322e-06,
+      "loss": 0.3034,
+      "step": 2607
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.652142262179501,
+      "learning_rate": 9.970720625870687e-06,
+      "loss": 0.3612,
+      "step": 2608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.999626891458472,
+      "learning_rate": 9.970678032905225e-06,
+      "loss": 0.3066,
+      "step": 2609
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1575367481344356,
+      "learning_rate": 9.970635409073201e-06,
+      "loss": 0.4201,
+      "step": 2610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.018505743620957,
+      "learning_rate": 9.970592754374879e-06,
+      "loss": 0.2688,
+      "step": 2611
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5818196231675703,
+      "learning_rate": 9.970550068810525e-06,
+      "loss": 0.1977,
+      "step": 2612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2305902023786373,
+      "learning_rate": 9.970507352380403e-06,
+      "loss": 0.4622,
+      "step": 2613
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.119977678499705,
+      "learning_rate": 9.970464605084778e-06,
+      "loss": 0.3856,
+      "step": 2614
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6587542818699803,
+      "learning_rate": 9.970421826923915e-06,
+      "loss": 0.1709,
+      "step": 2615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.298513489795439,
+      "learning_rate": 9.97037901789808e-06,
+      "loss": 0.7039,
+      "step": 2616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2840757581769515,
+      "learning_rate": 9.97033617800754e-06,
+      "loss": 0.1814,
+      "step": 2617
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7989715313749732,
+      "learning_rate": 9.970293307252561e-06,
+      "loss": 0.336,
+      "step": 2618
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3613500270663907,
+      "learning_rate": 9.970250405633407e-06,
+      "loss": 0.2263,
+      "step": 2619
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.710332445678015,
+      "learning_rate": 9.970207473150349e-06,
+      "loss": 0.4242,
+      "step": 2620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7186209066634819,
+      "learning_rate": 9.970164509803649e-06,
+      "loss": 0.3218,
+      "step": 2621
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7252502633041935,
+      "learning_rate": 9.970121515593573e-06,
+      "loss": 0.3798,
+      "step": 2622
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0903086777941695,
+      "learning_rate": 9.970078490520394e-06,
+      "loss": 0.2744,
+      "step": 2623
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7419695910354642,
+      "learning_rate": 9.970035434584374e-06,
+      "loss": 0.323,
+      "step": 2624
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8619980101976816,
+      "learning_rate": 9.969992347785781e-06,
+      "loss": 0.4412,
+      "step": 2625
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7922574712236001,
+      "learning_rate": 9.969949230124884e-06,
+      "loss": 0.4462,
+      "step": 2626
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.040835344301383,
+      "learning_rate": 9.96990608160195e-06,
+      "loss": 0.2037,
+      "step": 2627
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3989396349133165,
+      "learning_rate": 9.969862902217249e-06,
+      "loss": 0.3391,
+      "step": 2628
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8727440191420588,
+      "learning_rate": 9.969819691971045e-06,
+      "loss": 0.1929,
+      "step": 2629
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0007040155615763,
+      "learning_rate": 9.969776450863608e-06,
+      "loss": 0.5701,
+      "step": 2630
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7992502326139417,
+      "learning_rate": 9.969733178895209e-06,
+      "loss": 0.5312,
+      "step": 2631
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1588912143143664,
+      "learning_rate": 9.969689876066115e-06,
+      "loss": 0.3866,
+      "step": 2632
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.323342275047016,
+      "learning_rate": 9.969646542376593e-06,
+      "loss": 0.4401,
+      "step": 2633
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6645975260215045,
+      "learning_rate": 9.969603177826914e-06,
+      "loss": 0.2528,
+      "step": 2634
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6022107344857117,
+      "learning_rate": 9.969559782417349e-06,
+      "loss": 0.6135,
+      "step": 2635
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.67434986044368,
+      "learning_rate": 9.969516356148163e-06,
+      "loss": 0.2713,
+      "step": 2636
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.540782560058804,
+      "learning_rate": 9.96947289901963e-06,
+      "loss": 0.4749,
+      "step": 2637
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2282905675701827,
+      "learning_rate": 9.969429411032017e-06,
+      "loss": 0.2048,
+      "step": 2638
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6730880213880797,
+      "learning_rate": 9.969385892185596e-06,
+      "loss": 0.4021,
+      "step": 2639
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.36149093028824,
+      "learning_rate": 9.969342342480636e-06,
+      "loss": 0.513,
+      "step": 2640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1232726774022561,
+      "learning_rate": 9.969298761917407e-06,
+      "loss": 0.186,
+      "step": 2641
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4433141958815168,
+      "learning_rate": 9.969255150496182e-06,
+      "loss": 0.7466,
+      "step": 2642
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.049284375085607,
+      "learning_rate": 9.969211508217228e-06,
+      "loss": 0.2694,
+      "step": 2643
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5154157634461987,
+      "learning_rate": 9.96916783508082e-06,
+      "loss": 0.3491,
+      "step": 2644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.535289195664389,
+      "learning_rate": 9.969124131087227e-06,
+      "loss": 0.3719,
+      "step": 2645
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.575588288452619,
+      "learning_rate": 9.969080396236722e-06,
+      "loss": 0.179,
+      "step": 2646
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7370651019875052,
+      "learning_rate": 9.969036630529573e-06,
+      "loss": 0.351,
+      "step": 2647
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.025574233470322,
+      "learning_rate": 9.968992833966055e-06,
+      "loss": 0.4686,
+      "step": 2648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.168365021261506,
+      "learning_rate": 9.96894900654644e-06,
+      "loss": 0.3823,
+      "step": 2649
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7392892550882524,
+      "learning_rate": 9.968905148271e-06,
+      "loss": 0.2859,
+      "step": 2650
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7620251002250478,
+      "learning_rate": 9.968861259140005e-06,
+      "loss": 0.426,
+      "step": 2651
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.883346069299524,
+      "learning_rate": 9.968817339153729e-06,
+      "loss": 0.1743,
+      "step": 2652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8396029628629438,
+      "learning_rate": 9.968773388312446e-06,
+      "loss": 0.5048,
+      "step": 2653
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.893554282958768,
+      "learning_rate": 9.968729406616426e-06,
+      "loss": 0.3239,
+      "step": 2654
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.107687103825677,
+      "learning_rate": 9.968685394065947e-06,
+      "loss": 0.2103,
+      "step": 2655
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.931754217867457,
+      "learning_rate": 9.968641350661276e-06,
+      "loss": 0.4164,
+      "step": 2656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.029552608820773,
+      "learning_rate": 9.968597276402692e-06,
+      "loss": 0.3747,
+      "step": 2657
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3442685324427373,
+      "learning_rate": 9.968553171290465e-06,
+      "loss": 0.1551,
+      "step": 2658
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.302863442134093,
+      "learning_rate": 9.968509035324871e-06,
+      "loss": 0.5894,
+      "step": 2659
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1292592167257887,
+      "learning_rate": 9.968464868506184e-06,
+      "loss": 0.3954,
+      "step": 2660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.903177832049841,
+      "learning_rate": 9.968420670834677e-06,
+      "loss": 0.1959,
+      "step": 2661
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.522842195284828,
+      "learning_rate": 9.968376442310626e-06,
+      "loss": 0.3708,
+      "step": 2662
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9482170313690546,
+      "learning_rate": 9.968332182934304e-06,
+      "loss": 0.5146,
+      "step": 2663
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2091821838412289,
+      "learning_rate": 9.968287892705985e-06,
+      "loss": 0.2325,
+      "step": 2664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2537546949819025,
+      "learning_rate": 9.96824357162595e-06,
+      "loss": 0.38,
+      "step": 2665
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8447738104005746,
+      "learning_rate": 9.968199219694466e-06,
+      "loss": 0.3669,
+      "step": 2666
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8018950391992161,
+      "learning_rate": 9.968154836911811e-06,
+      "loss": 0.2,
+      "step": 2667
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1153607000527965,
+      "learning_rate": 9.968110423278265e-06,
+      "loss": 0.4103,
+      "step": 2668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5266206110662943,
+      "learning_rate": 9.9680659787941e-06,
+      "loss": 0.2192,
+      "step": 2669
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0739920397065887,
+      "learning_rate": 9.968021503459593e-06,
+      "loss": 0.3879,
+      "step": 2670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.965717674391118,
+      "learning_rate": 9.96797699727502e-06,
+      "loss": 0.4352,
+      "step": 2671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.494886604897138,
+      "learning_rate": 9.967932460240655e-06,
+      "loss": 0.7358,
+      "step": 2672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.294043145313456,
+      "learning_rate": 9.967887892356778e-06,
+      "loss": 0.3237,
+      "step": 2673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.347114529049936,
+      "learning_rate": 9.967843293623665e-06,
+      "loss": 0.4355,
+      "step": 2674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7980380150682151,
+      "learning_rate": 9.967798664041592e-06,
+      "loss": 0.3528,
+      "step": 2675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5053006840508476,
+      "learning_rate": 9.967754003610838e-06,
+      "loss": 0.4477,
+      "step": 2676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6744433571818655,
+      "learning_rate": 9.967709312331677e-06,
+      "loss": 0.4189,
+      "step": 2677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1051022382223343,
+      "learning_rate": 9.96766459020439e-06,
+      "loss": 0.3886,
+      "step": 2678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5226842044869349,
+      "learning_rate": 9.967619837229252e-06,
+      "loss": 0.2876,
+      "step": 2679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3356061847185123,
+      "learning_rate": 9.967575053406543e-06,
+      "loss": 0.8665,
+      "step": 2680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9841474455173862,
+      "learning_rate": 9.96753023873654e-06,
+      "loss": 0.3169,
+      "step": 2681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.729333727609508,
+      "learning_rate": 9.967485393219521e-06,
+      "loss": 0.305,
+      "step": 2682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7222354920023029,
+      "learning_rate": 9.967440516855764e-06,
+      "loss": 0.396,
+      "step": 2683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0596105174612154,
+      "learning_rate": 9.96739560964555e-06,
+      "loss": 0.2636,
+      "step": 2684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8737429577413278,
+      "learning_rate": 9.967350671589156e-06,
+      "loss": 0.4205,
+      "step": 2685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.383731349072958,
+      "learning_rate": 9.967305702686863e-06,
+      "loss": 0.5866,
+      "step": 2686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.248083812543113,
+      "learning_rate": 9.967260702938947e-06,
+      "loss": 0.1966,
+      "step": 2687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8742771463968573,
+      "learning_rate": 9.96721567234569e-06,
+      "loss": 0.2351,
+      "step": 2688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7008575692179717,
+      "learning_rate": 9.96717061090737e-06,
+      "loss": 0.1717,
+      "step": 2689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9945179865239575,
+      "learning_rate": 9.967125518624267e-06,
+      "loss": 0.4108,
+      "step": 2690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.335320075354738,
+      "learning_rate": 9.967080395496663e-06,
+      "loss": 0.7075,
+      "step": 2691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.861550791131902,
+      "learning_rate": 9.967035241524837e-06,
+      "loss": 0.2648,
+      "step": 2692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9569616400546266,
+      "learning_rate": 9.96699005670907e-06,
+      "loss": 0.5001,
+      "step": 2693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7162677548407683,
+      "learning_rate": 9.96694484104964e-06,
+      "loss": 0.3408,
+      "step": 2694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2918311551334094,
+      "learning_rate": 9.96689959454683e-06,
+      "loss": 0.5817,
+      "step": 2695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2536693296419115,
+      "learning_rate": 9.96685431720092e-06,
+      "loss": 0.6978,
+      "step": 2696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0567049920278997,
+      "learning_rate": 9.966809009012192e-06,
+      "loss": 0.5414,
+      "step": 2697
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.329594524906252,
+      "learning_rate": 9.96676366998093e-06,
+      "loss": 0.2003,
+      "step": 2698
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.354689094223971,
+      "learning_rate": 9.96671830010741e-06,
+      "loss": 0.2497,
+      "step": 2699
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.163023709678682,
+      "learning_rate": 9.966672899391915e-06,
+      "loss": 0.3118,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.713813289781116,
+      "learning_rate": 9.966627467834731e-06,
+      "loss": 0.3737,
+      "step": 2701
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9145346701424977,
+      "learning_rate": 9.966582005436137e-06,
+      "loss": 0.513,
+      "step": 2702
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.750911266771048,
+      "learning_rate": 9.966536512196414e-06,
+      "loss": 0.2664,
+      "step": 2703
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.948347930679787,
+      "learning_rate": 9.966490988115849e-06,
+      "loss": 0.8596,
+      "step": 2704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.962172700534835,
+      "learning_rate": 9.966445433194721e-06,
+      "loss": 0.2911,
+      "step": 2705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.640427124916118,
+      "learning_rate": 9.966399847433312e-06,
+      "loss": 0.2753,
+      "step": 2706
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5275003738293604,
+      "learning_rate": 9.966354230831907e-06,
+      "loss": 0.1672,
+      "step": 2707
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0634310672653386,
+      "learning_rate": 9.966308583390791e-06,
+      "loss": 0.201,
+      "step": 2708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6101583870434448,
+      "learning_rate": 9.966262905110245e-06,
+      "loss": 0.3048,
+      "step": 2709
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9482788388430612,
+      "learning_rate": 9.966217195990554e-06,
+      "loss": 0.4479,
+      "step": 2710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5321106292453375,
+      "learning_rate": 9.966171456031999e-06,
+      "loss": 0.2083,
+      "step": 2711
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1449123731801123,
+      "learning_rate": 9.966125685234866e-06,
+      "loss": 0.4066,
+      "step": 2712
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.950562195706456,
+      "learning_rate": 9.966079883599443e-06,
+      "loss": 0.4146,
+      "step": 2713
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0406366315805187,
+      "learning_rate": 9.966034051126006e-06,
+      "loss": 0.2474,
+      "step": 2714
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.484432339756313,
+      "learning_rate": 9.965988187814848e-06,
+      "loss": 0.4681,
+      "step": 2715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7487470492582806,
+      "learning_rate": 9.965942293666249e-06,
+      "loss": 0.1459,
+      "step": 2716
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4501940992161533,
+      "learning_rate": 9.965896368680494e-06,
+      "loss": 0.3347,
+      "step": 2717
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7986694125095783,
+      "learning_rate": 9.965850412857872e-06,
+      "loss": 0.3263,
+      "step": 2718
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7191311153607371,
+      "learning_rate": 9.965804426198662e-06,
+      "loss": 0.2415,
+      "step": 2719
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3622035572219326,
+      "learning_rate": 9.965758408703157e-06,
+      "loss": 0.6479,
+      "step": 2720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3198140665425138,
+      "learning_rate": 9.965712360371636e-06,
+      "loss": 0.3141,
+      "step": 2721
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2760733304705543,
+      "learning_rate": 9.965666281204388e-06,
+      "loss": 0.3884,
+      "step": 2722
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9969873014699628,
+      "learning_rate": 9.965620171201703e-06,
+      "loss": 0.3681,
+      "step": 2723
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0155840045249365,
+      "learning_rate": 9.96557403036386e-06,
+      "loss": 0.2163,
+      "step": 2724
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6800795895912226,
+      "learning_rate": 9.965527858691152e-06,
+      "loss": 0.3183,
+      "step": 2725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.026059833794923,
+      "learning_rate": 9.96548165618386e-06,
+      "loss": 0.5012,
+      "step": 2726
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6578362983627752,
+      "learning_rate": 9.965435422842273e-06,
+      "loss": 0.319,
+      "step": 2727
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.172642582476937,
+      "learning_rate": 9.96538915866668e-06,
+      "loss": 0.2689,
+      "step": 2728
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.516451391876936,
+      "learning_rate": 9.965342863657367e-06,
+      "loss": 0.5349,
+      "step": 2729
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2623672873210996,
+      "learning_rate": 9.965296537814622e-06,
+      "loss": 0.4149,
+      "step": 2730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.004050511750067,
+      "learning_rate": 9.96525018113873e-06,
+      "loss": 0.5777,
+      "step": 2731
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.85411595469858,
+      "learning_rate": 9.965203793629982e-06,
+      "loss": 0.3238,
+      "step": 2732
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1018792236358284,
+      "learning_rate": 9.965157375288667e-06,
+      "loss": 0.3783,
+      "step": 2733
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.970950955099591,
+      "learning_rate": 9.965110926115068e-06,
+      "loss": 0.4069,
+      "step": 2734
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4973431815689981,
+      "learning_rate": 9.96506444610948e-06,
+      "loss": 0.215,
+      "step": 2735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5928833746376743,
+      "learning_rate": 9.965017935272188e-06,
+      "loss": 0.2229,
+      "step": 2736
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.199210971527166,
+      "learning_rate": 9.964971393603479e-06,
+      "loss": 0.4937,
+      "step": 2737
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6542573358884174,
+      "learning_rate": 9.964924821103645e-06,
+      "loss": 0.3407,
+      "step": 2738
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8895428137621362,
+      "learning_rate": 9.964878217772975e-06,
+      "loss": 0.3187,
+      "step": 2739
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1330944336054434,
+      "learning_rate": 9.964831583611758e-06,
+      "loss": 0.335,
+      "step": 2740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4498399740623458,
+      "learning_rate": 9.964784918620284e-06,
+      "loss": 0.748,
+      "step": 2741
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9001870029270562,
+      "learning_rate": 9.96473822279884e-06,
+      "loss": 0.2602,
+      "step": 2742
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9082933090279968,
+      "learning_rate": 9.964691496147718e-06,
+      "loss": 0.4845,
+      "step": 2743
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9840991098748777,
+      "learning_rate": 9.964644738667211e-06,
+      "loss": 0.2208,
+      "step": 2744
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7589649999925112,
+      "learning_rate": 9.964597950357604e-06,
+      "loss": 0.2881,
+      "step": 2745
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.576955521371238,
+      "learning_rate": 9.964551131219192e-06,
+      "loss": 0.4959,
+      "step": 2746
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.383557828434349,
+      "learning_rate": 9.964504281252262e-06,
+      "loss": 0.4525,
+      "step": 2747
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6875108943693509,
+      "learning_rate": 9.964457400457108e-06,
+      "loss": 0.3299,
+      "step": 2748
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8477214777524058,
+      "learning_rate": 9.96441048883402e-06,
+      "loss": 0.2744,
+      "step": 2749
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6633722749836686,
+      "learning_rate": 9.96436354638329e-06,
+      "loss": 0.1782,
+      "step": 2750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.101623417279852,
+      "learning_rate": 9.964316573105207e-06,
+      "loss": 0.3074,
+      "step": 2751
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.864317851271421,
+      "learning_rate": 9.964269569000065e-06,
+      "loss": 0.1566,
+      "step": 2752
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.491146794132274,
+      "learning_rate": 9.964222534068155e-06,
+      "loss": 0.2233,
+      "step": 2753
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9428428768650483,
+      "learning_rate": 9.964175468309772e-06,
+      "loss": 0.4272,
+      "step": 2754
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0041339440030495,
+      "learning_rate": 9.964128371725203e-06,
+      "loss": 0.3373,
+      "step": 2755
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.061814363986585,
+      "learning_rate": 9.964081244314742e-06,
+      "loss": 0.3774,
+      "step": 2756
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7254928062880945,
+      "learning_rate": 9.964034086078686e-06,
+      "loss": 0.378,
+      "step": 2757
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.174248297460073,
+      "learning_rate": 9.963986897017324e-06,
+      "loss": 0.3753,
+      "step": 2758
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7962663780771808,
+      "learning_rate": 9.963939677130948e-06,
+      "loss": 0.2544,
+      "step": 2759
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2990762840771404,
+      "learning_rate": 9.963892426419855e-06,
+      "loss": 0.5198,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5461197006144942,
+      "learning_rate": 9.963845144884334e-06,
+      "loss": 0.2162,
+      "step": 2761
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3643643773929592,
+      "learning_rate": 9.963797832524682e-06,
+      "loss": 0.3294,
+      "step": 2762
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9116078577277642,
+      "learning_rate": 9.96375048934119e-06,
+      "loss": 0.3663,
+      "step": 2763
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1282803446554457,
+      "learning_rate": 9.963703115334157e-06,
+      "loss": 0.444,
+      "step": 2764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.004550740685008,
+      "learning_rate": 9.963655710503872e-06,
+      "loss": 0.4059,
+      "step": 2765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7476032685172964,
+      "learning_rate": 9.963608274850632e-06,
+      "loss": 0.2371,
+      "step": 2766
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1837452885380944,
+      "learning_rate": 9.963560808374729e-06,
+      "loss": 0.2687,
+      "step": 2767
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.16926867716584,
+      "learning_rate": 9.963513311076461e-06,
+      "loss": 0.2568,
+      "step": 2768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.361494178391949,
+      "learning_rate": 9.96346578295612e-06,
+      "loss": 0.3869,
+      "step": 2769
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.096531991149565,
+      "learning_rate": 9.963418224014005e-06,
+      "loss": 0.2994,
+      "step": 2770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.344363658359443,
+      "learning_rate": 9.963370634250407e-06,
+      "loss": 0.2622,
+      "step": 2771
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4358701121443913,
+      "learning_rate": 9.963323013665624e-06,
+      "loss": 0.2149,
+      "step": 2772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7944814247451804,
+      "learning_rate": 9.963275362259952e-06,
+      "loss": 0.1892,
+      "step": 2773
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2583765543004737,
+      "learning_rate": 9.963227680033685e-06,
+      "loss": 0.2723,
+      "step": 2774
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.942503528861789,
+      "learning_rate": 9.963179966987121e-06,
+      "loss": 0.3079,
+      "step": 2775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5787294924975426,
+      "learning_rate": 9.963132223120556e-06,
+      "loss": 0.2692,
+      "step": 2776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9722815280453723,
+      "learning_rate": 9.963084448434283e-06,
+      "loss": 0.3344,
+      "step": 2777
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1598350327464937,
+      "learning_rate": 9.963036642928605e-06,
+      "loss": 0.4246,
+      "step": 2778
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7016309898259583,
+      "learning_rate": 9.962988806603814e-06,
+      "loss": 0.5386,
+      "step": 2779
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4104685256494514,
+      "learning_rate": 9.962940939460209e-06,
+      "loss": 0.2678,
+      "step": 2780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.056988924824706,
+      "learning_rate": 9.962893041498085e-06,
+      "loss": 0.3818,
+      "step": 2781
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4505258324329224,
+      "learning_rate": 9.962845112717743e-06,
+      "loss": 0.5405,
+      "step": 2782
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8489763975380338,
+      "learning_rate": 9.962797153119478e-06,
+      "loss": 0.4287,
+      "step": 2783
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5345919593584478,
+      "learning_rate": 9.96274916270359e-06,
+      "loss": 0.1886,
+      "step": 2784
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7038829969352516,
+      "learning_rate": 9.962701141470375e-06,
+      "loss": 0.2257,
+      "step": 2785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.975125051150464,
+      "learning_rate": 9.962653089420132e-06,
+      "loss": 0.2264,
+      "step": 2786
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.508052945122697,
+      "learning_rate": 9.962605006553157e-06,
+      "loss": 0.6,
+      "step": 2787
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7689915733020791,
+      "learning_rate": 9.962556892869752e-06,
+      "loss": 0.3465,
+      "step": 2788
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.516240050669982,
+      "learning_rate": 9.962508748370216e-06,
+      "loss": 0.3825,
+      "step": 2789
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.768366933763923,
+      "learning_rate": 9.962460573054847e-06,
+      "loss": 0.2153,
+      "step": 2790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4958891807149866,
+      "learning_rate": 9.962412366923942e-06,
+      "loss": 0.8226,
+      "step": 2791
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.163328996698307,
+      "learning_rate": 9.962364129977802e-06,
+      "loss": 0.3896,
+      "step": 2792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.361742612747845,
+      "learning_rate": 9.962315862216727e-06,
+      "loss": 0.7168,
+      "step": 2793
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9767349982097058,
+      "learning_rate": 9.962267563641016e-06,
+      "loss": 0.5621,
+      "step": 2794
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8185000626401704,
+      "learning_rate": 9.96221923425097e-06,
+      "loss": 0.3994,
+      "step": 2795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.656160582746806,
+      "learning_rate": 9.962170874046887e-06,
+      "loss": 0.1197,
+      "step": 2796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.98071531307148,
+      "learning_rate": 9.96212248302907e-06,
+      "loss": 0.2133,
+      "step": 2797
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8318948633958028,
+      "learning_rate": 9.962074061197817e-06,
+      "loss": 0.3525,
+      "step": 2798
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1170043232634437,
+      "learning_rate": 9.962025608553432e-06,
+      "loss": 0.4254,
+      "step": 2799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3759697155210164,
+      "learning_rate": 9.961977125096212e-06,
+      "loss": 0.2484,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.224653694039584,
+      "learning_rate": 9.961928610826458e-06,
+      "loss": 0.225,
+      "step": 2801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4640378680747395,
+      "learning_rate": 9.961880065744475e-06,
+      "loss": 0.1852,
+      "step": 2802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.574474424767737,
+      "learning_rate": 9.961831489850564e-06,
+      "loss": 0.2321,
+      "step": 2803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8923515503075081,
+      "learning_rate": 9.961782883145024e-06,
+      "loss": 0.3974,
+      "step": 2804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.037702874456923,
+      "learning_rate": 9.961734245628157e-06,
+      "loss": 0.453,
+      "step": 2805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2211615190824405,
+      "learning_rate": 9.961685577300266e-06,
+      "loss": 0.5733,
+      "step": 2806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.211950521610738,
+      "learning_rate": 9.961636878161655e-06,
+      "loss": 0.2443,
+      "step": 2807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7233997803413936,
+      "learning_rate": 9.961588148212624e-06,
+      "loss": 0.2733,
+      "step": 2808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3651836942507827,
+      "learning_rate": 9.961539387453474e-06,
+      "loss": 0.1746,
+      "step": 2809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.307889963398315,
+      "learning_rate": 9.961490595884513e-06,
+      "loss": 0.5491,
+      "step": 2810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7836627754275722,
+      "learning_rate": 9.96144177350604e-06,
+      "loss": 0.1677,
+      "step": 2811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9188066815879232,
+      "learning_rate": 9.961392920318357e-06,
+      "loss": 0.4299,
+      "step": 2812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.004369184029411,
+      "learning_rate": 9.961344036321773e-06,
+      "loss": 0.2983,
+      "step": 2813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4762163732634237,
+      "learning_rate": 9.961295121516587e-06,
+      "loss": 0.4698,
+      "step": 2814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6114143348519878,
+      "learning_rate": 9.961246175903103e-06,
+      "loss": 0.3681,
+      "step": 2815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9942598297926077,
+      "learning_rate": 9.961197199481626e-06,
+      "loss": 0.3797,
+      "step": 2816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9585654133046675,
+      "learning_rate": 9.96114819225246e-06,
+      "loss": 0.3637,
+      "step": 2817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.250247445010441,
+      "learning_rate": 9.96109915421591e-06,
+      "loss": 0.5892,
+      "step": 2818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0606351896768205,
+      "learning_rate": 9.96105008537228e-06,
+      "loss": 0.5512,
+      "step": 2819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.300230889265562,
+      "learning_rate": 9.961000985721872e-06,
+      "loss": 0.1964,
+      "step": 2820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9920862845345984,
+      "learning_rate": 9.960951855264995e-06,
+      "loss": 0.3454,
+      "step": 2821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.04293980040015,
+      "learning_rate": 9.960902694001953e-06,
+      "loss": 0.2973,
+      "step": 2822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8461766367243067,
+      "learning_rate": 9.96085350193305e-06,
+      "loss": 0.2424,
+      "step": 2823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.110041724906592,
+      "learning_rate": 9.960804279058591e-06,
+      "loss": 0.4217,
+      "step": 2824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5844357555306896,
+      "learning_rate": 9.960755025378886e-06,
+      "loss": 0.1855,
+      "step": 2825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.26664299292216,
+      "learning_rate": 9.960705740894235e-06,
+      "loss": 0.4546,
+      "step": 2826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.967915933527969,
+      "learning_rate": 9.960656425604946e-06,
+      "loss": 0.4346,
+      "step": 2827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.727696076843414,
+      "learning_rate": 9.960607079511328e-06,
+      "loss": 0.3052,
+      "step": 2828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7484867022122264,
+      "learning_rate": 9.960557702613684e-06,
+      "loss": 0.2345,
+      "step": 2829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.079730614660211,
+      "learning_rate": 9.960508294912323e-06,
+      "loss": 0.2341,
+      "step": 2830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9223530570202336,
+      "learning_rate": 9.960458856407551e-06,
+      "loss": 0.4472,
+      "step": 2831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9806005821516461,
+      "learning_rate": 9.960409387099674e-06,
+      "loss": 0.5364,
+      "step": 2832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.237366494399438,
+      "learning_rate": 9.960359886989002e-06,
+      "loss": 0.5575,
+      "step": 2833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.987943050947291,
+      "learning_rate": 9.960310356075837e-06,
+      "loss": 0.2643,
+      "step": 2834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.377863928102971,
+      "learning_rate": 9.960260794360492e-06,
+      "loss": 0.3029,
+      "step": 2835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8736952632985118,
+      "learning_rate": 9.960211201843272e-06,
+      "loss": 0.2601,
+      "step": 2836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0869329341879492,
+      "learning_rate": 9.960161578524487e-06,
+      "loss": 0.4174,
+      "step": 2837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9474438592423804,
+      "learning_rate": 9.960111924404445e-06,
+      "loss": 0.3597,
+      "step": 2838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9052512002856572,
+      "learning_rate": 9.96006223948345e-06,
+      "loss": 0.2275,
+      "step": 2839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6390376404585303,
+      "learning_rate": 9.960012523761816e-06,
+      "loss": 0.2808,
+      "step": 2840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5349689055873887,
+      "learning_rate": 9.959962777239847e-06,
+      "loss": 0.393,
+      "step": 2841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3935731005772085,
+      "learning_rate": 9.959912999917856e-06,
+      "loss": 0.2288,
+      "step": 2842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.241596097408532,
+      "learning_rate": 9.95986319179615e-06,
+      "loss": 0.4175,
+      "step": 2843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3080854679320044,
+      "learning_rate": 9.95981335287504e-06,
+      "loss": 0.7481,
+      "step": 2844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7450617699771014,
+      "learning_rate": 9.959763483154831e-06,
+      "loss": 0.3604,
+      "step": 2845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2184767924256215,
+      "learning_rate": 9.959713582635839e-06,
+      "loss": 0.2029,
+      "step": 2846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5947650226176455,
+      "learning_rate": 9.95966365131837e-06,
+      "loss": 0.2029,
+      "step": 2847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.166848884922481,
+      "learning_rate": 9.959613689202734e-06,
+      "loss": 0.5821,
+      "step": 2848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.485224440645538,
+      "learning_rate": 9.959563696289244e-06,
+      "loss": 0.5042,
+      "step": 2849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0212083489863133,
+      "learning_rate": 9.959513672578206e-06,
+      "loss": 0.3597,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7364377551883625,
+      "learning_rate": 9.959463618069935e-06,
+      "loss": 0.7787,
+      "step": 2851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.58428973771713,
+      "learning_rate": 9.959413532764739e-06,
+      "loss": 0.2972,
+      "step": 2852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6204135996123818,
+      "learning_rate": 9.95936341666293e-06,
+      "loss": 0.226,
+      "step": 2853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7254523531594026,
+      "learning_rate": 9.95931326976482e-06,
+      "loss": 0.4064,
+      "step": 2854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.34204172185885,
+      "learning_rate": 9.95926309207072e-06,
+      "loss": 0.6716,
+      "step": 2855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.480164519280556,
+      "learning_rate": 9.95921288358094e-06,
+      "loss": 0.3504,
+      "step": 2856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9800100133834129,
+      "learning_rate": 9.959162644295793e-06,
+      "loss": 0.2982,
+      "step": 2857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.73904706287473,
+      "learning_rate": 9.959112374215592e-06,
+      "loss": 0.2221,
+      "step": 2858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.332223925785893,
+      "learning_rate": 9.959062073340646e-06,
+      "loss": 0.7001,
+      "step": 2859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.40357763965839,
+      "learning_rate": 9.959011741671272e-06,
+      "loss": 0.4848,
+      "step": 2860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0296125056380925,
+      "learning_rate": 9.95896137920778e-06,
+      "loss": 0.2278,
+      "step": 2861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0710049029172057,
+      "learning_rate": 9.95891098595048e-06,
+      "loss": 0.2834,
+      "step": 2862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.238300286789605,
+      "learning_rate": 9.95886056189969e-06,
+      "loss": 0.4186,
+      "step": 2863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.477156290116421,
+      "learning_rate": 9.95881010705572e-06,
+      "loss": 0.2298,
+      "step": 2864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6821402118895348,
+      "learning_rate": 9.958759621418883e-06,
+      "loss": 0.2532,
+      "step": 2865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.545509604262646,
+      "learning_rate": 9.958709104989496e-06,
+      "loss": 0.4784,
+      "step": 2866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7800637486312376,
+      "learning_rate": 9.958658557767869e-06,
+      "loss": 0.4337,
+      "step": 2867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.939402814905988,
+      "learning_rate": 9.958607979754318e-06,
+      "loss": 0.5567,
+      "step": 2868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.85826276323012,
+      "learning_rate": 9.958557370949154e-06,
+      "loss": 0.393,
+      "step": 2869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8333816106537546,
+      "learning_rate": 9.958506731352695e-06,
+      "loss": 0.4207,
+      "step": 2870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2458167991725486,
+      "learning_rate": 9.958456060965255e-06,
+      "loss": 0.6168,
+      "step": 2871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1971841754652166,
+      "learning_rate": 9.958405359787146e-06,
+      "loss": 0.4465,
+      "step": 2872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.088475238923138,
+      "learning_rate": 9.958354627818685e-06,
+      "loss": 0.1397,
+      "step": 2873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6305443299537627,
+      "learning_rate": 9.958303865060188e-06,
+      "loss": 0.1922,
+      "step": 2874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.376985184540273,
+      "learning_rate": 9.958253071511967e-06,
+      "loss": 0.717,
+      "step": 2875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.623359737241071,
+      "learning_rate": 9.95820224717434e-06,
+      "loss": 0.2172,
+      "step": 2876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8079089542892435,
+      "learning_rate": 9.95815139204762e-06,
+      "loss": 0.2445,
+      "step": 2877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3349258485213527,
+      "learning_rate": 9.958100506132127e-06,
+      "loss": 0.1865,
+      "step": 2878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.66815772203322,
+      "learning_rate": 9.958049589428172e-06,
+      "loss": 0.1777,
+      "step": 2879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.567755315611035,
+      "learning_rate": 9.957998641936075e-06,
+      "loss": 0.2928,
+      "step": 2880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6873538121727858,
+      "learning_rate": 9.957947663656152e-06,
+      "loss": 0.324,
+      "step": 2881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6197835890352874,
+      "learning_rate": 9.957896654588717e-06,
+      "loss": 0.2266,
+      "step": 2882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7189718710138533,
+      "learning_rate": 9.95784561473409e-06,
+      "loss": 0.3703,
+      "step": 2883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.545251892251099,
+      "learning_rate": 9.957794544092586e-06,
+      "loss": 0.2028,
+      "step": 2884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9255386280950064,
+      "learning_rate": 9.957743442664523e-06,
+      "loss": 0.2076,
+      "step": 2885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.108539412265815,
+      "learning_rate": 9.957692310450216e-06,
+      "loss": 0.3966,
+      "step": 2886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8464186406726588,
+      "learning_rate": 9.957641147449986e-06,
+      "loss": 0.277,
+      "step": 2887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0660828155191306,
+      "learning_rate": 9.957589953664147e-06,
+      "loss": 0.2271,
+      "step": 2888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3036078651630627,
+      "learning_rate": 9.95753872909302e-06,
+      "loss": 0.2006,
+      "step": 2889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5168363262963003,
+      "learning_rate": 9.957487473736922e-06,
+      "loss": 0.1919,
+      "step": 2890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1668314175751964,
+      "learning_rate": 9.957436187596172e-06,
+      "loss": 0.508,
+      "step": 2891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9475023255882165,
+      "learning_rate": 9.957384870671087e-06,
+      "loss": 0.2517,
+      "step": 2892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3710749780958196,
+      "learning_rate": 9.957333522961988e-06,
+      "loss": 0.5166,
+      "step": 2893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6727652897721086,
+      "learning_rate": 9.957282144469188e-06,
+      "loss": 0.2719,
+      "step": 2894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9258945307966582,
+      "learning_rate": 9.957230735193015e-06,
+      "loss": 0.4347,
+      "step": 2895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9296293829069766,
+      "learning_rate": 9.957179295133781e-06,
+      "loss": 0.3353,
+      "step": 2896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0081604014201924,
+      "learning_rate": 9.95712782429181e-06,
+      "loss": 0.2468,
+      "step": 2897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6581625377419134,
+      "learning_rate": 9.957076322667419e-06,
+      "loss": 0.2919,
+      "step": 2898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2189145405055495,
+      "learning_rate": 9.957024790260927e-06,
+      "loss": 0.116,
+      "step": 2899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9298552544452972,
+      "learning_rate": 9.956973227072658e-06,
+      "loss": 0.3951,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9484385455999804,
+      "learning_rate": 9.956921633102927e-06,
+      "loss": 0.4927,
+      "step": 2901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4187691369889106,
+      "learning_rate": 9.956870008352057e-06,
+      "loss": 0.3363,
+      "step": 2902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6260837577030065,
+      "learning_rate": 9.95681835282037e-06,
+      "loss": 0.5784,
+      "step": 2903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4099521272304885,
+      "learning_rate": 9.956766666508186e-06,
+      "loss": 0.3392,
+      "step": 2904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0872225999686562,
+      "learning_rate": 9.956714949415824e-06,
+      "loss": 0.4053,
+      "step": 2905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2356730605311226,
+      "learning_rate": 9.956663201543607e-06,
+      "loss": 0.4764,
+      "step": 2906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.226903858792762,
+      "learning_rate": 9.956611422891855e-06,
+      "loss": 0.2312,
+      "step": 2907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5915618301695786,
+      "learning_rate": 9.956559613460891e-06,
+      "loss": 0.2264,
+      "step": 2908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0739980964610525,
+      "learning_rate": 9.956507773251036e-06,
+      "loss": 0.2367,
+      "step": 2909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7285196207291278,
+      "learning_rate": 9.956455902262611e-06,
+      "loss": 0.366,
+      "step": 2910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2076730385783945,
+      "learning_rate": 9.95640400049594e-06,
+      "loss": 0.5466,
+      "step": 2911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.112525039586527,
+      "learning_rate": 9.956352067951345e-06,
+      "loss": 0.4217,
+      "step": 2912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5207310477788925,
+      "learning_rate": 9.956300104629148e-06,
+      "loss": 0.1874,
+      "step": 2913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7181055074386022,
+      "learning_rate": 9.95624811052967e-06,
+      "loss": 0.2992,
+      "step": 2914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4076987301013069,
+      "learning_rate": 9.956196085653237e-06,
+      "loss": 0.3023,
+      "step": 2915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.747917306882478,
+      "learning_rate": 9.95614403000017e-06,
+      "loss": 0.2487,
+      "step": 2916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6139405949081371,
+      "learning_rate": 9.956091943570791e-06,
+      "loss": 0.2966,
+      "step": 2917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9935076135724878,
+      "learning_rate": 9.956039826365427e-06,
+      "loss": 0.2428,
+      "step": 2918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.003281337868058,
+      "learning_rate": 9.9559876783844e-06,
+      "loss": 0.3549,
+      "step": 2919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7543088115387127,
+      "learning_rate": 9.955935499628033e-06,
+      "loss": 0.2226,
+      "step": 2920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5239756147309156,
+      "learning_rate": 9.95588329009665e-06,
+      "loss": 0.2522,
+      "step": 2921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3223918049559154,
+      "learning_rate": 9.955831049790576e-06,
+      "loss": 0.4448,
+      "step": 2922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.128046285980775,
+      "learning_rate": 9.955778778710137e-06,
+      "loss": 0.2736,
+      "step": 2923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.281783247377477,
+      "learning_rate": 9.955726476855654e-06,
+      "loss": 0.3903,
+      "step": 2924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.680621805496556,
+      "learning_rate": 9.955674144227453e-06,
+      "loss": 0.3625,
+      "step": 2925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1884405991928464,
+      "learning_rate": 9.955621780825861e-06,
+      "loss": 0.3394,
+      "step": 2926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.471014290070104,
+      "learning_rate": 9.955569386651203e-06,
+      "loss": 0.2088,
+      "step": 2927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9437953057482755,
+      "learning_rate": 9.955516961703802e-06,
+      "loss": 0.4295,
+      "step": 2928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9887801857748835,
+      "learning_rate": 9.955464505983984e-06,
+      "loss": 0.5395,
+      "step": 2929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6148854052049724,
+      "learning_rate": 9.955412019492075e-06,
+      "loss": 0.308,
+      "step": 2930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6486542974978529,
+      "learning_rate": 9.955359502228402e-06,
+      "loss": 0.3518,
+      "step": 2931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.848895012527935,
+      "learning_rate": 9.955306954193292e-06,
+      "loss": 0.3492,
+      "step": 2932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.810097115631275,
+      "learning_rate": 9.955254375387067e-06,
+      "loss": 0.4129,
+      "step": 2933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4915377315636247,
+      "learning_rate": 9.955201765810057e-06,
+      "loss": 0.2435,
+      "step": 2934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7538824824396513,
+      "learning_rate": 9.95514912546259e-06,
+      "loss": 0.2572,
+      "step": 2935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0440024100474203,
+      "learning_rate": 9.955096454344988e-06,
+      "loss": 0.4777,
+      "step": 2936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9961328595952856,
+      "learning_rate": 9.955043752457581e-06,
+      "loss": 0.2761,
+      "step": 2937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.53508141750208,
+      "learning_rate": 9.954991019800697e-06,
+      "loss": 0.2694,
+      "step": 2938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.735743562627638,
+      "learning_rate": 9.954938256374663e-06,
+      "loss": 0.3315,
+      "step": 2939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3321401825603694,
+      "learning_rate": 9.954885462179807e-06,
+      "loss": 0.1861,
+      "step": 2940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9588225384361244,
+      "learning_rate": 9.954832637216455e-06,
+      "loss": 0.6011,
+      "step": 2941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7893138713068584,
+      "learning_rate": 9.954779781484936e-06,
+      "loss": 0.2965,
+      "step": 2942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3441564016855354,
+      "learning_rate": 9.954726894985577e-06,
+      "loss": 0.1647,
+      "step": 2943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.875133700372283,
+      "learning_rate": 9.954673977718709e-06,
+      "loss": 0.3717,
+      "step": 2944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.157714191133669,
+      "learning_rate": 9.954621029684661e-06,
+      "loss": 0.2457,
+      "step": 2945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4129271981607945,
+      "learning_rate": 9.95456805088376e-06,
+      "loss": 0.1355,
+      "step": 2946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4059985264744292,
+      "learning_rate": 9.954515041316333e-06,
+      "loss": 0.3717,
+      "step": 2947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7589451531551177,
+      "learning_rate": 9.954462000982713e-06,
+      "loss": 0.1959,
+      "step": 2948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7558511769885181,
+      "learning_rate": 9.954408929883226e-06,
+      "loss": 0.3872,
+      "step": 2949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.992567377710216,
+      "learning_rate": 9.954355828018202e-06,
+      "loss": 0.4696,
+      "step": 2950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6298752085918955,
+      "learning_rate": 9.954302695387976e-06,
+      "loss": 0.2544,
+      "step": 2951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.668285062935082,
+      "learning_rate": 9.954249531992872e-06,
+      "loss": 0.3482,
+      "step": 2952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9988400329032365,
+      "learning_rate": 9.954196337833222e-06,
+      "loss": 0.4571,
+      "step": 2953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.242288161885857,
+      "learning_rate": 9.954143112909356e-06,
+      "loss": 0.6182,
+      "step": 2954
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8385973432033604,
+      "learning_rate": 9.954089857221605e-06,
+      "loss": 0.2697,
+      "step": 2955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8731869634244607,
+      "learning_rate": 9.9540365707703e-06,
+      "loss": 0.246,
+      "step": 2956
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.770748641756402,
+      "learning_rate": 9.95398325355577e-06,
+      "loss": 0.2496,
+      "step": 2957
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3059484749907764,
+      "learning_rate": 9.95392990557835e-06,
+      "loss": 0.4733,
+      "step": 2958
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5900808855766668,
+      "learning_rate": 9.953876526838368e-06,
+      "loss": 0.192,
+      "step": 2959
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.114045983113954,
+      "learning_rate": 9.953823117336156e-06,
+      "loss": 0.4355,
+      "step": 2960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9787236186261192,
+      "learning_rate": 9.953769677072046e-06,
+      "loss": 0.4946,
+      "step": 2961
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9705322028934453,
+      "learning_rate": 9.953716206046368e-06,
+      "loss": 0.1862,
+      "step": 2962
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4087464849060938,
+      "learning_rate": 9.95366270425946e-06,
+      "loss": 0.1926,
+      "step": 2963
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5977497306653987,
+      "learning_rate": 9.953609171711646e-06,
+      "loss": 0.2823,
+      "step": 2964
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.094120604894882,
+      "learning_rate": 9.953555608403265e-06,
+      "loss": 0.3584,
+      "step": 2965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7138451512324762,
+      "learning_rate": 9.953502014334646e-06,
+      "loss": 0.1972,
+      "step": 2966
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.219499615850327,
+      "learning_rate": 9.953448389506124e-06,
+      "loss": 0.6759,
+      "step": 2967
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0724584002628026,
+      "learning_rate": 9.953394733918028e-06,
+      "loss": 0.5307,
+      "step": 2968
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8129156212664688,
+      "learning_rate": 9.953341047570697e-06,
+      "loss": 0.2044,
+      "step": 2969
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.367638680294635,
+      "learning_rate": 9.953287330464463e-06,
+      "loss": 0.3851,
+      "step": 2970
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.291711033406085,
+      "learning_rate": 9.953233582599654e-06,
+      "loss": 0.4038,
+      "step": 2971
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7461862390246639,
+      "learning_rate": 9.95317980397661e-06,
+      "loss": 0.2994,
+      "step": 2972
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0950302953083044,
+      "learning_rate": 9.953125994595663e-06,
+      "loss": 0.1952,
+      "step": 2973
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9458930631458253,
+      "learning_rate": 9.953072154457145e-06,
+      "loss": 0.379,
+      "step": 2974
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.887600173565646,
+      "learning_rate": 9.953018283561393e-06,
+      "loss": 0.4371,
+      "step": 2975
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.44925616259751,
+      "learning_rate": 9.952964381908742e-06,
+      "loss": 0.2326,
+      "step": 2976
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1969189823951734,
+      "learning_rate": 9.952910449499524e-06,
+      "loss": 0.6358,
+      "step": 2977
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1693588694867236,
+      "learning_rate": 9.952856486334077e-06,
+      "loss": 0.2053,
+      "step": 2978
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4654167275387633,
+      "learning_rate": 9.952802492412734e-06,
+      "loss": 0.2218,
+      "step": 2979
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.081308500653386,
+      "learning_rate": 9.95274846773583e-06,
+      "loss": 0.2644,
+      "step": 2980
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.771606267892529,
+      "learning_rate": 9.952694412303701e-06,
+      "loss": 0.3485,
+      "step": 2981
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.012697489152867,
+      "learning_rate": 9.952640326116683e-06,
+      "loss": 0.6087,
+      "step": 2982
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9888181448655824,
+      "learning_rate": 9.952586209175112e-06,
+      "loss": 0.2988,
+      "step": 2983
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.729289464638853,
+      "learning_rate": 9.952532061479326e-06,
+      "loss": 0.2401,
+      "step": 2984
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.824896018939536,
+      "learning_rate": 9.952477883029656e-06,
+      "loss": 0.3528,
+      "step": 2985
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8392476069716233,
+      "learning_rate": 9.952423673826444e-06,
+      "loss": 0.1546,
+      "step": 2986
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.331412445710245,
+      "learning_rate": 9.952369433870022e-06,
+      "loss": 0.3146,
+      "step": 2987
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8176892087380294,
+      "learning_rate": 9.95231516316073e-06,
+      "loss": 0.4276,
+      "step": 2988
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.034559568772463,
+      "learning_rate": 9.952260861698903e-06,
+      "loss": 0.4657,
+      "step": 2989
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9862557344551925,
+      "learning_rate": 9.95220652948488e-06,
+      "loss": 0.3218,
+      "step": 2990
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2068614963801325,
+      "learning_rate": 9.952152166518998e-06,
+      "loss": 0.2659,
+      "step": 2991
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0812329986095515,
+      "learning_rate": 9.952097772801593e-06,
+      "loss": 0.5416,
+      "step": 2992
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1445578762826507,
+      "learning_rate": 9.952043348333004e-06,
+      "loss": 0.5458,
+      "step": 2993
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.468820989952696,
+      "learning_rate": 9.951988893113568e-06,
+      "loss": 0.2113,
+      "step": 2994
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9642366728489546,
+      "learning_rate": 9.951934407143627e-06,
+      "loss": 0.2411,
+      "step": 2995
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3894708658029011,
+      "learning_rate": 9.951879890423514e-06,
+      "loss": 0.2038,
+      "step": 2996
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6245739784154416,
+      "learning_rate": 9.951825342953571e-06,
+      "loss": 0.2044,
+      "step": 2997
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7769608011706766,
+      "learning_rate": 9.951770764734135e-06,
+      "loss": 0.2407,
+      "step": 2998
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.569463375735414,
+      "learning_rate": 9.951716155765544e-06,
+      "loss": 0.2154,
+      "step": 2999
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.342998904632525,
+      "learning_rate": 9.951661516048142e-06,
+      "loss": 0.1851,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4785861675007965,
+      "learning_rate": 9.951606845582262e-06,
+      "loss": 0.3316,
+      "step": 3001
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.995878241564023,
+      "learning_rate": 9.951552144368248e-06,
+      "loss": 0.4456,
+      "step": 3002
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9438588923743045,
+      "learning_rate": 9.951497412406437e-06,
+      "loss": 0.3612,
+      "step": 3003
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.060329393236291,
+      "learning_rate": 9.95144264969717e-06,
+      "loss": 0.441,
+      "step": 3004
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0920992578382713,
+      "learning_rate": 9.951387856240789e-06,
+      "loss": 0.4162,
+      "step": 3005
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.181958959947891,
+      "learning_rate": 9.95133303203763e-06,
+      "loss": 0.698,
+      "step": 3006
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7943493931876233,
+      "learning_rate": 9.951278177088037e-06,
+      "loss": 0.6673,
+      "step": 3007
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9928058263084594,
+      "learning_rate": 9.951223291392348e-06,
+      "loss": 0.3638,
+      "step": 3008
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.091743280810167,
+      "learning_rate": 9.951168374950906e-06,
+      "loss": 0.5327,
+      "step": 3009
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9680676640226538,
+      "learning_rate": 9.951113427764052e-06,
+      "loss": 0.3911,
+      "step": 3010
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3984674644462296,
+      "learning_rate": 9.951058449832124e-06,
+      "loss": 0.616,
+      "step": 3011
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.313726711959741,
+      "learning_rate": 9.951003441155469e-06,
+      "loss": 0.7069,
+      "step": 3012
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0327646916930213,
+      "learning_rate": 9.950948401734425e-06,
+      "loss": 0.3727,
+      "step": 3013
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.511890025847314,
+      "learning_rate": 9.950893331569332e-06,
+      "loss": 0.6399,
+      "step": 3014
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.316109451205721,
+      "learning_rate": 9.950838230660535e-06,
+      "loss": 0.2339,
+      "step": 3015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.785426947772924,
+      "learning_rate": 9.950783099008375e-06,
+      "loss": 0.5097,
+      "step": 3016
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.326873267775545,
+      "learning_rate": 9.950727936613195e-06,
+      "loss": 0.3692,
+      "step": 3017
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.6641723180014,
+      "learning_rate": 9.950672743475337e-06,
+      "loss": 0.5346,
+      "step": 3018
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.698619725074781,
+      "learning_rate": 9.950617519595144e-06,
+      "loss": 0.3673,
+      "step": 3019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.808908179780409,
+      "learning_rate": 9.950562264972958e-06,
+      "loss": 0.2718,
+      "step": 3020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9649510515464568,
+      "learning_rate": 9.950506979609125e-06,
+      "loss": 0.3999,
+      "step": 3021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3427177889303297,
+      "learning_rate": 9.950451663503984e-06,
+      "loss": 0.2165,
+      "step": 3022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.687046153356113,
+      "learning_rate": 9.950396316657883e-06,
+      "loss": 0.3334,
+      "step": 3023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7266741798495477,
+      "learning_rate": 9.950340939071162e-06,
+      "loss": 0.328,
+      "step": 3024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9044093719901019,
+      "learning_rate": 9.950285530744168e-06,
+      "loss": 0.5291,
+      "step": 3025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9512493222123517,
+      "learning_rate": 9.950230091677243e-06,
+      "loss": 0.2515,
+      "step": 3026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0040105322903674,
+      "learning_rate": 9.95017462187073e-06,
+      "loss": 0.5776,
+      "step": 3027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8790125549126881,
+      "learning_rate": 9.950119121324976e-06,
+      "loss": 0.2026,
+      "step": 3028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6746940631865976,
+      "learning_rate": 9.950063590040324e-06,
+      "loss": 0.2518,
+      "step": 3029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6510496821244036,
+      "learning_rate": 9.95000802801712e-06,
+      "loss": 0.1827,
+      "step": 3030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6960899240803218,
+      "learning_rate": 9.94995243525571e-06,
+      "loss": 0.2478,
+      "step": 3031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7167023425499954,
+      "learning_rate": 9.949896811756436e-06,
+      "loss": 0.3269,
+      "step": 3032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4012519802361108,
+      "learning_rate": 9.949841157519646e-06,
+      "loss": 0.178,
+      "step": 3033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2096346463839183,
+      "learning_rate": 9.949785472545685e-06,
+      "loss": 0.5297,
+      "step": 3034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.279389444087867,
+      "learning_rate": 9.949729756834899e-06,
+      "loss": 0.2676,
+      "step": 3035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2734450467301464,
+      "learning_rate": 9.949674010387634e-06,
+      "loss": 0.2142,
+      "step": 3036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1472771980408498,
+      "learning_rate": 9.949618233204235e-06,
+      "loss": 0.1058,
+      "step": 3037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7984319301315146,
+      "learning_rate": 9.94956242528505e-06,
+      "loss": 0.2422,
+      "step": 3038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4138475908127293,
+      "learning_rate": 9.949506586630423e-06,
+      "loss": 0.2657,
+      "step": 3039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7876582819211335,
+      "learning_rate": 9.949450717240703e-06,
+      "loss": 0.3881,
+      "step": 3040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.254636437993784,
+      "learning_rate": 9.949394817116237e-06,
+      "loss": 0.5945,
+      "step": 3041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2230754307398581,
+      "learning_rate": 9.949338886257371e-06,
+      "loss": 0.1834,
+      "step": 3042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5113914736764182,
+      "learning_rate": 9.949282924664454e-06,
+      "loss": 0.2673,
+      "step": 3043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3754144859025152,
+      "learning_rate": 9.949226932337831e-06,
+      "loss": 0.5893,
+      "step": 3044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1702658675026454,
+      "learning_rate": 9.94917090927785e-06,
+      "loss": 0.3918,
+      "step": 3045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0356551600357413,
+      "learning_rate": 9.949114855484863e-06,
+      "loss": 0.513,
+      "step": 3046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4319215787690356,
+      "learning_rate": 9.949058770959212e-06,
+      "loss": 0.2339,
+      "step": 3047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.191877241849387,
+      "learning_rate": 9.94900265570125e-06,
+      "loss": 0.3258,
+      "step": 3048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8638250089348232,
+      "learning_rate": 9.948946509711324e-06,
+      "loss": 0.4237,
+      "step": 3049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.105409423190451,
+      "learning_rate": 9.948890332989782e-06,
+      "loss": 0.4592,
+      "step": 3050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4057507979938388,
+      "learning_rate": 9.948834125536973e-06,
+      "loss": 0.2352,
+      "step": 3051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8229240426981517,
+      "learning_rate": 9.948777887353244e-06,
+      "loss": 0.1862,
+      "step": 3052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0208867281483007,
+      "learning_rate": 9.94872161843895e-06,
+      "loss": 0.5653,
+      "step": 3053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.232292352484334,
+      "learning_rate": 9.948665318794435e-06,
+      "loss": 0.8832,
+      "step": 3054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8987574915961014,
+      "learning_rate": 9.948608988420051e-06,
+      "loss": 0.3965,
+      "step": 3055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5943371705584728,
+      "learning_rate": 9.948552627316148e-06,
+      "loss": 0.1733,
+      "step": 3056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8802784213654555,
+      "learning_rate": 9.948496235483074e-06,
+      "loss": 0.257,
+      "step": 3057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1605196124674,
+      "learning_rate": 9.94843981292118e-06,
+      "loss": 0.6052,
+      "step": 3058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.141431601658616,
+      "learning_rate": 9.948383359630819e-06,
+      "loss": 0.1617,
+      "step": 3059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5259836292188662,
+      "learning_rate": 9.948326875612337e-06,
+      "loss": 0.3659,
+      "step": 3060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6279001879751023,
+      "learning_rate": 9.948270360866089e-06,
+      "loss": 0.1976,
+      "step": 3061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6258317900120292,
+      "learning_rate": 9.948213815392424e-06,
+      "loss": 0.2615,
+      "step": 3062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8658491429709345,
+      "learning_rate": 9.94815723919169e-06,
+      "loss": 0.2347,
+      "step": 3063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1926967088960219,
+      "learning_rate": 9.948100632264246e-06,
+      "loss": 0.1991,
+      "step": 3064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0497853122357053,
+      "learning_rate": 9.948043994610437e-06,
+      "loss": 0.2107,
+      "step": 3065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.421569392703618,
+      "learning_rate": 9.947987326230618e-06,
+      "loss": 0.2892,
+      "step": 3066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3229936496454457,
+      "learning_rate": 9.947930627125138e-06,
+      "loss": 0.4406,
+      "step": 3067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6939610662951614,
+      "learning_rate": 9.94787389729435e-06,
+      "loss": 0.1981,
+      "step": 3068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0122552100867805,
+      "learning_rate": 9.947817136738609e-06,
+      "loss": 0.2317,
+      "step": 3069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.292128418692131,
+      "learning_rate": 9.947760345458265e-06,
+      "loss": 0.209,
+      "step": 3070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2653413726591674,
+      "learning_rate": 9.94770352345367e-06,
+      "loss": 0.489,
+      "step": 3071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.23671345483703,
+      "learning_rate": 9.94764667072518e-06,
+      "loss": 0.1594,
+      "step": 3072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5821719448493547,
+      "learning_rate": 9.947589787273144e-06,
+      "loss": 0.2321,
+      "step": 3073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4359880560940068,
+      "learning_rate": 9.947532873097918e-06,
+      "loss": 0.3817,
+      "step": 3074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7804706776011379,
+      "learning_rate": 9.947475928199855e-06,
+      "loss": 0.3544,
+      "step": 3075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.928561689179417,
+      "learning_rate": 9.947418952579309e-06,
+      "loss": 0.4838,
+      "step": 3076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5799538223621536,
+      "learning_rate": 9.947361946236632e-06,
+      "loss": 0.3896,
+      "step": 3077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8549557599106226,
+      "learning_rate": 9.947304909172178e-06,
+      "loss": 0.3454,
+      "step": 3078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.927857569938935,
+      "learning_rate": 9.947247841386304e-06,
+      "loss": 0.2979,
+      "step": 3079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9275786680993146,
+      "learning_rate": 9.947190742879362e-06,
+      "loss": 0.3996,
+      "step": 3080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.652893585907055,
+      "learning_rate": 9.947133613651708e-06,
+      "loss": 0.5115,
+      "step": 3081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0057563346687637,
+      "learning_rate": 9.947076453703694e-06,
+      "loss": 0.3211,
+      "step": 3082
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5058456796302886,
+      "learning_rate": 9.947019263035679e-06,
+      "loss": 0.29,
+      "step": 3083
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5465634109253898,
+      "learning_rate": 9.946962041648015e-06,
+      "loss": 0.2568,
+      "step": 3084
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2281270469880834,
+      "learning_rate": 9.946904789541059e-06,
+      "loss": 0.6042,
+      "step": 3085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.636446932524902,
+      "learning_rate": 9.946847506715165e-06,
+      "loss": 0.1565,
+      "step": 3086
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9292084976255355,
+      "learning_rate": 9.94679019317069e-06,
+      "loss": 0.5934,
+      "step": 3087
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5779338215706384,
+      "learning_rate": 9.94673284890799e-06,
+      "loss": 0.2545,
+      "step": 3088
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6468208213595046,
+      "learning_rate": 9.94667547392742e-06,
+      "loss": 0.2536,
+      "step": 3089
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1953303618060795,
+      "learning_rate": 9.946618068229337e-06,
+      "loss": 0.5586,
+      "step": 3090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8471362005088954,
+      "learning_rate": 9.946560631814098e-06,
+      "loss": 0.4467,
+      "step": 3091
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0021703164982894,
+      "learning_rate": 9.946503164682058e-06,
+      "loss": 0.407,
+      "step": 3092
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7049373850466343,
+      "learning_rate": 9.946445666833576e-06,
+      "loss": 0.1939,
+      "step": 3093
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9798259413701734,
+      "learning_rate": 9.946388138269007e-06,
+      "loss": 0.1805,
+      "step": 3094
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.752419106699519,
+      "learning_rate": 9.946330578988709e-06,
+      "loss": 0.5694,
+      "step": 3095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3388452580644365,
+      "learning_rate": 9.946272988993042e-06,
+      "loss": 0.3726,
+      "step": 3096
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.000547192816185,
+      "learning_rate": 9.946215368282359e-06,
+      "loss": 0.446,
+      "step": 3097
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.503233697295942,
+      "learning_rate": 9.94615771685702e-06,
+      "loss": 0.3533,
+      "step": 3098
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.716685252417912,
+      "learning_rate": 9.946100034717383e-06,
+      "loss": 0.2421,
+      "step": 3099
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7706517299394382,
+      "learning_rate": 9.946042321863807e-06,
+      "loss": 0.3552,
+      "step": 3100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.668053872752746,
+      "learning_rate": 9.945984578296648e-06,
+      "loss": 0.2965,
+      "step": 3101
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.126555837222646,
+      "learning_rate": 9.945926804016267e-06,
+      "loss": 0.1917,
+      "step": 3102
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7741429956927537,
+      "learning_rate": 9.945868999023023e-06,
+      "loss": 0.2185,
+      "step": 3103
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7041162222420232,
+      "learning_rate": 9.945811163317274e-06,
+      "loss": 0.2234,
+      "step": 3104
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0074898900336255,
+      "learning_rate": 9.945753296899377e-06,
+      "loss": 0.229,
+      "step": 3105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9448605360911269,
+      "learning_rate": 9.945695399769695e-06,
+      "loss": 0.4961,
+      "step": 3106
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0048523801639098,
+      "learning_rate": 9.945637471928585e-06,
+      "loss": 0.2266,
+      "step": 3107
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3648097557243304,
+      "learning_rate": 9.94557951337641e-06,
+      "loss": 0.5136,
+      "step": 3108
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7175900846006784,
+      "learning_rate": 9.945521524113524e-06,
+      "loss": 0.3398,
+      "step": 3109
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.68679201474057,
+      "learning_rate": 9.945463504140292e-06,
+      "loss": 0.5598,
+      "step": 3110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6978887110048901,
+      "learning_rate": 9.945405453457072e-06,
+      "loss": 0.339,
+      "step": 3111
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6652637855819357,
+      "learning_rate": 9.945347372064226e-06,
+      "loss": 0.3419,
+      "step": 3112
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6867623965393235,
+      "learning_rate": 9.945289259962115e-06,
+      "loss": 0.3932,
+      "step": 3113
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2796152606732427,
+      "learning_rate": 9.945231117151098e-06,
+      "loss": 0.3428,
+      "step": 3114
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.443159282733628,
+      "learning_rate": 9.945172943631538e-06,
+      "loss": 0.499,
+      "step": 3115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.033691589428732,
+      "learning_rate": 9.945114739403793e-06,
+      "loss": 0.3812,
+      "step": 3116
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5354843031799212,
+      "learning_rate": 9.94505650446823e-06,
+      "loss": 0.225,
+      "step": 3117
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.974661272464115,
+      "learning_rate": 9.944998238825205e-06,
+      "loss": 0.434,
+      "step": 3118
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1027835468153326,
+      "learning_rate": 9.944939942475082e-06,
+      "loss": 0.3086,
+      "step": 3119
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.895492804707057,
+      "learning_rate": 9.944881615418222e-06,
+      "loss": 0.5208,
+      "step": 3120
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 43.188268453162394,
+      "learning_rate": 9.944823257654991e-06,
+      "loss": 0.7154,
+      "step": 3121
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0290094691604814,
+      "learning_rate": 9.944764869185747e-06,
+      "loss": 0.2226,
+      "step": 3122
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9337424327058688,
+      "learning_rate": 9.944706450010854e-06,
+      "loss": 0.3045,
+      "step": 3123
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.069846195490793,
+      "learning_rate": 9.944648000130676e-06,
+      "loss": 0.3057,
+      "step": 3124
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 82.58434533372615,
+      "learning_rate": 9.944589519545576e-06,
+      "loss": 0.5501,
+      "step": 3125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.374661013537227,
+      "learning_rate": 9.944531008255912e-06,
+      "loss": 0.4949,
+      "step": 3126
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 116.61913957825662,
+      "learning_rate": 9.944472466262057e-06,
+      "loss": 1.0232,
+      "step": 3127
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.2350989908588925,
+      "learning_rate": 9.944413893564365e-06,
+      "loss": 0.4319,
+      "step": 3128
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.538062993437981,
+      "learning_rate": 9.944355290163204e-06,
+      "loss": 0.3185,
+      "step": 3129
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.436839505730756,
+      "learning_rate": 9.944296656058939e-06,
+      "loss": 0.7627,
+      "step": 3130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.71285395731048,
+      "learning_rate": 9.944237991251931e-06,
+      "loss": 0.6416,
+      "step": 3131
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.879475289113219,
+      "learning_rate": 9.944179295742548e-06,
+      "loss": 0.5848,
+      "step": 3132
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.332701889470311,
+      "learning_rate": 9.94412056953115e-06,
+      "loss": 0.3474,
+      "step": 3133
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.925564726086609,
+      "learning_rate": 9.944061812618107e-06,
+      "loss": 0.5582,
+      "step": 3134
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 26.668959044233254,
+      "learning_rate": 9.944003025003779e-06,
+      "loss": 0.4135,
+      "step": 3135
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.921122292656005,
+      "learning_rate": 9.943944206688535e-06,
+      "loss": 0.8144,
+      "step": 3136
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.856775292889266,
+      "learning_rate": 9.943885357672736e-06,
+      "loss": 0.7989,
+      "step": 3137
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.252076601142408,
+      "learning_rate": 9.94382647795675e-06,
+      "loss": 0.3533,
+      "step": 3138
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.465687699481652,
+      "learning_rate": 9.943767567540944e-06,
+      "loss": 0.3812,
+      "step": 3139
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.425263488655563,
+      "learning_rate": 9.94370862642568e-06,
+      "loss": 0.3598,
+      "step": 3140
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.320169583842034,
+      "learning_rate": 9.943649654611329e-06,
+      "loss": 0.4531,
+      "step": 3141
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 26.231457328204268,
+      "learning_rate": 9.943590652098254e-06,
+      "loss": 0.8233,
+      "step": 3142
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.354072983508124,
+      "learning_rate": 9.943531618886819e-06,
+      "loss": 0.3605,
+      "step": 3143
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.985862447476746,
+      "learning_rate": 9.943472554977395e-06,
+      "loss": 0.2353,
+      "step": 3144
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.84957428222727,
+      "learning_rate": 9.943413460370348e-06,
+      "loss": 0.8442,
+      "step": 3145
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.406758097526466,
+      "learning_rate": 9.943354335066044e-06,
+      "loss": 0.6837,
+      "step": 3146
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.05361884807542,
+      "learning_rate": 9.943295179064849e-06,
+      "loss": 0.6161,
+      "step": 3147
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5008265082742756,
+      "learning_rate": 9.943235992367132e-06,
+      "loss": 0.2664,
+      "step": 3148
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6266759790055363,
+      "learning_rate": 9.94317677497326e-06,
+      "loss": 0.6602,
+      "step": 3149
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.387706238637477,
+      "learning_rate": 9.943117526883602e-06,
+      "loss": 0.7325,
+      "step": 3150
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2058386496042095,
+      "learning_rate": 9.943058248098523e-06,
+      "loss": 0.4486,
+      "step": 3151
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 84.15132534096314,
+      "learning_rate": 9.942998938618394e-06,
+      "loss": 0.6619,
+      "step": 3152
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.48748409199573,
+      "learning_rate": 9.942939598443581e-06,
+      "loss": 0.7408,
+      "step": 3153
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.763518250246666,
+      "learning_rate": 9.942880227574456e-06,
+      "loss": 0.7066,
+      "step": 3154
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.017632996532159,
+      "learning_rate": 9.942820826011384e-06,
+      "loss": 0.2049,
+      "step": 3155
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9372972816889533,
+      "learning_rate": 9.942761393754737e-06,
+      "loss": 0.2192,
+      "step": 3156
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.823515998496424,
+      "learning_rate": 9.942701930804878e-06,
+      "loss": 0.4305,
+      "step": 3157
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7254168910800254,
+      "learning_rate": 9.942642437162185e-06,
+      "loss": 0.3517,
+      "step": 3158
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.404183045736552,
+      "learning_rate": 9.94258291282702e-06,
+      "loss": 0.19,
+      "step": 3159
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2428991800564044,
+      "learning_rate": 9.942523357799756e-06,
+      "loss": 0.6921,
+      "step": 3160
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4886425085481956,
+      "learning_rate": 9.942463772080764e-06,
+      "loss": 0.157,
+      "step": 3161
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8497849345625603,
+      "learning_rate": 9.942404155670411e-06,
+      "loss": 0.4687,
+      "step": 3162
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7806701913026424,
+      "learning_rate": 9.942344508569068e-06,
+      "loss": 0.5165,
+      "step": 3163
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1280395256166071,
+      "learning_rate": 9.942284830777109e-06,
+      "loss": 0.2208,
+      "step": 3164
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.656730206948277,
+      "learning_rate": 9.942225122294898e-06,
+      "loss": 0.8003,
+      "step": 3165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.490707882013047,
+      "learning_rate": 9.942165383122811e-06,
+      "loss": 0.2386,
+      "step": 3166
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.315211716604277,
+      "learning_rate": 9.942105613261216e-06,
+      "loss": 0.6111,
+      "step": 3167
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.943424408921313,
+      "learning_rate": 9.942045812710488e-06,
+      "loss": 0.3044,
+      "step": 3168
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.911104259560314,
+      "learning_rate": 9.941985981470996e-06,
+      "loss": 0.2072,
+      "step": 3169
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9879332876846532,
+      "learning_rate": 9.941926119543108e-06,
+      "loss": 0.3119,
+      "step": 3170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3387555939114952,
+      "learning_rate": 9.941866226927202e-06,
+      "loss": 0.1798,
+      "step": 3171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3361874545528605,
+      "learning_rate": 9.941806303623647e-06,
+      "loss": 0.2688,
+      "step": 3172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6363742631590434,
+      "learning_rate": 9.941746349632813e-06,
+      "loss": 0.7029,
+      "step": 3173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4724494220837037,
+      "learning_rate": 9.941686364955077e-06,
+      "loss": 0.5134,
+      "step": 3174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.70881503809514,
+      "learning_rate": 9.941626349590807e-06,
+      "loss": 0.4046,
+      "step": 3175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4919439992065744,
+      "learning_rate": 9.941566303540377e-06,
+      "loss": 0.4313,
+      "step": 3176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.9231999187906,
+      "learning_rate": 9.941506226804162e-06,
+      "loss": 0.5978,
+      "step": 3177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.82886343494018,
+      "learning_rate": 9.941446119382533e-06,
+      "loss": 0.385,
+      "step": 3178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9417602630767372,
+      "learning_rate": 9.941385981275863e-06,
+      "loss": 0.3047,
+      "step": 3179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5933359459072802,
+      "learning_rate": 9.941325812484527e-06,
+      "loss": 0.5032,
+      "step": 3180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2846853178522863,
+      "learning_rate": 9.9412656130089e-06,
+      "loss": 0.4015,
+      "step": 3181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.191345968794315,
+      "learning_rate": 9.94120538284935e-06,
+      "loss": 0.4834,
+      "step": 3182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3641278198831404,
+      "learning_rate": 9.941145122006257e-06,
+      "loss": 0.2536,
+      "step": 3183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5481366108661847,
+      "learning_rate": 9.94108483047999e-06,
+      "loss": 0.2375,
+      "step": 3184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9412239342172015,
+      "learning_rate": 9.941024508270928e-06,
+      "loss": 0.2998,
+      "step": 3185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.08487002601124,
+      "learning_rate": 9.940964155379446e-06,
+      "loss": 0.2029,
+      "step": 3186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6118824508157543,
+      "learning_rate": 9.940903771805915e-06,
+      "loss": 0.2053,
+      "step": 3187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6517885867208004,
+      "learning_rate": 9.94084335755071e-06,
+      "loss": 0.4006,
+      "step": 3188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.776918565804417,
+      "learning_rate": 9.94078291261421e-06,
+      "loss": 0.2339,
+      "step": 3189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6004092320834737,
+      "learning_rate": 9.940722436996788e-06,
+      "loss": 0.5698,
+      "step": 3190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9873396661852776,
+      "learning_rate": 9.940661930698818e-06,
+      "loss": 0.3753,
+      "step": 3191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0417022134320377,
+      "learning_rate": 9.940601393720677e-06,
+      "loss": 0.2575,
+      "step": 3192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2656107606111493,
+      "learning_rate": 9.940540826062743e-06,
+      "loss": 0.3722,
+      "step": 3193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6101005085528468,
+      "learning_rate": 9.940480227725392e-06,
+      "loss": 0.3346,
+      "step": 3194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4616876858825996,
+      "learning_rate": 9.940419598708995e-06,
+      "loss": 0.2047,
+      "step": 3195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6373028947551223,
+      "learning_rate": 9.940358939013933e-06,
+      "loss": 0.1843,
+      "step": 3196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4658951864110694,
+      "learning_rate": 9.940298248640584e-06,
+      "loss": 0.3929,
+      "step": 3197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.064465716710663,
+      "learning_rate": 9.940237527589321e-06,
+      "loss": 0.263,
+      "step": 3198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.977548353385293,
+      "learning_rate": 9.940176775860522e-06,
+      "loss": 0.3383,
+      "step": 3199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4663569646338988,
+      "learning_rate": 9.940115993454567e-06,
+      "loss": 0.5644,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9779483550569377,
+      "learning_rate": 9.940055180371831e-06,
+      "loss": 0.4345,
+      "step": 3201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3894182038947553,
+      "learning_rate": 9.93999433661269e-06,
+      "loss": 0.346,
+      "step": 3202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.290230297927582,
+      "learning_rate": 9.939933462177527e-06,
+      "loss": 0.1454,
+      "step": 3203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6585060355180434,
+      "learning_rate": 9.939872557066715e-06,
+      "loss": 0.4129,
+      "step": 3204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6517657854100456,
+      "learning_rate": 9.939811621280635e-06,
+      "loss": 0.3091,
+      "step": 3205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8435638463071375,
+      "learning_rate": 9.939750654819665e-06,
+      "loss": 0.2288,
+      "step": 3206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6552521121482973,
+      "learning_rate": 9.939689657684182e-06,
+      "loss": 0.5081,
+      "step": 3207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7951614502260305,
+      "learning_rate": 9.939628629874566e-06,
+      "loss": 0.3761,
+      "step": 3208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5096317863607152,
+      "learning_rate": 9.939567571391196e-06,
+      "loss": 0.3174,
+      "step": 3209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3647308947751933,
+      "learning_rate": 9.939506482234451e-06,
+      "loss": 0.2856,
+      "step": 3210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.112870569569435,
+      "learning_rate": 9.939445362404712e-06,
+      "loss": 0.3724,
+      "step": 3211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.241836426940015,
+      "learning_rate": 9.939384211902354e-06,
+      "loss": 0.6273,
+      "step": 3212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.331400135962809,
+      "learning_rate": 9.939323030727763e-06,
+      "loss": 0.2595,
+      "step": 3213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7576919079847284,
+      "learning_rate": 9.939261818881312e-06,
+      "loss": 0.1821,
+      "step": 3214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.131377757625665,
+      "learning_rate": 9.939200576363387e-06,
+      "loss": 0.3568,
+      "step": 3215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7692921194808597,
+      "learning_rate": 9.939139303174365e-06,
+      "loss": 0.2925,
+      "step": 3216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6211302328026023,
+      "learning_rate": 9.939077999314628e-06,
+      "loss": 0.3312,
+      "step": 3217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.343113214294503,
+      "learning_rate": 9.939016664784557e-06,
+      "loss": 0.2771,
+      "step": 3218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9357642658919425,
+      "learning_rate": 9.938955299584531e-06,
+      "loss": 0.2973,
+      "step": 3219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6022525231306852,
+      "learning_rate": 9.938893903714932e-06,
+      "loss": 0.2149,
+      "step": 3220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8930817318842783,
+      "learning_rate": 9.938832477176141e-06,
+      "loss": 0.397,
+      "step": 3221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2269224578614053,
+      "learning_rate": 9.93877101996854e-06,
+      "loss": 0.4889,
+      "step": 3222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5224481769293712,
+      "learning_rate": 9.938709532092513e-06,
+      "loss": 0.2037,
+      "step": 3223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.845245728508078,
+      "learning_rate": 9.938648013548436e-06,
+      "loss": 0.3335,
+      "step": 3224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.011404582656385,
+      "learning_rate": 9.938586464336694e-06,
+      "loss": 0.4356,
+      "step": 3225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7669096783985105,
+      "learning_rate": 9.93852488445767e-06,
+      "loss": 0.2451,
+      "step": 3226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.105141375397188,
+      "learning_rate": 9.938463273911748e-06,
+      "loss": 0.3966,
+      "step": 3227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.059249049779393,
+      "learning_rate": 9.938401632699306e-06,
+      "loss": 0.4346,
+      "step": 3228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3416328022726622,
+      "learning_rate": 9.93833996082073e-06,
+      "loss": 0.4339,
+      "step": 3229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1898544719541073,
+      "learning_rate": 9.938278258276401e-06,
+      "loss": 0.339,
+      "step": 3230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.839744891475338,
+      "learning_rate": 9.938216525066706e-06,
+      "loss": 0.5979,
+      "step": 3231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9496765573796462,
+      "learning_rate": 9.938154761192023e-06,
+      "loss": 0.2009,
+      "step": 3232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.089171271106691,
+      "learning_rate": 9.93809296665274e-06,
+      "loss": 0.4036,
+      "step": 3233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7143044221036794,
+      "learning_rate": 9.938031141449239e-06,
+      "loss": 0.2551,
+      "step": 3234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.386362889009004,
+      "learning_rate": 9.937969285581904e-06,
+      "loss": 0.5229,
+      "step": 3235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2727655388721906,
+      "learning_rate": 9.937907399051117e-06,
+      "loss": 0.2476,
+      "step": 3236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.436980363280678,
+      "learning_rate": 9.937845481857265e-06,
+      "loss": 0.5857,
+      "step": 3237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5943930908561486,
+      "learning_rate": 9.93778353400073e-06,
+      "loss": 0.314,
+      "step": 3238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2412776629585887,
+      "learning_rate": 9.937721555481903e-06,
+      "loss": 0.3665,
+      "step": 3239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3653582024253894,
+      "learning_rate": 9.937659546301161e-06,
+      "loss": 0.2927,
+      "step": 3240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6853383091047267,
+      "learning_rate": 9.937597506458893e-06,
+      "loss": 0.2089,
+      "step": 3241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9161695288267666,
+      "learning_rate": 9.937535435955485e-06,
+      "loss": 0.371,
+      "step": 3242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6269529784570014,
+      "learning_rate": 9.93747333479132e-06,
+      "loss": 0.3041,
+      "step": 3243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7390230203484642,
+      "learning_rate": 9.937411202966783e-06,
+      "loss": 0.3278,
+      "step": 3244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5476362810753117,
+      "learning_rate": 9.937349040482262e-06,
+      "loss": 0.2261,
+      "step": 3245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1328974137502605,
+      "learning_rate": 9.937286847338146e-06,
+      "loss": 0.5,
+      "step": 3246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.070446924128094,
+      "learning_rate": 9.937224623534813e-06,
+      "loss": 0.3366,
+      "step": 3247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5187339909910704,
+      "learning_rate": 9.937162369072656e-06,
+      "loss": 0.2059,
+      "step": 3248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1298538059111687,
+      "learning_rate": 9.937100083952061e-06,
+      "loss": 0.3105,
+      "step": 3249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.734087112521571,
+      "learning_rate": 9.937037768173412e-06,
+      "loss": 0.5365,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6322320426546741,
+      "learning_rate": 9.936975421737098e-06,
+      "loss": 0.3608,
+      "step": 3251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3918227070218685,
+      "learning_rate": 9.936913044643505e-06,
+      "loss": 0.2328,
+      "step": 3252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7959717014256167,
+      "learning_rate": 9.936850636893022e-06,
+      "loss": 0.4042,
+      "step": 3253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7966199237725498,
+      "learning_rate": 9.936788198486035e-06,
+      "loss": 0.2296,
+      "step": 3254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7400162428475947,
+      "learning_rate": 9.936725729422932e-06,
+      "loss": 0.2405,
+      "step": 3255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0832586517458753,
+      "learning_rate": 9.936663229704102e-06,
+      "loss": 0.2292,
+      "step": 3256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.917434054581085,
+      "learning_rate": 9.93660069932993e-06,
+      "loss": 0.2194,
+      "step": 3257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9441113881186667,
+      "learning_rate": 9.936538138300808e-06,
+      "loss": 0.405,
+      "step": 3258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.310598235796903,
+      "learning_rate": 9.936475546617124e-06,
+      "loss": 0.4161,
+      "step": 3259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9862943213183026,
+      "learning_rate": 9.936412924279264e-06,
+      "loss": 0.5132,
+      "step": 3260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7317842689864995,
+      "learning_rate": 9.93635027128762e-06,
+      "loss": 0.5894,
+      "step": 3261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8718315055774164,
+      "learning_rate": 9.936287587642576e-06,
+      "loss": 0.4906,
+      "step": 3262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.378013408839847,
+      "learning_rate": 9.936224873344529e-06,
+      "loss": 0.2882,
+      "step": 3263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1624225641178745,
+      "learning_rate": 9.936162128393863e-06,
+      "loss": 0.5186,
+      "step": 3264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.235683174917416,
+      "learning_rate": 9.93609935279097e-06,
+      "loss": 0.3845,
+      "step": 3265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9997017384693008,
+      "learning_rate": 9.936036546536237e-06,
+      "loss": 0.4713,
+      "step": 3266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6011472373010536,
+      "learning_rate": 9.935973709630055e-06,
+      "loss": 0.2069,
+      "step": 3267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.343051497959274,
+      "learning_rate": 9.935910842072815e-06,
+      "loss": 0.2116,
+      "step": 3268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0615749813486683,
+      "learning_rate": 9.93584794386491e-06,
+      "loss": 0.2023,
+      "step": 3269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.339482821849827,
+      "learning_rate": 9.935785015006726e-06,
+      "loss": 0.2039,
+      "step": 3270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9579470503742158,
+      "learning_rate": 9.935722055498655e-06,
+      "loss": 0.4754,
+      "step": 3271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1881275775396296,
+      "learning_rate": 9.93565906534109e-06,
+      "loss": 0.3587,
+      "step": 3272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8023672877024413,
+      "learning_rate": 9.935596044534422e-06,
+      "loss": 0.2542,
+      "step": 3273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8870824538316093,
+      "learning_rate": 9.935532993079038e-06,
+      "loss": 0.4757,
+      "step": 3274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9254453880739624,
+      "learning_rate": 9.935469910975334e-06,
+      "loss": 0.5148,
+      "step": 3275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.048809440534869,
+      "learning_rate": 9.935406798223701e-06,
+      "loss": 0.512,
+      "step": 3276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4201060941588306,
+      "learning_rate": 9.935343654824531e-06,
+      "loss": 0.7031,
+      "step": 3277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3375433565902495,
+      "learning_rate": 9.935280480778213e-06,
+      "loss": 0.4806,
+      "step": 3278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8285837005545096,
+      "learning_rate": 9.935217276085144e-06,
+      "loss": 0.3638,
+      "step": 3279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7411552073684415,
+      "learning_rate": 9.935154040745714e-06,
+      "loss": 0.3599,
+      "step": 3280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.173769735013099,
+      "learning_rate": 9.935090774760313e-06,
+      "loss": 0.4433,
+      "step": 3281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5020908396749812,
+      "learning_rate": 9.935027478129339e-06,
+      "loss": 0.277,
+      "step": 3282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6333396644696645,
+      "learning_rate": 9.934964150853184e-06,
+      "loss": 0.4115,
+      "step": 3283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5239908961667303,
+      "learning_rate": 9.934900792932238e-06,
+      "loss": 0.2623,
+      "step": 3284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.141165946912992,
+      "learning_rate": 9.934837404366896e-06,
+      "loss": 0.2529,
+      "step": 3285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.931620081839561,
+      "learning_rate": 9.934773985157556e-06,
+      "loss": 0.4254,
+      "step": 3286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.486021047744415,
+      "learning_rate": 9.934710535304603e-06,
+      "loss": 0.3178,
+      "step": 3287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.744540297966952,
+      "learning_rate": 9.934647054808439e-06,
+      "loss": 0.4396,
+      "step": 3288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2977455915600307,
+      "learning_rate": 9.934583543669454e-06,
+      "loss": 0.4932,
+      "step": 3289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7858091742134439,
+      "learning_rate": 9.934520001888044e-06,
+      "loss": 0.3628,
+      "step": 3290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2924481244390282,
+      "learning_rate": 9.934456429464602e-06,
+      "loss": 0.6995,
+      "step": 3291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6946108196565954,
+      "learning_rate": 9.934392826399525e-06,
+      "loss": 0.2728,
+      "step": 3292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8338376271907657,
+      "learning_rate": 9.934329192693206e-06,
+      "loss": 0.3388,
+      "step": 3293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.595778116897745,
+      "learning_rate": 9.934265528346042e-06,
+      "loss": 0.4827,
+      "step": 3294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9920571826871867,
+      "learning_rate": 9.934201833358425e-06,
+      "loss": 0.3396,
+      "step": 3295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4690599759978187,
+      "learning_rate": 9.934138107730755e-06,
+      "loss": 0.4357,
+      "step": 3296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7143741338712224,
+      "learning_rate": 9.934074351463425e-06,
+      "loss": 0.3316,
+      "step": 3297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0055217960298357,
+      "learning_rate": 9.934010564556832e-06,
+      "loss": 0.413,
+      "step": 3298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8717680096499278,
+      "learning_rate": 9.93394674701137e-06,
+      "loss": 0.2865,
+      "step": 3299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4457107439030614,
+      "learning_rate": 9.933882898827439e-06,
+      "loss": 0.6729,
+      "step": 3300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5156053814804205,
+      "learning_rate": 9.933819020005432e-06,
+      "loss": 0.5901,
+      "step": 3301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.079891543907819,
+      "learning_rate": 9.933755110545746e-06,
+      "loss": 0.3034,
+      "step": 3302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.688182419086753,
+      "learning_rate": 9.93369117044878e-06,
+      "loss": 0.3065,
+      "step": 3303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.92569753375904,
+      "learning_rate": 9.93362719971493e-06,
+      "loss": 0.3095,
+      "step": 3304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6981937781043046,
+      "learning_rate": 9.933563198344594e-06,
+      "loss": 0.3213,
+      "step": 3305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5828852926727126,
+      "learning_rate": 9.933499166338168e-06,
+      "loss": 0.3271,
+      "step": 3306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8608602994636818,
+      "learning_rate": 9.933435103696048e-06,
+      "loss": 0.4435,
+      "step": 3307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7046279367535733,
+      "learning_rate": 9.933371010418635e-06,
+      "loss": 0.2249,
+      "step": 3308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.498492812387802,
+      "learning_rate": 9.933306886506328e-06,
+      "loss": 0.1828,
+      "step": 3309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.24200810651846,
+      "learning_rate": 9.933242731959522e-06,
+      "loss": 0.2269,
+      "step": 3310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5572670378592708,
+      "learning_rate": 9.933178546778615e-06,
+      "loss": 0.1808,
+      "step": 3311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8545558129073285,
+      "learning_rate": 9.933114330964009e-06,
+      "loss": 0.5279,
+      "step": 3312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9118057181309038,
+      "learning_rate": 9.9330500845161e-06,
+      "loss": 0.3923,
+      "step": 3313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0506089892019586,
+      "learning_rate": 9.932985807435289e-06,
+      "loss": 0.3901,
+      "step": 3314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9842888520531905,
+      "learning_rate": 9.932921499721973e-06,
+      "loss": 0.2131,
+      "step": 3315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7550188138182947,
+      "learning_rate": 9.932857161376553e-06,
+      "loss": 0.17,
+      "step": 3316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0907032063885698,
+      "learning_rate": 9.932792792399428e-06,
+      "loss": 0.518,
+      "step": 3317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3888571130101774,
+      "learning_rate": 9.932728392790996e-06,
+      "loss": 0.5715,
+      "step": 3318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4380709638933822,
+      "learning_rate": 9.93266396255166e-06,
+      "loss": 0.1758,
+      "step": 3319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0728226756986703,
+      "learning_rate": 9.932599501681819e-06,
+      "loss": 0.5287,
+      "step": 3320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6688396940521875,
+      "learning_rate": 9.932535010181872e-06,
+      "loss": 0.3242,
+      "step": 3321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.80269892855065,
+      "learning_rate": 9.932470488052219e-06,
+      "loss": 0.2888,
+      "step": 3322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6461090076320586,
+      "learning_rate": 9.932405935293264e-06,
+      "loss": 0.2693,
+      "step": 3323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.474537639262505,
+      "learning_rate": 9.932341351905405e-06,
+      "loss": 0.1736,
+      "step": 3324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.064538035455592,
+      "learning_rate": 9.932276737889044e-06,
+      "loss": 0.4105,
+      "step": 3325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7981398629478218,
+      "learning_rate": 9.932212093244583e-06,
+      "loss": 0.3379,
+      "step": 3326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7124584298674677,
+      "learning_rate": 9.93214741797242e-06,
+      "loss": 0.1912,
+      "step": 3327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7722416366389768,
+      "learning_rate": 9.932082712072962e-06,
+      "loss": 0.3331,
+      "step": 3328
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.374253097243737,
+      "learning_rate": 9.932017975546606e-06,
+      "loss": 0.2858,
+      "step": 3329
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.490791306725128,
+      "learning_rate": 9.931953208393758e-06,
+      "loss": 0.2347,
+      "step": 3330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.855201407325057,
+      "learning_rate": 9.931888410614816e-06,
+      "loss": 0.2648,
+      "step": 3331
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7956047502498522,
+      "learning_rate": 9.931823582210187e-06,
+      "loss": 0.3282,
+      "step": 3332
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0022846492325685,
+      "learning_rate": 9.93175872318027e-06,
+      "loss": 0.531,
+      "step": 3333
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8829267000253367,
+      "learning_rate": 9.931693833525468e-06,
+      "loss": 0.3605,
+      "step": 3334
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2613128416636423,
+      "learning_rate": 9.931628913246185e-06,
+      "loss": 0.2002,
+      "step": 3335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0870156270126627,
+      "learning_rate": 9.931563962342824e-06,
+      "loss": 0.2957,
+      "step": 3336
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.842462728775567,
+      "learning_rate": 9.931498980815788e-06,
+      "loss": 0.2608,
+      "step": 3337
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1836166568465682,
+      "learning_rate": 9.931433968665482e-06,
+      "loss": 0.1711,
+      "step": 3338
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7810267551129728,
+      "learning_rate": 9.931368925892308e-06,
+      "loss": 0.3312,
+      "step": 3339
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7204599197794392,
+      "learning_rate": 9.93130385249667e-06,
+      "loss": 0.3543,
+      "step": 3340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2401358092181771,
+      "learning_rate": 9.931238748478973e-06,
+      "loss": 0.1787,
+      "step": 3341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.013193023607743,
+      "learning_rate": 9.93117361383962e-06,
+      "loss": 0.496,
+      "step": 3342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0276234228062644,
+      "learning_rate": 9.931108448579017e-06,
+      "loss": 0.2242,
+      "step": 3343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2730375115017685,
+      "learning_rate": 9.931043252697568e-06,
+      "loss": 0.4061,
+      "step": 3344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8513785323161527,
+      "learning_rate": 9.930978026195676e-06,
+      "loss": 0.2983,
+      "step": 3345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4488076544907202,
+      "learning_rate": 9.93091276907375e-06,
+      "loss": 0.2253,
+      "step": 3346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7281092236024644,
+      "learning_rate": 9.930847481332193e-06,
+      "loss": 0.2471,
+      "step": 3347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.665254470449809,
+      "learning_rate": 9.93078216297141e-06,
+      "loss": 0.2537,
+      "step": 3348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.588065218740476,
+      "learning_rate": 9.930716813991805e-06,
+      "loss": 0.5368,
+      "step": 3349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0567082070660976,
+      "learning_rate": 9.930651434393788e-06,
+      "loss": 0.3201,
+      "step": 3350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6155934932078915,
+      "learning_rate": 9.930586024177764e-06,
+      "loss": 0.3133,
+      "step": 3351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3139108046381172,
+      "learning_rate": 9.930520583344136e-06,
+      "loss": 0.1567,
+      "step": 3352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0705404875837825,
+      "learning_rate": 9.930455111893311e-06,
+      "loss": 0.524,
+      "step": 3353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.445582028782796,
+      "learning_rate": 9.9303896098257e-06,
+      "loss": 0.1727,
+      "step": 3354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5257886202722655,
+      "learning_rate": 9.930324077141704e-06,
+      "loss": 0.3591,
+      "step": 3355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.087639553579175,
+      "learning_rate": 9.930258513841735e-06,
+      "loss": 0.5613,
+      "step": 3356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0974419340478225,
+      "learning_rate": 9.930192919926195e-06,
+      "loss": 0.3636,
+      "step": 3357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.192243066084853,
+      "learning_rate": 9.930127295395497e-06,
+      "loss": 0.6006,
+      "step": 3358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5940852852703342,
+      "learning_rate": 9.930061640250045e-06,
+      "loss": 0.2699,
+      "step": 3359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6607532370301916,
+      "learning_rate": 9.929995954490245e-06,
+      "loss": 0.3386,
+      "step": 3360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.682200479654931,
+      "learning_rate": 9.92993023811651e-06,
+      "loss": 0.5274,
+      "step": 3361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9660742681417704,
+      "learning_rate": 9.929864491129243e-06,
+      "loss": 0.4591,
+      "step": 3362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.83987735911973,
+      "learning_rate": 9.929798713528858e-06,
+      "loss": 0.3698,
+      "step": 3363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2484350192524873,
+      "learning_rate": 9.929732905315756e-06,
+      "loss": 0.3628,
+      "step": 3364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2329947870375606,
+      "learning_rate": 9.929667066490352e-06,
+      "loss": 0.5228,
+      "step": 3365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9152087418989996,
+      "learning_rate": 9.929601197053052e-06,
+      "loss": 0.3494,
+      "step": 3366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9653871346087988,
+      "learning_rate": 9.929535297004264e-06,
+      "loss": 0.2752,
+      "step": 3367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7867415753213052,
+      "learning_rate": 9.929469366344401e-06,
+      "loss": 0.2211,
+      "step": 3368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6535490793673429,
+      "learning_rate": 9.92940340507387e-06,
+      "loss": 0.2004,
+      "step": 3369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5162328516300194,
+      "learning_rate": 9.929337413193078e-06,
+      "loss": 0.178,
+      "step": 3370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5756615423521916,
+      "learning_rate": 9.92927139070244e-06,
+      "loss": 0.3409,
+      "step": 3371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7239106021974147,
+      "learning_rate": 9.929205337602364e-06,
+      "loss": 0.1916,
+      "step": 3372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0591060869607578,
+      "learning_rate": 9.929139253893258e-06,
+      "loss": 0.2263,
+      "step": 3373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.382907402775191,
+      "learning_rate": 9.929073139575535e-06,
+      "loss": 0.4559,
+      "step": 3374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8923880668981381,
+      "learning_rate": 9.929006994649603e-06,
+      "loss": 0.1287,
+      "step": 3375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6329399907838524,
+      "learning_rate": 9.928940819115878e-06,
+      "loss": 0.3196,
+      "step": 3376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.776888118233227,
+      "learning_rate": 9.928874612974764e-06,
+      "loss": 0.3174,
+      "step": 3377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.136681069389756,
+      "learning_rate": 9.928808376226676e-06,
+      "loss": 0.4533,
+      "step": 3378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9555221381159633,
+      "learning_rate": 9.928742108872025e-06,
+      "loss": 0.4348,
+      "step": 3379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.393934408858469,
+      "learning_rate": 9.928675810911222e-06,
+      "loss": 0.4516,
+      "step": 3380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8071076650332294,
+      "learning_rate": 9.92860948234468e-06,
+      "loss": 0.3102,
+      "step": 3381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3866352872099843,
+      "learning_rate": 9.92854312317281e-06,
+      "loss": 0.4094,
+      "step": 3382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3094475224196815,
+      "learning_rate": 9.928476733396023e-06,
+      "loss": 0.2421,
+      "step": 3383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4332517896880923,
+      "learning_rate": 9.928410313014731e-06,
+      "loss": 0.2777,
+      "step": 3384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.245221328477105,
+      "learning_rate": 9.928343862029348e-06,
+      "loss": 0.2451,
+      "step": 3385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.943134590068819,
+      "learning_rate": 9.928277380440288e-06,
+      "loss": 0.3425,
+      "step": 3386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9848632812500002,
+      "learning_rate": 9.92821086824796e-06,
+      "loss": 0.5358,
+      "step": 3387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5338261377227194,
+      "learning_rate": 9.92814432545278e-06,
+      "loss": 0.3361,
+      "step": 3388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7798783056846115,
+      "learning_rate": 9.928077752055161e-06,
+      "loss": 0.4376,
+      "step": 3389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6526645367048463,
+      "learning_rate": 9.928011148055514e-06,
+      "loss": 0.3092,
+      "step": 3390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7043470123946085,
+      "learning_rate": 9.927944513454255e-06,
+      "loss": 0.3517,
+      "step": 3391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.133446015949611,
+      "learning_rate": 9.927877848251798e-06,
+      "loss": 0.305,
+      "step": 3392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.256537978584912,
+      "learning_rate": 9.927811152448555e-06,
+      "loss": 0.3638,
+      "step": 3393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3897687574399866,
+      "learning_rate": 9.927744426044942e-06,
+      "loss": 0.2273,
+      "step": 3394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7795000975676591,
+      "learning_rate": 9.92767766904137e-06,
+      "loss": 0.3503,
+      "step": 3395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5594516080756544,
+      "learning_rate": 9.92761088143826e-06,
+      "loss": 0.2174,
+      "step": 3396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9802381693521902,
+      "learning_rate": 9.92754406323602e-06,
+      "loss": 0.3905,
+      "step": 3397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7564188412792838,
+      "learning_rate": 9.92747721443507e-06,
+      "loss": 0.2445,
+      "step": 3398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7792271203647945,
+      "learning_rate": 9.927410335035822e-06,
+      "loss": 0.3908,
+      "step": 3399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6047862930647194,
+      "learning_rate": 9.927343425038692e-06,
+      "loss": 0.41,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.15649264808585,
+      "learning_rate": 9.927276484444096e-06,
+      "loss": 0.5447,
+      "step": 3401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5487031008275114,
+      "learning_rate": 9.927209513252449e-06,
+      "loss": 0.4375,
+      "step": 3402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2248705441875924,
+      "learning_rate": 9.927142511464166e-06,
+      "loss": 0.4323,
+      "step": 3403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.809092401431036,
+      "learning_rate": 9.927075479079667e-06,
+      "loss": 0.3248,
+      "step": 3404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.676975329447816,
+      "learning_rate": 9.927008416099364e-06,
+      "loss": 0.3133,
+      "step": 3405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.935864914194053,
+      "learning_rate": 9.926941322523673e-06,
+      "loss": 0.6656,
+      "step": 3406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0085642702305138,
+      "learning_rate": 9.926874198353016e-06,
+      "loss": 0.3776,
+      "step": 3407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9800986220622843,
+      "learning_rate": 9.926807043587804e-06,
+      "loss": 0.6461,
+      "step": 3408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.80484999183316,
+      "learning_rate": 9.926739858228457e-06,
+      "loss": 0.9087,
+      "step": 3409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9339828099795555,
+      "learning_rate": 9.926672642275393e-06,
+      "loss": 0.425,
+      "step": 3410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3819439983912993,
+      "learning_rate": 9.926605395729027e-06,
+      "loss": 0.2059,
+      "step": 3411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.019640006525558,
+      "learning_rate": 9.926538118589777e-06,
+      "loss": 0.2318,
+      "step": 3412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9895373949935518,
+      "learning_rate": 9.926470810858063e-06,
+      "loss": 0.2378,
+      "step": 3413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.127588371644675,
+      "learning_rate": 9.926403472534301e-06,
+      "loss": 0.3594,
+      "step": 3414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4910429964259053,
+      "learning_rate": 9.926336103618908e-06,
+      "loss": 0.2022,
+      "step": 3415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.407814167722685,
+      "learning_rate": 9.926268704112305e-06,
+      "loss": 0.5126,
+      "step": 3416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5828936828098743,
+      "learning_rate": 9.926201274014909e-06,
+      "loss": 0.199,
+      "step": 3417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8526639277918502,
+      "learning_rate": 9.92613381332714e-06,
+      "loss": 0.4357,
+      "step": 3418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8336785440298073,
+      "learning_rate": 9.926066322049414e-06,
+      "loss": 0.4991,
+      "step": 3419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8259782372284499,
+      "learning_rate": 9.925998800182155e-06,
+      "loss": 0.2604,
+      "step": 3420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1198870807672145,
+      "learning_rate": 9.925931247725777e-06,
+      "loss": 0.5783,
+      "step": 3421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4627092093623697,
+      "learning_rate": 9.925863664680704e-06,
+      "loss": 0.3693,
+      "step": 3422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6347365359654507,
+      "learning_rate": 9.925796051047352e-06,
+      "loss": 0.4429,
+      "step": 3423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1451702138804842,
+      "learning_rate": 9.925728406826142e-06,
+      "loss": 0.6069,
+      "step": 3424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.092711575475384,
+      "learning_rate": 9.925660732017497e-06,
+      "loss": 0.2329,
+      "step": 3425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5396337465627963,
+      "learning_rate": 9.925593026621833e-06,
+      "loss": 0.2659,
+      "step": 3426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7063121284799405,
+      "learning_rate": 9.925525290639573e-06,
+      "loss": 0.3422,
+      "step": 3427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0919686320532178,
+      "learning_rate": 9.925457524071135e-06,
+      "loss": 0.1734,
+      "step": 3428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.963237887357893,
+      "learning_rate": 9.925389726916943e-06,
+      "loss": 0.3125,
+      "step": 3429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9848093623759255,
+      "learning_rate": 9.925321899177416e-06,
+      "loss": 0.2258,
+      "step": 3430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.365077611411926,
+      "learning_rate": 9.925254040852976e-06,
+      "loss": 0.3224,
+      "step": 3431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6832100603603033,
+      "learning_rate": 9.925186151944046e-06,
+      "loss": 0.27,
+      "step": 3432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.58810022045176,
+      "learning_rate": 9.925118232451043e-06,
+      "loss": 0.3857,
+      "step": 3433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5743916135019675,
+      "learning_rate": 9.925050282374394e-06,
+      "loss": 0.2836,
+      "step": 3434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7135151504264101,
+      "learning_rate": 9.924982301714515e-06,
+      "loss": 0.2095,
+      "step": 3435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.790905702601875,
+      "learning_rate": 9.924914290471835e-06,
+      "loss": 0.43,
+      "step": 3436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5933500026381406,
+      "learning_rate": 9.924846248646772e-06,
+      "loss": 0.3215,
+      "step": 3437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.075742183657952,
+      "learning_rate": 9.924778176239748e-06,
+      "loss": 0.4572,
+      "step": 3438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7606030079636408,
+      "learning_rate": 9.924710073251188e-06,
+      "loss": 0.3728,
+      "step": 3439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.181331096879544,
+      "learning_rate": 9.924641939681515e-06,
+      "loss": 0.2715,
+      "step": 3440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.953614978365267,
+      "learning_rate": 9.92457377553115e-06,
+      "loss": 0.5526,
+      "step": 3441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2704634795424052,
+      "learning_rate": 9.924505580800517e-06,
+      "loss": 0.7963,
+      "step": 3442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4699238666903112,
+      "learning_rate": 9.92443735549004e-06,
+      "loss": 0.3189,
+      "step": 3443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.454202200159187,
+      "learning_rate": 9.924369099600144e-06,
+      "loss": 0.2126,
+      "step": 3444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.255686412307161,
+      "learning_rate": 9.92430081313125e-06,
+      "loss": 0.3892,
+      "step": 3445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.545654629478485,
+      "learning_rate": 9.924232496083785e-06,
+      "loss": 0.2024,
+      "step": 3446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4971953499691126,
+      "learning_rate": 9.924164148458171e-06,
+      "loss": 0.1982,
+      "step": 3447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5349558063619841,
+      "learning_rate": 9.924095770254833e-06,
+      "loss": 0.2733,
+      "step": 3448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7949447547359267,
+      "learning_rate": 9.924027361474196e-06,
+      "loss": 0.3313,
+      "step": 3449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9473117738724788,
+      "learning_rate": 9.923958922116684e-06,
+      "loss": 0.4065,
+      "step": 3450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.866890073574272,
+      "learning_rate": 9.923890452182722e-06,
+      "loss": 0.2893,
+      "step": 3451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.140420221711065,
+      "learning_rate": 9.923821951672737e-06,
+      "loss": 0.731,
+      "step": 3452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1187187337396973,
+      "learning_rate": 9.923753420587154e-06,
+      "loss": 0.4891,
+      "step": 3453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5900833174330637,
+      "learning_rate": 9.923684858926395e-06,
+      "loss": 0.2602,
+      "step": 3454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6126222580760086,
+      "learning_rate": 9.92361626669089e-06,
+      "loss": 0.1987,
+      "step": 3455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0121744102790444,
+      "learning_rate": 9.923547643881063e-06,
+      "loss": 0.5545,
+      "step": 3456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1383023643914623,
+      "learning_rate": 9.923478990497339e-06,
+      "loss": 0.1858,
+      "step": 3457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5364651135656036,
+      "learning_rate": 9.923410306540147e-06,
+      "loss": 0.195,
+      "step": 3458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4021067113495547,
+      "learning_rate": 9.923341592009913e-06,
+      "loss": 0.2752,
+      "step": 3459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5770929757868672,
+      "learning_rate": 9.923272846907063e-06,
+      "loss": 0.35,
+      "step": 3460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4291047031440545,
+      "learning_rate": 9.923204071232023e-06,
+      "loss": 0.1903,
+      "step": 3461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7841615263584083,
+      "learning_rate": 9.92313526498522e-06,
+      "loss": 0.239,
+      "step": 3462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1765316366073666,
+      "learning_rate": 9.923066428167085e-06,
+      "loss": 0.3744,
+      "step": 3463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.096969861631757,
+      "learning_rate": 9.92299756077804e-06,
+      "loss": 0.1953,
+      "step": 3464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.213234052414049,
+      "learning_rate": 9.922928662818515e-06,
+      "loss": 0.4194,
+      "step": 3465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.386199233296775,
+      "learning_rate": 9.922859734288941e-06,
+      "loss": 0.3351,
+      "step": 3466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.660148726053579,
+      "learning_rate": 9.92279077518974e-06,
+      "loss": 0.2572,
+      "step": 3467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.734101823823886,
+      "learning_rate": 9.922721785521345e-06,
+      "loss": 0.394,
+      "step": 3468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4967878489518598,
+      "learning_rate": 9.922652765284184e-06,
+      "loss": 0.3175,
+      "step": 3469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7808033881033483,
+      "learning_rate": 9.922583714478682e-06,
+      "loss": 0.2447,
+      "step": 3470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0202156397707816,
+      "learning_rate": 9.922514633105271e-06,
+      "loss": 0.4114,
+      "step": 3471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.212065531250885,
+      "learning_rate": 9.922445521164381e-06,
+      "loss": 0.187,
+      "step": 3472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.207061051699201,
+      "learning_rate": 9.922376378656438e-06,
+      "loss": 0.2874,
+      "step": 3473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.378020112121948,
+      "learning_rate": 9.922307205581871e-06,
+      "loss": 0.2236,
+      "step": 3474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4441357796969245,
+      "learning_rate": 9.922238001941113e-06,
+      "loss": 0.2265,
+      "step": 3475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6936718169604332,
+      "learning_rate": 9.922168767734592e-06,
+      "loss": 0.3832,
+      "step": 3476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.541208262423697,
+      "learning_rate": 9.922099502962738e-06,
+      "loss": 0.3392,
+      "step": 3477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8180100487516835,
+      "learning_rate": 9.92203020762598e-06,
+      "loss": 0.4481,
+      "step": 3478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1259443703621013,
+      "learning_rate": 9.92196088172475e-06,
+      "loss": 0.1791,
+      "step": 3479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7003120741609652,
+      "learning_rate": 9.921891525259477e-06,
+      "loss": 0.1905,
+      "step": 3480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.149609993451169,
+      "learning_rate": 9.921822138230594e-06,
+      "loss": 0.4551,
+      "step": 3481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8702388952904874,
+      "learning_rate": 9.921752720638529e-06,
+      "loss": 0.4123,
+      "step": 3482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7357013416431362,
+      "learning_rate": 9.921683272483715e-06,
+      "loss": 0.4669,
+      "step": 3483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4473101754481008,
+      "learning_rate": 9.921613793766583e-06,
+      "loss": 0.3685,
+      "step": 3484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1464554963382074,
+      "learning_rate": 9.921544284487564e-06,
+      "loss": 0.1797,
+      "step": 3485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5194905123606641,
+      "learning_rate": 9.921474744647088e-06,
+      "loss": 0.2077,
+      "step": 3486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.044353519071442,
+      "learning_rate": 9.92140517424559e-06,
+      "loss": 0.5871,
+      "step": 3487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.163015346365519,
+      "learning_rate": 9.921335573283501e-06,
+      "loss": 0.5715,
+      "step": 3488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7228002509656113,
+      "learning_rate": 9.921265941761253e-06,
+      "loss": 0.1986,
+      "step": 3489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.098953005437305,
+      "learning_rate": 9.921196279679278e-06,
+      "loss": 0.1833,
+      "step": 3490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6739291513372458,
+      "learning_rate": 9.921126587038009e-06,
+      "loss": 0.2943,
+      "step": 3491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7071529929877123,
+      "learning_rate": 9.921056863837876e-06,
+      "loss": 0.4043,
+      "step": 3492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.661215462217969,
+      "learning_rate": 9.920987110079316e-06,
+      "loss": 0.2304,
+      "step": 3493
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.161751422007898,
+      "learning_rate": 9.920917325762763e-06,
+      "loss": 0.4311,
+      "step": 3494
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.350080926369159,
+      "learning_rate": 9.920847510888644e-06,
+      "loss": 0.341,
+      "step": 3495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.821405047814119,
+      "learning_rate": 9.9207776654574e-06,
+      "loss": 0.4,
+      "step": 3496
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8690660914246418,
+      "learning_rate": 9.92070778946946e-06,
+      "loss": 0.5041,
+      "step": 3497
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6034004039930316,
+      "learning_rate": 9.92063788292526e-06,
+      "loss": 0.3702,
+      "step": 3498
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.652017186858251,
+      "learning_rate": 9.920567945825232e-06,
+      "loss": 0.2143,
+      "step": 3499
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.955771075240766,
+      "learning_rate": 9.920497978169812e-06,
+      "loss": 0.2173,
+      "step": 3500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0110187127313526,
+      "learning_rate": 9.920427979959433e-06,
+      "loss": 0.4966,
+      "step": 3501
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.907189798074007,
+      "learning_rate": 9.920357951194532e-06,
+      "loss": 0.4906,
+      "step": 3502
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7082130915108746,
+      "learning_rate": 9.920287891875542e-06,
+      "loss": 0.3322,
+      "step": 3503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.397200156883805,
+      "learning_rate": 9.920217802002899e-06,
+      "loss": 0.2327,
+      "step": 3504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8808911440692588,
+      "learning_rate": 9.920147681577037e-06,
+      "loss": 0.2507,
+      "step": 3505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8071235547402649,
+      "learning_rate": 9.920077530598393e-06,
+      "loss": 0.2463,
+      "step": 3506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9276911664022,
+      "learning_rate": 9.920007349067403e-06,
+      "loss": 0.2814,
+      "step": 3507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6552126172800323,
+      "learning_rate": 9.9199371369845e-06,
+      "loss": 0.3854,
+      "step": 3508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.816904172023424,
+      "learning_rate": 9.919866894350122e-06,
+      "loss": 0.4322,
+      "step": 3509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9474175125160742,
+      "learning_rate": 9.919796621164705e-06,
+      "loss": 0.3213,
+      "step": 3510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.841753068162287,
+      "learning_rate": 9.919726317428686e-06,
+      "loss": 0.4379,
+      "step": 3511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.529639910813939,
+      "learning_rate": 9.9196559831425e-06,
+      "loss": 0.2762,
+      "step": 3512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8833654628463325,
+      "learning_rate": 9.919585618306584e-06,
+      "loss": 0.2911,
+      "step": 3513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9138299177900844,
+      "learning_rate": 9.919515222921377e-06,
+      "loss": 0.3319,
+      "step": 3514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4732860210218865,
+      "learning_rate": 9.919444796987314e-06,
+      "loss": 0.2931,
+      "step": 3515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.428944535365628,
+      "learning_rate": 9.919374340504833e-06,
+      "loss": 0.2993,
+      "step": 3516
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7035489560833152,
+      "learning_rate": 9.919303853474372e-06,
+      "loss": 0.4925,
+      "step": 3517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6488466252290204,
+      "learning_rate": 9.919233335896369e-06,
+      "loss": 0.3184,
+      "step": 3518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.087438507659138,
+      "learning_rate": 9.919162787771259e-06,
+      "loss": 0.3483,
+      "step": 3519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6934387909989366,
+      "learning_rate": 9.919092209099483e-06,
+      "loss": 0.2274,
+      "step": 3520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.473982336362879,
+      "learning_rate": 9.919021599881478e-06,
+      "loss": 0.5351,
+      "step": 3521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.02404728100279,
+      "learning_rate": 9.918950960117684e-06,
+      "loss": 0.6738,
+      "step": 3522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9817994065967266,
+      "learning_rate": 9.91888028980854e-06,
+      "loss": 0.4417,
+      "step": 3523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9107921054936312,
+      "learning_rate": 9.918809588954482e-06,
+      "loss": 0.23,
+      "step": 3524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8221594471882718,
+      "learning_rate": 9.918738857555949e-06,
+      "loss": 0.235,
+      "step": 3525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3155897593445585,
+      "learning_rate": 9.918668095613382e-06,
+      "loss": 0.2461,
+      "step": 3526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6735475053435673,
+      "learning_rate": 9.918597303127222e-06,
+      "loss": 0.3125,
+      "step": 3527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.804753364252838,
+      "learning_rate": 9.918526480097905e-06,
+      "loss": 0.4916,
+      "step": 3528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5003187095623147,
+      "learning_rate": 9.918455626525873e-06,
+      "loss": 0.312,
+      "step": 3529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0086956364070936,
+      "learning_rate": 9.918384742411567e-06,
+      "loss": 0.4671,
+      "step": 3530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0005044179968063,
+      "learning_rate": 9.918313827755423e-06,
+      "loss": 0.2234,
+      "step": 3531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4010488141833597,
+      "learning_rate": 9.918242882557884e-06,
+      "loss": 0.2384,
+      "step": 3532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.208146048347533,
+      "learning_rate": 9.918171906819394e-06,
+      "loss": 0.3569,
+      "step": 3533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8075635212701837,
+      "learning_rate": 9.918100900540387e-06,
+      "loss": 0.4086,
+      "step": 3534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1475783693715274,
+      "learning_rate": 9.918029863721309e-06,
+      "loss": 0.5257,
+      "step": 3535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 39.71021306864768,
+      "learning_rate": 9.917958796362598e-06,
+      "loss": 0.2663,
+      "step": 3536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.284162856452192,
+      "learning_rate": 9.917887698464698e-06,
+      "loss": 0.5554,
+      "step": 3537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7476087394057742,
+      "learning_rate": 9.91781657002805e-06,
+      "loss": 0.3024,
+      "step": 3538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3451904956412277,
+      "learning_rate": 9.917745411053092e-06,
+      "loss": 0.5874,
+      "step": 3539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7785776487344012,
+      "learning_rate": 9.91767422154027e-06,
+      "loss": 0.3265,
+      "step": 3540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6434656392191143,
+      "learning_rate": 9.917603001490025e-06,
+      "loss": 0.3516,
+      "step": 3541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.774398830523734,
+      "learning_rate": 9.9175317509028e-06,
+      "loss": 0.4031,
+      "step": 3542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.18198121980156,
+      "learning_rate": 9.917460469779035e-06,
+      "loss": 0.2128,
+      "step": 3543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8511419341738353,
+      "learning_rate": 9.917389158119175e-06,
+      "loss": 0.3164,
+      "step": 3544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.013658203463389,
+      "learning_rate": 9.917317815923662e-06,
+      "loss": 0.4359,
+      "step": 3545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0325136047239376,
+      "learning_rate": 9.917246443192939e-06,
+      "loss": 0.1781,
+      "step": 3546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7627556164430496,
+      "learning_rate": 9.917175039927448e-06,
+      "loss": 0.2318,
+      "step": 3547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.832105916703882,
+      "learning_rate": 9.917103606127634e-06,
+      "loss": 0.1998,
+      "step": 3548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6253012689672928,
+      "learning_rate": 9.917032141793941e-06,
+      "loss": 0.3342,
+      "step": 3549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1145427802539705,
+      "learning_rate": 9.916960646926811e-06,
+      "loss": 0.5297,
+      "step": 3550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6307584285902776,
+      "learning_rate": 9.91688912152669e-06,
+      "loss": 0.2076,
+      "step": 3551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6722301391172656,
+      "learning_rate": 9.91681756559402e-06,
+      "loss": 0.3572,
+      "step": 3552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3838723061485205,
+      "learning_rate": 9.916745979129247e-06,
+      "loss": 0.595,
+      "step": 3553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8662034518153132,
+      "learning_rate": 9.916674362132814e-06,
+      "loss": 0.4296,
+      "step": 3554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5549440363951246,
+      "learning_rate": 9.916602714605168e-06,
+      "loss": 0.2739,
+      "step": 3555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3282946569407605,
+      "learning_rate": 9.916531036546753e-06,
+      "loss": 0.1807,
+      "step": 3556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.209260570825989,
+      "learning_rate": 9.916459327958013e-06,
+      "loss": 0.1762,
+      "step": 3557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1656098222131868,
+      "learning_rate": 9.916387588839393e-06,
+      "loss": 0.4629,
+      "step": 3558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5524880939678247,
+      "learning_rate": 9.91631581919134e-06,
+      "loss": 0.3442,
+      "step": 3559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1371407462916103,
+      "learning_rate": 9.9162440190143e-06,
+      "loss": 0.2124,
+      "step": 3560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9809867888621011,
+      "learning_rate": 9.916172188308718e-06,
+      "loss": 0.3174,
+      "step": 3561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.758818448767476,
+      "learning_rate": 9.916100327075038e-06,
+      "loss": 0.3351,
+      "step": 3562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6007882456442775,
+      "learning_rate": 9.91602843531371e-06,
+      "loss": 0.3196,
+      "step": 3563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.476215426977764,
+      "learning_rate": 9.915956513025179e-06,
+      "loss": 0.3088,
+      "step": 3564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1208477474752065,
+      "learning_rate": 9.915884560209889e-06,
+      "loss": 0.2372,
+      "step": 3565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2711425812743213,
+      "learning_rate": 9.91581257686829e-06,
+      "loss": 0.5152,
+      "step": 3566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4393222367515253,
+      "learning_rate": 9.915740563000828e-06,
+      "loss": 0.4264,
+      "step": 3567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8537542432735183,
+      "learning_rate": 9.915668518607953e-06,
+      "loss": 0.437,
+      "step": 3568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.082639344179904,
+      "learning_rate": 9.915596443690106e-06,
+      "loss": 0.4256,
+      "step": 3569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7524654542347164,
+      "learning_rate": 9.915524338247741e-06,
+      "loss": 0.6343,
+      "step": 3570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 132.16311507886937,
+      "learning_rate": 9.9154522022813e-06,
+      "loss": 0.3209,
+      "step": 3571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5758471627750166,
+      "learning_rate": 9.915380035791239e-06,
+      "loss": 0.679,
+      "step": 3572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1025197532748736,
+      "learning_rate": 9.915307838777997e-06,
+      "loss": 0.5068,
+      "step": 3573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.028827348569629,
+      "learning_rate": 9.915235611242026e-06,
+      "loss": 0.4323,
+      "step": 3574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.224066489729171,
+      "learning_rate": 9.915163353183777e-06,
+      "loss": 0.5139,
+      "step": 3575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7515095347989729,
+      "learning_rate": 9.915091064603694e-06,
+      "loss": 0.4046,
+      "step": 3576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.676545909006698,
+      "learning_rate": 9.915018745502232e-06,
+      "loss": 0.2141,
+      "step": 3577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7694051499730437,
+      "learning_rate": 9.914946395879835e-06,
+      "loss": 0.3348,
+      "step": 3578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.726151374268712,
+      "learning_rate": 9.914874015736953e-06,
+      "loss": 0.316,
+      "step": 3579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8456965370666156,
+      "learning_rate": 9.914801605074037e-06,
+      "loss": 0.4735,
+      "step": 3580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2868257660187763,
+      "learning_rate": 9.914729163891537e-06,
+      "loss": 0.1984,
+      "step": 3581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4140946469434965,
+      "learning_rate": 9.9146566921899e-06,
+      "loss": 0.1983,
+      "step": 3582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0740907705017513,
+      "learning_rate": 9.914584189969577e-06,
+      "loss": 0.647,
+      "step": 3583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2837219626416376,
+      "learning_rate": 9.914511657231022e-06,
+      "loss": 0.1163,
+      "step": 3584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6100688475643665,
+      "learning_rate": 9.914439093974681e-06,
+      "loss": 0.3138,
+      "step": 3585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0256811170917337,
+      "learning_rate": 9.914366500201007e-06,
+      "loss": 0.6154,
+      "step": 3586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.588772962331126,
+      "learning_rate": 9.914293875910449e-06,
+      "loss": 0.4589,
+      "step": 3587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3583127083180337,
+      "learning_rate": 9.91422122110346e-06,
+      "loss": 0.1959,
+      "step": 3588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5446722887599476,
+      "learning_rate": 9.914148535780488e-06,
+      "loss": 0.2601,
+      "step": 3589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0073334979726916,
+      "learning_rate": 9.914075819941986e-06,
+      "loss": 0.4913,
+      "step": 3590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0381985109145324,
+      "learning_rate": 9.914003073588407e-06,
+      "loss": 0.4476,
+      "step": 3591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4435370417205313,
+      "learning_rate": 9.913930296720202e-06,
+      "loss": 0.3265,
+      "step": 3592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5819300112885968,
+      "learning_rate": 9.913857489337824e-06,
+      "loss": 0.3041,
+      "step": 3593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1814849784333017,
+      "learning_rate": 9.913784651441721e-06,
+      "loss": 0.5784,
+      "step": 3594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2953989464842415,
+      "learning_rate": 9.913711783032349e-06,
+      "loss": 0.1968,
+      "step": 3595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9552215272061633,
+      "learning_rate": 9.91363888411016e-06,
+      "loss": 0.1954,
+      "step": 3596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.767348229226625,
+      "learning_rate": 9.913565954675606e-06,
+      "loss": 0.3177,
+      "step": 3597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4804361454284758,
+      "learning_rate": 9.91349299472914e-06,
+      "loss": 0.1999,
+      "step": 3598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4632147976052619,
+      "learning_rate": 9.913420004271213e-06,
+      "loss": 0.2601,
+      "step": 3599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2120136171214653,
+      "learning_rate": 9.913346983302283e-06,
+      "loss": 0.2239,
+      "step": 3600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8358220084368189,
+      "learning_rate": 9.9132739318228e-06,
+      "loss": 0.379,
+      "step": 3601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.501765241871736,
+      "learning_rate": 9.913200849833218e-06,
+      "loss": 0.7279,
+      "step": 3602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9500384088548266,
+      "learning_rate": 9.913127737333992e-06,
+      "loss": 0.3503,
+      "step": 3603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0206662167110285,
+      "learning_rate": 9.913054594325574e-06,
+      "loss": 0.3865,
+      "step": 3604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.469607823033269,
+      "learning_rate": 9.91298142080842e-06,
+      "loss": 0.3111,
+      "step": 3605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8446792123916316,
+      "learning_rate": 9.912908216782983e-06,
+      "loss": 0.2255,
+      "step": 3606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0952898170785224,
+      "learning_rate": 9.91283498224972e-06,
+      "loss": 0.5321,
+      "step": 3607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.624294737762664,
+      "learning_rate": 9.912761717209084e-06,
+      "loss": 0.3038,
+      "step": 3608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.217037776007468,
+      "learning_rate": 9.912688421661529e-06,
+      "loss": 0.5959,
+      "step": 3609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.008465386157474,
+      "learning_rate": 9.912615095607512e-06,
+      "loss": 0.5067,
+      "step": 3610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0647103511183187,
+      "learning_rate": 9.912541739047487e-06,
+      "loss": 0.2827,
+      "step": 3611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8518269024157252,
+      "learning_rate": 9.912468351981912e-06,
+      "loss": 0.5035,
+      "step": 3612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4210511613811148,
+      "learning_rate": 9.912394934411239e-06,
+      "loss": 0.3727,
+      "step": 3613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5286744703722235,
+      "learning_rate": 9.912321486335926e-06,
+      "loss": 0.221,
+      "step": 3614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6135824767415512,
+      "learning_rate": 9.912248007756429e-06,
+      "loss": 0.3511,
+      "step": 3615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1720720951910308,
+      "learning_rate": 9.912174498673204e-06,
+      "loss": 0.2577,
+      "step": 3616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9359750137246854,
+      "learning_rate": 9.912100959086708e-06,
+      "loss": 0.71,
+      "step": 3617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5239632292365954,
+      "learning_rate": 9.912027388997397e-06,
+      "loss": 0.215,
+      "step": 3618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6525126386684312,
+      "learning_rate": 9.911953788405729e-06,
+      "loss": 0.429,
+      "step": 3619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6080692787671946,
+      "learning_rate": 9.91188015731216e-06,
+      "loss": 0.4127,
+      "step": 3620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9934207913526292,
+      "learning_rate": 9.911806495717144e-06,
+      "loss": 0.5188,
+      "step": 3621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8186847383385167,
+      "learning_rate": 9.911732803621146e-06,
+      "loss": 0.5848,
+      "step": 3622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.468220672115196,
+      "learning_rate": 9.911659081024617e-06,
+      "loss": 0.6318,
+      "step": 3623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5545506632792137,
+      "learning_rate": 9.911585327928019e-06,
+      "loss": 0.2822,
+      "step": 3624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5343299648858495,
+      "learning_rate": 9.911511544331807e-06,
+      "loss": 0.2833,
+      "step": 3625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9725363364456634,
+      "learning_rate": 9.911437730236439e-06,
+      "loss": 0.5316,
+      "step": 3626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5492941214289955,
+      "learning_rate": 9.911363885642378e-06,
+      "loss": 0.3375,
+      "step": 3627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.076111542335581,
+      "learning_rate": 9.911290010550077e-06,
+      "loss": 0.5268,
+      "step": 3628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6411048550717182,
+      "learning_rate": 9.911216104959998e-06,
+      "loss": 0.3598,
+      "step": 3629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6591362692106713,
+      "learning_rate": 9.911142168872597e-06,
+      "loss": 0.2784,
+      "step": 3630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7170574909376037,
+      "learning_rate": 9.911068202288337e-06,
+      "loss": 0.2749,
+      "step": 3631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3158420335591445,
+      "learning_rate": 9.910994205207674e-06,
+      "loss": 0.2197,
+      "step": 3632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3353682931094921,
+      "learning_rate": 9.910920177631071e-06,
+      "loss": 0.1664,
+      "step": 3633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0266757812838803,
+      "learning_rate": 9.910846119558983e-06,
+      "loss": 0.4768,
+      "step": 3634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4422806661630836,
+      "learning_rate": 9.910772030991875e-06,
+      "loss": 0.5178,
+      "step": 3635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6336158095101354,
+      "learning_rate": 9.910697911930203e-06,
+      "loss": 0.3067,
+      "step": 3636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.578748978582763,
+      "learning_rate": 9.91062376237443e-06,
+      "loss": 0.3284,
+      "step": 3637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7826152200062104,
+      "learning_rate": 9.910549582325012e-06,
+      "loss": 0.2955,
+      "step": 3638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0974982360200434,
+      "learning_rate": 9.910475371782418e-06,
+      "loss": 0.4568,
+      "step": 3639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6167275977286626,
+      "learning_rate": 9.9104011307471e-06,
+      "loss": 0.236,
+      "step": 3640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6830665975652046,
+      "learning_rate": 9.910326859219524e-06,
+      "loss": 0.3755,
+      "step": 3641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7732537212179504,
+      "learning_rate": 9.91025255720015e-06,
+      "loss": 0.473,
+      "step": 3642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.135939164708681,
+      "learning_rate": 9.910178224689437e-06,
+      "loss": 0.5378,
+      "step": 3643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6886236701454351,
+      "learning_rate": 9.910103861687852e-06,
+      "loss": 0.3686,
+      "step": 3644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.597785862839218,
+      "learning_rate": 9.910029468195852e-06,
+      "loss": 0.3143,
+      "step": 3645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8813541391450581,
+      "learning_rate": 9.909955044213902e-06,
+      "loss": 0.493,
+      "step": 3646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.064147424936094,
+      "learning_rate": 9.909880589742463e-06,
+      "loss": 0.5676,
+      "step": 3647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9117218458729734,
+      "learning_rate": 9.909806104781995e-06,
+      "loss": 0.3869,
+      "step": 3648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7004139610072082,
+      "learning_rate": 9.909731589332964e-06,
+      "loss": 0.3453,
+      "step": 3649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1633645595105304,
+      "learning_rate": 9.90965704339583e-06,
+      "loss": 0.433,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.549670787524235,
+      "learning_rate": 9.90958246697106e-06,
+      "loss": 0.2823,
+      "step": 3651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.682871173368848,
+      "learning_rate": 9.909507860059113e-06,
+      "loss": 0.2716,
+      "step": 3652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5401462875609802,
+      "learning_rate": 9.909433222660453e-06,
+      "loss": 0.2842,
+      "step": 3653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5975842308096766,
+      "learning_rate": 9.909358554775544e-06,
+      "loss": 0.4434,
+      "step": 3654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.613973774876318,
+      "learning_rate": 9.90928385640485e-06,
+      "loss": 0.1926,
+      "step": 3655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.777205195633145,
+      "learning_rate": 9.909209127548836e-06,
+      "loss": 0.4118,
+      "step": 3656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8580929028417004,
+      "learning_rate": 9.909134368207964e-06,
+      "loss": 0.2453,
+      "step": 3657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.510582775276544,
+      "learning_rate": 9.9090595783827e-06,
+      "loss": 0.7003,
+      "step": 3658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8898878157841108,
+      "learning_rate": 9.908984758073506e-06,
+      "loss": 0.305,
+      "step": 3659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7585936548443992,
+      "learning_rate": 9.908909907280848e-06,
+      "loss": 0.2573,
+      "step": 3660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9654912507382358,
+      "learning_rate": 9.90883502600519e-06,
+      "loss": 0.463,
+      "step": 3661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6463375682249528,
+      "learning_rate": 9.908760114247e-06,
+      "loss": 0.3129,
+      "step": 3662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.257619221902355,
+      "learning_rate": 9.90868517200674e-06,
+      "loss": 0.185,
+      "step": 3663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4937802788562164,
+      "learning_rate": 9.908610199284876e-06,
+      "loss": 0.498,
+      "step": 3664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.067694495000262,
+      "learning_rate": 9.908535196081874e-06,
+      "loss": 0.4959,
+      "step": 3665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9133977747831665,
+      "learning_rate": 9.9084601623982e-06,
+      "loss": 0.2574,
+      "step": 3666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.25144823804931,
+      "learning_rate": 9.90838509823432e-06,
+      "loss": 0.705,
+      "step": 3667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5205109950861029,
+      "learning_rate": 9.9083100035907e-06,
+      "loss": 0.2344,
+      "step": 3668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.759088417685001,
+      "learning_rate": 9.908234878467804e-06,
+      "loss": 0.3369,
+      "step": 3669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8862299355020542,
+      "learning_rate": 9.908159722866102e-06,
+      "loss": 0.264,
+      "step": 3670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3522929274046211,
+      "learning_rate": 9.90808453678606e-06,
+      "loss": 0.2071,
+      "step": 3671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9717592579475665,
+      "learning_rate": 9.908009320228143e-06,
+      "loss": 0.3659,
+      "step": 3672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9602391188594033,
+      "learning_rate": 9.90793407319282e-06,
+      "loss": 0.1125,
+      "step": 3673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7449596827065859,
+      "learning_rate": 9.907858795680556e-06,
+      "loss": 0.3637,
+      "step": 3674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.132593437095413,
+      "learning_rate": 9.90778348769182e-06,
+      "loss": 0.6171,
+      "step": 3675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4953187772963892,
+      "learning_rate": 9.907708149227082e-06,
+      "loss": 0.323,
+      "step": 3676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.020600816604572,
+      "learning_rate": 9.907632780286806e-06,
+      "loss": 0.14,
+      "step": 3677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7916684557292195,
+      "learning_rate": 9.90755738087146e-06,
+      "loss": 0.3318,
+      "step": 3678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4232193072989927,
+      "learning_rate": 9.907481950981516e-06,
+      "loss": 0.2911,
+      "step": 3679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6412191359689599,
+      "learning_rate": 9.90740649061744e-06,
+      "loss": 0.1942,
+      "step": 3680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7961915001604485,
+      "learning_rate": 9.9073309997797e-06,
+      "loss": 0.3506,
+      "step": 3681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.391932186203637,
+      "learning_rate": 9.907255478468764e-06,
+      "loss": 0.446,
+      "step": 3682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8550227020589054,
+      "learning_rate": 9.907179926685101e-06,
+      "loss": 0.3735,
+      "step": 3683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.502192449662728,
+      "learning_rate": 9.907104344429187e-06,
+      "loss": 0.3066,
+      "step": 3684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.287912079689496,
+      "learning_rate": 9.907028731701482e-06,
+      "loss": 0.3274,
+      "step": 3685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4492869425356105,
+      "learning_rate": 9.90695308850246e-06,
+      "loss": 0.163,
+      "step": 3686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7609703198626454,
+      "learning_rate": 9.906877414832589e-06,
+      "loss": 0.1852,
+      "step": 3687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.875412916509835,
+      "learning_rate": 9.906801710692341e-06,
+      "loss": 0.257,
+      "step": 3688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8361093171865157,
+      "learning_rate": 9.906725976082187e-06,
+      "loss": 0.2898,
+      "step": 3689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7996733312749544,
+      "learning_rate": 9.906650211002593e-06,
+      "loss": 0.3111,
+      "step": 3690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9495596769094334,
+      "learning_rate": 9.906574415454033e-06,
+      "loss": 0.447,
+      "step": 3691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7289551494007678,
+      "learning_rate": 9.906498589436977e-06,
+      "loss": 0.3201,
+      "step": 3692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4244590571327884,
+      "learning_rate": 9.906422732951894e-06,
+      "loss": 0.6831,
+      "step": 3693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.393225028885733,
+      "learning_rate": 9.906346845999257e-06,
+      "loss": 0.455,
+      "step": 3694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.316318671129272,
+      "learning_rate": 9.906270928579538e-06,
+      "loss": 0.4194,
+      "step": 3695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1689365740380127,
+      "learning_rate": 9.906194980693207e-06,
+      "loss": 0.1487,
+      "step": 3696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3993400479620448,
+      "learning_rate": 9.906119002340735e-06,
+      "loss": 0.3186,
+      "step": 3697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7829038487935178,
+      "learning_rate": 9.906042993522594e-06,
+      "loss": 0.2258,
+      "step": 3698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9325147411763788,
+      "learning_rate": 9.905966954239256e-06,
+      "loss": 0.4714,
+      "step": 3699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4226369257352418,
+      "learning_rate": 9.905890884491196e-06,
+      "loss": 0.2781,
+      "step": 3700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6416903578998079,
+      "learning_rate": 9.905814784278883e-06,
+      "loss": 0.1657,
+      "step": 3701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7854175877169975,
+      "learning_rate": 9.90573865360279e-06,
+      "loss": 0.2445,
+      "step": 3702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8567014113222324,
+      "learning_rate": 9.905662492463392e-06,
+      "loss": 0.6163,
+      "step": 3703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7752252591948352,
+      "learning_rate": 9.90558630086116e-06,
+      "loss": 0.2493,
+      "step": 3704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.765972444293952,
+      "learning_rate": 9.905510078796566e-06,
+      "loss": 0.4831,
+      "step": 3705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9440394519895396,
+      "learning_rate": 9.905433826270086e-06,
+      "loss": 0.2937,
+      "step": 3706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.850794967078812,
+      "learning_rate": 9.905357543282193e-06,
+      "loss": 0.3566,
+      "step": 3707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3157317926914558,
+      "learning_rate": 9.905281229833359e-06,
+      "loss": 0.2038,
+      "step": 3708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4679393704976023,
+      "learning_rate": 9.905204885924059e-06,
+      "loss": 0.7507,
+      "step": 3709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9013774109395576,
+      "learning_rate": 9.905128511554767e-06,
+      "loss": 0.336,
+      "step": 3710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8256400083495006,
+      "learning_rate": 9.905052106725958e-06,
+      "loss": 0.2006,
+      "step": 3711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.259674954058395,
+      "learning_rate": 9.904975671438104e-06,
+      "loss": 0.5881,
+      "step": 3712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1599063521798185,
+      "learning_rate": 9.904899205691683e-06,
+      "loss": 0.5587,
+      "step": 3713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1913489029940454,
+      "learning_rate": 9.904822709487167e-06,
+      "loss": 0.426,
+      "step": 3714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.858157093948999,
+      "learning_rate": 9.904746182825034e-06,
+      "loss": 0.4014,
+      "step": 3715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1494576892992194,
+      "learning_rate": 9.904669625705756e-06,
+      "loss": 0.5923,
+      "step": 3716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.886721609046302,
+      "learning_rate": 9.90459303812981e-06,
+      "loss": 0.1564,
+      "step": 3717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2260140228667367,
+      "learning_rate": 9.904516420097671e-06,
+      "loss": 0.512,
+      "step": 3718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5911284189915373,
+      "learning_rate": 9.904439771609816e-06,
+      "loss": 0.2488,
+      "step": 3719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5944560652593305,
+      "learning_rate": 9.90436309266672e-06,
+      "loss": 0.5921,
+      "step": 3720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.864562082500817,
+      "learning_rate": 9.904286383268859e-06,
+      "loss": 0.6219,
+      "step": 3721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5830010349851058,
+      "learning_rate": 9.904209643416708e-06,
+      "loss": 0.2308,
+      "step": 3722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3733997854603042,
+      "learning_rate": 9.904132873110749e-06,
+      "loss": 0.178,
+      "step": 3723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6529371940931854,
+      "learning_rate": 9.904056072351452e-06,
+      "loss": 0.238,
+      "step": 3724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.445591822261945,
+      "learning_rate": 9.903979241139299e-06,
+      "loss": 0.4573,
+      "step": 3725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3997401651484738,
+      "learning_rate": 9.903902379474764e-06,
+      "loss": 0.1768,
+      "step": 3726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6260355768633734,
+      "learning_rate": 9.903825487358323e-06,
+      "loss": 0.421,
+      "step": 3727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.267461998830323,
+      "learning_rate": 9.903748564790459e-06,
+      "loss": 0.5497,
+      "step": 3728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.39278225653219,
+      "learning_rate": 9.903671611771644e-06,
+      "loss": 0.1922,
+      "step": 3729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7875191224015698,
+      "learning_rate": 9.903594628302358e-06,
+      "loss": 0.2683,
+      "step": 3730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.550785064694413,
+      "learning_rate": 9.90351761438308e-06,
+      "loss": 0.3116,
+      "step": 3731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9749035715151113,
+      "learning_rate": 9.903440570014288e-06,
+      "loss": 0.4651,
+      "step": 3732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.677316269385167,
+      "learning_rate": 9.90336349519646e-06,
+      "loss": 0.4125,
+      "step": 3733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9371711024953269,
+      "learning_rate": 9.903286389930075e-06,
+      "loss": 0.3897,
+      "step": 3734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7232651294027224,
+      "learning_rate": 9.90320925421561e-06,
+      "loss": 0.1951,
+      "step": 3735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.491939463522158,
+      "learning_rate": 9.903132088053546e-06,
+      "loss": 0.1953,
+      "step": 3736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.983568793699857,
+      "learning_rate": 9.903054891444361e-06,
+      "loss": 0.5047,
+      "step": 3737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3043742460532903,
+      "learning_rate": 9.902977664388536e-06,
+      "loss": 0.2877,
+      "step": 3738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5804983365376886,
+      "learning_rate": 9.902900406886548e-06,
+      "loss": 0.2108,
+      "step": 3739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1525933178113643,
+      "learning_rate": 9.90282311893888e-06,
+      "loss": 0.6104,
+      "step": 3740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7808369149182612,
+      "learning_rate": 9.902745800546008e-06,
+      "loss": 0.3282,
+      "step": 3741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0476834687974934,
+      "learning_rate": 9.902668451708415e-06,
+      "loss": 0.6053,
+      "step": 3742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0882859582844358,
+      "learning_rate": 9.90259107242658e-06,
+      "loss": 0.2519,
+      "step": 3743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3309279652290038,
+      "learning_rate": 9.902513662700986e-06,
+      "loss": 0.5015,
+      "step": 3744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4392032225204716,
+      "learning_rate": 9.90243622253211e-06,
+      "loss": 0.2828,
+      "step": 3745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.62694991478941,
+      "learning_rate": 9.902358751920436e-06,
+      "loss": 0.3668,
+      "step": 3746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6545711709636404,
+      "learning_rate": 9.902281250866442e-06,
+      "loss": 0.2105,
+      "step": 3747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.606305257813538,
+      "learning_rate": 9.902203719370613e-06,
+      "loss": 0.4284,
+      "step": 3748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4694321803313504,
+      "learning_rate": 9.902126157433427e-06,
+      "loss": 0.1923,
+      "step": 3749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.010544547195821,
+      "learning_rate": 9.902048565055366e-06,
+      "loss": 0.5736,
+      "step": 3750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4120328359451455,
+      "learning_rate": 9.901970942236915e-06,
+      "loss": 0.2003,
+      "step": 3751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.499347118758082,
+      "learning_rate": 9.901893288978553e-06,
+      "loss": 0.3393,
+      "step": 3752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.602122339233501,
+      "learning_rate": 9.901815605280764e-06,
+      "loss": 0.3124,
+      "step": 3753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0066646085126068,
+      "learning_rate": 9.901737891144028e-06,
+      "loss": 0.5158,
+      "step": 3754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7474776891722146,
+      "learning_rate": 9.90166014656883e-06,
+      "loss": 0.4586,
+      "step": 3755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5326726530307426,
+      "learning_rate": 9.901582371555652e-06,
+      "loss": 0.1765,
+      "step": 3756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.846871457362771,
+      "learning_rate": 9.901504566104976e-06,
+      "loss": 0.3576,
+      "step": 3757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7321662198114798,
+      "learning_rate": 9.901426730217288e-06,
+      "loss": 0.3435,
+      "step": 3758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8303101350815592,
+      "learning_rate": 9.901348863893066e-06,
+      "loss": 0.4717,
+      "step": 3759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.07683913977079,
+      "learning_rate": 9.9012709671328e-06,
+      "loss": 0.1689,
+      "step": 3760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0508625341382865,
+      "learning_rate": 9.90119303993697e-06,
+      "loss": 0.3766,
+      "step": 3761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7922695631264391,
+      "learning_rate": 9.90111508230606e-06,
+      "loss": 0.1889,
+      "step": 3762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1163661954404955,
+      "learning_rate": 9.901037094240554e-06,
+      "loss": 0.5804,
+      "step": 3763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0458593469495128,
+      "learning_rate": 9.900959075740936e-06,
+      "loss": 0.3598,
+      "step": 3764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9269757734816289,
+      "learning_rate": 9.900881026807692e-06,
+      "loss": 0.5155,
+      "step": 3765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.210397381403282,
+      "learning_rate": 9.900802947441307e-06,
+      "loss": 0.2474,
+      "step": 3766
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3633334360272262,
+      "learning_rate": 9.900724837642265e-06,
+      "loss": 0.2392,
+      "step": 3767
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8652199909394553,
+      "learning_rate": 9.900646697411049e-06,
+      "loss": 0.3955,
+      "step": 3768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.988240792438062,
+      "learning_rate": 9.900568526748147e-06,
+      "loss": 0.3319,
+      "step": 3769
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3425639075922335,
+      "learning_rate": 9.900490325654045e-06,
+      "loss": 0.3103,
+      "step": 3770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8206750443230135,
+      "learning_rate": 9.900412094129225e-06,
+      "loss": 0.4606,
+      "step": 3771
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7720955926670061,
+      "learning_rate": 9.900333832174176e-06,
+      "loss": 0.4522,
+      "step": 3772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2616428237319806,
+      "learning_rate": 9.900255539789384e-06,
+      "loss": 0.6157,
+      "step": 3773
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5614709087838408,
+      "learning_rate": 9.900177216975332e-06,
+      "loss": 0.2888,
+      "step": 3774
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3436121398602419,
+      "learning_rate": 9.900098863732508e-06,
+      "loss": 0.3147,
+      "step": 3775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7750355708560814,
+      "learning_rate": 9.900020480061401e-06,
+      "loss": 0.2365,
+      "step": 3776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0631654633948586,
+      "learning_rate": 9.899942065962496e-06,
+      "loss": 0.6289,
+      "step": 3777
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3237033817566584,
+      "learning_rate": 9.89986362143628e-06,
+      "loss": 0.4469,
+      "step": 3778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4203426250851097,
+      "learning_rate": 9.899785146483235e-06,
+      "loss": 0.5307,
+      "step": 3779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8093643812418105,
+      "learning_rate": 9.899706641103858e-06,
+      "loss": 0.2558,
+      "step": 3780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8395407117339293,
+      "learning_rate": 9.89962810529863e-06,
+      "loss": 0.3491,
+      "step": 3781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8519481969452798,
+      "learning_rate": 9.89954953906804e-06,
+      "loss": 0.373,
+      "step": 3782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4078248742138728,
+      "learning_rate": 9.899470942412578e-06,
+      "loss": 0.2134,
+      "step": 3783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7574195804144301,
+      "learning_rate": 9.899392315332728e-06,
+      "loss": 0.2481,
+      "step": 3784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8880041178697802,
+      "learning_rate": 9.899313657828981e-06,
+      "loss": 0.2506,
+      "step": 3785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4567952367607984,
+      "learning_rate": 9.899234969901825e-06,
+      "loss": 0.144,
+      "step": 3786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5756056287241267,
+      "learning_rate": 9.89915625155175e-06,
+      "loss": 0.3881,
+      "step": 3787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5369186464804399,
+      "learning_rate": 9.899077502779242e-06,
+      "loss": 0.1937,
+      "step": 3788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1500460524771101,
+      "learning_rate": 9.898998723584793e-06,
+      "loss": 0.1897,
+      "step": 3789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4032661783266798,
+      "learning_rate": 9.898919913968887e-06,
+      "loss": 0.1684,
+      "step": 3790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5376764775773328,
+      "learning_rate": 9.89884107393202e-06,
+      "loss": 0.3708,
+      "step": 3791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.075696023045859,
+      "learning_rate": 9.898762203474678e-06,
+      "loss": 0.2059,
+      "step": 3792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6260109837643177,
+      "learning_rate": 9.898683302597351e-06,
+      "loss": 0.3043,
+      "step": 3793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0516438295930493,
+      "learning_rate": 9.89860437130053e-06,
+      "loss": 0.4782,
+      "step": 3794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3121359439441427,
+      "learning_rate": 9.898525409584703e-06,
+      "loss": 0.2587,
+      "step": 3795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9307824833180587,
+      "learning_rate": 9.898446417450363e-06,
+      "loss": 0.2938,
+      "step": 3796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6473586407407876,
+      "learning_rate": 9.898367394897998e-06,
+      "loss": 0.2353,
+      "step": 3797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5730352603075286,
+      "learning_rate": 9.898288341928099e-06,
+      "loss": 0.2245,
+      "step": 3798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0923100829447923,
+      "learning_rate": 9.89820925854116e-06,
+      "loss": 0.5301,
+      "step": 3799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8964957915570417,
+      "learning_rate": 9.89813014473767e-06,
+      "loss": 0.5989,
+      "step": 3800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5801111177788696,
+      "learning_rate": 9.89805100051812e-06,
+      "loss": 0.3473,
+      "step": 3801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8453140010068572,
+      "learning_rate": 9.897971825883e-06,
+      "loss": 0.3868,
+      "step": 3802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0201723088468473,
+      "learning_rate": 9.897892620832806e-06,
+      "loss": 0.3641,
+      "step": 3803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7504943294234447,
+      "learning_rate": 9.897813385368025e-06,
+      "loss": 0.505,
+      "step": 3804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2860795692887426,
+      "learning_rate": 9.897734119489152e-06,
+      "loss": 0.2331,
+      "step": 3805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.200004065608744,
+      "learning_rate": 9.897654823196679e-06,
+      "loss": 0.318,
+      "step": 3806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.393707843239982,
+      "learning_rate": 9.897575496491096e-06,
+      "loss": 0.6655,
+      "step": 3807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2600744591954567,
+      "learning_rate": 9.897496139372899e-06,
+      "loss": 0.7711,
+      "step": 3808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4093834489706407,
+      "learning_rate": 9.897416751842578e-06,
+      "loss": 0.4695,
+      "step": 3809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.03995357058553,
+      "learning_rate": 9.897337333900629e-06,
+      "loss": 0.4534,
+      "step": 3810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3049398768976546,
+      "learning_rate": 9.897257885547542e-06,
+      "loss": 0.4019,
+      "step": 3811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8990604532383644,
+      "learning_rate": 9.897178406783813e-06,
+      "loss": 0.3986,
+      "step": 3812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6948352792478423,
+      "learning_rate": 9.897098897609933e-06,
+      "loss": 0.4657,
+      "step": 3813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5826712173415505,
+      "learning_rate": 9.897019358026397e-06,
+      "loss": 0.32,
+      "step": 3814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2792254308439333,
+      "learning_rate": 9.8969397880337e-06,
+      "loss": 0.7594,
+      "step": 3815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8991195332680668,
+      "learning_rate": 9.896860187632334e-06,
+      "loss": 0.4059,
+      "step": 3816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.170015142515412,
+      "learning_rate": 9.896780556822793e-06,
+      "loss": 0.7009,
+      "step": 3817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5617015038091924,
+      "learning_rate": 9.896700895605575e-06,
+      "loss": 0.3308,
+      "step": 3818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9768655491885527,
+      "learning_rate": 9.896621203981172e-06,
+      "loss": 0.5539,
+      "step": 3819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7256332740702767,
+      "learning_rate": 9.89654148195008e-06,
+      "loss": 0.2285,
+      "step": 3820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5978435672363671,
+      "learning_rate": 9.896461729512792e-06,
+      "loss": 0.2294,
+      "step": 3821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4995753126655524,
+      "learning_rate": 9.896381946669806e-06,
+      "loss": 0.1913,
+      "step": 3822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8967857219638866,
+      "learning_rate": 9.896302133421615e-06,
+      "loss": 0.2822,
+      "step": 3823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3826598321754036,
+      "learning_rate": 9.896222289768715e-06,
+      "loss": 0.4761,
+      "step": 3824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7817720936363095,
+      "learning_rate": 9.896142415711604e-06,
+      "loss": 0.3645,
+      "step": 3825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4127934153908397,
+      "learning_rate": 9.896062511250775e-06,
+      "loss": 0.2782,
+      "step": 3826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8209351960313769,
+      "learning_rate": 9.895982576386727e-06,
+      "loss": 0.3904,
+      "step": 3827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5299763540235867,
+      "learning_rate": 9.895902611119955e-06,
+      "loss": 0.2956,
+      "step": 3828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.382823746037416,
+      "learning_rate": 9.895822615450954e-06,
+      "loss": 0.252,
+      "step": 3829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6542927788117374,
+      "learning_rate": 9.895742589380223e-06,
+      "loss": 0.1953,
+      "step": 3830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5733980637942968,
+      "learning_rate": 9.89566253290826e-06,
+      "loss": 0.3927,
+      "step": 3831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9872708104500818,
+      "learning_rate": 9.895582446035559e-06,
+      "loss": 0.3119,
+      "step": 3832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3248426064930015,
+      "learning_rate": 9.895502328762619e-06,
+      "loss": 0.4242,
+      "step": 3833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8352791884270307,
+      "learning_rate": 9.895422181089937e-06,
+      "loss": 0.1861,
+      "step": 3834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5812759197508113,
+      "learning_rate": 9.895342003018012e-06,
+      "loss": 0.3709,
+      "step": 3835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9051421671137017,
+      "learning_rate": 9.89526179454734e-06,
+      "loss": 0.3439,
+      "step": 3836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.454006337979761,
+      "learning_rate": 9.895181555678419e-06,
+      "loss": 0.1844,
+      "step": 3837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9336545568595913,
+      "learning_rate": 9.895101286411749e-06,
+      "loss": 0.3602,
+      "step": 3838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.008390484454412,
+      "learning_rate": 9.895020986747827e-06,
+      "loss": 0.4901,
+      "step": 3839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4393703483476563,
+      "learning_rate": 9.894940656687153e-06,
+      "loss": 0.1705,
+      "step": 3840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7285620128204566,
+      "learning_rate": 9.894860296230225e-06,
+      "loss": 0.3732,
+      "step": 3841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5215626238346487,
+      "learning_rate": 9.894779905377542e-06,
+      "loss": 0.6629,
+      "step": 3842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2873122152398648,
+      "learning_rate": 9.894699484129603e-06,
+      "loss": 0.2682,
+      "step": 3843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.298433806405383,
+      "learning_rate": 9.894619032486909e-06,
+      "loss": 0.465,
+      "step": 3844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7288852835175548,
+      "learning_rate": 9.894538550449957e-06,
+      "loss": 0.439,
+      "step": 3845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5455341391826134,
+      "learning_rate": 9.894458038019248e-06,
+      "loss": 0.2322,
+      "step": 3846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0015864822713865,
+      "learning_rate": 9.894377495195282e-06,
+      "loss": 0.368,
+      "step": 3847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0553535281241047,
+      "learning_rate": 9.894296921978558e-06,
+      "loss": 0.7187,
+      "step": 3848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.754402268921128,
+      "learning_rate": 9.894216318369581e-06,
+      "loss": 0.49,
+      "step": 3849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2667292436091384,
+      "learning_rate": 9.894135684368845e-06,
+      "loss": 0.6077,
+      "step": 3850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6742040062992563,
+      "learning_rate": 9.894055019976855e-06,
+      "loss": 0.1432,
+      "step": 3851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7874292488659662,
+      "learning_rate": 9.89397432519411e-06,
+      "loss": 0.2008,
+      "step": 3852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2902982741588185,
+      "learning_rate": 9.893893600021112e-06,
+      "loss": 0.1928,
+      "step": 3853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.864225344327704,
+      "learning_rate": 9.893812844458361e-06,
+      "loss": 0.4917,
+      "step": 3854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7024920144713072,
+      "learning_rate": 9.89373205850636e-06,
+      "loss": 0.3825,
+      "step": 3855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9663519881759917,
+      "learning_rate": 9.893651242165612e-06,
+      "loss": 0.2149,
+      "step": 3856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5457479050603742,
+      "learning_rate": 9.893570395436616e-06,
+      "loss": 0.2463,
+      "step": 3857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.049925045934301,
+      "learning_rate": 9.893489518319875e-06,
+      "loss": 0.397,
+      "step": 3858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1588128518882064,
+      "learning_rate": 9.89340861081589e-06,
+      "loss": 0.335,
+      "step": 3859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3547920907155948,
+      "learning_rate": 9.893327672925165e-06,
+      "loss": 0.2914,
+      "step": 3860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2245766545598484,
+      "learning_rate": 9.893246704648203e-06,
+      "loss": 0.2018,
+      "step": 3861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7081369729604723,
+      "learning_rate": 9.893165705985505e-06,
+      "loss": 0.1554,
+      "step": 3862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1523145122374985,
+      "learning_rate": 9.893084676937576e-06,
+      "loss": 0.1874,
+      "step": 3863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6668932169679376,
+      "learning_rate": 9.89300361750492e-06,
+      "loss": 0.4499,
+      "step": 3864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4980378646628436,
+      "learning_rate": 9.892922527688036e-06,
+      "loss": 0.3885,
+      "step": 3865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.884301077956923,
+      "learning_rate": 9.89284140748743e-06,
+      "loss": 0.1854,
+      "step": 3866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.96936042344423,
+      "learning_rate": 9.892760256903607e-06,
+      "loss": 0.3944,
+      "step": 3867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6637793510493382,
+      "learning_rate": 9.89267907593707e-06,
+      "loss": 0.1686,
+      "step": 3868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.399660098162582,
+      "learning_rate": 9.892597864588323e-06,
+      "loss": 0.3709,
+      "step": 3869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.67787683612621,
+      "learning_rate": 9.892516622857869e-06,
+      "loss": 0.3372,
+      "step": 3870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.146579691720455,
+      "learning_rate": 9.892435350746214e-06,
+      "loss": 0.4969,
+      "step": 3871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1665098334665236,
+      "learning_rate": 9.892354048253862e-06,
+      "loss": 0.215,
+      "step": 3872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.100088222813301,
+      "learning_rate": 9.892272715381319e-06,
+      "loss": 0.561,
+      "step": 3873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9734473577813456,
+      "learning_rate": 9.892191352129088e-06,
+      "loss": 0.218,
+      "step": 3874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.231960858946485,
+      "learning_rate": 9.892109958497676e-06,
+      "loss": 0.2682,
+      "step": 3875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2751896716269058,
+      "learning_rate": 9.89202853448759e-06,
+      "loss": 0.5396,
+      "step": 3876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3198896233811572,
+      "learning_rate": 9.891947080099331e-06,
+      "loss": 0.1996,
+      "step": 3877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9435281268131475,
+      "learning_rate": 9.891865595333407e-06,
+      "loss": 0.43,
+      "step": 3878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7325043599512855,
+      "learning_rate": 9.891784080190327e-06,
+      "loss": 0.322,
+      "step": 3879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4425937366647974,
+      "learning_rate": 9.891702534670592e-06,
+      "loss": 0.3435,
+      "step": 3880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9680546599835291,
+      "learning_rate": 9.891620958774713e-06,
+      "loss": 0.5831,
+      "step": 3881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8723902461287354,
+      "learning_rate": 9.891539352503195e-06,
+      "loss": 0.3635,
+      "step": 3882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.158526031365228,
+      "learning_rate": 9.891457715856542e-06,
+      "loss": 0.5168,
+      "step": 3883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0478321087249287,
+      "learning_rate": 9.891376048835264e-06,
+      "loss": 0.35,
+      "step": 3884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6424836370844815,
+      "learning_rate": 9.891294351439867e-06,
+      "loss": 0.2254,
+      "step": 3885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.450215211485682,
+      "learning_rate": 9.891212623670858e-06,
+      "loss": 0.3293,
+      "step": 3886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6343004392024019,
+      "learning_rate": 9.891130865528746e-06,
+      "loss": 0.3446,
+      "step": 3887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5975788687641257,
+      "learning_rate": 9.89104907701404e-06,
+      "loss": 0.203,
+      "step": 3888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0105945445876645,
+      "learning_rate": 9.890967258127242e-06,
+      "loss": 0.5964,
+      "step": 3889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.82917038054368,
+      "learning_rate": 9.890885408868866e-06,
+      "loss": 0.4078,
+      "step": 3890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.667096385593259,
+      "learning_rate": 9.890803529239418e-06,
+      "loss": 0.5018,
+      "step": 3891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.530711064572903,
+      "learning_rate": 9.890721619239407e-06,
+      "loss": 0.1946,
+      "step": 3892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9889606015968382,
+      "learning_rate": 9.89063967886934e-06,
+      "loss": 0.2375,
+      "step": 3893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9303735139111515,
+      "learning_rate": 9.890557708129726e-06,
+      "loss": 0.5327,
+      "step": 3894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5021508679948745,
+      "learning_rate": 9.890475707021077e-06,
+      "loss": 0.784,
+      "step": 3895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7930316768565753,
+      "learning_rate": 9.890393675543899e-06,
+      "loss": 0.2074,
+      "step": 3896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0248429046141623,
+      "learning_rate": 9.890311613698704e-06,
+      "loss": 0.5132,
+      "step": 3897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4892079079192992,
+      "learning_rate": 9.890229521485998e-06,
+      "loss": 0.3991,
+      "step": 3898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9966018817574391,
+      "learning_rate": 9.890147398906293e-06,
+      "loss": 0.3813,
+      "step": 3899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8504802945463075,
+      "learning_rate": 9.8900652459601e-06,
+      "loss": 0.3767,
+      "step": 3900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7270584041566355,
+      "learning_rate": 9.88998306264793e-06,
+      "loss": 0.2775,
+      "step": 3901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.951637388186754,
+      "learning_rate": 9.889900848970288e-06,
+      "loss": 0.3738,
+      "step": 3902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8349030116562803,
+      "learning_rate": 9.88981860492769e-06,
+      "loss": 0.4612,
+      "step": 3903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5192830243393654,
+      "learning_rate": 9.889736330520645e-06,
+      "loss": 0.3196,
+      "step": 3904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.575265095777514,
+      "learning_rate": 9.889654025749661e-06,
+      "loss": 0.2903,
+      "step": 3905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.199395154040594,
+      "learning_rate": 9.889571690615254e-06,
+      "loss": 0.2136,
+      "step": 3906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1500656198291141,
+      "learning_rate": 9.88948932511793e-06,
+      "loss": 0.2214,
+      "step": 3907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8791926635570764,
+      "learning_rate": 9.889406929258207e-06,
+      "loss": 0.3836,
+      "step": 3908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6174392129972726,
+      "learning_rate": 9.889324503036592e-06,
+      "loss": 0.1747,
+      "step": 3909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.319030446061189,
+      "learning_rate": 9.889242046453598e-06,
+      "loss": 0.5647,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8583600816562167,
+      "learning_rate": 9.889159559509736e-06,
+      "loss": 0.2098,
+      "step": 3911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9075490949344769,
+      "learning_rate": 9.889077042205518e-06,
+      "loss": 0.4729,
+      "step": 3912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.831006530128441,
+      "learning_rate": 9.88899449454146e-06,
+      "loss": 0.2737,
+      "step": 3913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7051851866014691,
+      "learning_rate": 9.88891191651807e-06,
+      "loss": 0.3258,
+      "step": 3914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9388254999847847,
+      "learning_rate": 9.888829308135863e-06,
+      "loss": 0.4565,
+      "step": 3915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7858961156036797,
+      "learning_rate": 9.888746669395354e-06,
+      "loss": 0.2239,
+      "step": 3916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9506719290488415,
+      "learning_rate": 9.888664000297052e-06,
+      "loss": 0.5319,
+      "step": 3917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1188785179933314,
+      "learning_rate": 9.888581300841473e-06,
+      "loss": 0.2335,
+      "step": 3918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7875934018311335,
+      "learning_rate": 9.88849857102913e-06,
+      "loss": 0.232,
+      "step": 3919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.645292192136341,
+      "learning_rate": 9.888415810860538e-06,
+      "loss": 0.4247,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3372834923347237,
+      "learning_rate": 9.888333020336207e-06,
+      "loss": 0.1633,
+      "step": 3921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9704431942725502,
+      "learning_rate": 9.888250199456657e-06,
+      "loss": 0.4442,
+      "step": 3922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5748968817241895,
+      "learning_rate": 9.888167348222396e-06,
+      "loss": 0.2506,
+      "step": 3923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8290637389680207,
+      "learning_rate": 9.888084466633943e-06,
+      "loss": 0.2304,
+      "step": 3924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1538209536450097,
+      "learning_rate": 9.888001554691811e-06,
+      "loss": 0.2593,
+      "step": 3925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.200922588864964,
+      "learning_rate": 9.887918612396515e-06,
+      "loss": 0.5837,
+      "step": 3926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0395737807219336,
+      "learning_rate": 9.887835639748569e-06,
+      "loss": 0.4618,
+      "step": 3927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.367963731300128,
+      "learning_rate": 9.88775263674849e-06,
+      "loss": 0.1982,
+      "step": 3928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.138636865721221,
+      "learning_rate": 9.887669603396792e-06,
+      "loss": 0.2576,
+      "step": 3929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8209634443784724,
+      "learning_rate": 9.887586539693993e-06,
+      "loss": 0.5522,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4316287228868874,
+      "learning_rate": 9.887503445640608e-06,
+      "loss": 0.2443,
+      "step": 3931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.08603797984911,
+      "learning_rate": 9.88742032123715e-06,
+      "loss": 0.5326,
+      "step": 3932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8467142998544686,
+      "learning_rate": 9.887337166484137e-06,
+      "loss": 0.5089,
+      "step": 3933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.897598662117049,
+      "learning_rate": 9.887253981382086e-06,
+      "loss": 0.471,
+      "step": 3934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4408258520482404,
+      "learning_rate": 9.887170765931513e-06,
+      "loss": 0.1989,
+      "step": 3935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.050613437326126,
+      "learning_rate": 9.887087520132936e-06,
+      "loss": 0.2838,
+      "step": 3936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.71930071070875,
+      "learning_rate": 9.887004243986871e-06,
+      "loss": 0.374,
+      "step": 3937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6741463304201962,
+      "learning_rate": 9.886920937493835e-06,
+      "loss": 0.3435,
+      "step": 3938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.331263046637224,
+      "learning_rate": 9.886837600654344e-06,
+      "loss": 0.5238,
+      "step": 3939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5562650952698625,
+      "learning_rate": 9.886754233468919e-06,
+      "loss": 0.3411,
+      "step": 3940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7559019015702917,
+      "learning_rate": 9.886670835938074e-06,
+      "loss": 0.1923,
+      "step": 3941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.438501609667054,
+      "learning_rate": 9.88658740806233e-06,
+      "loss": 0.168,
+      "step": 3942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2568219992035137,
+      "learning_rate": 9.8865039498422e-06,
+      "loss": 0.539,
+      "step": 3943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5873333632314,
+      "learning_rate": 9.886420461278209e-06,
+      "loss": 0.2813,
+      "step": 3944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1629060416406776,
+      "learning_rate": 9.886336942370872e-06,
+      "loss": 0.5601,
+      "step": 3945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.010398635305736,
+      "learning_rate": 9.886253393120707e-06,
+      "loss": 0.4798,
+      "step": 3946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5560159279310235,
+      "learning_rate": 9.886169813528233e-06,
+      "loss": 0.3374,
+      "step": 3947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3071620975105045,
+      "learning_rate": 9.88608620359397e-06,
+      "loss": 0.5856,
+      "step": 3948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3574310329869312,
+      "learning_rate": 9.886002563318437e-06,
+      "loss": 0.2868,
+      "step": 3949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.356061705798859,
+      "learning_rate": 9.885918892702154e-06,
+      "loss": 0.2747,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5367279258031303,
+      "learning_rate": 9.885835191745639e-06,
+      "loss": 0.3321,
+      "step": 3951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0795025741132878,
+      "learning_rate": 9.885751460449413e-06,
+      "loss": 0.2042,
+      "step": 3952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0834860380156695,
+      "learning_rate": 9.885667698813995e-06,
+      "loss": 0.5042,
+      "step": 3953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.467928575551718,
+      "learning_rate": 9.885583906839903e-06,
+      "loss": 0.3907,
+      "step": 3954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.747224496256671,
+      "learning_rate": 9.885500084527665e-06,
+      "loss": 0.5501,
+      "step": 3955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9270351073098804,
+      "learning_rate": 9.885416231877794e-06,
+      "loss": 0.4838,
+      "step": 3956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0717409855810005,
+      "learning_rate": 9.885332348890812e-06,
+      "loss": 0.3738,
+      "step": 3957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1197948588566717,
+      "learning_rate": 9.885248435567243e-06,
+      "loss": 0.5433,
+      "step": 3958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.651420567584262,
+      "learning_rate": 9.885164491907606e-06,
+      "loss": 0.1938,
+      "step": 3959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1424161907193393,
+      "learning_rate": 9.885080517912422e-06,
+      "loss": 0.4307,
+      "step": 3960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0623783704266696,
+      "learning_rate": 9.884996513582212e-06,
+      "loss": 0.3576,
+      "step": 3961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6707587392744834,
+      "learning_rate": 9.884912478917499e-06,
+      "loss": 0.2474,
+      "step": 3962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1108858278671914,
+      "learning_rate": 9.884828413918805e-06,
+      "loss": 0.1918,
+      "step": 3963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4196631275544784,
+      "learning_rate": 9.884744318586651e-06,
+      "loss": 0.2607,
+      "step": 3964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0436642048410847,
+      "learning_rate": 9.88466019292156e-06,
+      "loss": 0.3249,
+      "step": 3965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.030850686658033,
+      "learning_rate": 9.884576036924053e-06,
+      "loss": 0.5299,
+      "step": 3966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9740692818763548,
+      "learning_rate": 9.884491850594653e-06,
+      "loss": 0.5116,
+      "step": 3967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.234769199517173,
+      "learning_rate": 9.884407633933885e-06,
+      "loss": 0.6668,
+      "step": 3968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1776732158198433,
+      "learning_rate": 9.88432338694227e-06,
+      "loss": 0.8091,
+      "step": 3969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5038015081418437,
+      "learning_rate": 9.88423910962033e-06,
+      "loss": 0.2536,
+      "step": 3970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.330436806622904,
+      "learning_rate": 9.884154801968592e-06,
+      "loss": 0.1758,
+      "step": 3971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9103857304312195,
+      "learning_rate": 9.884070463987576e-06,
+      "loss": 0.3075,
+      "step": 3972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.061753419654705,
+      "learning_rate": 9.883986095677807e-06,
+      "loss": 0.5636,
+      "step": 3973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5572275086862564,
+      "learning_rate": 9.883901697039809e-06,
+      "loss": 0.2611,
+      "step": 3974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5621790425151212,
+      "learning_rate": 9.883817268074107e-06,
+      "loss": 0.2337,
+      "step": 3975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0699456258715263,
+      "learning_rate": 9.883732808781223e-06,
+      "loss": 0.2215,
+      "step": 3976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8197507687365406,
+      "learning_rate": 9.883648319161684e-06,
+      "loss": 0.3844,
+      "step": 3977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3205447803176167,
+      "learning_rate": 9.883563799216011e-06,
+      "loss": 0.4697,
+      "step": 3978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9408859691741074,
+      "learning_rate": 9.883479248944734e-06,
+      "loss": 0.4049,
+      "step": 3979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4307392139681474,
+      "learning_rate": 9.883394668348373e-06,
+      "loss": 0.1664,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.697842358404383,
+      "learning_rate": 9.883310057427458e-06,
+      "loss": 0.2592,
+      "step": 3981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1910583505529995,
+      "learning_rate": 9.883225416182509e-06,
+      "loss": 0.5673,
+      "step": 3982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8990612418211534,
+      "learning_rate": 9.883140744614056e-06,
+      "loss": 0.4495,
+      "step": 3983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.057192934761535,
+      "learning_rate": 9.883056042722624e-06,
+      "loss": 0.2604,
+      "step": 3984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0352958807024395,
+      "learning_rate": 9.882971310508737e-06,
+      "loss": 0.4229,
+      "step": 3985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3378603743871353,
+      "learning_rate": 9.882886547972924e-06,
+      "loss": 0.2642,
+      "step": 3986
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1470641192673847,
+      "learning_rate": 9.882801755115709e-06,
+      "loss": 0.1742,
+      "step": 3987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3987366233881504,
+      "learning_rate": 9.882716931937619e-06,
+      "loss": 0.2004,
+      "step": 3988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8051345940189598,
+      "learning_rate": 9.882632078439183e-06,
+      "loss": 0.4142,
+      "step": 3989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.633171147671682,
+      "learning_rate": 9.882547194620923e-06,
+      "loss": 0.2667,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.108177222728489,
+      "learning_rate": 9.882462280483371e-06,
+      "loss": 0.4658,
+      "step": 3991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0440321246364292,
+      "learning_rate": 9.882377336027053e-06,
+      "loss": 0.5326,
+      "step": 3992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.409032411252573,
+      "learning_rate": 9.882292361252495e-06,
+      "loss": 0.3764,
+      "step": 3993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4678009727581243,
+      "learning_rate": 9.882207356160225e-06,
+      "loss": 0.2557,
+      "step": 3994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1723714439494244,
+      "learning_rate": 9.88212232075077e-06,
+      "loss": 0.6515,
+      "step": 3995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.708538596733063,
+      "learning_rate": 9.882037255024664e-06,
+      "loss": 0.3733,
+      "step": 3996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.069444193580436,
+      "learning_rate": 9.881952158982428e-06,
+      "loss": 0.4695,
+      "step": 3997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4112568850239513,
+      "learning_rate": 9.881867032624593e-06,
+      "loss": 0.2275,
+      "step": 3998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0442562203988515,
+      "learning_rate": 9.881781875951688e-06,
+      "loss": 0.6629,
+      "step": 3999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.215197849392975,
+      "learning_rate": 9.88169668896424e-06,
+      "loss": 0.5051,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.967078256035557,
+      "learning_rate": 9.881611471662782e-06,
+      "loss": 0.3884,
+      "step": 4001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4706407320490094,
+      "learning_rate": 9.88152622404784e-06,
+      "loss": 0.6663,
+      "step": 4002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4909954752271346,
+      "learning_rate": 9.881440946119942e-06,
+      "loss": 0.2477,
+      "step": 4003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7592095185439216,
+      "learning_rate": 9.881355637879622e-06,
+      "loss": 0.3622,
+      "step": 4004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.509512965292277,
+      "learning_rate": 9.881270299327408e-06,
+      "loss": 0.2722,
+      "step": 4005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4892365551041153,
+      "learning_rate": 9.881184930463828e-06,
+      "loss": 0.2224,
+      "step": 4006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4209603951463343,
+      "learning_rate": 9.881099531289414e-06,
+      "loss": 0.5109,
+      "step": 4007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.127967013117167,
+      "learning_rate": 9.881014101804694e-06,
+      "loss": 0.1354,
+      "step": 4008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.376822819895593,
+      "learning_rate": 9.880928642010202e-06,
+      "loss": 0.4317,
+      "step": 4009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.688288429417983,
+      "learning_rate": 9.880843151906465e-06,
+      "loss": 0.2725,
+      "step": 4010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2765443543203217,
+      "learning_rate": 9.880757631494017e-06,
+      "loss": 0.2069,
+      "step": 4011
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8044587233886338,
+      "learning_rate": 9.880672080773387e-06,
+      "loss": 0.2962,
+      "step": 4012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.226295220913367,
+      "learning_rate": 9.880586499745109e-06,
+      "loss": 0.3638,
+      "step": 4013
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8939563350822903,
+      "learning_rate": 9.88050088840971e-06,
+      "loss": 0.4704,
+      "step": 4014
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5411392375069635,
+      "learning_rate": 9.880415246767725e-06,
+      "loss": 0.2202,
+      "step": 4015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4153164669995377,
+      "learning_rate": 9.880329574819686e-06,
+      "loss": 0.1597,
+      "step": 4016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2833321194720932,
+      "learning_rate": 9.880243872566124e-06,
+      "loss": 0.3933,
+      "step": 4017
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3316313795024461,
+      "learning_rate": 9.880158140007569e-06,
+      "loss": 0.2568,
+      "step": 4018
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4103494839348536,
+      "learning_rate": 9.880072377144558e-06,
+      "loss": 0.174,
+      "step": 4019
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5266294638433298,
+      "learning_rate": 9.87998658397762e-06,
+      "loss": 0.367,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6834259226142265,
+      "learning_rate": 9.879900760507288e-06,
+      "loss": 0.1866,
+      "step": 4021
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7567539630656275,
+      "learning_rate": 9.879814906734096e-06,
+      "loss": 0.4102,
+      "step": 4022
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.478960025521896,
+      "learning_rate": 9.879729022658578e-06,
+      "loss": 0.2143,
+      "step": 4023
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7557563836398415,
+      "learning_rate": 9.879643108281265e-06,
+      "loss": 0.3806,
+      "step": 4024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1652229860719583,
+      "learning_rate": 9.879557163602693e-06,
+      "loss": 0.4131,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.082488251391926,
+      "learning_rate": 9.879471188623393e-06,
+      "loss": 0.2474,
+      "step": 4026
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8367663845828375,
+      "learning_rate": 9.879385183343901e-06,
+      "loss": 0.2897,
+      "step": 4027
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5125089740118445,
+      "learning_rate": 9.87929914776475e-06,
+      "loss": 0.6549,
+      "step": 4028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.241884742792217,
+      "learning_rate": 9.879213081886476e-06,
+      "loss": 0.6928,
+      "step": 4029
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9403902154448922,
+      "learning_rate": 9.879126985709611e-06,
+      "loss": 0.3662,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4986479736367615,
+      "learning_rate": 9.87904085923469e-06,
+      "loss": 0.3046,
+      "step": 4031
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1147006773751291,
+      "learning_rate": 9.878954702462249e-06,
+      "loss": 0.1587,
+      "step": 4032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2332141524103684,
+      "learning_rate": 9.878868515392821e-06,
+      "loss": 0.1853,
+      "step": 4033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.411237545152751,
+      "learning_rate": 9.878782298026945e-06,
+      "loss": 0.2937,
+      "step": 4034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.559297923627606,
+      "learning_rate": 9.878696050365153e-06,
+      "loss": 0.1478,
+      "step": 4035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3469867771105182,
+      "learning_rate": 9.878609772407983e-06,
+      "loss": 0.205,
+      "step": 4036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9567223607783673,
+      "learning_rate": 9.87852346415597e-06,
+      "loss": 0.4355,
+      "step": 4037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.156075346134411,
+      "learning_rate": 9.878437125609646e-06,
+      "loss": 0.3866,
+      "step": 4038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2828386520735509,
+      "learning_rate": 9.878350756769552e-06,
+      "loss": 0.2105,
+      "step": 4039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4531426992671577,
+      "learning_rate": 9.878264357636224e-06,
+      "loss": 0.3686,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7438659793471978,
+      "learning_rate": 9.878177928210196e-06,
+      "loss": 0.3293,
+      "step": 4041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0205561695239833,
+      "learning_rate": 9.878091468492007e-06,
+      "loss": 0.4265,
+      "step": 4042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5372478708695105,
+      "learning_rate": 9.878004978482191e-06,
+      "loss": 0.3322,
+      "step": 4043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.879338204899311,
+      "learning_rate": 9.87791845818129e-06,
+      "loss": 0.4823,
+      "step": 4044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3244351615288368,
+      "learning_rate": 9.877831907589837e-06,
+      "loss": 0.2311,
+      "step": 4045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9226661988008324,
+      "learning_rate": 9.877745326708372e-06,
+      "loss": 0.2391,
+      "step": 4046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1615799715797626,
+      "learning_rate": 9.877658715537429e-06,
+      "loss": 0.2185,
+      "step": 4047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.331638560764887,
+      "learning_rate": 9.87757207407755e-06,
+      "loss": 0.3208,
+      "step": 4048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9404021569781835,
+      "learning_rate": 9.87748540232927e-06,
+      "loss": 0.2439,
+      "step": 4049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.2863093141192214,
+      "learning_rate": 9.87739870029313e-06,
+      "loss": 0.2288,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.92479793289989,
+      "learning_rate": 9.877311967969665e-06,
+      "loss": 0.5613,
+      "step": 4051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9175737988222503,
+      "learning_rate": 9.877225205359417e-06,
+      "loss": 0.1912,
+      "step": 4052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1173388683923515,
+      "learning_rate": 9.877138412462922e-06,
+      "loss": 0.5605,
+      "step": 4053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.42853102525771,
+      "learning_rate": 9.877051589280721e-06,
+      "loss": 0.3509,
+      "step": 4054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.886648574272258,
+      "learning_rate": 9.876964735813353e-06,
+      "loss": 0.1814,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9183725370974298,
+      "learning_rate": 9.876877852061356e-06,
+      "loss": 0.5127,
+      "step": 4056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.928243529264023,
+      "learning_rate": 9.87679093802527e-06,
+      "loss": 0.5466,
+      "step": 4057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0508640249895014,
+      "learning_rate": 9.876703993705635e-06,
+      "loss": 0.2474,
+      "step": 4058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8851650743716664,
+      "learning_rate": 9.87661701910299e-06,
+      "loss": 0.373,
+      "step": 4059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.286785079089533,
+      "learning_rate": 9.876530014217875e-06,
+      "loss": 0.1463,
+      "step": 4060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9056518663768913,
+      "learning_rate": 9.876442979050834e-06,
+      "loss": 0.4141,
+      "step": 4061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6113429340491208,
+      "learning_rate": 9.876355913602403e-06,
+      "loss": 0.2107,
+      "step": 4062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2445345168549675,
+      "learning_rate": 9.876268817873125e-06,
+      "loss": 0.274,
+      "step": 4063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3537434004733147,
+      "learning_rate": 9.87618169186354e-06,
+      "loss": 0.706,
+      "step": 4064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.749508674868066,
+      "learning_rate": 9.87609453557419e-06,
+      "loss": 0.2316,
+      "step": 4065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4661389395484994,
+      "learning_rate": 9.876007349005614e-06,
+      "loss": 0.7405,
+      "step": 4066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9392913979503437,
+      "learning_rate": 9.875920132158354e-06,
+      "loss": 0.2352,
+      "step": 4067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.706323987802887,
+      "learning_rate": 9.875832885032955e-06,
+      "loss": 0.4923,
+      "step": 4068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9831983539553213,
+      "learning_rate": 9.875745607629955e-06,
+      "loss": 0.4481,
+      "step": 4069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3902868954292944,
+      "learning_rate": 9.875658299949896e-06,
+      "loss": 0.4687,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.778185066457735,
+      "learning_rate": 9.875570961993321e-06,
+      "loss": 0.4952,
+      "step": 4071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.849016439852516,
+      "learning_rate": 9.875483593760774e-06,
+      "loss": 0.2094,
+      "step": 4072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3264229695478205,
+      "learning_rate": 9.875396195252795e-06,
+      "loss": 0.2198,
+      "step": 4073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.631417785342884,
+      "learning_rate": 9.87530876646993e-06,
+      "loss": 0.1775,
+      "step": 4074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2462000637084043,
+      "learning_rate": 9.875221307412718e-06,
+      "loss": 0.232,
+      "step": 4075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9621458320184968,
+      "learning_rate": 9.875133818081703e-06,
+      "loss": 0.3373,
+      "step": 4076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9911528462406132,
+      "learning_rate": 9.875046298477431e-06,
+      "loss": 0.5824,
+      "step": 4077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.901668107447077,
+      "learning_rate": 9.874958748600444e-06,
+      "loss": 0.3195,
+      "step": 4078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1009517388412395,
+      "learning_rate": 9.874871168451284e-06,
+      "loss": 0.3668,
+      "step": 4079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.150234050403514,
+      "learning_rate": 9.874783558030495e-06,
+      "loss": 0.3006,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.103241655217035,
+      "learning_rate": 9.874695917338625e-06,
+      "loss": 0.1988,
+      "step": 4081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1832027615710028,
+      "learning_rate": 9.874608246376215e-06,
+      "loss": 0.2231,
+      "step": 4082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4803532276022906,
+      "learning_rate": 9.874520545143808e-06,
+      "loss": 0.2469,
+      "step": 4083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1979803795733273,
+      "learning_rate": 9.874432813641952e-06,
+      "loss": 0.1503,
+      "step": 4084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2261059900114997,
+      "learning_rate": 9.874345051871191e-06,
+      "loss": 0.1931,
+      "step": 4085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.193089714016466,
+      "learning_rate": 9.874257259832068e-06,
+      "loss": 0.3958,
+      "step": 4086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7123838521036745,
+      "learning_rate": 9.874169437525128e-06,
+      "loss": 0.226,
+      "step": 4087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8548931855394388,
+      "learning_rate": 9.874081584950918e-06,
+      "loss": 0.4182,
+      "step": 4088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.189219993770592,
+      "learning_rate": 9.873993702109986e-06,
+      "loss": 0.1979,
+      "step": 4089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7800502082622924,
+      "learning_rate": 9.873905789002872e-06,
+      "loss": 0.2686,
+      "step": 4090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2382063194296637,
+      "learning_rate": 9.873817845630127e-06,
+      "loss": 0.1969,
+      "step": 4091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.358774932419384,
+      "learning_rate": 9.873729871992294e-06,
+      "loss": 0.1824,
+      "step": 4092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9381066303249885,
+      "learning_rate": 9.87364186808992e-06,
+      "loss": 0.3883,
+      "step": 4093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8427666450217621,
+      "learning_rate": 9.873553833923553e-06,
+      "loss": 0.5573,
+      "step": 4094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8023506987875535,
+      "learning_rate": 9.873465769493736e-06,
+      "loss": 0.4058,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9286415184179448,
+      "learning_rate": 9.87337767480102e-06,
+      "loss": 0.5073,
+      "step": 4096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6009515434377433,
+      "learning_rate": 9.87328954984595e-06,
+      "loss": 0.3609,
+      "step": 4097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9739473593380807,
+      "learning_rate": 9.873201394629076e-06,
+      "loss": 0.2877,
+      "step": 4098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5142513224328464,
+      "learning_rate": 9.87311320915094e-06,
+      "loss": 0.2146,
+      "step": 4099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.847645523653086,
+      "learning_rate": 9.873024993412093e-06,
+      "loss": 0.5565,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0086492991472364,
+      "learning_rate": 9.872936747413082e-06,
+      "loss": 0.4956,
+      "step": 4101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2002728889541794,
+      "learning_rate": 9.872848471154456e-06,
+      "loss": 0.2175,
+      "step": 4102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7099934242077484,
+      "learning_rate": 9.872760164636763e-06,
+      "loss": 0.2765,
+      "step": 4103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8564482583028106,
+      "learning_rate": 9.87267182786055e-06,
+      "loss": 0.4728,
+      "step": 4104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9921784195038534,
+      "learning_rate": 9.872583460826369e-06,
+      "loss": 0.368,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7549997629331227,
+      "learning_rate": 9.872495063534763e-06,
+      "loss": 0.3188,
+      "step": 4106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8089682971863432,
+      "learning_rate": 9.872406635986284e-06,
+      "loss": 0.4215,
+      "step": 4107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6677748904706036,
+      "learning_rate": 9.872318178181484e-06,
+      "loss": 0.3499,
+      "step": 4108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.203856258913923,
+      "learning_rate": 9.872229690120908e-06,
+      "loss": 0.4995,
+      "step": 4109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5135478541111431,
+      "learning_rate": 9.872141171805108e-06,
+      "loss": 0.2135,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.270688272503417,
+      "learning_rate": 9.872052623234632e-06,
+      "loss": 0.325,
+      "step": 4111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6342027759436921,
+      "learning_rate": 9.87196404441003e-06,
+      "loss": 0.2762,
+      "step": 4112
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 41101,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2056,
+  "total_flos": 339812383801344.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}