diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,129549 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4502080241356658,
+  "eval_steps": 500,
+  "global_step": 18504,
+  "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
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1163639986728873,
+      "learning_rate": 9.871875435331852e-06,
+      "loss": 0.2875,
+      "step": 4113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8644153973981983,
+      "learning_rate": 9.871786796000651e-06,
+      "loss": 0.3945,
+      "step": 4114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.803237522455453,
+      "learning_rate": 9.871698126416972e-06,
+      "loss": 0.4481,
+      "step": 4115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2155200066494052,
+      "learning_rate": 9.871609426581373e-06,
+      "loss": 0.2207,
+      "step": 4116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4144202406841886,
+      "learning_rate": 9.871520696494398e-06,
+      "loss": 0.1882,
+      "step": 4117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5813831697333918,
+      "learning_rate": 9.871431936156602e-06,
+      "loss": 0.3986,
+      "step": 4118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4450008623407686,
+      "learning_rate": 9.871343145568532e-06,
+      "loss": 0.2678,
+      "step": 4119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7631044457206737,
+      "learning_rate": 9.871254324730745e-06,
+      "loss": 0.3861,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9832416671560202,
+      "learning_rate": 9.87116547364379e-06,
+      "loss": 0.2659,
+      "step": 4121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2776466351421494,
+      "learning_rate": 9.871076592308218e-06,
+      "loss": 0.5165,
+      "step": 4122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9250975646802344,
+      "learning_rate": 9.870987680724581e-06,
+      "loss": 0.2459,
+      "step": 4123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.916496721971009,
+      "learning_rate": 9.87089873889343e-06,
+      "loss": 0.3589,
+      "step": 4124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8900581329730195,
+      "learning_rate": 9.87080976681532e-06,
+      "loss": 0.532,
+      "step": 4125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.985772453686031,
+      "learning_rate": 9.870720764490804e-06,
+      "loss": 0.2137,
+      "step": 4126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9609348821432566,
+      "learning_rate": 9.870631731920433e-06,
+      "loss": 0.3597,
+      "step": 4127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.600526714334836,
+      "learning_rate": 9.870542669104759e-06,
+      "loss": 0.3852,
+      "step": 4128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6234390848937705,
+      "learning_rate": 9.870453576044335e-06,
+      "loss": 0.4213,
+      "step": 4129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.315010698975543,
+      "learning_rate": 9.870364452739716e-06,
+      "loss": 0.7482,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9674064047822881,
+      "learning_rate": 9.870275299191455e-06,
+      "loss": 0.2684,
+      "step": 4131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7353177558383799,
+      "learning_rate": 9.870186115400105e-06,
+      "loss": 0.342,
+      "step": 4132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.609716929990769,
+      "learning_rate": 9.87009690136622e-06,
+      "loss": 0.2997,
+      "step": 4133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7482342326597275,
+      "learning_rate": 9.870007657090354e-06,
+      "loss": 0.3772,
+      "step": 4134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4647899990599025,
+      "learning_rate": 9.86991838257306e-06,
+      "loss": 0.3702,
+      "step": 4135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.020474169503135,
+      "learning_rate": 9.869829077814897e-06,
+      "loss": 0.2602,
+      "step": 4136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.599460012614585,
+      "learning_rate": 9.869739742816414e-06,
+      "loss": 0.3875,
+      "step": 4137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8524302212841368,
+      "learning_rate": 9.869650377578166e-06,
+      "loss": 0.2605,
+      "step": 4138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9190802529139253,
+      "learning_rate": 9.869560982100713e-06,
+      "loss": 0.5329,
+      "step": 4139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8242954290798717,
+      "learning_rate": 9.869471556384604e-06,
+      "loss": 0.0751,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.953800319269532,
+      "learning_rate": 9.869382100430401e-06,
+      "loss": 0.2997,
+      "step": 4141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4069134127421856,
+      "learning_rate": 9.869292614238653e-06,
+      "loss": 0.3419,
+      "step": 4142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.487129858288207,
+      "learning_rate": 9.869203097809919e-06,
+      "loss": 0.5059,
+      "step": 4143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1732307051863287,
+      "learning_rate": 9.869113551144754e-06,
+      "loss": 0.409,
+      "step": 4144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7155870985837014,
+      "learning_rate": 9.869023974243716e-06,
+      "loss": 0.2495,
+      "step": 4145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3560213919212178,
+      "learning_rate": 9.868934367107358e-06,
+      "loss": 0.2623,
+      "step": 4146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1244629990452006,
+      "learning_rate": 9.86884472973624e-06,
+      "loss": 0.2212,
+      "step": 4147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7408247951166094,
+      "learning_rate": 9.868755062130915e-06,
+      "loss": 0.2455,
+      "step": 4148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.452269696602912,
+      "learning_rate": 9.86866536429194e-06,
+      "loss": 0.5922,
+      "step": 4149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4617403394775565,
+      "learning_rate": 9.868575636219876e-06,
+      "loss": 0.1712,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0976374570264005,
+      "learning_rate": 9.868485877915277e-06,
+      "loss": 0.3938,
+      "step": 4151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8417257900011248,
+      "learning_rate": 9.868396089378702e-06,
+      "loss": 0.3556,
+      "step": 4152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8641100385796177,
+      "learning_rate": 9.868306270610706e-06,
+      "loss": 0.4432,
+      "step": 4153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3520719160191983,
+      "learning_rate": 9.868216421611849e-06,
+      "loss": 0.693,
+      "step": 4154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.21545474749781,
+      "learning_rate": 9.86812654238269e-06,
+      "loss": 0.2002,
+      "step": 4155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7235650241046372,
+      "learning_rate": 9.868036632923784e-06,
+      "loss": 0.1766,
+      "step": 4156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.031068470986352,
+      "learning_rate": 9.867946693235691e-06,
+      "loss": 0.5386,
+      "step": 4157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.362074280533798,
+      "learning_rate": 9.867856723318968e-06,
+      "loss": 0.8183,
+      "step": 4158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1990866003088723,
+      "learning_rate": 9.867766723174177e-06,
+      "loss": 0.5982,
+      "step": 4159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.421255866251876,
+      "learning_rate": 9.867676692801873e-06,
+      "loss": 0.4998,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.235742077394644,
+      "learning_rate": 9.867586632202617e-06,
+      "loss": 0.5475,
+      "step": 4161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.748233648796483,
+      "learning_rate": 9.867496541376969e-06,
+      "loss": 0.3516,
+      "step": 4162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.910112556004585,
+      "learning_rate": 9.867406420325487e-06,
+      "loss": 0.5117,
+      "step": 4163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.868695229067738,
+      "learning_rate": 9.867316269048731e-06,
+      "loss": 0.4862,
+      "step": 4164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9743670759196,
+      "learning_rate": 9.867226087547263e-06,
+      "loss": 0.3441,
+      "step": 4165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8293583084338059,
+      "learning_rate": 9.867135875821639e-06,
+      "loss": 0.4285,
+      "step": 4166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.030914843858886,
+      "learning_rate": 9.867045633872422e-06,
+      "loss": 0.5391,
+      "step": 4167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8893770165894825,
+      "learning_rate": 9.866955361700171e-06,
+      "loss": 0.5167,
+      "step": 4168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7758344299100297,
+      "learning_rate": 9.866865059305448e-06,
+      "loss": 0.2903,
+      "step": 4169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9653724050225099,
+      "learning_rate": 9.866774726688812e-06,
+      "loss": 0.5518,
+      "step": 4170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.985934892813279,
+      "learning_rate": 9.866684363850825e-06,
+      "loss": 0.194,
+      "step": 4171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7931229207195356,
+      "learning_rate": 9.866593970792049e-06,
+      "loss": 0.3618,
+      "step": 4172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8244737257011043,
+      "learning_rate": 9.866503547513045e-06,
+      "loss": 0.2665,
+      "step": 4173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.825657116151357,
+      "learning_rate": 9.86641309401437e-06,
+      "loss": 0.3721,
+      "step": 4174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.644085315425116,
+      "learning_rate": 9.866322610296594e-06,
+      "loss": 0.409,
+      "step": 4175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2980730744621358,
+      "learning_rate": 9.866232096360272e-06,
+      "loss": 0.2956,
+      "step": 4176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.700192126644158,
+      "learning_rate": 9.866141552205968e-06,
+      "loss": 0.5079,
+      "step": 4177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8147242007905398,
+      "learning_rate": 9.866050977834247e-06,
+      "loss": 0.3212,
+      "step": 4178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6214500702069026,
+      "learning_rate": 9.865960373245667e-06,
+      "loss": 0.3677,
+      "step": 4179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.20172296376509,
+      "learning_rate": 9.865869738440793e-06,
+      "loss": 0.7487,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.633042552018668,
+      "learning_rate": 9.865779073420188e-06,
+      "loss": 0.3159,
+      "step": 4181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.802840758725579,
+      "learning_rate": 9.865688378184415e-06,
+      "loss": 0.2496,
+      "step": 4182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8435104262960127,
+      "learning_rate": 9.865597652734036e-06,
+      "loss": 0.5157,
+      "step": 4183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1007810614640055,
+      "learning_rate": 9.865506897069617e-06,
+      "loss": 0.2134,
+      "step": 4184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7849604313502505,
+      "learning_rate": 9.865416111191717e-06,
+      "loss": 0.443,
+      "step": 4185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8527621957812228,
+      "learning_rate": 9.865325295100906e-06,
+      "loss": 0.2264,
+      "step": 4186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3360191649299766,
+      "learning_rate": 9.86523444879774e-06,
+      "loss": 0.2448,
+      "step": 4187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.178689735135025,
+      "learning_rate": 9.865143572282789e-06,
+      "loss": 0.5322,
+      "step": 4188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5014875506413081,
+      "learning_rate": 9.865052665556618e-06,
+      "loss": 0.1611,
+      "step": 4189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5451124618683678,
+      "learning_rate": 9.864961728619789e-06,
+      "loss": 0.1589,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.76304021615249,
+      "learning_rate": 9.864870761472864e-06,
+      "loss": 0.2411,
+      "step": 4191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0501619402958595,
+      "learning_rate": 9.864779764116414e-06,
+      "loss": 0.5327,
+      "step": 4192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.397771798086435,
+      "learning_rate": 9.864688736551e-06,
+      "loss": 0.2016,
+      "step": 4193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4157795108393887,
+      "learning_rate": 9.864597678777188e-06,
+      "loss": 0.229,
+      "step": 4194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.306457576869312,
+      "learning_rate": 9.864506590795545e-06,
+      "loss": 0.2576,
+      "step": 4195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.327859351344334,
+      "learning_rate": 9.864415472606635e-06,
+      "loss": 0.6558,
+      "step": 4196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.130806465086448,
+      "learning_rate": 9.864324324211022e-06,
+      "loss": 0.7449,
+      "step": 4197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.33284309607318,
+      "learning_rate": 9.864233145609277e-06,
+      "loss": 0.5935,
+      "step": 4198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8596208433904937,
+      "learning_rate": 9.864141936801963e-06,
+      "loss": 0.3302,
+      "step": 4199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9787910397768587,
+      "learning_rate": 9.864050697789646e-06,
+      "loss": 0.4521,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1899098801848889,
+      "learning_rate": 9.863959428572893e-06,
+      "loss": 0.2247,
+      "step": 4201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7033824572945457,
+      "learning_rate": 9.863868129152272e-06,
+      "loss": 0.3373,
+      "step": 4202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1807597541227968,
+      "learning_rate": 9.86377679952835e-06,
+      "loss": 0.4512,
+      "step": 4203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5193939833433925,
+      "learning_rate": 9.863685439701688e-06,
+      "loss": 0.3364,
+      "step": 4204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1803660541223906,
+      "learning_rate": 9.863594049672865e-06,
+      "loss": 0.147,
+      "step": 4205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2949448334627887,
+      "learning_rate": 9.863502629442438e-06,
+      "loss": 0.1479,
+      "step": 4206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.75504781958784,
+      "learning_rate": 9.86341117901098e-06,
+      "loss": 0.4085,
+      "step": 4207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7158380213319195,
+      "learning_rate": 9.863319698379058e-06,
+      "loss": 0.2851,
+      "step": 4208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2001957256318447,
+      "learning_rate": 9.863228187547237e-06,
+      "loss": 0.4476,
+      "step": 4209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8667460123154669,
+      "learning_rate": 9.863136646516092e-06,
+      "loss": 0.4967,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8253477307436388,
+      "learning_rate": 9.863045075286186e-06,
+      "loss": 0.2937,
+      "step": 4211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4662140807218083,
+      "learning_rate": 9.862953473858088e-06,
+      "loss": 0.4448,
+      "step": 4212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.706013690475553,
+      "learning_rate": 9.862861842232368e-06,
+      "loss": 0.2892,
+      "step": 4213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0646065372570668,
+      "learning_rate": 9.862770180409596e-06,
+      "loss": 0.3467,
+      "step": 4214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7047485303924197,
+      "learning_rate": 9.862678488390339e-06,
+      "loss": 0.3562,
+      "step": 4215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5902155715516608,
+      "learning_rate": 9.862586766175166e-06,
+      "loss": 0.4011,
+      "step": 4216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.875400170697433,
+      "learning_rate": 9.86249501376465e-06,
+      "loss": 0.5647,
+      "step": 4217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0070272052883023,
+      "learning_rate": 9.862403231159358e-06,
+      "loss": 0.4482,
+      "step": 4218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3795935050120267,
+      "learning_rate": 9.86231141835986e-06,
+      "loss": 0.129,
+      "step": 4219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9521513228516774,
+      "learning_rate": 9.862219575366727e-06,
+      "loss": 0.2931,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5155983534171409,
+      "learning_rate": 9.86212770218053e-06,
+      "loss": 0.1756,
+      "step": 4221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.547540430032305,
+      "learning_rate": 9.86203579880184e-06,
+      "loss": 0.3671,
+      "step": 4222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7722335249188452,
+      "learning_rate": 9.861943865231222e-06,
+      "loss": 0.3681,
+      "step": 4223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8506916649903347,
+      "learning_rate": 9.861851901469255e-06,
+      "loss": 0.7134,
+      "step": 4224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.073673708295872,
+      "learning_rate": 9.861759907516504e-06,
+      "loss": 0.4336,
+      "step": 4225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7806728519726125,
+      "learning_rate": 9.861667883373545e-06,
+      "loss": 0.4042,
+      "step": 4226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4569342455647392,
+      "learning_rate": 9.861575829040945e-06,
+      "loss": 0.2149,
+      "step": 4227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9005170037177082,
+      "learning_rate": 9.861483744519278e-06,
+      "loss": 0.5144,
+      "step": 4228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7802512700946065,
+      "learning_rate": 9.861391629809116e-06,
+      "loss": 0.3602,
+      "step": 4229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0427586336198513,
+      "learning_rate": 9.861299484911028e-06,
+      "loss": 0.2497,
+      "step": 4230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7676725246449392,
+      "learning_rate": 9.861207309825593e-06,
+      "loss": 0.2126,
+      "step": 4231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.916900756956424,
+      "learning_rate": 9.861115104553375e-06,
+      "loss": 0.457,
+      "step": 4232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4166867711122586,
+      "learning_rate": 9.861022869094953e-06,
+      "loss": 0.2421,
+      "step": 4233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8528337963081565,
+      "learning_rate": 9.860930603450896e-06,
+      "loss": 0.2341,
+      "step": 4234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6199149798303305,
+      "learning_rate": 9.860838307621778e-06,
+      "loss": 0.2382,
+      "step": 4235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4732376210833393,
+      "learning_rate": 9.860745981608175e-06,
+      "loss": 0.2594,
+      "step": 4236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0055844266092024,
+      "learning_rate": 9.860653625410654e-06,
+      "loss": 0.3399,
+      "step": 4237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.463815744805448,
+      "learning_rate": 9.860561239029796e-06,
+      "loss": 0.2403,
+      "step": 4238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.794705690268146,
+      "learning_rate": 9.860468822466168e-06,
+      "loss": 0.5007,
+      "step": 4239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6654160293783156,
+      "learning_rate": 9.860376375720348e-06,
+      "loss": 0.2006,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0955261127271227,
+      "learning_rate": 9.860283898792908e-06,
+      "loss": 0.3404,
+      "step": 4241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6811450888294097,
+      "learning_rate": 9.860191391684424e-06,
+      "loss": 0.3675,
+      "step": 4242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6782807502980237,
+      "learning_rate": 9.86009885439547e-06,
+      "loss": 0.1829,
+      "step": 4243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0406128166464486,
+      "learning_rate": 9.86000628692662e-06,
+      "loss": 0.2549,
+      "step": 4244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.340536026288558,
+      "learning_rate": 9.859913689278449e-06,
+      "loss": 0.2642,
+      "step": 4245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7434649231875834,
+      "learning_rate": 9.859821061451531e-06,
+      "loss": 0.2999,
+      "step": 4246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2774049806297003,
+      "learning_rate": 9.859728403446443e-06,
+      "loss": 0.6175,
+      "step": 4247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4669602045385197,
+      "learning_rate": 9.85963571526376e-06,
+      "loss": 0.278,
+      "step": 4248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7914801154398559,
+      "learning_rate": 9.859542996904058e-06,
+      "loss": 0.1493,
+      "step": 4249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6717227661486755,
+      "learning_rate": 9.85945024836791e-06,
+      "loss": 0.3887,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5412093936369549,
+      "learning_rate": 9.859357469655895e-06,
+      "loss": 0.2199,
+      "step": 4251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7694091923198192,
+      "learning_rate": 9.859264660768587e-06,
+      "loss": 0.2326,
+      "step": 4252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.46358452876138,
+      "learning_rate": 9.859171821706563e-06,
+      "loss": 0.5937,
+      "step": 4253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1054254361087095,
+      "learning_rate": 9.859078952470402e-06,
+      "loss": 0.4302,
+      "step": 4254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6463209458012087,
+      "learning_rate": 9.858986053060678e-06,
+      "loss": 0.293,
+      "step": 4255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6478184672160252,
+      "learning_rate": 9.858893123477967e-06,
+      "loss": 0.3284,
+      "step": 4256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5826812656785805,
+      "learning_rate": 9.858800163722849e-06,
+      "loss": 0.2834,
+      "step": 4257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.042553569553451,
+      "learning_rate": 9.858707173795899e-06,
+      "loss": 0.5318,
+      "step": 4258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.359827229485923,
+      "learning_rate": 9.858614153697695e-06,
+      "loss": 0.2384,
+      "step": 4259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6934246415844847,
+      "learning_rate": 9.858521103428816e-06,
+      "loss": 0.3179,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7062038821495893,
+      "learning_rate": 9.858428022989836e-06,
+      "loss": 0.2043,
+      "step": 4261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3260563391357945,
+      "learning_rate": 9.858334912381337e-06,
+      "loss": 0.1858,
+      "step": 4262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.100190735993516,
+      "learning_rate": 9.858241771603896e-06,
+      "loss": 0.4964,
+      "step": 4263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8422167088148076,
+      "learning_rate": 9.858148600658091e-06,
+      "loss": 0.2753,
+      "step": 4264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.837989188296862,
+      "learning_rate": 9.858055399544501e-06,
+      "loss": 0.5081,
+      "step": 4265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8080699895539936,
+      "learning_rate": 9.857962168263705e-06,
+      "loss": 0.3372,
+      "step": 4266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8254944141415321,
+      "learning_rate": 9.857868906816282e-06,
+      "loss": 0.2326,
+      "step": 4267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2596109517090028,
+      "learning_rate": 9.857775615202809e-06,
+      "loss": 0.1842,
+      "step": 4268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0207949111318633,
+      "learning_rate": 9.857682293423868e-06,
+      "loss": 0.1861,
+      "step": 4269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7036338820213677,
+      "learning_rate": 9.857588941480035e-06,
+      "loss": 0.2063,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.023213628218807,
+      "learning_rate": 9.857495559371895e-06,
+      "loss": 0.407,
+      "step": 4271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.137637190292217,
+      "learning_rate": 9.857402147100023e-06,
+      "loss": 0.6392,
+      "step": 4272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8408169167080022,
+      "learning_rate": 9.857308704665003e-06,
+      "loss": 0.4759,
+      "step": 4273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0587017458070718,
+      "learning_rate": 9.857215232067412e-06,
+      "loss": 0.4635,
+      "step": 4274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3797819839107361,
+      "learning_rate": 9.85712172930783e-06,
+      "loss": 0.1685,
+      "step": 4275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0881182454629923,
+      "learning_rate": 9.857028196386843e-06,
+      "loss": 0.2989,
+      "step": 4276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.832137470226494,
+      "learning_rate": 9.856934633305027e-06,
+      "loss": 0.7622,
+      "step": 4277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.597012444915321,
+      "learning_rate": 9.856841040062963e-06,
+      "loss": 0.2955,
+      "step": 4278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3276148072090443,
+      "learning_rate": 9.856747416661235e-06,
+      "loss": 0.2001,
+      "step": 4279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6626293072929712,
+      "learning_rate": 9.856653763100422e-06,
+      "loss": 0.3742,
+      "step": 4280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8000278427699203,
+      "learning_rate": 9.856560079381104e-06,
+      "loss": 0.2433,
+      "step": 4281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8669888378743391,
+      "learning_rate": 9.856466365503868e-06,
+      "loss": 0.4677,
+      "step": 4282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5564400461121926,
+      "learning_rate": 9.856372621469291e-06,
+      "loss": 0.2481,
+      "step": 4283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.194185239991819,
+      "learning_rate": 9.856278847277957e-06,
+      "loss": 0.7057,
+      "step": 4284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.789690199621485,
+      "learning_rate": 9.85618504293045e-06,
+      "loss": 0.2299,
+      "step": 4285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9093724809558192,
+      "learning_rate": 9.856091208427352e-06,
+      "loss": 0.2086,
+      "step": 4286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.634350245530288,
+      "learning_rate": 9.855997343769242e-06,
+      "loss": 0.9821,
+      "step": 4287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1442593400406427,
+      "learning_rate": 9.855903448956707e-06,
+      "loss": 0.2524,
+      "step": 4288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7199976792396092,
+      "learning_rate": 9.855809523990329e-06,
+      "loss": 0.2056,
+      "step": 4289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.178644376769929,
+      "learning_rate": 9.855715568870689e-06,
+      "loss": 0.385,
+      "step": 4290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8462882148969135,
+      "learning_rate": 9.855621583598374e-06,
+      "loss": 0.287,
+      "step": 4291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7530457009308542,
+      "learning_rate": 9.855527568173964e-06,
+      "loss": 0.2092,
+      "step": 4292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8525098558053044,
+      "learning_rate": 9.855433522598045e-06,
+      "loss": 0.519,
+      "step": 4293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8603742983572658,
+      "learning_rate": 9.855339446871202e-06,
+      "loss": 0.3595,
+      "step": 4294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6158282054411859,
+      "learning_rate": 9.855245340994018e-06,
+      "loss": 0.3277,
+      "step": 4295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.051927104721794,
+      "learning_rate": 9.855151204967077e-06,
+      "loss": 0.3821,
+      "step": 4296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7367999162124392,
+      "learning_rate": 9.855057038790962e-06,
+      "loss": 0.2506,
+      "step": 4297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9281663420339457,
+      "learning_rate": 9.85496284246626e-06,
+      "loss": 0.5425,
+      "step": 4298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5705959843484973,
+      "learning_rate": 9.854868615993557e-06,
+      "loss": 0.2389,
+      "step": 4299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0986102401634255,
+      "learning_rate": 9.854774359373436e-06,
+      "loss": 0.4259,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9043968137779543,
+      "learning_rate": 9.854680072606481e-06,
+      "loss": 0.2102,
+      "step": 4301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8801937470350227,
+      "learning_rate": 9.854585755693283e-06,
+      "loss": 0.4076,
+      "step": 4302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3810252092754718,
+      "learning_rate": 9.85449140863442e-06,
+      "loss": 0.2052,
+      "step": 4303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7087738038172173,
+      "learning_rate": 9.854397031430484e-06,
+      "loss": 0.4935,
+      "step": 4304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8543610167353817,
+      "learning_rate": 9.85430262408206e-06,
+      "loss": 0.4912,
+      "step": 4305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.817030214996135,
+      "learning_rate": 9.854208186589732e-06,
+      "loss": 0.4637,
+      "step": 4306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6101388646223256,
+      "learning_rate": 9.854113718954088e-06,
+      "loss": 0.3799,
+      "step": 4307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4540763606852323,
+      "learning_rate": 9.854019221175715e-06,
+      "loss": 0.3627,
+      "step": 4308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.157568176215653,
+      "learning_rate": 9.853924693255196e-06,
+      "loss": 0.3587,
+      "step": 4309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4177318932142675,
+      "learning_rate": 9.853830135193125e-06,
+      "loss": 0.3371,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1618653786585607,
+      "learning_rate": 9.853735546990083e-06,
+      "loss": 0.2243,
+      "step": 4311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2743183655452992,
+      "learning_rate": 9.85364092864666e-06,
+      "loss": 0.1682,
+      "step": 4312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.337606582283182,
+      "learning_rate": 9.853546280163444e-06,
+      "loss": 0.2394,
+      "step": 4313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.077017654010568,
+      "learning_rate": 9.853451601541023e-06,
+      "loss": 0.281,
+      "step": 4314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6604897242698795,
+      "learning_rate": 9.853356892779981e-06,
+      "loss": 0.4433,
+      "step": 4315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9350754772973489,
+      "learning_rate": 9.85326215388091e-06,
+      "loss": 0.3784,
+      "step": 4316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0345138114238677,
+      "learning_rate": 9.853167384844397e-06,
+      "loss": 0.4118,
+      "step": 4317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.531901993787722,
+      "learning_rate": 9.85307258567103e-06,
+      "loss": 0.3192,
+      "step": 4318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.865326708360709,
+      "learning_rate": 9.8529777563614e-06,
+      "loss": 0.3494,
+      "step": 4319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5480571477926015,
+      "learning_rate": 9.852882896916094e-06,
+      "loss": 0.2278,
+      "step": 4320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4328502993241292,
+      "learning_rate": 9.852788007335702e-06,
+      "loss": 0.3498,
+      "step": 4321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0325488302891905,
+      "learning_rate": 9.852693087620812e-06,
+      "loss": 0.414,
+      "step": 4322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.171038209136791,
+      "learning_rate": 9.852598137772014e-06,
+      "loss": 0.3661,
+      "step": 4323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0421272864363087,
+      "learning_rate": 9.852503157789896e-06,
+      "loss": 0.4061,
+      "step": 4324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.339781951591271,
+      "learning_rate": 9.852408147675052e-06,
+      "loss": 0.1776,
+      "step": 4325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6895907869646911,
+      "learning_rate": 9.852313107428068e-06,
+      "loss": 0.2671,
+      "step": 4326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5298180989049488,
+      "learning_rate": 9.852218037049536e-06,
+      "loss": 0.2916,
+      "step": 4327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.804870356754238,
+      "learning_rate": 9.852122936540046e-06,
+      "loss": 0.5251,
+      "step": 4328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8676936469127383,
+      "learning_rate": 9.85202780590019e-06,
+      "loss": 0.1926,
+      "step": 4329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7784732960188783,
+      "learning_rate": 9.851932645130553e-06,
+      "loss": 0.4073,
+      "step": 4330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0383585960887016,
+      "learning_rate": 9.851837454231734e-06,
+      "loss": 0.6209,
+      "step": 4331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5548327368560042,
+      "learning_rate": 9.851742233204317e-06,
+      "loss": 0.3978,
+      "step": 4332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7401135506520113,
+      "learning_rate": 9.8516469820489e-06,
+      "loss": 0.2671,
+      "step": 4333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.01489601148828,
+      "learning_rate": 9.85155170076607e-06,
+      "loss": 0.2301,
+      "step": 4334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6226636156365875,
+      "learning_rate": 9.85145638935642e-06,
+      "loss": 0.289,
+      "step": 4335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.34872578183758,
+      "learning_rate": 9.85136104782054e-06,
+      "loss": 0.211,
+      "step": 4336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6241888350850706,
+      "learning_rate": 9.851265676159025e-06,
+      "loss": 0.3843,
+      "step": 4337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7288118830532495,
+      "learning_rate": 9.851170274372465e-06,
+      "loss": 0.5326,
+      "step": 4338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4423036263023432,
+      "learning_rate": 9.851074842461455e-06,
+      "loss": 0.2072,
+      "step": 4339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.243544968452566,
+      "learning_rate": 9.850979380426585e-06,
+      "loss": 0.3967,
+      "step": 4340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1768023835894823,
+      "learning_rate": 9.850883888268447e-06,
+      "loss": 0.6553,
+      "step": 4341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0906502495511066,
+      "learning_rate": 9.850788365987635e-06,
+      "loss": 0.4655,
+      "step": 4342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6562150983012072,
+      "learning_rate": 9.850692813584746e-06,
+      "loss": 0.3754,
+      "step": 4343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6994593702054317,
+      "learning_rate": 9.85059723106037e-06,
+      "loss": 0.5401,
+      "step": 4344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9245278959252032,
+      "learning_rate": 9.850501618415098e-06,
+      "loss": 0.547,
+      "step": 4345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9130899761287787,
+      "learning_rate": 9.850405975649529e-06,
+      "loss": 0.5349,
+      "step": 4346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9415944044708195,
+      "learning_rate": 9.850310302764253e-06,
+      "loss": 0.5402,
+      "step": 4347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1553304549757093,
+      "learning_rate": 9.850214599759867e-06,
+      "loss": 0.5853,
+      "step": 4348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.439808026665956,
+      "learning_rate": 9.850118866636962e-06,
+      "loss": 0.2585,
+      "step": 4349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1129830613582474,
+      "learning_rate": 9.850023103396135e-06,
+      "loss": 0.245,
+      "step": 4350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3054211860680582,
+      "learning_rate": 9.849927310037978e-06,
+      "loss": 0.1995,
+      "step": 4351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3612520629015092,
+      "learning_rate": 9.84983148656309e-06,
+      "loss": 0.1653,
+      "step": 4352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0815671650471033,
+      "learning_rate": 9.849735632972064e-06,
+      "loss": 0.2251,
+      "step": 4353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.000451530582658,
+      "learning_rate": 9.849639749265494e-06,
+      "loss": 0.2818,
+      "step": 4354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8730757395854203,
+      "learning_rate": 9.849543835443978e-06,
+      "loss": 0.2107,
+      "step": 4355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6005838660645129,
+      "learning_rate": 9.849447891508107e-06,
+      "loss": 0.2196,
+      "step": 4356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5703396011974573,
+      "learning_rate": 9.849351917458482e-06,
+      "loss": 0.3813,
+      "step": 4357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6834530638840457,
+      "learning_rate": 9.849255913295697e-06,
+      "loss": 0.2363,
+      "step": 4358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.849721387554993,
+      "learning_rate": 9.849159879020347e-06,
+      "loss": 0.582,
+      "step": 4359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3907448577662362,
+      "learning_rate": 9.84906381463303e-06,
+      "loss": 0.1531,
+      "step": 4360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.25765508875189,
+      "learning_rate": 9.848967720134342e-06,
+      "loss": 0.4969,
+      "step": 4361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0016504896345273,
+      "learning_rate": 9.848871595524878e-06,
+      "loss": 0.2987,
+      "step": 4362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8238559215042718,
+      "learning_rate": 9.848775440805239e-06,
+      "loss": 0.38,
+      "step": 4363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.915419751290601,
+      "learning_rate": 9.848679255976017e-06,
+      "loss": 0.3337,
+      "step": 4364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9944822894854728,
+      "learning_rate": 9.848583041037812e-06,
+      "loss": 0.3752,
+      "step": 4365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8551352351754342,
+      "learning_rate": 9.848486795991222e-06,
+      "loss": 0.2629,
+      "step": 4366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7773701487666211,
+      "learning_rate": 9.848390520836844e-06,
+      "loss": 0.3057,
+      "step": 4367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1928158956928605,
+      "learning_rate": 9.848294215575276e-06,
+      "loss": 0.1944,
+      "step": 4368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.455441684235769,
+      "learning_rate": 9.848197880207115e-06,
+      "loss": 0.2194,
+      "step": 4369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6644148666203344,
+      "learning_rate": 9.848101514732961e-06,
+      "loss": 0.2645,
+      "step": 4370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6412016673434207,
+      "learning_rate": 9.84800511915341e-06,
+      "loss": 0.8585,
+      "step": 4371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9600060328991646,
+      "learning_rate": 9.847908693469063e-06,
+      "loss": 0.3112,
+      "step": 4372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.437609787563221,
+      "learning_rate": 9.847812237680518e-06,
+      "loss": 0.2268,
+      "step": 4373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8145560860549892,
+      "learning_rate": 9.847715751788375e-06,
+      "loss": 0.4959,
+      "step": 4374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8557686814276566,
+      "learning_rate": 9.847619235793231e-06,
+      "loss": 0.2061,
+      "step": 4375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5376670435949604,
+      "learning_rate": 9.847522689695685e-06,
+      "loss": 0.5687,
+      "step": 4376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9572564488508257,
+      "learning_rate": 9.84742611349634e-06,
+      "loss": 0.5728,
+      "step": 4377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4904722283907916,
+      "learning_rate": 9.847329507195791e-06,
+      "loss": 0.2529,
+      "step": 4378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9707480388623373,
+      "learning_rate": 9.847232870794643e-06,
+      "loss": 0.4656,
+      "step": 4379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7227570813484228,
+      "learning_rate": 9.84713620429349e-06,
+      "loss": 0.4052,
+      "step": 4380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0501756482354643,
+      "learning_rate": 9.847039507692938e-06,
+      "loss": 0.3515,
+      "step": 4381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.389077122088904,
+      "learning_rate": 9.846942780993585e-06,
+      "loss": 0.2419,
+      "step": 4382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2488508749088836,
+      "learning_rate": 9.846846024196033e-06,
+      "loss": 0.4394,
+      "step": 4383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8475504354940722,
+      "learning_rate": 9.84674923730088e-06,
+      "loss": 0.4,
+      "step": 4384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7058651168653232,
+      "learning_rate": 9.846652420308728e-06,
+      "loss": 0.1357,
+      "step": 4385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8005028300497479,
+      "learning_rate": 9.84655557322018e-06,
+      "loss": 0.4338,
+      "step": 4386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6654510972669063,
+      "learning_rate": 9.846458696035838e-06,
+      "loss": 0.4194,
+      "step": 4387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0329079170909234,
+      "learning_rate": 9.8463617887563e-06,
+      "loss": 0.3702,
+      "step": 4388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5844222367880096,
+      "learning_rate": 9.84626485138217e-06,
+      "loss": 0.2964,
+      "step": 4389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3129436766232243,
+      "learning_rate": 9.846167883914051e-06,
+      "loss": 0.2572,
+      "step": 4390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3977762369091018,
+      "learning_rate": 9.846070886352542e-06,
+      "loss": 0.1997,
+      "step": 4391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.655585560586302,
+      "learning_rate": 9.845973858698247e-06,
+      "loss": 0.2866,
+      "step": 4392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6705560851841132,
+      "learning_rate": 9.84587680095177e-06,
+      "loss": 0.3596,
+      "step": 4393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.551675939590702,
+      "learning_rate": 9.84577971311371e-06,
+      "loss": 0.2488,
+      "step": 4394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9604485878759297,
+      "learning_rate": 9.845682595184674e-06,
+      "loss": 0.5132,
+      "step": 4395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0436135764188013,
+      "learning_rate": 9.845585447165263e-06,
+      "loss": 0.6113,
+      "step": 4396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.69972433702206,
+      "learning_rate": 9.84548826905608e-06,
+      "loss": 0.438,
+      "step": 4397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2685208011415166,
+      "learning_rate": 9.84539106085773e-06,
+      "loss": 0.2451,
+      "step": 4398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5238488106691386,
+      "learning_rate": 9.845293822570815e-06,
+      "loss": 0.3423,
+      "step": 4399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9546377007950702,
+      "learning_rate": 9.84519655419594e-06,
+      "loss": 0.2511,
+      "step": 4400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3654113744839882,
+      "learning_rate": 9.845099255733708e-06,
+      "loss": 0.1638,
+      "step": 4401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.890222573603167,
+      "learning_rate": 9.845001927184723e-06,
+      "loss": 0.3016,
+      "step": 4402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1797630816518954,
+      "learning_rate": 9.844904568549592e-06,
+      "loss": 0.5977,
+      "step": 4403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6667755647686753,
+      "learning_rate": 9.844807179828915e-06,
+      "loss": 0.3818,
+      "step": 4404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7514719435178643,
+      "learning_rate": 9.844709761023301e-06,
+      "loss": 0.4461,
+      "step": 4405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5118584247991766,
+      "learning_rate": 9.844612312133352e-06,
+      "loss": 0.2535,
+      "step": 4406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3256079580181874,
+      "learning_rate": 9.844514833159676e-06,
+      "loss": 0.2109,
+      "step": 4407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.377157749467764,
+      "learning_rate": 9.844417324102876e-06,
+      "loss": 0.4637,
+      "step": 4408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0425877334330798,
+      "learning_rate": 9.844319784963559e-06,
+      "loss": 0.4123,
+      "step": 4409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.187003712896031,
+      "learning_rate": 9.844222215742329e-06,
+      "loss": 0.6581,
+      "step": 4410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9422168286779855,
+      "learning_rate": 9.844124616439795e-06,
+      "loss": 0.415,
+      "step": 4411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2910557955611062,
+      "learning_rate": 9.844026987056557e-06,
+      "loss": 0.2348,
+      "step": 4412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2833407871079674,
+      "learning_rate": 9.843929327593228e-06,
+      "loss": 0.2519,
+      "step": 4413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9126484221167444,
+      "learning_rate": 9.84383163805041e-06,
+      "loss": 0.6984,
+      "step": 4414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7703976667915682,
+      "learning_rate": 9.843733918428713e-06,
+      "loss": 0.233,
+      "step": 4415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8493548402399738,
+      "learning_rate": 9.843636168728741e-06,
+      "loss": 0.1959,
+      "step": 4416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7690766510585292,
+      "learning_rate": 9.843538388951101e-06,
+      "loss": 0.3041,
+      "step": 4417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.14819528774934,
+      "learning_rate": 9.843440579096405e-06,
+      "loss": 0.5622,
+      "step": 4418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8360100021656036,
+      "learning_rate": 9.84334273916525e-06,
+      "loss": 0.5345,
+      "step": 4419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.189693393506078,
+      "learning_rate": 9.843244869158255e-06,
+      "loss": 0.5659,
+      "step": 4420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9727263919710645,
+      "learning_rate": 9.843146969076022e-06,
+      "loss": 0.1872,
+      "step": 4421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7845243007376639,
+      "learning_rate": 9.843049038919161e-06,
+      "loss": 0.3509,
+      "step": 4422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.58720599966427,
+      "learning_rate": 9.842951078688278e-06,
+      "loss": 0.4795,
+      "step": 4423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.544037308740999,
+      "learning_rate": 9.842853088383982e-06,
+      "loss": 0.5968,
+      "step": 4424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3816941292993654,
+      "learning_rate": 9.842755068006881e-06,
+      "loss": 0.1548,
+      "step": 4425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6515929678539951,
+      "learning_rate": 9.842657017557586e-06,
+      "loss": 0.3146,
+      "step": 4426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4615578864153036,
+      "learning_rate": 9.842558937036703e-06,
+      "loss": 0.3066,
+      "step": 4427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3305951393594753,
+      "learning_rate": 9.842460826444842e-06,
+      "loss": 0.2067,
+      "step": 4428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8666870452060835,
+      "learning_rate": 9.842362685782614e-06,
+      "loss": 0.5208,
+      "step": 4429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0442315514955385,
+      "learning_rate": 9.842264515050625e-06,
+      "loss": 0.4396,
+      "step": 4430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9806009432823448,
+      "learning_rate": 9.84216631424949e-06,
+      "loss": 0.5389,
+      "step": 4431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7053819409730857,
+      "learning_rate": 9.842068083379812e-06,
+      "loss": 0.1887,
+      "step": 4432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6907119074640475,
+      "learning_rate": 9.841969822442207e-06,
+      "loss": 0.4263,
+      "step": 4433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.27638596606286,
+      "learning_rate": 9.84187153143728e-06,
+      "loss": 0.2311,
+      "step": 4434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.304514673630083,
+      "learning_rate": 9.841773210365646e-06,
+      "loss": 0.2744,
+      "step": 4435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2106008076970785,
+      "learning_rate": 9.841674859227913e-06,
+      "loss": 0.1806,
+      "step": 4436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2798665665964615,
+      "learning_rate": 9.841576478024692e-06,
+      "loss": 0.2931,
+      "step": 4437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0176911520895207,
+      "learning_rate": 9.841478066756595e-06,
+      "loss": 0.5869,
+      "step": 4438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5435825334286806,
+      "learning_rate": 9.841379625424232e-06,
+      "loss": 0.3053,
+      "step": 4439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3039791311674103,
+      "learning_rate": 9.841281154028213e-06,
+      "loss": 0.4819,
+      "step": 4440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9239295661633062,
+      "learning_rate": 9.841182652569152e-06,
+      "loss": 0.4786,
+      "step": 4441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1302712430424178,
+      "learning_rate": 9.84108412104766e-06,
+      "loss": 0.5223,
+      "step": 4442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.550328440878724,
+      "learning_rate": 9.840985559464349e-06,
+      "loss": 0.3031,
+      "step": 4443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2651555133951793,
+      "learning_rate": 9.84088696781983e-06,
+      "loss": 0.4949,
+      "step": 4444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3421198898135898,
+      "learning_rate": 9.840788346114716e-06,
+      "loss": 0.4564,
+      "step": 4445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.857946743899389,
+      "learning_rate": 9.840689694349618e-06,
+      "loss": 0.1614,
+      "step": 4446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5948716358696768,
+      "learning_rate": 9.840591012525151e-06,
+      "loss": 0.2093,
+      "step": 4447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9656936142291377,
+      "learning_rate": 9.840492300641927e-06,
+      "loss": 0.3549,
+      "step": 4448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.717413647193845,
+      "learning_rate": 9.840393558700558e-06,
+      "loss": 0.4236,
+      "step": 4449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2469099833931137,
+      "learning_rate": 9.840294786701657e-06,
+      "loss": 0.607,
+      "step": 4450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8773960698362235,
+      "learning_rate": 9.840195984645839e-06,
+      "loss": 0.3921,
+      "step": 4451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6507323908441762,
+      "learning_rate": 9.840097152533716e-06,
+      "loss": 0.3205,
+      "step": 4452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5584218879347607,
+      "learning_rate": 9.839998290365903e-06,
+      "loss": 0.282,
+      "step": 4453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3613593491476874,
+      "learning_rate": 9.839899398143013e-06,
+      "loss": 0.4598,
+      "step": 4454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.383447442054569,
+      "learning_rate": 9.839800475865661e-06,
+      "loss": 0.1508,
+      "step": 4455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7100257599818607,
+      "learning_rate": 9.839701523534459e-06,
+      "loss": 0.1936,
+      "step": 4456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8884952190352287,
+      "learning_rate": 9.839602541150026e-06,
+      "loss": 0.3122,
+      "step": 4457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6541429693516854,
+      "learning_rate": 9.839503528712971e-06,
+      "loss": 0.3607,
+      "step": 4458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.242434564462914,
+      "learning_rate": 9.839404486223914e-06,
+      "loss": 0.4104,
+      "step": 4459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6282708143004716,
+      "learning_rate": 9.839305413683466e-06,
+      "loss": 0.3984,
+      "step": 4460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7653715681808306,
+      "learning_rate": 9.839206311092244e-06,
+      "loss": 0.1699,
+      "step": 4461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3161500592911977,
+      "learning_rate": 9.839107178450861e-06,
+      "loss": 0.6976,
+      "step": 4462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.200057572727717,
+      "learning_rate": 9.83900801575994e-06,
+      "loss": 0.7828,
+      "step": 4463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1228404361476447,
+      "learning_rate": 9.838908823020089e-06,
+      "loss": 0.6619,
+      "step": 4464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3899391225120326,
+      "learning_rate": 9.838809600231925e-06,
+      "loss": 0.5669,
+      "step": 4465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9342117786104647,
+      "learning_rate": 9.838710347396067e-06,
+      "loss": 0.4871,
+      "step": 4466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8336916934747203,
+      "learning_rate": 9.83861106451313e-06,
+      "loss": 0.2548,
+      "step": 4467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.359957597567648,
+      "learning_rate": 9.838511751583731e-06,
+      "loss": 0.2964,
+      "step": 4468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6724997093074985,
+      "learning_rate": 9.838412408608485e-06,
+      "loss": 0.2068,
+      "step": 4469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6958639592803655,
+      "learning_rate": 9.838313035588012e-06,
+      "loss": 0.2586,
+      "step": 4470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.455901941926364,
+      "learning_rate": 9.838213632522926e-06,
+      "loss": 0.1844,
+      "step": 4471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8431057491377263,
+      "learning_rate": 9.838114199413846e-06,
+      "loss": 0.3922,
+      "step": 4472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8980990371084063,
+      "learning_rate": 9.838014736261388e-06,
+      "loss": 0.58,
+      "step": 4473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2037185746692953,
+      "learning_rate": 9.837915243066173e-06,
+      "loss": 0.22,
+      "step": 4474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.801286239732663,
+      "learning_rate": 9.837815719828814e-06,
+      "loss": 0.3507,
+      "step": 4475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2417116820819578,
+      "learning_rate": 9.837716166549931e-06,
+      "loss": 0.2346,
+      "step": 4476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9870911392086057,
+      "learning_rate": 9.837616583230146e-06,
+      "loss": 0.287,
+      "step": 4477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5374747057855802,
+      "learning_rate": 9.83751696987007e-06,
+      "loss": 0.2207,
+      "step": 4478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.694077096565524,
+      "learning_rate": 9.837417326470328e-06,
+      "loss": 0.461,
+      "step": 4479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8544678809494957,
+      "learning_rate": 9.837317653031534e-06,
+      "loss": 0.4355,
+      "step": 4480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5351540878513632,
+      "learning_rate": 9.837217949554313e-06,
+      "loss": 0.3094,
+      "step": 4481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0036298029467976,
+      "learning_rate": 9.837118216039279e-06,
+      "loss": 0.2904,
+      "step": 4482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5373508210421338,
+      "learning_rate": 9.837018452487052e-06,
+      "loss": 0.3774,
+      "step": 4483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.610003190175661,
+      "learning_rate": 9.836918658898253e-06,
+      "loss": 0.5909,
+      "step": 4484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7429465107373152,
+      "learning_rate": 9.8368188352735e-06,
+      "loss": 0.3217,
+      "step": 4485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1506398876803443,
+      "learning_rate": 9.836718981613414e-06,
+      "loss": 0.306,
+      "step": 4486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2274896054862057,
+      "learning_rate": 9.836619097918617e-06,
+      "loss": 0.1879,
+      "step": 4487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.380731126152422,
+      "learning_rate": 9.836519184189723e-06,
+      "loss": 0.3054,
+      "step": 4488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.276889662713276,
+      "learning_rate": 9.83641924042736e-06,
+      "loss": 0.2184,
+      "step": 4489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4917097125644214,
+      "learning_rate": 9.836319266632144e-06,
+      "loss": 0.2386,
+      "step": 4490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.081561425163239,
+      "learning_rate": 9.836219262804695e-06,
+      "loss": 0.2341,
+      "step": 4491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7003175799969263,
+      "learning_rate": 9.836119228945639e-06,
+      "loss": 0.4877,
+      "step": 4492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.596535774184786,
+      "learning_rate": 9.836019165055595e-06,
+      "loss": 0.248,
+      "step": 4493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4045833754985648,
+      "learning_rate": 9.835919071135181e-06,
+      "loss": 0.1877,
+      "step": 4494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5994035264661675,
+      "learning_rate": 9.835818947185023e-06,
+      "loss": 0.5097,
+      "step": 4495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0622033137006923,
+      "learning_rate": 9.835718793205739e-06,
+      "loss": 0.5257,
+      "step": 4496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.738051817250758,
+      "learning_rate": 9.835618609197953e-06,
+      "loss": 0.3319,
+      "step": 4497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0862653417520307,
+      "learning_rate": 9.835518395162286e-06,
+      "loss": 0.4987,
+      "step": 4498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7780365107328238,
+      "learning_rate": 9.835418151099365e-06,
+      "loss": 0.2179,
+      "step": 4499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6967417278256764,
+      "learning_rate": 9.835317877009806e-06,
+      "loss": 0.4068,
+      "step": 4500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.003132018048732,
+      "learning_rate": 9.835217572894234e-06,
+      "loss": 0.4552,
+      "step": 4501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2085409081867393,
+      "learning_rate": 9.835117238753273e-06,
+      "loss": 0.4127,
+      "step": 4502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5293015690679115,
+      "learning_rate": 9.835016874587546e-06,
+      "loss": 0.1874,
+      "step": 4503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.897482711384346,
+      "learning_rate": 9.834916480397674e-06,
+      "loss": 0.2343,
+      "step": 4504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3319348806608131,
+      "learning_rate": 9.834816056184283e-06,
+      "loss": 0.1514,
+      "step": 4505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.435153475898433,
+      "learning_rate": 9.834715601947994e-06,
+      "loss": 0.3701,
+      "step": 4506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.22738088993617,
+      "learning_rate": 9.834615117689434e-06,
+      "loss": 0.385,
+      "step": 4507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.759242230821156,
+      "learning_rate": 9.834514603409225e-06,
+      "loss": 0.2158,
+      "step": 4508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2749872504443012,
+      "learning_rate": 9.834414059107992e-06,
+      "loss": 0.2225,
+      "step": 4509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8990981989159348,
+      "learning_rate": 9.834313484786357e-06,
+      "loss": 0.513,
+      "step": 4510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.673555443184386,
+      "learning_rate": 9.834212880444949e-06,
+      "loss": 0.3606,
+      "step": 4511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.229453502157323,
+      "learning_rate": 9.834112246084388e-06,
+      "loss": 0.3733,
+      "step": 4512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7444028646545915,
+      "learning_rate": 9.834011581705302e-06,
+      "loss": 0.2469,
+      "step": 4513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.588323897667798,
+      "learning_rate": 9.833910887308315e-06,
+      "loss": 0.29,
+      "step": 4514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9805340502204363,
+      "learning_rate": 9.833810162894051e-06,
+      "loss": 0.3714,
+      "step": 4515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7544691461185868,
+      "learning_rate": 9.83370940846314e-06,
+      "loss": 0.462,
+      "step": 4516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3086721482207457,
+      "learning_rate": 9.833608624016201e-06,
+      "loss": 0.1533,
+      "step": 4517
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3795560029980802,
+      "learning_rate": 9.833507809553864e-06,
+      "loss": 0.3316,
+      "step": 4518
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8915971156251676,
+      "learning_rate": 9.833406965076754e-06,
+      "loss": 0.1858,
+      "step": 4519
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7096031899670014,
+      "learning_rate": 9.8333060905855e-06,
+      "loss": 0.2342,
+      "step": 4520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.730945048001539,
+      "learning_rate": 9.833205186080726e-06,
+      "loss": 0.2025,
+      "step": 4521
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.679308797594771,
+      "learning_rate": 9.833104251563058e-06,
+      "loss": 0.4091,
+      "step": 4522
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0993755370045983,
+      "learning_rate": 9.833003287033122e-06,
+      "loss": 0.3021,
+      "step": 4523
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7441872593710068,
+      "learning_rate": 9.832902292491547e-06,
+      "loss": 0.3376,
+      "step": 4524
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6640633372631564,
+      "learning_rate": 9.83280126793896e-06,
+      "loss": 0.3387,
+      "step": 4525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5823682708625781,
+      "learning_rate": 9.832700213375987e-06,
+      "loss": 0.2889,
+      "step": 4526
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4104156255600813,
+      "learning_rate": 9.832599128803257e-06,
+      "loss": 0.3385,
+      "step": 4527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2398004884936125,
+      "learning_rate": 9.832498014221396e-06,
+      "loss": 0.8185,
+      "step": 4528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.838338231907068,
+      "learning_rate": 9.832396869631034e-06,
+      "loss": 0.6274,
+      "step": 4529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.29546126974056,
+      "learning_rate": 9.832295695032799e-06,
+      "loss": 0.2102,
+      "step": 4530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0476481528123887,
+      "learning_rate": 9.832194490427315e-06,
+      "loss": 0.5836,
+      "step": 4531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7122756306343339,
+      "learning_rate": 9.832093255815217e-06,
+      "loss": 0.3963,
+      "step": 4532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.659408827435084,
+      "learning_rate": 9.83199199119713e-06,
+      "loss": 0.2407,
+      "step": 4533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3381715636012985,
+      "learning_rate": 9.831890696573683e-06,
+      "loss": 0.3111,
+      "step": 4534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6497970348752853,
+      "learning_rate": 9.831789371945504e-06,
+      "loss": 0.2742,
+      "step": 4535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.500638254791087,
+      "learning_rate": 9.831688017313223e-06,
+      "loss": 0.2048,
+      "step": 4536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4129561085827838,
+      "learning_rate": 9.831586632677473e-06,
+      "loss": 0.1377,
+      "step": 4537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.833847308902184,
+      "learning_rate": 9.831485218038877e-06,
+      "loss": 0.2087,
+      "step": 4538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2830057268130102,
+      "learning_rate": 9.831383773398068e-06,
+      "loss": 0.2115,
+      "step": 4539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.842900733210103,
+      "learning_rate": 9.831282298755679e-06,
+      "loss": 0.1998,
+      "step": 4540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6470430948873591,
+      "learning_rate": 9.831180794112336e-06,
+      "loss": 0.3049,
+      "step": 4541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6678461011868702,
+      "learning_rate": 9.83107925946867e-06,
+      "loss": 0.3642,
+      "step": 4542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8958372667991439,
+      "learning_rate": 9.830977694825313e-06,
+      "loss": 0.4495,
+      "step": 4543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.873977441904108,
+      "learning_rate": 9.830876100182893e-06,
+      "loss": 0.2093,
+      "step": 4544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6704600398363145,
+      "learning_rate": 9.830774475542045e-06,
+      "loss": 0.2984,
+      "step": 4545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.942338397331924,
+      "learning_rate": 9.830672820903396e-06,
+      "loss": 0.394,
+      "step": 4546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2641288645336126,
+      "learning_rate": 9.830571136267578e-06,
+      "loss": 0.1743,
+      "step": 4547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3177585419474067,
+      "learning_rate": 9.830469421635224e-06,
+      "loss": 0.1812,
+      "step": 4548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0419934900791388,
+      "learning_rate": 9.830367677006966e-06,
+      "loss": 0.2623,
+      "step": 4549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.301548677637324,
+      "learning_rate": 9.830265902383434e-06,
+      "loss": 0.2888,
+      "step": 4550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.657541891297927,
+      "learning_rate": 9.83016409776526e-06,
+      "loss": 0.4069,
+      "step": 4551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3631224546612057,
+      "learning_rate": 9.83006226315308e-06,
+      "loss": 0.163,
+      "step": 4552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7126339814755447,
+      "learning_rate": 9.829960398547521e-06,
+      "loss": 0.367,
+      "step": 4553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2026238919939374,
+      "learning_rate": 9.829858503949217e-06,
+      "loss": 0.4687,
+      "step": 4554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2975020882552393,
+      "learning_rate": 9.829756579358803e-06,
+      "loss": 0.319,
+      "step": 4555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.682368294965615,
+      "learning_rate": 9.82965462477691e-06,
+      "loss": 0.3972,
+      "step": 4556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2596470926201557,
+      "learning_rate": 9.829552640204171e-06,
+      "loss": 0.559,
+      "step": 4557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.900129839419771,
+      "learning_rate": 9.829450625641221e-06,
+      "loss": 0.5359,
+      "step": 4558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2719821269766203,
+      "learning_rate": 9.829348581088692e-06,
+      "loss": 0.5161,
+      "step": 4559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0144743333328496,
+      "learning_rate": 9.829246506547219e-06,
+      "loss": 0.2974,
+      "step": 4560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4346119567560354,
+      "learning_rate": 9.829144402017435e-06,
+      "loss": 0.2704,
+      "step": 4561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7001095727711832,
+      "learning_rate": 9.829042267499971e-06,
+      "loss": 0.4449,
+      "step": 4562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.164375888428821,
+      "learning_rate": 9.828940102995467e-06,
+      "loss": 0.5872,
+      "step": 4563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6662919725372605,
+      "learning_rate": 9.828837908504553e-06,
+      "loss": 0.2073,
+      "step": 4564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.562658559383444,
+      "learning_rate": 9.828735684027867e-06,
+      "loss": 0.4123,
+      "step": 4565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8182026009834862,
+      "learning_rate": 9.828633429566042e-06,
+      "loss": 0.4801,
+      "step": 4566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7893722507576937,
+      "learning_rate": 9.828531145119711e-06,
+      "loss": 0.3539,
+      "step": 4567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7310302869524703,
+      "learning_rate": 9.828428830689511e-06,
+      "loss": 0.3259,
+      "step": 4568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7819987509413504,
+      "learning_rate": 9.82832648627608e-06,
+      "loss": 0.2807,
+      "step": 4569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2530549240258113,
+      "learning_rate": 9.82822411188005e-06,
+      "loss": 0.1182,
+      "step": 4570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5985843975027338,
+      "learning_rate": 9.828121707502056e-06,
+      "loss": 0.2305,
+      "step": 4571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4059777658593193,
+      "learning_rate": 9.828019273142737e-06,
+      "loss": 0.3002,
+      "step": 4572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.10179229458479,
+      "learning_rate": 9.827916808802728e-06,
+      "loss": 0.4431,
+      "step": 4573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9586807321881197,
+      "learning_rate": 9.827814314482665e-06,
+      "loss": 0.1465,
+      "step": 4574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5873457007626617,
+      "learning_rate": 9.827711790183185e-06,
+      "loss": 0.3948,
+      "step": 4575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3698382313663386,
+      "learning_rate": 9.827609235904922e-06,
+      "loss": 0.5429,
+      "step": 4576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0521459943060285,
+      "learning_rate": 9.827506651648516e-06,
+      "loss": 0.3906,
+      "step": 4577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.253311415728855,
+      "learning_rate": 9.827404037414603e-06,
+      "loss": 0.2386,
+      "step": 4578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9310652837407591,
+      "learning_rate": 9.82730139320382e-06,
+      "loss": 0.4086,
+      "step": 4579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.061658351572914,
+      "learning_rate": 9.827198719016807e-06,
+      "loss": 0.275,
+      "step": 4580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7646933182042375,
+      "learning_rate": 9.827096014854197e-06,
+      "loss": 0.3795,
+      "step": 4581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1056319016071465,
+      "learning_rate": 9.826993280716627e-06,
+      "loss": 0.4366,
+      "step": 4582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.492202743732022,
+      "learning_rate": 9.826890516604742e-06,
+      "loss": 0.2758,
+      "step": 4583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.119901065362281,
+      "learning_rate": 9.826787722519175e-06,
+      "loss": 0.6023,
+      "step": 4584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.426052378650704,
+      "learning_rate": 9.826684898460566e-06,
+      "loss": 0.219,
+      "step": 4585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6343991724564413,
+      "learning_rate": 9.82658204442955e-06,
+      "loss": 0.3378,
+      "step": 4586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1774315456706543,
+      "learning_rate": 9.826479160426772e-06,
+      "loss": 0.5289,
+      "step": 4587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1322478929528246,
+      "learning_rate": 9.826376246452865e-06,
+      "loss": 0.3552,
+      "step": 4588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3210386530950025,
+      "learning_rate": 9.826273302508472e-06,
+      "loss": 0.1495,
+      "step": 4589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9653973946453058,
+      "learning_rate": 9.826170328594228e-06,
+      "loss": 0.3486,
+      "step": 4590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4157957340379905,
+      "learning_rate": 9.826067324710777e-06,
+      "loss": 0.1918,
+      "step": 4591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6372716830054566,
+      "learning_rate": 9.825964290858758e-06,
+      "loss": 0.3162,
+      "step": 4592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4228829306428632,
+      "learning_rate": 9.825861227038808e-06,
+      "loss": 0.3696,
+      "step": 4593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0541009972413677,
+      "learning_rate": 9.82575813325157e-06,
+      "loss": 0.4095,
+      "step": 4594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9684758790231318,
+      "learning_rate": 9.82565500949768e-06,
+      "loss": 0.1965,
+      "step": 4595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4399124702598978,
+      "learning_rate": 9.825551855777784e-06,
+      "loss": 0.7065,
+      "step": 4596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7217557287706626,
+      "learning_rate": 9.82544867209252e-06,
+      "loss": 0.4395,
+      "step": 4597
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0323659399027614,
+      "learning_rate": 9.825345458442526e-06,
+      "loss": 0.7702,
+      "step": 4598
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8067086758555821,
+      "learning_rate": 9.825242214828447e-06,
+      "loss": 0.2896,
+      "step": 4599
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1206265848534733,
+      "learning_rate": 9.825138941250922e-06,
+      "loss": 0.1511,
+      "step": 4600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0414641403124474,
+      "learning_rate": 9.825035637710592e-06,
+      "loss": 0.5488,
+      "step": 4601
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9087379978804353,
+      "learning_rate": 9.8249323042081e-06,
+      "loss": 0.2363,
+      "step": 4602
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3093897334455726,
+      "learning_rate": 9.824828940744086e-06,
+      "loss": 0.1584,
+      "step": 4603
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4283114922437776,
+      "learning_rate": 9.824725547319195e-06,
+      "loss": 0.2724,
+      "step": 4604
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2997367703073555,
+      "learning_rate": 9.824622123934065e-06,
+      "loss": 0.5717,
+      "step": 4605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5948349728384343,
+      "learning_rate": 9.824518670589342e-06,
+      "loss": 0.4309,
+      "step": 4606
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8689608392779093,
+      "learning_rate": 9.824415187285666e-06,
+      "loss": 0.4204,
+      "step": 4607
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.83232921137539,
+      "learning_rate": 9.82431167402368e-06,
+      "loss": 0.5347,
+      "step": 4608
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0959270931774439,
+      "learning_rate": 9.824208130804025e-06,
+      "loss": 0.1884,
+      "step": 4609
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7774593776777103,
+      "learning_rate": 9.824104557627348e-06,
+      "loss": 0.4762,
+      "step": 4610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7979898767348104,
+      "learning_rate": 9.824000954494289e-06,
+      "loss": 0.4192,
+      "step": 4611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6719558910159653,
+      "learning_rate": 9.82389732140549e-06,
+      "loss": 0.399,
+      "step": 4612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6221124780216714,
+      "learning_rate": 9.823793658361601e-06,
+      "loss": 0.231,
+      "step": 4613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9827276604751134,
+      "learning_rate": 9.82368996536326e-06,
+      "loss": 0.2355,
+      "step": 4614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.576337044072285,
+      "learning_rate": 9.823586242411113e-06,
+      "loss": 0.2822,
+      "step": 4615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7632099488749293,
+      "learning_rate": 9.823482489505803e-06,
+      "loss": 0.2972,
+      "step": 4616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5219806123879864,
+      "learning_rate": 9.823378706647975e-06,
+      "loss": 0.3287,
+      "step": 4617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.16725855011595,
+      "learning_rate": 9.823274893838274e-06,
+      "loss": 0.4872,
+      "step": 4618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7683854298560764,
+      "learning_rate": 9.823171051077344e-06,
+      "loss": 0.5107,
+      "step": 4619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2566018290366607,
+      "learning_rate": 9.823067178365829e-06,
+      "loss": 0.1804,
+      "step": 4620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.630976476291293,
+      "learning_rate": 9.822963275704375e-06,
+      "loss": 0.2019,
+      "step": 4621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8168115102227276,
+      "learning_rate": 9.822859343093628e-06,
+      "loss": 0.4481,
+      "step": 4622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9074336895097321,
+      "learning_rate": 9.822755380534233e-06,
+      "loss": 0.5954,
+      "step": 4623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7275013809019166,
+      "learning_rate": 9.822651388026833e-06,
+      "loss": 0.4917,
+      "step": 4624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.859608923999296,
+      "learning_rate": 9.822547365572078e-06,
+      "loss": 0.3484,
+      "step": 4625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2274799173350133,
+      "learning_rate": 9.822443313170609e-06,
+      "loss": 0.1806,
+      "step": 4626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.05601239005532,
+      "learning_rate": 9.822339230823077e-06,
+      "loss": 0.4107,
+      "step": 4627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8022824399761406,
+      "learning_rate": 9.822235118530126e-06,
+      "loss": 0.4142,
+      "step": 4628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3678251509727417,
+      "learning_rate": 9.822130976292402e-06,
+      "loss": 0.2237,
+      "step": 4629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9474111825916733,
+      "learning_rate": 9.822026804110553e-06,
+      "loss": 0.3971,
+      "step": 4630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1906578023416383,
+      "learning_rate": 9.821922601985225e-06,
+      "loss": 0.4414,
+      "step": 4631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6476155386839846,
+      "learning_rate": 9.821818369917066e-06,
+      "loss": 0.6308,
+      "step": 4632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6646389164395543,
+      "learning_rate": 9.821714107906722e-06,
+      "loss": 0.1988,
+      "step": 4633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6384299535030045,
+      "learning_rate": 9.821609815954841e-06,
+      "loss": 0.2944,
+      "step": 4634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8579898451208348,
+      "learning_rate": 9.821505494062071e-06,
+      "loss": 0.2263,
+      "step": 4635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1637154724389815,
+      "learning_rate": 9.82140114222906e-06,
+      "loss": 0.5318,
+      "step": 4636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5930857840366008,
+      "learning_rate": 9.821296760456455e-06,
+      "loss": 0.2064,
+      "step": 4637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9330870031759764,
+      "learning_rate": 9.821192348744904e-06,
+      "loss": 0.3136,
+      "step": 4638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.807675476618249,
+      "learning_rate": 9.821087907095057e-06,
+      "loss": 0.5333,
+      "step": 4639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3189239197437566,
+      "learning_rate": 9.820983435507561e-06,
+      "loss": 0.2749,
+      "step": 4640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4688435471449799,
+      "learning_rate": 9.820878933983067e-06,
+      "loss": 0.2332,
+      "step": 4641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.921573981692656,
+      "learning_rate": 9.820774402522221e-06,
+      "loss": 0.2091,
+      "step": 4642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8547281011750705,
+      "learning_rate": 9.820669841125674e-06,
+      "loss": 0.2018,
+      "step": 4643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.33170051721465,
+      "learning_rate": 9.820565249794075e-06,
+      "loss": 0.5477,
+      "step": 4644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.071391720455743,
+      "learning_rate": 9.82046062852807e-06,
+      "loss": 0.434,
+      "step": 4645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.179866593704473,
+      "learning_rate": 9.820355977328316e-06,
+      "loss": 0.3314,
+      "step": 4646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.822483622113696,
+      "learning_rate": 9.820251296195456e-06,
+      "loss": 0.3025,
+      "step": 4647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3752213646688798,
+      "learning_rate": 9.820146585130143e-06,
+      "loss": 0.1932,
+      "step": 4648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.970508259570823,
+      "learning_rate": 9.820041844133028e-06,
+      "loss": 0.5499,
+      "step": 4649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.920288437071003,
+      "learning_rate": 9.81993707320476e-06,
+      "loss": 0.2794,
+      "step": 4650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.171032166577066,
+      "learning_rate": 9.81983227234599e-06,
+      "loss": 0.243,
+      "step": 4651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.586444603951267,
+      "learning_rate": 9.81972744155737e-06,
+      "loss": 0.1835,
+      "step": 4652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6139582305789337,
+      "learning_rate": 9.819622580839547e-06,
+      "loss": 0.3388,
+      "step": 4653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8127962684200594,
+      "learning_rate": 9.819517690193177e-06,
+      "loss": 0.2396,
+      "step": 4654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0679981755611996,
+      "learning_rate": 9.819412769618907e-06,
+      "loss": 0.4084,
+      "step": 4655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9124213208738066,
+      "learning_rate": 9.819307819117394e-06,
+      "loss": 0.4805,
+      "step": 4656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.860143802085283,
+      "learning_rate": 9.819202838689284e-06,
+      "loss": 0.2742,
+      "step": 4657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9163318338109059,
+      "learning_rate": 9.819097828335233e-06,
+      "loss": 0.6341,
+      "step": 4658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.407635874787608,
+      "learning_rate": 9.81899278805589e-06,
+      "loss": 0.1897,
+      "step": 4659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7154240983926035,
+      "learning_rate": 9.81888771785191e-06,
+      "loss": 0.3462,
+      "step": 4660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.209421183792707,
+      "learning_rate": 9.818782617723943e-06,
+      "loss": 0.2017,
+      "step": 4661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.661306794848318,
+      "learning_rate": 9.818677487672645e-06,
+      "loss": 0.3648,
+      "step": 4662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5480346595461245,
+      "learning_rate": 9.818572327698665e-06,
+      "loss": 0.2683,
+      "step": 4663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6247783885743743,
+      "learning_rate": 9.818467137802658e-06,
+      "loss": 0.3899,
+      "step": 4664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4594636204051534,
+      "learning_rate": 9.818361917985274e-06,
+      "loss": 0.1775,
+      "step": 4665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.622481969134091,
+      "learning_rate": 9.818256668247174e-06,
+      "loss": 0.5774,
+      "step": 4666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.13759178165345,
+      "learning_rate": 9.818151388589005e-06,
+      "loss": 0.283,
+      "step": 4667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.262398163139593,
+      "learning_rate": 9.818046079011424e-06,
+      "loss": 0.2593,
+      "step": 4668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.793146905795158,
+      "learning_rate": 9.81794073951508e-06,
+      "loss": 0.3338,
+      "step": 4669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7316938459641014,
+      "learning_rate": 9.817835370100633e-06,
+      "loss": 0.4464,
+      "step": 4670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0055762381776328,
+      "learning_rate": 9.817729970768735e-06,
+      "loss": 0.27,
+      "step": 4671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8802106061236594,
+      "learning_rate": 9.817624541520038e-06,
+      "loss": 0.4941,
+      "step": 4672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1655015438568563,
+      "learning_rate": 9.817519082355203e-06,
+      "loss": 0.2571,
+      "step": 4673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.327583288368135,
+      "learning_rate": 9.817413593274879e-06,
+      "loss": 0.638,
+      "step": 4674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.524146225553579,
+      "learning_rate": 9.817308074279723e-06,
+      "loss": 0.6309,
+      "step": 4675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5369887480366575,
+      "learning_rate": 9.817202525370392e-06,
+      "loss": 0.3312,
+      "step": 4676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9516068967376496,
+      "learning_rate": 9.817096946547537e-06,
+      "loss": 0.3294,
+      "step": 4677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1471187909321858,
+      "learning_rate": 9.816991337811818e-06,
+      "loss": 0.1648,
+      "step": 4678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5776543375706198,
+      "learning_rate": 9.81688569916389e-06,
+      "loss": 0.1894,
+      "step": 4679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8909736313832926,
+      "learning_rate": 9.816780030604408e-06,
+      "loss": 0.2423,
+      "step": 4680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7076873064179596,
+      "learning_rate": 9.816674332134027e-06,
+      "loss": 0.2634,
+      "step": 4681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8639216322243755,
+      "learning_rate": 9.816568603753405e-06,
+      "loss": 0.2426,
+      "step": 4682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7634778512748683,
+      "learning_rate": 9.8164628454632e-06,
+      "loss": 0.4981,
+      "step": 4683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4901390270163026,
+      "learning_rate": 9.816357057264066e-06,
+      "loss": 0.3581,
+      "step": 4684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6499032243250438,
+      "learning_rate": 9.81625123915666e-06,
+      "loss": 0.2913,
+      "step": 4685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8132765191775386,
+      "learning_rate": 9.81614539114164e-06,
+      "loss": 0.2397,
+      "step": 4686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9414842731908222,
+      "learning_rate": 9.816039513219664e-06,
+      "loss": 0.3948,
+      "step": 4687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7179106353395137,
+      "learning_rate": 9.81593360539139e-06,
+      "loss": 0.3513,
+      "step": 4688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.007057058956146,
+      "learning_rate": 9.815827667657472e-06,
+      "loss": 0.3515,
+      "step": 4689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5227213753639597,
+      "learning_rate": 9.815721700018572e-06,
+      "loss": 0.2093,
+      "step": 4690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8008563197742469,
+      "learning_rate": 9.815615702475346e-06,
+      "loss": 0.2573,
+      "step": 4691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7195098671110083,
+      "learning_rate": 9.815509675028451e-06,
+      "loss": 0.4043,
+      "step": 4692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1632410273699256,
+      "learning_rate": 9.815403617678549e-06,
+      "loss": 0.4214,
+      "step": 4693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5185674672211147,
+      "learning_rate": 9.815297530426295e-06,
+      "loss": 0.2176,
+      "step": 4694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7288544101932883,
+      "learning_rate": 9.815191413272351e-06,
+      "loss": 0.512,
+      "step": 4695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0728783380297173,
+      "learning_rate": 9.815085266217374e-06,
+      "loss": 0.5087,
+      "step": 4696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6027737619393607,
+      "learning_rate": 9.814979089262022e-06,
+      "loss": 0.2767,
+      "step": 4697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6732653959447417,
+      "learning_rate": 9.814872882406956e-06,
+      "loss": 0.3695,
+      "step": 4698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.027755375788094,
+      "learning_rate": 9.814766645652836e-06,
+      "loss": 0.3049,
+      "step": 4699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6516635793468764,
+      "learning_rate": 9.81466037900032e-06,
+      "loss": 0.3034,
+      "step": 4700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.413603254417938,
+      "learning_rate": 9.81455408245007e-06,
+      "loss": 0.8957,
+      "step": 4701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2704346281993213,
+      "learning_rate": 9.814447756002744e-06,
+      "loss": 0.7486,
+      "step": 4702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2180154047752234,
+      "learning_rate": 9.814341399659004e-06,
+      "loss": 0.3506,
+      "step": 4703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0888873576812874,
+      "learning_rate": 9.814235013419508e-06,
+      "loss": 0.4478,
+      "step": 4704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.60038057072251,
+      "learning_rate": 9.81412859728492e-06,
+      "loss": 0.3214,
+      "step": 4705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8757630742537785,
+      "learning_rate": 9.814022151255898e-06,
+      "loss": 0.2827,
+      "step": 4706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9331836149873411,
+      "learning_rate": 9.813915675333105e-06,
+      "loss": 0.3353,
+      "step": 4707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3305707626971401,
+      "learning_rate": 9.8138091695172e-06,
+      "loss": 0.2065,
+      "step": 4708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2775820471923205,
+      "learning_rate": 9.813702633808846e-06,
+      "loss": 0.3647,
+      "step": 4709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7317340435814546,
+      "learning_rate": 9.813596068208705e-06,
+      "loss": 0.3539,
+      "step": 4710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0622013067213087,
+      "learning_rate": 9.813489472717436e-06,
+      "loss": 0.1959,
+      "step": 4711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.675599959822041,
+      "learning_rate": 9.813382847335705e-06,
+      "loss": 0.239,
+      "step": 4712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.16972587792158,
+      "learning_rate": 9.813276192064171e-06,
+      "loss": 0.4254,
+      "step": 4713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9676258644199796,
+      "learning_rate": 9.813169506903497e-06,
+      "loss": 0.2534,
+      "step": 4714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.061436263228582,
+      "learning_rate": 9.813062791854347e-06,
+      "loss": 0.3285,
+      "step": 4715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6902513679935565,
+      "learning_rate": 9.81295604691738e-06,
+      "loss": 0.4177,
+      "step": 4716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.11919608985641,
+      "learning_rate": 9.812849272093264e-06,
+      "loss": 0.1958,
+      "step": 4717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6259078644224272,
+      "learning_rate": 9.812742467382657e-06,
+      "loss": 0.212,
+      "step": 4718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3997390792904916,
+      "learning_rate": 9.812635632786225e-06,
+      "loss": 0.2137,
+      "step": 4719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7660868302342771,
+      "learning_rate": 9.812528768304633e-06,
+      "loss": 0.3663,
+      "step": 4720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.554055682189794,
+      "learning_rate": 9.81242187393854e-06,
+      "loss": 0.305,
+      "step": 4721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6332779088721046,
+      "learning_rate": 9.812314949688613e-06,
+      "loss": 0.3614,
+      "step": 4722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7609844411366267,
+      "learning_rate": 9.812207995555516e-06,
+      "loss": 0.4101,
+      "step": 4723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3662086443169459,
+      "learning_rate": 9.812101011539912e-06,
+      "loss": 0.2138,
+      "step": 4724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5931782614395675,
+      "learning_rate": 9.811993997642466e-06,
+      "loss": 0.2078,
+      "step": 4725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.490158670334852,
+      "learning_rate": 9.811886953863841e-06,
+      "loss": 0.2284,
+      "step": 4726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.173329737478433,
+      "learning_rate": 9.811779880204705e-06,
+      "loss": 0.6642,
+      "step": 4727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0961755324997142,
+      "learning_rate": 9.811672776665719e-06,
+      "loss": 0.5349,
+      "step": 4728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1200931886837235,
+      "learning_rate": 9.811565643247552e-06,
+      "loss": 0.4658,
+      "step": 4729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.039257434866492,
+      "learning_rate": 9.811458479950866e-06,
+      "loss": 0.3972,
+      "step": 4730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7655734385920236,
+      "learning_rate": 9.81135128677633e-06,
+      "loss": 0.2064,
+      "step": 4731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7249635222627775,
+      "learning_rate": 9.811244063724605e-06,
+      "loss": 0.327,
+      "step": 4732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5224822534092963,
+      "learning_rate": 9.81113681079636e-06,
+      "loss": 0.3009,
+      "step": 4733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6193339528071753,
+      "learning_rate": 9.811029527992259e-06,
+      "loss": 0.3689,
+      "step": 4734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7891784780299305,
+      "learning_rate": 9.810922215312971e-06,
+      "loss": 0.3435,
+      "step": 4735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6964832385834734,
+      "learning_rate": 9.81081487275916e-06,
+      "loss": 0.4064,
+      "step": 4736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.645824186887154,
+      "learning_rate": 9.810707500331493e-06,
+      "loss": 0.2787,
+      "step": 4737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5185025116257065,
+      "learning_rate": 9.810600098030639e-06,
+      "loss": 0.3906,
+      "step": 4738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9138549575391546,
+      "learning_rate": 9.810492665857262e-06,
+      "loss": 0.466,
+      "step": 4739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5417960720480586,
+      "learning_rate": 9.81038520381203e-06,
+      "loss": 0.291,
+      "step": 4740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.72116794434918,
+      "learning_rate": 9.81027771189561e-06,
+      "loss": 0.3241,
+      "step": 4741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5636109527262163,
+      "learning_rate": 9.81017019010867e-06,
+      "loss": 0.3109,
+      "step": 4742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9889396371825454,
+      "learning_rate": 9.810062638451879e-06,
+      "loss": 0.2808,
+      "step": 4743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.807455046320292,
+      "learning_rate": 9.809955056925904e-06,
+      "loss": 0.3804,
+      "step": 4744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1895993274941787,
+      "learning_rate": 9.809847445531411e-06,
+      "loss": 0.5706,
+      "step": 4745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.758618532393399,
+      "learning_rate": 9.80973980426907e-06,
+      "loss": 0.2329,
+      "step": 4746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4875529852814828,
+      "learning_rate": 9.80963213313955e-06,
+      "loss": 0.2402,
+      "step": 4747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3418562677151926,
+      "learning_rate": 9.809524432143518e-06,
+      "loss": 0.3481,
+      "step": 4748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1743769437019713,
+      "learning_rate": 9.809416701281644e-06,
+      "loss": 0.4358,
+      "step": 4749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3809929369011067,
+      "learning_rate": 9.809308940554597e-06,
+      "loss": 0.1747,
+      "step": 4750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.644298999679694,
+      "learning_rate": 9.809201149963045e-06,
+      "loss": 0.4352,
+      "step": 4751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.033867836359629,
+      "learning_rate": 9.809093329507658e-06,
+      "loss": 0.506,
+      "step": 4752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.711370008487679,
+      "learning_rate": 9.808985479189106e-06,
+      "loss": 0.2039,
+      "step": 4753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9704836977487836,
+      "learning_rate": 9.80887759900806e-06,
+      "loss": 0.388,
+      "step": 4754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4961106819808225,
+      "learning_rate": 9.808769688965187e-06,
+      "loss": 0.271,
+      "step": 4755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.303642637397354,
+      "learning_rate": 9.808661749061157e-06,
+      "loss": 0.3415,
+      "step": 4756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9764769388610313,
+      "learning_rate": 9.808553779296643e-06,
+      "loss": 0.4141,
+      "step": 4757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5025725085880883,
+      "learning_rate": 9.808445779672316e-06,
+      "loss": 0.2186,
+      "step": 4758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7384515786033752,
+      "learning_rate": 9.808337750188841e-06,
+      "loss": 0.5149,
+      "step": 4759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6478811473209471,
+      "learning_rate": 9.808229690846895e-06,
+      "loss": 0.3251,
+      "step": 4760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3051716318860156,
+      "learning_rate": 9.808121601647148e-06,
+      "loss": 0.1972,
+      "step": 4761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.70653175710507,
+      "learning_rate": 9.808013482590268e-06,
+      "loss": 0.2004,
+      "step": 4762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9590817937845577,
+      "learning_rate": 9.807905333676929e-06,
+      "loss": 0.379,
+      "step": 4763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5638523382904261,
+      "learning_rate": 9.8077971549078e-06,
+      "loss": 0.2067,
+      "step": 4764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9419451758034818,
+      "learning_rate": 9.807688946283555e-06,
+      "loss": 0.4535,
+      "step": 4765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9397211571501014,
+      "learning_rate": 9.807580707804867e-06,
+      "loss": 0.3591,
+      "step": 4766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7788784946358807,
+      "learning_rate": 9.807472439472407e-06,
+      "loss": 0.1726,
+      "step": 4767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.992349490891225,
+      "learning_rate": 9.807364141286846e-06,
+      "loss": 0.2438,
+      "step": 4768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4778875745477766,
+      "learning_rate": 9.807255813248857e-06,
+      "loss": 0.2201,
+      "step": 4769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6677641352237607,
+      "learning_rate": 9.807147455359114e-06,
+      "loss": 0.3732,
+      "step": 4770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0428705283016075,
+      "learning_rate": 9.807039067618289e-06,
+      "loss": 0.6575,
+      "step": 4771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1683521756489212,
+      "learning_rate": 9.806930650027054e-06,
+      "loss": 0.5392,
+      "step": 4772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3245023336558268,
+      "learning_rate": 9.806822202586084e-06,
+      "loss": 0.1717,
+      "step": 4773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0798875470680254,
+      "learning_rate": 9.806713725296053e-06,
+      "loss": 0.2381,
+      "step": 4774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7517203492118267,
+      "learning_rate": 9.806605218157632e-06,
+      "loss": 0.5193,
+      "step": 4775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7003255155556318,
+      "learning_rate": 9.806496681171496e-06,
+      "loss": 0.5013,
+      "step": 4776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.39821702734963,
+      "learning_rate": 9.80638811433832e-06,
+      "loss": 0.1436,
+      "step": 4777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9926271439673524,
+      "learning_rate": 9.806279517658777e-06,
+      "loss": 0.5563,
+      "step": 4778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.311242696122326,
+      "learning_rate": 9.806170891133542e-06,
+      "loss": 0.5638,
+      "step": 4779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7924756491941016,
+      "learning_rate": 9.80606223476329e-06,
+      "loss": 0.3914,
+      "step": 4780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.035471344496155,
+      "learning_rate": 9.805953548548693e-06,
+      "loss": 0.5615,
+      "step": 4781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8147314821053393,
+      "learning_rate": 9.805844832490429e-06,
+      "loss": 0.3256,
+      "step": 4782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0304570594383535,
+      "learning_rate": 9.805736086589171e-06,
+      "loss": 0.6293,
+      "step": 4783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5853641535357785,
+      "learning_rate": 9.805627310845596e-06,
+      "loss": 0.3283,
+      "step": 4784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.009203822110655,
+      "learning_rate": 9.805518505260378e-06,
+      "loss": 0.3902,
+      "step": 4785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9412438230668028,
+      "learning_rate": 9.805409669834196e-06,
+      "loss": 0.492,
+      "step": 4786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9197555695043251,
+      "learning_rate": 9.80530080456772e-06,
+      "loss": 0.3585,
+      "step": 4787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6360732105150968,
+      "learning_rate": 9.805191909461631e-06,
+      "loss": 0.2394,
+      "step": 4788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7279453780931229,
+      "learning_rate": 9.805082984516602e-06,
+      "loss": 0.3644,
+      "step": 4789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8996407498610866,
+      "learning_rate": 9.804974029733314e-06,
+      "loss": 0.4098,
+      "step": 4790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0477675246503044,
+      "learning_rate": 9.804865045112437e-06,
+      "loss": 0.4967,
+      "step": 4791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8947842537174127,
+      "learning_rate": 9.804756030654654e-06,
+      "loss": 0.438,
+      "step": 4792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5913036459898253,
+      "learning_rate": 9.804646986360638e-06,
+      "loss": 0.2478,
+      "step": 4793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4665919591244072,
+      "learning_rate": 9.804537912231066e-06,
+      "loss": 0.3293,
+      "step": 4794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6043968659944177,
+      "learning_rate": 9.804428808266618e-06,
+      "loss": 0.3946,
+      "step": 4795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8557895443319772,
+      "learning_rate": 9.804319674467969e-06,
+      "loss": 0.2769,
+      "step": 4796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4681390311126563,
+      "learning_rate": 9.804210510835798e-06,
+      "loss": 0.2768,
+      "step": 4797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8339812935832729,
+      "learning_rate": 9.804101317370783e-06,
+      "loss": 0.2849,
+      "step": 4798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9041102039103115,
+      "learning_rate": 9.803992094073603e-06,
+      "loss": 0.1934,
+      "step": 4799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.765365611071038,
+      "learning_rate": 9.803882840944934e-06,
+      "loss": 0.3666,
+      "step": 4800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.234888840920071,
+      "learning_rate": 9.803773557985455e-06,
+      "loss": 0.3525,
+      "step": 4801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.967704054222146,
+      "learning_rate": 9.803664245195845e-06,
+      "loss": 0.2251,
+      "step": 4802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8904484368864662,
+      "learning_rate": 9.803554902576783e-06,
+      "loss": 0.2267,
+      "step": 4803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7062155042382092,
+      "learning_rate": 9.803445530128948e-06,
+      "loss": 0.2274,
+      "step": 4804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.421681055528706,
+      "learning_rate": 9.803336127853018e-06,
+      "loss": 0.2343,
+      "step": 4805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.04891140662174,
+      "learning_rate": 9.803226695749675e-06,
+      "loss": 0.7306,
+      "step": 4806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6735676370235524,
+      "learning_rate": 9.803117233819597e-06,
+      "loss": 0.3436,
+      "step": 4807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.079917579150387,
+      "learning_rate": 9.803007742063461e-06,
+      "loss": 0.7219,
+      "step": 4808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5463735750324548,
+      "learning_rate": 9.802898220481953e-06,
+      "loss": 0.3494,
+      "step": 4809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4946817887876502,
+      "learning_rate": 9.802788669075747e-06,
+      "loss": 0.3814,
+      "step": 4810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.105146590446894,
+      "learning_rate": 9.802679087845527e-06,
+      "loss": 0.3238,
+      "step": 4811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6284645428495668,
+      "learning_rate": 9.802569476791971e-06,
+      "loss": 0.3609,
+      "step": 4812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5405319528163148,
+      "learning_rate": 9.802459835915763e-06,
+      "loss": 0.2534,
+      "step": 4813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9316991714229605,
+      "learning_rate": 9.802350165217582e-06,
+      "loss": 0.3553,
+      "step": 4814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3385809391348684,
+      "learning_rate": 9.802240464698108e-06,
+      "loss": 0.2256,
+      "step": 4815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5349962524824963,
+      "learning_rate": 9.802130734358023e-06,
+      "loss": 0.4144,
+      "step": 4816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.483547226461111,
+      "learning_rate": 9.802020974198009e-06,
+      "loss": 0.1356,
+      "step": 4817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9782789414514905,
+      "learning_rate": 9.801911184218746e-06,
+      "loss": 0.4612,
+      "step": 4818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2542480882193259,
+      "learning_rate": 9.801801364420919e-06,
+      "loss": 0.2489,
+      "step": 4819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.656781669068801,
+      "learning_rate": 9.801691514805206e-06,
+      "loss": 0.1932,
+      "step": 4820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6373978005068248,
+      "learning_rate": 9.801581635372291e-06,
+      "loss": 0.3691,
+      "step": 4821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2620869031515396,
+      "learning_rate": 9.801471726122858e-06,
+      "loss": 0.2344,
+      "step": 4822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7843694726888426,
+      "learning_rate": 9.801361787057586e-06,
+      "loss": 0.4461,
+      "step": 4823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5847187621099015,
+      "learning_rate": 9.80125181817716e-06,
+      "loss": 0.3803,
+      "step": 4824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9195799188339946,
+      "learning_rate": 9.801141819482263e-06,
+      "loss": 0.4344,
+      "step": 4825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9441039609450663,
+      "learning_rate": 9.801031790973576e-06,
+      "loss": 0.4425,
+      "step": 4826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.386288508980211,
+      "learning_rate": 9.800921732651785e-06,
+      "loss": 0.2269,
+      "step": 4827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.613477133896388,
+      "learning_rate": 9.800811644517572e-06,
+      "loss": 0.4877,
+      "step": 4828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.602224752693578,
+      "learning_rate": 9.800701526571618e-06,
+      "loss": 0.2665,
+      "step": 4829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7329086057559826,
+      "learning_rate": 9.800591378814612e-06,
+      "loss": 0.2744,
+      "step": 4830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4742248224317,
+      "learning_rate": 9.800481201247236e-06,
+      "loss": 0.4785,
+      "step": 4831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.11795545916466,
+      "learning_rate": 9.800370993870173e-06,
+      "loss": 0.5694,
+      "step": 4832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8640386633456092,
+      "learning_rate": 9.800260756684107e-06,
+      "loss": 0.2787,
+      "step": 4833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.274480379505721,
+      "learning_rate": 9.800150489689726e-06,
+      "loss": 0.1962,
+      "step": 4834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9604389252160832,
+      "learning_rate": 9.80004019288771e-06,
+      "loss": 0.1806,
+      "step": 4835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3753820506259002,
+      "learning_rate": 9.799929866278747e-06,
+      "loss": 0.2166,
+      "step": 4836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8152085406059826,
+      "learning_rate": 9.799819509863521e-06,
+      "loss": 0.2771,
+      "step": 4837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.994838790403344,
+      "learning_rate": 9.799709123642719e-06,
+      "loss": 0.4136,
+      "step": 4838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6559335748237582,
+      "learning_rate": 9.799598707617024e-06,
+      "loss": 0.4042,
+      "step": 4839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5454522559531805,
+      "learning_rate": 9.799488261787122e-06,
+      "loss": 0.3394,
+      "step": 4840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7078148813047846,
+      "learning_rate": 9.799377786153701e-06,
+      "loss": 0.4916,
+      "step": 4841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.577810385946851,
+      "learning_rate": 9.799267280717443e-06,
+      "loss": 0.381,
+      "step": 4842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5056196311365653,
+      "learning_rate": 9.79915674547904e-06,
+      "loss": 0.2375,
+      "step": 4843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7632955992541155,
+      "learning_rate": 9.799046180439173e-06,
+      "loss": 0.2148,
+      "step": 4844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.141900446473629,
+      "learning_rate": 9.798935585598533e-06,
+      "loss": 0.2807,
+      "step": 4845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5398979775505601,
+      "learning_rate": 9.798824960957802e-06,
+      "loss": 0.3838,
+      "step": 4846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5360263222500938,
+      "learning_rate": 9.798714306517671e-06,
+      "loss": 0.2992,
+      "step": 4847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.023159416708524,
+      "learning_rate": 9.798603622278826e-06,
+      "loss": 0.3903,
+      "step": 4848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8965305417036948,
+      "learning_rate": 9.798492908241953e-06,
+      "loss": 0.4207,
+      "step": 4849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8175580054554492,
+      "learning_rate": 9.79838216440774e-06,
+      "loss": 0.4279,
+      "step": 4850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1164427710811284,
+      "learning_rate": 9.798271390776876e-06,
+      "loss": 0.6095,
+      "step": 4851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4643091810039084,
+      "learning_rate": 9.798160587350048e-06,
+      "loss": 0.6702,
+      "step": 4852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.013837894437163,
+      "learning_rate": 9.798049754127943e-06,
+      "loss": 0.7265,
+      "step": 4853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9187148870678612,
+      "learning_rate": 9.797938891111251e-06,
+      "loss": 0.2192,
+      "step": 4854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3864358837371076,
+      "learning_rate": 9.79782799830066e-06,
+      "loss": 0.3938,
+      "step": 4855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7567719304213014,
+      "learning_rate": 9.797717075696857e-06,
+      "loss": 0.3174,
+      "step": 4856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0431757543076343,
+      "learning_rate": 9.797606123300534e-06,
+      "loss": 0.6075,
+      "step": 4857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7504437641777164,
+      "learning_rate": 9.797495141112377e-06,
+      "loss": 0.5143,
+      "step": 4858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6582146634081218,
+      "learning_rate": 9.797384129133078e-06,
+      "loss": 0.2206,
+      "step": 4859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.637918251254042,
+      "learning_rate": 9.797273087363323e-06,
+      "loss": 0.4262,
+      "step": 4860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.805175137384293,
+      "learning_rate": 9.797162015803803e-06,
+      "loss": 0.3471,
+      "step": 4861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6368824688004395,
+      "learning_rate": 9.797050914455208e-06,
+      "loss": 0.2891,
+      "step": 4862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.06361798128153,
+      "learning_rate": 9.796939783318229e-06,
+      "loss": 0.5285,
+      "step": 4863
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3851181154654912,
+      "learning_rate": 9.796828622393554e-06,
+      "loss": 0.2501,
+      "step": 4864
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2901067752090507,
+      "learning_rate": 9.796717431681875e-06,
+      "loss": 0.2583,
+      "step": 4865
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8919904808287977,
+      "learning_rate": 9.79660621118388e-06,
+      "loss": 0.3161,
+      "step": 4866
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0398096053247317,
+      "learning_rate": 9.796494960900263e-06,
+      "loss": 0.5217,
+      "step": 4867
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.063183345603686,
+      "learning_rate": 9.796383680831712e-06,
+      "loss": 0.1685,
+      "step": 4868
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5308703126976535,
+      "learning_rate": 9.79627237097892e-06,
+      "loss": 0.3079,
+      "step": 4869
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4189598381829147,
+      "learning_rate": 9.796161031342577e-06,
+      "loss": 0.6992,
+      "step": 4870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1231935129349164,
+      "learning_rate": 9.796049661923375e-06,
+      "loss": 0.6828,
+      "step": 4871
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6035303191837118,
+      "learning_rate": 9.795938262722005e-06,
+      "loss": 0.2915,
+      "step": 4872
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9685564513661744,
+      "learning_rate": 9.79582683373916e-06,
+      "loss": 0.2727,
+      "step": 4873
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9330475180145024,
+      "learning_rate": 9.795715374975529e-06,
+      "loss": 0.2361,
+      "step": 4874
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9014703441471943,
+      "learning_rate": 9.795603886431807e-06,
+      "loss": 0.1979,
+      "step": 4875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.545229227765382,
+      "learning_rate": 9.795492368108685e-06,
+      "loss": 0.237,
+      "step": 4876
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2632181408393277,
+      "learning_rate": 9.795380820006858e-06,
+      "loss": 0.2215,
+      "step": 4877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5546714887321698,
+      "learning_rate": 9.795269242127015e-06,
+      "loss": 0.1684,
+      "step": 4878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1410057571537293,
+      "learning_rate": 9.795157634469851e-06,
+      "loss": 0.3772,
+      "step": 4879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9491387152055637,
+      "learning_rate": 9.795045997036058e-06,
+      "loss": 0.23,
+      "step": 4880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.517814654713571,
+      "learning_rate": 9.794934329826332e-06,
+      "loss": 0.3017,
+      "step": 4881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2301195920833676,
+      "learning_rate": 9.794822632841363e-06,
+      "loss": 0.2057,
+      "step": 4882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.580661233456134,
+      "learning_rate": 9.794710906081844e-06,
+      "loss": 0.2223,
+      "step": 4883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3867953985864636,
+      "learning_rate": 9.794599149548471e-06,
+      "loss": 0.262,
+      "step": 4884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4467843397050535,
+      "learning_rate": 9.79448736324194e-06,
+      "loss": 0.2132,
+      "step": 4885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4839806635228572,
+      "learning_rate": 9.794375547162942e-06,
+      "loss": 0.2284,
+      "step": 4886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0015821085199432,
+      "learning_rate": 9.794263701312171e-06,
+      "loss": 0.3829,
+      "step": 4887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7989560888379852,
+      "learning_rate": 9.794151825690324e-06,
+      "loss": 0.3238,
+      "step": 4888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.213263042912686,
+      "learning_rate": 9.794039920298093e-06,
+      "loss": 0.2898,
+      "step": 4889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9195110933257031,
+      "learning_rate": 9.793927985136176e-06,
+      "loss": 0.3532,
+      "step": 4890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9016328693618991,
+      "learning_rate": 9.793816020205264e-06,
+      "loss": 0.2673,
+      "step": 4891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.529896480229174,
+      "learning_rate": 9.793704025506056e-06,
+      "loss": 0.6098,
+      "step": 4892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7071347347587724,
+      "learning_rate": 9.793592001039247e-06,
+      "loss": 0.7263,
+      "step": 4893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6991733150327166,
+      "learning_rate": 9.79347994680553e-06,
+      "loss": 0.3711,
+      "step": 4894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6733963274766022,
+      "learning_rate": 9.793367862805604e-06,
+      "loss": 0.205,
+      "step": 4895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9556701404046968,
+      "learning_rate": 9.793255749040163e-06,
+      "loss": 0.3807,
+      "step": 4896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4940300719805821,
+      "learning_rate": 9.793143605509902e-06,
+      "loss": 0.2301,
+      "step": 4897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0664129383288996,
+      "learning_rate": 9.793031432215521e-06,
+      "loss": 0.4268,
+      "step": 4898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.880749925061138,
+      "learning_rate": 9.792919229157716e-06,
+      "loss": 0.2035,
+      "step": 4899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.194490102217181,
+      "learning_rate": 9.79280699633718e-06,
+      "loss": 0.3637,
+      "step": 4900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2026605576612384,
+      "learning_rate": 9.792694733754614e-06,
+      "loss": 0.2034,
+      "step": 4901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.473811882570549,
+      "learning_rate": 9.792582441410713e-06,
+      "loss": 0.1627,
+      "step": 4902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8911110749688589,
+      "learning_rate": 9.792470119306174e-06,
+      "loss": 0.49,
+      "step": 4903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5734126533205712,
+      "learning_rate": 9.792357767441697e-06,
+      "loss": 0.3272,
+      "step": 4904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6526709642915662,
+      "learning_rate": 9.792245385817977e-06,
+      "loss": 0.3004,
+      "step": 4905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5570909191054334,
+      "learning_rate": 9.792132974435713e-06,
+      "loss": 0.3762,
+      "step": 4906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0048926307225994,
+      "learning_rate": 9.792020533295603e-06,
+      "loss": 0.1931,
+      "step": 4907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4560390397375311,
+      "learning_rate": 9.791908062398344e-06,
+      "loss": 0.2205,
+      "step": 4908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7755267780331632,
+      "learning_rate": 9.791795561744637e-06,
+      "loss": 0.3012,
+      "step": 4909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.226219036983621,
+      "learning_rate": 9.791683031335179e-06,
+      "loss": 0.7167,
+      "step": 4910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.547531762161607,
+      "learning_rate": 9.791570471170668e-06,
+      "loss": 0.5424,
+      "step": 4911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4323935200860236,
+      "learning_rate": 9.791457881251804e-06,
+      "loss": 0.1933,
+      "step": 4912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.13104202858304,
+      "learning_rate": 9.791345261579287e-06,
+      "loss": 0.4466,
+      "step": 4913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9301591366301847,
+      "learning_rate": 9.791232612153814e-06,
+      "loss": 0.5144,
+      "step": 4914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.94507081089222,
+      "learning_rate": 9.791119932976087e-06,
+      "loss": 0.5163,
+      "step": 4915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9056222147685085,
+      "learning_rate": 9.791007224046805e-06,
+      "loss": 0.5735,
+      "step": 4916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2632285553576383,
+      "learning_rate": 9.790894485366667e-06,
+      "loss": 0.171,
+      "step": 4917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.919523079357527,
+      "learning_rate": 9.790781716936373e-06,
+      "loss": 0.4097,
+      "step": 4918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0474236386420688,
+      "learning_rate": 9.790668918756624e-06,
+      "loss": 0.6214,
+      "step": 4919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7129140741359112,
+      "learning_rate": 9.790556090828121e-06,
+      "loss": 0.2301,
+      "step": 4920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9960839108935664,
+      "learning_rate": 9.790443233151563e-06,
+      "loss": 0.21,
+      "step": 4921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.468745723677558,
+      "learning_rate": 9.790330345727652e-06,
+      "loss": 0.2223,
+      "step": 4922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2999648975850118,
+      "learning_rate": 9.79021742855709e-06,
+      "loss": 0.4055,
+      "step": 4923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3860422324375783,
+      "learning_rate": 9.790104481640575e-06,
+      "loss": 0.2247,
+      "step": 4924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7244156307123237,
+      "learning_rate": 9.78999150497881e-06,
+      "loss": 0.3626,
+      "step": 4925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3250710942880448,
+      "learning_rate": 9.7898784985725e-06,
+      "loss": 0.1986,
+      "step": 4926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2578334606686927,
+      "learning_rate": 9.78976546242234e-06,
+      "loss": 0.1909,
+      "step": 4927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.723687589817798,
+      "learning_rate": 9.789652396529037e-06,
+      "loss": 0.219,
+      "step": 4928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.419372074354307,
+      "learning_rate": 9.789539300893294e-06,
+      "loss": 0.3172,
+      "step": 4929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5448120767806017,
+      "learning_rate": 9.789426175515809e-06,
+      "loss": 0.2981,
+      "step": 4930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0773733974032615,
+      "learning_rate": 9.789313020397284e-06,
+      "loss": 0.6946,
+      "step": 4931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4798953951577238,
+      "learning_rate": 9.789199835538427e-06,
+      "loss": 0.3188,
+      "step": 4932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4628330443475133,
+      "learning_rate": 9.789086620939936e-06,
+      "loss": 0.1678,
+      "step": 4933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.36316269274348,
+      "learning_rate": 9.788973376602517e-06,
+      "loss": 0.3159,
+      "step": 4934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.779225774021591,
+      "learning_rate": 9.788860102526871e-06,
+      "loss": 0.1434,
+      "step": 4935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.975060657583775,
+      "learning_rate": 9.788746798713705e-06,
+      "loss": 0.3909,
+      "step": 4936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5028667473952342,
+      "learning_rate": 9.788633465163718e-06,
+      "loss": 0.2083,
+      "step": 4937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8498925190710873,
+      "learning_rate": 9.788520101877617e-06,
+      "loss": 0.4802,
+      "step": 4938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4132464273298475,
+      "learning_rate": 9.788406708856104e-06,
+      "loss": 0.26,
+      "step": 4939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9217660531899117,
+      "learning_rate": 9.788293286099884e-06,
+      "loss": 0.5404,
+      "step": 4940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8966534555887966,
+      "learning_rate": 9.788179833609662e-06,
+      "loss": 0.3439,
+      "step": 4941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1040023764752793,
+      "learning_rate": 9.788066351386142e-06,
+      "loss": 0.4832,
+      "step": 4942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.417755014951763,
+      "learning_rate": 9.787952839430027e-06,
+      "loss": 0.2139,
+      "step": 4943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1472419255853543,
+      "learning_rate": 9.787839297742025e-06,
+      "loss": 0.4284,
+      "step": 4944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.515435745280248,
+      "learning_rate": 9.787725726322839e-06,
+      "loss": 0.2857,
+      "step": 4945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.027863283419426,
+      "learning_rate": 9.787612125173174e-06,
+      "loss": 0.5714,
+      "step": 4946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2925186625296137,
+      "learning_rate": 9.787498494293737e-06,
+      "loss": 0.368,
+      "step": 4947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5852519771160232,
+      "learning_rate": 9.787384833685234e-06,
+      "loss": 0.2353,
+      "step": 4948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7991252057417217,
+      "learning_rate": 9.787271143348367e-06,
+      "loss": 0.452,
+      "step": 4949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7984845292001563,
+      "learning_rate": 9.787157423283847e-06,
+      "loss": 0.3919,
+      "step": 4950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.694681852356963,
+      "learning_rate": 9.787043673492375e-06,
+      "loss": 0.3087,
+      "step": 4951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6767012109462132,
+      "learning_rate": 9.786929893974662e-06,
+      "loss": 0.3414,
+      "step": 4952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8734132072362688,
+      "learning_rate": 9.786816084731412e-06,
+      "loss": 0.24,
+      "step": 4953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0722859316018782,
+      "learning_rate": 9.786702245763333e-06,
+      "loss": 0.6684,
+      "step": 4954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7593978683855545,
+      "learning_rate": 9.78658837707113e-06,
+      "loss": 0.2306,
+      "step": 4955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.618532201068276,
+      "learning_rate": 9.786474478655512e-06,
+      "loss": 0.2553,
+      "step": 4956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.117080133120615,
+      "learning_rate": 9.786360550517186e-06,
+      "loss": 0.2233,
+      "step": 4957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.794296219945587,
+      "learning_rate": 9.786246592656858e-06,
+      "loss": 0.2193,
+      "step": 4958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7873150524747203,
+      "learning_rate": 9.786132605075238e-06,
+      "loss": 0.3374,
+      "step": 4959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3283138198576414,
+      "learning_rate": 9.786018587773032e-06,
+      "loss": 0.7449,
+      "step": 4960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9185704105280104,
+      "learning_rate": 9.785904540750948e-06,
+      "loss": 0.3603,
+      "step": 4961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.547589936393077,
+      "learning_rate": 9.785790464009697e-06,
+      "loss": 0.4177,
+      "step": 4962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2777670777094283,
+      "learning_rate": 9.785676357549983e-06,
+      "loss": 0.1117,
+      "step": 4963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2425549540204797,
+      "learning_rate": 9.785562221372516e-06,
+      "loss": 0.3753,
+      "step": 4964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9673136285126078,
+      "learning_rate": 9.785448055478006e-06,
+      "loss": 0.3875,
+      "step": 4965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8812505145443436,
+      "learning_rate": 9.785333859867163e-06,
+      "loss": 0.1719,
+      "step": 4966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.785384249161061,
+      "learning_rate": 9.785219634540693e-06,
+      "loss": 0.447,
+      "step": 4967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9095705885846466,
+      "learning_rate": 9.78510537949931e-06,
+      "loss": 0.582,
+      "step": 4968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6650697856317123,
+      "learning_rate": 9.784991094743717e-06,
+      "loss": 0.2098,
+      "step": 4969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8023358624719361,
+      "learning_rate": 9.784876780274628e-06,
+      "loss": 0.2023,
+      "step": 4970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2894849061997413,
+      "learning_rate": 9.784762436092752e-06,
+      "loss": 0.4253,
+      "step": 4971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4924172479206033,
+      "learning_rate": 9.784648062198799e-06,
+      "loss": 0.4233,
+      "step": 4972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.668380985642546,
+      "learning_rate": 9.784533658593478e-06,
+      "loss": 0.3667,
+      "step": 4973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.615362264927831,
+      "learning_rate": 9.784419225277502e-06,
+      "loss": 0.1845,
+      "step": 4974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9536884563220014,
+      "learning_rate": 9.78430476225158e-06,
+      "loss": 0.4764,
+      "step": 4975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3103242875571326,
+      "learning_rate": 9.784190269516422e-06,
+      "loss": 0.2057,
+      "step": 4976
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.118868839396962,
+      "learning_rate": 9.784075747072741e-06,
+      "loss": 0.4957,
+      "step": 4977
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5822396623827055,
+      "learning_rate": 9.783961194921246e-06,
+      "loss": 0.2785,
+      "step": 4978
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6539715353541407,
+      "learning_rate": 9.78384661306265e-06,
+      "loss": 0.3309,
+      "step": 4979
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.036785293797261,
+      "learning_rate": 9.783732001497664e-06,
+      "loss": 0.5285,
+      "step": 4980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.019281546754563,
+      "learning_rate": 9.783617360227e-06,
+      "loss": 0.1648,
+      "step": 4981
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8210433211231984,
+      "learning_rate": 9.78350268925137e-06,
+      "loss": 0.2159,
+      "step": 4982
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.937313814023451,
+      "learning_rate": 9.783387988571485e-06,
+      "loss": 0.2388,
+      "step": 4983
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8877282818136765,
+      "learning_rate": 9.783273258188058e-06,
+      "loss": 0.2466,
+      "step": 4984
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2628473289242184,
+      "learning_rate": 9.783158498101802e-06,
+      "loss": 0.674,
+      "step": 4985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3818361423883185,
+      "learning_rate": 9.783043708313427e-06,
+      "loss": 0.393,
+      "step": 4986
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3673884107924372,
+      "learning_rate": 9.782928888823648e-06,
+      "loss": 0.2759,
+      "step": 4987
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7802217677952281,
+      "learning_rate": 9.782814039633179e-06,
+      "loss": 0.227,
+      "step": 4988
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5067620138749072,
+      "learning_rate": 9.782699160742733e-06,
+      "loss": 0.219,
+      "step": 4989
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.375580097014391,
+      "learning_rate": 9.782584252153021e-06,
+      "loss": 0.1895,
+      "step": 4990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2623886243396796,
+      "learning_rate": 9.782469313864757e-06,
+      "loss": 0.3302,
+      "step": 4991
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5699866464450922,
+      "learning_rate": 9.782354345878656e-06,
+      "loss": 0.2648,
+      "step": 4992
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0775344697959426,
+      "learning_rate": 9.782239348195432e-06,
+      "loss": 0.1898,
+      "step": 4993
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.628102225329238,
+      "learning_rate": 9.782124320815799e-06,
+      "loss": 0.3613,
+      "step": 4994
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.823074807164974,
+      "learning_rate": 9.782009263740471e-06,
+      "loss": 0.5112,
+      "step": 4995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9300242373446763,
+      "learning_rate": 9.781894176970162e-06,
+      "loss": 0.5581,
+      "step": 4996
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.486069339675652,
+      "learning_rate": 9.781779060505585e-06,
+      "loss": 0.357,
+      "step": 4997
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7911546043794284,
+      "learning_rate": 9.78166391434746e-06,
+      "loss": 0.4984,
+      "step": 4998
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7239687994690172,
+      "learning_rate": 9.781548738496498e-06,
+      "loss": 0.3943,
+      "step": 4999
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.280313005279691,
+      "learning_rate": 9.781433532953415e-06,
+      "loss": 0.3141,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7103977092160083,
+      "learning_rate": 9.781318297718926e-06,
+      "loss": 0.2738,
+      "step": 5001
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2315698420023349,
+      "learning_rate": 9.781203032793747e-06,
+      "loss": 0.1196,
+      "step": 5002
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.840274873667251,
+      "learning_rate": 9.781087738178595e-06,
+      "loss": 0.3592,
+      "step": 5003
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8534205525880372,
+      "learning_rate": 9.780972413874183e-06,
+      "loss": 0.5632,
+      "step": 5004
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.827599618912419,
+      "learning_rate": 9.780857059881232e-06,
+      "loss": 0.354,
+      "step": 5005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0891705507169553,
+      "learning_rate": 9.780741676200454e-06,
+      "loss": 0.3836,
+      "step": 5006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1959004412415184,
+      "learning_rate": 9.780626262832564e-06,
+      "loss": 0.5533,
+      "step": 5007
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9560159116606821,
+      "learning_rate": 9.780510819778285e-06,
+      "loss": 0.3464,
+      "step": 5008
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9361226133048866,
+      "learning_rate": 9.78039534703833e-06,
+      "loss": 0.2661,
+      "step": 5009
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4711115980357097,
+      "learning_rate": 9.780279844613414e-06,
+      "loss": 0.2968,
+      "step": 5010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.07106334304359,
+      "learning_rate": 9.78016431250426e-06,
+      "loss": 0.1286,
+      "step": 5011
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6431738082903242,
+      "learning_rate": 9.780048750711578e-06,
+      "loss": 0.2552,
+      "step": 5012
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5213120917544634,
+      "learning_rate": 9.779933159236092e-06,
+      "loss": 0.2173,
+      "step": 5013
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5737433366429927,
+      "learning_rate": 9.779817538078517e-06,
+      "loss": 0.3431,
+      "step": 5014
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7284532999127975,
+      "learning_rate": 9.779701887239572e-06,
+      "loss": 0.428,
+      "step": 5015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6220464779310158,
+      "learning_rate": 9.779586206719973e-06,
+      "loss": 0.3554,
+      "step": 5016
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.635440213527204,
+      "learning_rate": 9.779470496520442e-06,
+      "loss": 0.2486,
+      "step": 5017
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5356786157073066,
+      "learning_rate": 9.779354756641695e-06,
+      "loss": 0.1453,
+      "step": 5018
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3654916498955627,
+      "learning_rate": 9.77923898708445e-06,
+      "loss": 0.2938,
+      "step": 5019
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2412208148554282,
+      "learning_rate": 9.779123187849428e-06,
+      "loss": 0.1899,
+      "step": 5020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6814368830126256,
+      "learning_rate": 9.779007358937347e-06,
+      "loss": 0.3371,
+      "step": 5021
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7553181699963787,
+      "learning_rate": 9.778891500348926e-06,
+      "loss": 0.2197,
+      "step": 5022
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.512288782055077,
+      "learning_rate": 9.778775612084885e-06,
+      "loss": 0.2835,
+      "step": 5023
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.616821932526773,
+      "learning_rate": 9.778659694145943e-06,
+      "loss": 0.2394,
+      "step": 5024
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5958646321360666,
+      "learning_rate": 9.77854374653282e-06,
+      "loss": 0.1483,
+      "step": 5025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6412136781509143,
+      "learning_rate": 9.778427769246239e-06,
+      "loss": 0.2262,
+      "step": 5026
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1545977516376724,
+      "learning_rate": 9.778311762286916e-06,
+      "loss": 0.578,
+      "step": 5027
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5005093593836218,
+      "learning_rate": 9.77819572565557e-06,
+      "loss": 0.1647,
+      "step": 5028
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.194683549932559,
+      "learning_rate": 9.778079659352928e-06,
+      "loss": 0.2469,
+      "step": 5029
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2140635472002264,
+      "learning_rate": 9.777963563379707e-06,
+      "loss": 0.1473,
+      "step": 5030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9888174265248475,
+      "learning_rate": 9.777847437736627e-06,
+      "loss": 0.2329,
+      "step": 5031
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8480421735872674,
+      "learning_rate": 9.777731282424411e-06,
+      "loss": 0.2559,
+      "step": 5032
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6075500052499576,
+      "learning_rate": 9.777615097443779e-06,
+      "loss": 0.2089,
+      "step": 5033
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9624319537022537,
+      "learning_rate": 9.777498882795453e-06,
+      "loss": 0.5204,
+      "step": 5034
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5277680407711134,
+      "learning_rate": 9.777382638480154e-06,
+      "loss": 0.2211,
+      "step": 5035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2180758849118516,
+      "learning_rate": 9.777266364498605e-06,
+      "loss": 0.3967,
+      "step": 5036
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.054377049210655,
+      "learning_rate": 9.777150060851528e-06,
+      "loss": 0.2474,
+      "step": 5037
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0066111454980566,
+      "learning_rate": 9.777033727539644e-06,
+      "loss": 0.4489,
+      "step": 5038
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8189535962684906,
+      "learning_rate": 9.776917364563676e-06,
+      "loss": 0.2616,
+      "step": 5039
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9317296454086799,
+      "learning_rate": 9.776800971924348e-06,
+      "loss": 0.24,
+      "step": 5040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2581230035863307,
+      "learning_rate": 9.77668454962238e-06,
+      "loss": 0.3501,
+      "step": 5041
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8752063637658025,
+      "learning_rate": 9.776568097658497e-06,
+      "loss": 0.644,
+      "step": 5042
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3953649963006047,
+      "learning_rate": 9.776451616033422e-06,
+      "loss": 0.2425,
+      "step": 5043
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2602141816338333,
+      "learning_rate": 9.776335104747878e-06,
+      "loss": 0.2724,
+      "step": 5044
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.824953511214272,
+      "learning_rate": 9.776218563802588e-06,
+      "loss": 0.2953,
+      "step": 5045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3823410081669127,
+      "learning_rate": 9.776101993198275e-06,
+      "loss": 0.2164,
+      "step": 5046
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.218957071984157,
+      "learning_rate": 9.775985392935665e-06,
+      "loss": 0.229,
+      "step": 5047
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.228588556935827,
+      "learning_rate": 9.775868763015481e-06,
+      "loss": 0.2037,
+      "step": 5048
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0142723542377174,
+      "learning_rate": 9.775752103438448e-06,
+      "loss": 0.2581,
+      "step": 5049
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9169671343168562,
+      "learning_rate": 9.775635414205289e-06,
+      "loss": 0.2507,
+      "step": 5050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2470741719997969,
+      "learning_rate": 9.775518695316728e-06,
+      "loss": 0.2191,
+      "step": 5051
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3635548392917647,
+      "learning_rate": 9.775401946773493e-06,
+      "loss": 0.1563,
+      "step": 5052
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4388839370795512,
+      "learning_rate": 9.775285168576306e-06,
+      "loss": 0.2622,
+      "step": 5053
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6314653040086116,
+      "learning_rate": 9.775168360725892e-06,
+      "loss": 0.4337,
+      "step": 5054
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.24699269757105,
+      "learning_rate": 9.77505152322298e-06,
+      "loss": 0.3979,
+      "step": 5055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8785532661421982,
+      "learning_rate": 9.774934656068292e-06,
+      "loss": 0.6001,
+      "step": 5056
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4490397412156357,
+      "learning_rate": 9.774817759262555e-06,
+      "loss": 0.3242,
+      "step": 5057
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.718307854395366,
+      "learning_rate": 9.774700832806493e-06,
+      "loss": 0.3502,
+      "step": 5058
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9722131722528793,
+      "learning_rate": 9.774583876700837e-06,
+      "loss": 0.4197,
+      "step": 5059
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.816424794256239,
+      "learning_rate": 9.774466890946308e-06,
+      "loss": 0.3705,
+      "step": 5060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.791034452948827,
+      "learning_rate": 9.774349875543634e-06,
+      "loss": 0.4612,
+      "step": 5061
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5747536491176746,
+      "learning_rate": 9.774232830493543e-06,
+      "loss": 0.2393,
+      "step": 5062
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4837004283218191,
+      "learning_rate": 9.77411575579676e-06,
+      "loss": 0.3453,
+      "step": 5063
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1135387816702393,
+      "learning_rate": 9.773998651454015e-06,
+      "loss": 0.1958,
+      "step": 5064
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.282556239394088,
+      "learning_rate": 9.77388151746603e-06,
+      "loss": 0.1809,
+      "step": 5065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3605444019064057,
+      "learning_rate": 9.773764353833539e-06,
+      "loss": 0.2557,
+      "step": 5066
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.55646978706823,
+      "learning_rate": 9.773647160557265e-06,
+      "loss": 0.2605,
+      "step": 5067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0316976824176196,
+      "learning_rate": 9.773529937637935e-06,
+      "loss": 0.2674,
+      "step": 5068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.687388177981464,
+      "learning_rate": 9.77341268507628e-06,
+      "loss": 0.3457,
+      "step": 5069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7978130462847233,
+      "learning_rate": 9.773295402873027e-06,
+      "loss": 0.5798,
+      "step": 5070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7780310046152588,
+      "learning_rate": 9.773178091028903e-06,
+      "loss": 0.3576,
+      "step": 5071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6075467111032329,
+      "learning_rate": 9.773060749544639e-06,
+      "loss": 0.2179,
+      "step": 5072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5346216212382495,
+      "learning_rate": 9.772943378420961e-06,
+      "loss": 0.1712,
+      "step": 5073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5679104116066513,
+      "learning_rate": 9.7728259776586e-06,
+      "loss": 0.1921,
+      "step": 5074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.971383414036836,
+      "learning_rate": 9.772708547258284e-06,
+      "loss": 0.2883,
+      "step": 5075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9737461742704687,
+      "learning_rate": 9.772591087220742e-06,
+      "loss": 0.4096,
+      "step": 5076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0701581069669617,
+      "learning_rate": 9.772473597546704e-06,
+      "loss": 0.672,
+      "step": 5077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8976247130749893,
+      "learning_rate": 9.772356078236899e-06,
+      "loss": 0.3439,
+      "step": 5078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3736765712853165,
+      "learning_rate": 9.772238529292058e-06,
+      "loss": 0.4663,
+      "step": 5079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2386692369931755,
+      "learning_rate": 9.77212095071291e-06,
+      "loss": 0.5408,
+      "step": 5080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5417439247009792,
+      "learning_rate": 9.772003342500183e-06,
+      "loss": 0.4119,
+      "step": 5081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0932179582780925,
+      "learning_rate": 9.771885704654611e-06,
+      "loss": 0.4364,
+      "step": 5082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7895714772941538,
+      "learning_rate": 9.771768037176922e-06,
+      "loss": 0.3456,
+      "step": 5083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6334559721839115,
+      "learning_rate": 9.771650340067848e-06,
+      "loss": 0.2197,
+      "step": 5084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6868997031534063,
+      "learning_rate": 9.771532613328121e-06,
+      "loss": 0.3954,
+      "step": 5085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8922586594611646,
+      "learning_rate": 9.77141485695847e-06,
+      "loss": 0.2717,
+      "step": 5086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4239594068037154,
+      "learning_rate": 9.771297070959625e-06,
+      "loss": 0.2154,
+      "step": 5087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7663754370666378,
+      "learning_rate": 9.771179255332322e-06,
+      "loss": 0.2575,
+      "step": 5088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.687445167251112,
+      "learning_rate": 9.771061410077287e-06,
+      "loss": 0.4134,
+      "step": 5089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9616408511502026,
+      "learning_rate": 9.770943535195256e-06,
+      "loss": 0.3364,
+      "step": 5090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.367978276034944,
+      "learning_rate": 9.77082563068696e-06,
+      "loss": 0.3197,
+      "step": 5091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4676861562558254,
+      "learning_rate": 9.77070769655313e-06,
+      "loss": 0.3199,
+      "step": 5092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5428797901438949,
+      "learning_rate": 9.7705897327945e-06,
+      "loss": 0.2561,
+      "step": 5093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0376994009094846,
+      "learning_rate": 9.7704717394118e-06,
+      "loss": 0.5169,
+      "step": 5094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.583284392719393,
+      "learning_rate": 9.770353716405764e-06,
+      "loss": 0.3398,
+      "step": 5095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.385962645892204,
+      "learning_rate": 9.770235663777128e-06,
+      "loss": 0.2659,
+      "step": 5096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.988032160521743,
+      "learning_rate": 9.770117581526619e-06,
+      "loss": 0.333,
+      "step": 5097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8767987295949138,
+      "learning_rate": 9.769999469654977e-06,
+      "loss": 0.2185,
+      "step": 5098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3222470437157192,
+      "learning_rate": 9.76988132816293e-06,
+      "loss": 0.2459,
+      "step": 5099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7672891875943697,
+      "learning_rate": 9.769763157051213e-06,
+      "loss": 0.4894,
+      "step": 5100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6327080238961735,
+      "learning_rate": 9.769644956320561e-06,
+      "loss": 0.5563,
+      "step": 5101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.366474850083114,
+      "learning_rate": 9.769526725971707e-06,
+      "loss": 0.3696,
+      "step": 5102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.492897751627114,
+      "learning_rate": 9.769408466005386e-06,
+      "loss": 0.7084,
+      "step": 5103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.300346184416403,
+      "learning_rate": 9.769290176422332e-06,
+      "loss": 0.5123,
+      "step": 5104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8307144622067502,
+      "learning_rate": 9.769171857223279e-06,
+      "loss": 0.2533,
+      "step": 5105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8973511621405363,
+      "learning_rate": 9.769053508408962e-06,
+      "loss": 0.3451,
+      "step": 5106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.652221724781815,
+      "learning_rate": 9.768935129980118e-06,
+      "loss": 0.222,
+      "step": 5107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7462962692379973,
+      "learning_rate": 9.768816721937478e-06,
+      "loss": 0.2307,
+      "step": 5108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.783431507539683,
+      "learning_rate": 9.76869828428178e-06,
+      "loss": 0.3626,
+      "step": 5109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0721082708531795,
+      "learning_rate": 9.76857981701376e-06,
+      "loss": 0.2762,
+      "step": 5110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6007005442190267,
+      "learning_rate": 9.76846132013415e-06,
+      "loss": 0.2349,
+      "step": 5111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4951640445036845,
+      "learning_rate": 9.76834279364369e-06,
+      "loss": 0.3134,
+      "step": 5112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0552903096087336,
+      "learning_rate": 9.768224237543115e-06,
+      "loss": 0.2799,
+      "step": 5113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2858662426790237,
+      "learning_rate": 9.768105651833162e-06,
+      "loss": 0.1874,
+      "step": 5114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4210227885677325,
+      "learning_rate": 9.767987036514565e-06,
+      "loss": 0.2882,
+      "step": 5115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.157164914867272,
+      "learning_rate": 9.76786839158806e-06,
+      "loss": 0.2715,
+      "step": 5116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.522755661014624,
+      "learning_rate": 9.767749717054387e-06,
+      "loss": 0.2177,
+      "step": 5117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1246681339914733,
+      "learning_rate": 9.76763101291428e-06,
+      "loss": 0.5397,
+      "step": 5118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7005689868759117,
+      "learning_rate": 9.76751227916848e-06,
+      "loss": 0.2937,
+      "step": 5119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.649969823788921,
+      "learning_rate": 9.76739351581772e-06,
+      "loss": 0.3733,
+      "step": 5120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.762118325943264,
+      "learning_rate": 9.767274722862738e-06,
+      "loss": 0.2525,
+      "step": 5121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7585803580060182,
+      "learning_rate": 9.767155900304275e-06,
+      "loss": 0.4522,
+      "step": 5122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.53330009299595,
+      "learning_rate": 9.767037048143067e-06,
+      "loss": 0.3341,
+      "step": 5123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.817234859357823,
+      "learning_rate": 9.76691816637985e-06,
+      "loss": 0.4924,
+      "step": 5124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.383867597321605,
+      "learning_rate": 9.766799255015364e-06,
+      "loss": 0.2426,
+      "step": 5125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.651540770789257,
+      "learning_rate": 9.766680314050348e-06,
+      "loss": 0.3366,
+      "step": 5126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9211443524019787,
+      "learning_rate": 9.76656134348554e-06,
+      "loss": 0.2712,
+      "step": 5127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8638913377390278,
+      "learning_rate": 9.76644234332168e-06,
+      "loss": 0.4083,
+      "step": 5128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7588362298261149,
+      "learning_rate": 9.766323313559504e-06,
+      "loss": 0.3838,
+      "step": 5129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1954858226165257,
+      "learning_rate": 9.766204254199753e-06,
+      "loss": 0.1933,
+      "step": 5130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6205225978212223,
+      "learning_rate": 9.766085165243167e-06,
+      "loss": 0.1988,
+      "step": 5131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4741546333897195,
+      "learning_rate": 9.765966046690485e-06,
+      "loss": 0.1796,
+      "step": 5132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9613944391353726,
+      "learning_rate": 9.765846898542445e-06,
+      "loss": 0.5601,
+      "step": 5133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6200285203801486,
+      "learning_rate": 9.76572772079979e-06,
+      "loss": 0.2575,
+      "step": 5134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9390422082249623,
+      "learning_rate": 9.765608513463258e-06,
+      "loss": 0.5682,
+      "step": 5135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3348938674654298,
+      "learning_rate": 9.765489276533589e-06,
+      "loss": 0.2614,
+      "step": 5136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9937663441618898,
+      "learning_rate": 9.765370010011524e-06,
+      "loss": 0.3581,
+      "step": 5137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1577328493222074,
+      "learning_rate": 9.765250713897806e-06,
+      "loss": 0.3674,
+      "step": 5138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6953731802878782,
+      "learning_rate": 9.765131388193169e-06,
+      "loss": 0.2845,
+      "step": 5139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8587508396180135,
+      "learning_rate": 9.765012032898362e-06,
+      "loss": 0.5718,
+      "step": 5140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5618825598522712,
+      "learning_rate": 9.764892648014122e-06,
+      "loss": 0.3247,
+      "step": 5141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.443967692958823,
+      "learning_rate": 9.76477323354119e-06,
+      "loss": 0.2231,
+      "step": 5142
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5109834447559847,
+      "learning_rate": 9.764653789480307e-06,
+      "loss": 0.269,
+      "step": 5143
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2248583332200453,
+      "learning_rate": 9.764534315832219e-06,
+      "loss": 0.2447,
+      "step": 5144
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8899541168976022,
+      "learning_rate": 9.764414812597664e-06,
+      "loss": 0.5829,
+      "step": 5145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6874189919083915,
+      "learning_rate": 9.764295279777385e-06,
+      "loss": 0.5028,
+      "step": 5146
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7495251994953769,
+      "learning_rate": 9.764175717372124e-06,
+      "loss": 0.3576,
+      "step": 5147
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6613746256254935,
+      "learning_rate": 9.764056125382626e-06,
+      "loss": 0.3116,
+      "step": 5148
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.540595310362286,
+      "learning_rate": 9.76393650380963e-06,
+      "loss": 0.2163,
+      "step": 5149
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3757792562181104,
+      "learning_rate": 9.76381685265388e-06,
+      "loss": 0.2101,
+      "step": 5150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.841263379270446,
+      "learning_rate": 9.76369717191612e-06,
+      "loss": 0.3043,
+      "step": 5151
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7709883191969031,
+      "learning_rate": 9.76357746159709e-06,
+      "loss": 0.3337,
+      "step": 5152
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.650853657522343,
+      "learning_rate": 9.76345772169754e-06,
+      "loss": 0.3411,
+      "step": 5153
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9957005921708209,
+      "learning_rate": 9.763337952218206e-06,
+      "loss": 0.3606,
+      "step": 5154
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5640674644338075,
+      "learning_rate": 9.763218153159838e-06,
+      "loss": 0.3562,
+      "step": 5155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8467282299437813,
+      "learning_rate": 9.763098324523177e-06,
+      "loss": 0.2938,
+      "step": 5156
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5853168348865585,
+      "learning_rate": 9.762978466308964e-06,
+      "loss": 0.3053,
+      "step": 5157
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.614631874882715,
+      "learning_rate": 9.76285857851795e-06,
+      "loss": 0.3874,
+      "step": 5158
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4646563934672703,
+      "learning_rate": 9.762738661150874e-06,
+      "loss": 0.2236,
+      "step": 5159
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.884442012063464,
+      "learning_rate": 9.762618714208482e-06,
+      "loss": 0.4905,
+      "step": 5160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6577774463007258,
+      "learning_rate": 9.762498737691522e-06,
+      "loss": 0.3159,
+      "step": 5161
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4645388832735822,
+      "learning_rate": 9.762378731600735e-06,
+      "loss": 0.2223,
+      "step": 5162
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7877398915975264,
+      "learning_rate": 9.762258695936865e-06,
+      "loss": 0.2129,
+      "step": 5163
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.053765329773476,
+      "learning_rate": 9.762138630700665e-06,
+      "loss": 0.2343,
+      "step": 5164
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.271656683923357,
+      "learning_rate": 9.762018535892872e-06,
+      "loss": 0.2634,
+      "step": 5165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.402218941949626,
+      "learning_rate": 9.761898411514237e-06,
+      "loss": 0.1655,
+      "step": 5166
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5386049478816253,
+      "learning_rate": 9.761778257565504e-06,
+      "loss": 0.2142,
+      "step": 5167
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3981094668228498,
+      "learning_rate": 9.76165807404742e-06,
+      "loss": 0.2608,
+      "step": 5168
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.02442748474294,
+      "learning_rate": 9.76153786096073e-06,
+      "loss": 0.5489,
+      "step": 5169
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8424366136255865,
+      "learning_rate": 9.761417618306182e-06,
+      "loss": 0.4062,
+      "step": 5170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6391805648101385,
+      "learning_rate": 9.761297346084522e-06,
+      "loss": 0.4916,
+      "step": 5171
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7929613532950555,
+      "learning_rate": 9.761177044296495e-06,
+      "loss": 0.6081,
+      "step": 5172
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5346433605390046,
+      "learning_rate": 9.761056712942851e-06,
+      "loss": 0.2128,
+      "step": 5173
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.551497205261199,
+      "learning_rate": 9.760936352024337e-06,
+      "loss": 0.3389,
+      "step": 5174
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.598728547582392,
+      "learning_rate": 9.760815961541699e-06,
+      "loss": 0.1938,
+      "step": 5175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.700652859037981,
+      "learning_rate": 9.760695541495684e-06,
+      "loss": 0.3554,
+      "step": 5176
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1858026426584478,
+      "learning_rate": 9.760575091887041e-06,
+      "loss": 0.1557,
+      "step": 5177
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9952163933032894,
+      "learning_rate": 9.760454612716519e-06,
+      "loss": 0.4499,
+      "step": 5178
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4051284271189024,
+      "learning_rate": 9.760334103984863e-06,
+      "loss": 0.1834,
+      "step": 5179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5939094192582568,
+      "learning_rate": 9.760213565692825e-06,
+      "loss": 0.3472,
+      "step": 5180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7999263004109658,
+      "learning_rate": 9.760092997841151e-06,
+      "loss": 0.1956,
+      "step": 5181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9930168968495037,
+      "learning_rate": 9.75997240043059e-06,
+      "loss": 0.3399,
+      "step": 5182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9127862733573897,
+      "learning_rate": 9.759851773461892e-06,
+      "loss": 0.3066,
+      "step": 5183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8436755386567336,
+      "learning_rate": 9.759731116935806e-06,
+      "loss": 0.1639,
+      "step": 5184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4559794201192275,
+      "learning_rate": 9.75961043085308e-06,
+      "loss": 0.3528,
+      "step": 5185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3645599948061757,
+      "learning_rate": 9.759489715214465e-06,
+      "loss": 0.1727,
+      "step": 5186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7535849194039634,
+      "learning_rate": 9.759368970020709e-06,
+      "loss": 0.2556,
+      "step": 5187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.336867510933314,
+      "learning_rate": 9.759248195272561e-06,
+      "loss": 0.273,
+      "step": 5188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3719349573780095,
+      "learning_rate": 9.759127390970774e-06,
+      "loss": 0.2156,
+      "step": 5189
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.920599421086592,
+      "learning_rate": 9.759006557116096e-06,
+      "loss": 0.6398,
+      "step": 5190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4969227631649702,
+      "learning_rate": 9.758885693709279e-06,
+      "loss": 0.2035,
+      "step": 5191
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4905372539321526,
+      "learning_rate": 9.75876480075107e-06,
+      "loss": 0.2608,
+      "step": 5192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3678999842685708,
+      "learning_rate": 9.758643878242226e-06,
+      "loss": 0.3521,
+      "step": 5193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.861596899651303,
+      "learning_rate": 9.758522926183492e-06,
+      "loss": 0.2275,
+      "step": 5194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4542430710894365,
+      "learning_rate": 9.75840194457562e-06,
+      "loss": 0.411,
+      "step": 5195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6972548964043919,
+      "learning_rate": 9.758280933419364e-06,
+      "loss": 0.5738,
+      "step": 5196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7410146717105728,
+      "learning_rate": 9.758159892715474e-06,
+      "loss": 0.4002,
+      "step": 5197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5161690190338915,
+      "learning_rate": 9.758038822464702e-06,
+      "loss": 0.2865,
+      "step": 5198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7261869919773856,
+      "learning_rate": 9.757917722667797e-06,
+      "loss": 0.2651,
+      "step": 5199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0765757665791615,
+      "learning_rate": 9.757796593325515e-06,
+      "loss": 0.3729,
+      "step": 5200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5019154686541596,
+      "learning_rate": 9.757675434438607e-06,
+      "loss": 0.3448,
+      "step": 5201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1063387246503846,
+      "learning_rate": 9.757554246007823e-06,
+      "loss": 0.3961,
+      "step": 5202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0571934390862405,
+      "learning_rate": 9.75743302803392e-06,
+      "loss": 0.4027,
+      "step": 5203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4919847218306355,
+      "learning_rate": 9.757311780517646e-06,
+      "loss": 0.1609,
+      "step": 5204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8251344926802917,
+      "learning_rate": 9.757190503459756e-06,
+      "loss": 0.4817,
+      "step": 5205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2774732214963427,
+      "learning_rate": 9.757069196861005e-06,
+      "loss": 0.231,
+      "step": 5206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4088687765692227,
+      "learning_rate": 9.756947860722143e-06,
+      "loss": 0.3374,
+      "step": 5207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.407179504305978,
+      "learning_rate": 9.756826495043926e-06,
+      "loss": 0.8926,
+      "step": 5208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2817870561311193,
+      "learning_rate": 9.756705099827107e-06,
+      "loss": 0.1502,
+      "step": 5209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.464612439118879,
+      "learning_rate": 9.756583675072437e-06,
+      "loss": 0.4941,
+      "step": 5210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5129763783529788,
+      "learning_rate": 9.756462220780674e-06,
+      "loss": 0.1833,
+      "step": 5211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9132818031232353,
+      "learning_rate": 9.75634073695257e-06,
+      "loss": 0.2185,
+      "step": 5212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9611825789842672,
+      "learning_rate": 9.756219223588882e-06,
+      "loss": 0.4212,
+      "step": 5213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1408442990539562,
+      "learning_rate": 9.75609768069036e-06,
+      "loss": 0.6255,
+      "step": 5214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9125593384640909,
+      "learning_rate": 9.75597610825776e-06,
+      "loss": 0.2589,
+      "step": 5215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5619996652618053,
+      "learning_rate": 9.755854506291842e-06,
+      "loss": 0.2594,
+      "step": 5216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5304635702106864,
+      "learning_rate": 9.755732874793356e-06,
+      "loss": 0.1987,
+      "step": 5217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3086415894147616,
+      "learning_rate": 9.755611213763057e-06,
+      "loss": 0.2329,
+      "step": 5218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5683189923703857,
+      "learning_rate": 9.755489523201704e-06,
+      "loss": 0.3148,
+      "step": 5219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6740425822661258,
+      "learning_rate": 9.75536780311005e-06,
+      "loss": 0.2217,
+      "step": 5220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9231688555879591,
+      "learning_rate": 9.755246053488852e-06,
+      "loss": 0.5257,
+      "step": 5221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.653165383016949,
+      "learning_rate": 9.755124274338865e-06,
+      "loss": 0.3045,
+      "step": 5222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7080601400099227,
+      "learning_rate": 9.755002465660848e-06,
+      "loss": 0.3587,
+      "step": 5223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.731732546352028,
+      "learning_rate": 9.754880627455553e-06,
+      "loss": 0.474,
+      "step": 5224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0033503127476076,
+      "learning_rate": 9.754758759723739e-06,
+      "loss": 0.6086,
+      "step": 5225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.28307964944386,
+      "learning_rate": 9.754636862466164e-06,
+      "loss": 0.1678,
+      "step": 5226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4499082338957696,
+      "learning_rate": 9.754514935683583e-06,
+      "loss": 0.4088,
+      "step": 5227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9362510512671134,
+      "learning_rate": 9.754392979376752e-06,
+      "loss": 0.3115,
+      "step": 5228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7283762169779384,
+      "learning_rate": 9.754270993546434e-06,
+      "loss": 0.2564,
+      "step": 5229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.336989202122411,
+      "learning_rate": 9.754148978193378e-06,
+      "loss": 0.2188,
+      "step": 5230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2069683112552645,
+      "learning_rate": 9.754026933318348e-06,
+      "loss": 0.199,
+      "step": 5231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7910869338754403,
+      "learning_rate": 9.7539048589221e-06,
+      "loss": 0.562,
+      "step": 5232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2372944988743515,
+      "learning_rate": 9.753782755005395e-06,
+      "loss": 0.2375,
+      "step": 5233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5219203485942765,
+      "learning_rate": 9.753660621568984e-06,
+      "loss": 0.3028,
+      "step": 5234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5886408451783993,
+      "learning_rate": 9.753538458613632e-06,
+      "loss": 0.3193,
+      "step": 5235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9630458813088372,
+      "learning_rate": 9.753416266140094e-06,
+      "loss": 0.2195,
+      "step": 5236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7924665233845831,
+      "learning_rate": 9.753294044149133e-06,
+      "loss": 0.4615,
+      "step": 5237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0898827337313728,
+      "learning_rate": 9.753171792641503e-06,
+      "loss": 0.2229,
+      "step": 5238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.833923172259611,
+      "learning_rate": 9.753049511617965e-06,
+      "loss": 0.3606,
+      "step": 5239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5235394174767942,
+      "learning_rate": 9.752927201079279e-06,
+      "loss": 0.26,
+      "step": 5240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7189131225859111,
+      "learning_rate": 9.752804861026204e-06,
+      "loss": 0.2325,
+      "step": 5241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5783550311975099,
+      "learning_rate": 9.7526824914595e-06,
+      "loss": 0.2428,
+      "step": 5242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4285834745129142,
+      "learning_rate": 9.752560092379928e-06,
+      "loss": 0.2385,
+      "step": 5243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0593420547521974,
+      "learning_rate": 9.752437663788246e-06,
+      "loss": 0.1673,
+      "step": 5244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.531790623304083,
+      "learning_rate": 9.752315205685213e-06,
+      "loss": 0.2752,
+      "step": 5245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6027000992578033,
+      "learning_rate": 9.752192718071594e-06,
+      "loss": 0.3725,
+      "step": 5246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5678974910902987,
+      "learning_rate": 9.752070200948145e-06,
+      "loss": 0.3175,
+      "step": 5247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9763870613734265,
+      "learning_rate": 9.75194765431563e-06,
+      "loss": 0.5436,
+      "step": 5248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7569436384334483,
+      "learning_rate": 9.751825078174809e-06,
+      "loss": 0.3968,
+      "step": 5249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.353443457591003,
+      "learning_rate": 9.751702472526443e-06,
+      "loss": 0.2797,
+      "step": 5250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4481502344285544,
+      "learning_rate": 9.751579837371292e-06,
+      "loss": 0.2482,
+      "step": 5251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8788216506406756,
+      "learning_rate": 9.751457172710122e-06,
+      "loss": 0.5607,
+      "step": 5252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2898454311459249,
+      "learning_rate": 9.751334478543688e-06,
+      "loss": 0.1693,
+      "step": 5253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8935061011295424,
+      "learning_rate": 9.751211754872758e-06,
+      "loss": 0.5258,
+      "step": 5254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.755706326409577,
+      "learning_rate": 9.75108900169809e-06,
+      "loss": 0.2336,
+      "step": 5255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5248994610429696,
+      "learning_rate": 9.750966219020447e-06,
+      "loss": 0.4053,
+      "step": 5256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.301837243348958,
+      "learning_rate": 9.750843406840595e-06,
+      "loss": 0.1551,
+      "step": 5257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1386567764097637,
+      "learning_rate": 9.750720565159293e-06,
+      "loss": 0.495,
+      "step": 5258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6514133534895423,
+      "learning_rate": 9.750597693977304e-06,
+      "loss": 0.2427,
+      "step": 5259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9044963243158093,
+      "learning_rate": 9.750474793295393e-06,
+      "loss": 0.5658,
+      "step": 5260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4614364159704054,
+      "learning_rate": 9.75035186311432e-06,
+      "loss": 0.3717,
+      "step": 5261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.40043124854381,
+      "learning_rate": 9.75022890343485e-06,
+      "loss": 0.8048,
+      "step": 5262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7988692950943608,
+      "learning_rate": 9.75010591425775e-06,
+      "loss": 0.4995,
+      "step": 5263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6162738188528525,
+      "learning_rate": 9.749982895583779e-06,
+      "loss": 0.2197,
+      "step": 5264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.08823947115463,
+      "learning_rate": 9.749859847413701e-06,
+      "loss": 0.5398,
+      "step": 5265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7777155141515162,
+      "learning_rate": 9.749736769748283e-06,
+      "loss": 0.3802,
+      "step": 5266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7153815032400952,
+      "learning_rate": 9.749613662588287e-06,
+      "loss": 0.3693,
+      "step": 5267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8407285065360397,
+      "learning_rate": 9.749490525934478e-06,
+      "loss": 0.3472,
+      "step": 5268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9967273968035406,
+      "learning_rate": 9.749367359787622e-06,
+      "loss": 0.438,
+      "step": 5269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.292716508765133,
+      "learning_rate": 9.749244164148483e-06,
+      "loss": 0.1661,
+      "step": 5270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5553601980531282,
+      "learning_rate": 9.749120939017824e-06,
+      "loss": 0.3855,
+      "step": 5271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5390473185918443,
+      "learning_rate": 9.748997684396414e-06,
+      "loss": 0.2889,
+      "step": 5272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6840622143949577,
+      "learning_rate": 9.748874400285016e-06,
+      "loss": 0.296,
+      "step": 5273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9246482262968676,
+      "learning_rate": 9.748751086684396e-06,
+      "loss": 0.3855,
+      "step": 5274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1105069141531236,
+      "learning_rate": 9.748627743595317e-06,
+      "loss": 0.1305,
+      "step": 5275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9660759857646675,
+      "learning_rate": 9.748504371018552e-06,
+      "loss": 0.3075,
+      "step": 5276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5981170052856906,
+      "learning_rate": 9.74838096895486e-06,
+      "loss": 0.2344,
+      "step": 5277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7624065490222678,
+      "learning_rate": 9.74825753740501e-06,
+      "loss": 0.3401,
+      "step": 5278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8417519071394863,
+      "learning_rate": 9.748134076369768e-06,
+      "loss": 0.4069,
+      "step": 5279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.612828870160643,
+      "learning_rate": 9.748010585849904e-06,
+      "loss": 0.2875,
+      "step": 5280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5569351090045374,
+      "learning_rate": 9.747887065846178e-06,
+      "loss": 0.3887,
+      "step": 5281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7594156107695942,
+      "learning_rate": 9.747763516359362e-06,
+      "loss": 0.2128,
+      "step": 5282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7496440218854663,
+      "learning_rate": 9.747639937390224e-06,
+      "loss": 0.274,
+      "step": 5283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4918841757729895,
+      "learning_rate": 9.747516328939527e-06,
+      "loss": 0.2888,
+      "step": 5284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0734509333626487,
+      "learning_rate": 9.747392691008044e-06,
+      "loss": 0.1484,
+      "step": 5285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6218373879534818,
+      "learning_rate": 9.74726902359654e-06,
+      "loss": 0.1646,
+      "step": 5286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6736107354561613,
+      "learning_rate": 9.74714532670578e-06,
+      "loss": 0.4964,
+      "step": 5287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4959009384929283,
+      "learning_rate": 9.747021600336537e-06,
+      "loss": 0.2753,
+      "step": 5288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.967317483863125,
+      "learning_rate": 9.746897844489578e-06,
+      "loss": 0.5202,
+      "step": 5289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8816098063722949,
+      "learning_rate": 9.746774059165668e-06,
+      "loss": 0.1699,
+      "step": 5290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5052569677378222,
+      "learning_rate": 9.74665024436558e-06,
+      "loss": 0.2537,
+      "step": 5291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9116703616598227,
+      "learning_rate": 9.746526400090083e-06,
+      "loss": 0.4467,
+      "step": 5292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.11418416201118,
+      "learning_rate": 9.746402526339942e-06,
+      "loss": 0.4255,
+      "step": 5293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1589522970277708,
+      "learning_rate": 9.74627862311593e-06,
+      "loss": 0.1691,
+      "step": 5294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.851664057493589,
+      "learning_rate": 9.746154690418814e-06,
+      "loss": 0.2272,
+      "step": 5295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.73983358441018,
+      "learning_rate": 9.746030728249365e-06,
+      "loss": 0.4087,
+      "step": 5296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8067056386436091,
+      "learning_rate": 9.745906736608353e-06,
+      "loss": 0.42,
+      "step": 5297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5407221216071618,
+      "learning_rate": 9.745782715496545e-06,
+      "loss": 0.2579,
+      "step": 5298
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8100179498398903,
+      "learning_rate": 9.745658664914716e-06,
+      "loss": 0.4801,
+      "step": 5299
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.452269450505568,
+      "learning_rate": 9.745534584863632e-06,
+      "loss": 0.3139,
+      "step": 5300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.508263238925582,
+      "learning_rate": 9.745410475344066e-06,
+      "loss": 0.1956,
+      "step": 5301
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.928000739960855,
+      "learning_rate": 9.745286336356788e-06,
+      "loss": 0.4522,
+      "step": 5302
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.920350881554574,
+      "learning_rate": 9.74516216790257e-06,
+      "loss": 0.3168,
+      "step": 5303
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7843605121084518,
+      "learning_rate": 9.74503796998218e-06,
+      "loss": 0.4595,
+      "step": 5304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1495475814231098,
+      "learning_rate": 9.744913742596391e-06,
+      "loss": 0.196,
+      "step": 5305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6475511219822305,
+      "learning_rate": 9.744789485745975e-06,
+      "loss": 0.4391,
+      "step": 5306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2814610935397068,
+      "learning_rate": 9.744665199431704e-06,
+      "loss": 0.3397,
+      "step": 5307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7256855442074308,
+      "learning_rate": 9.744540883654348e-06,
+      "loss": 0.3635,
+      "step": 5308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.58997264780054,
+      "learning_rate": 9.74441653841468e-06,
+      "loss": 0.3814,
+      "step": 5309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9343986608386234,
+      "learning_rate": 9.744292163713473e-06,
+      "loss": 0.5596,
+      "step": 5310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5920043388860488,
+      "learning_rate": 9.744167759551497e-06,
+      "loss": 0.3865,
+      "step": 5311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4191359577632465,
+      "learning_rate": 9.744043325929526e-06,
+      "loss": 0.1886,
+      "step": 5312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7387868656583105,
+      "learning_rate": 9.743918862848334e-06,
+      "loss": 0.905,
+      "step": 5313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1214730100361567,
+      "learning_rate": 9.743794370308691e-06,
+      "loss": 0.2482,
+      "step": 5314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6545428483898543,
+      "learning_rate": 9.743669848311372e-06,
+      "loss": 0.3697,
+      "step": 5315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.586726574899602,
+      "learning_rate": 9.743545296857149e-06,
+      "loss": 0.2706,
+      "step": 5316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6671337186045219,
+      "learning_rate": 9.743420715946798e-06,
+      "loss": 0.2363,
+      "step": 5317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9611105481576367,
+      "learning_rate": 9.74329610558109e-06,
+      "loss": 0.4276,
+      "step": 5318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7586419246791098,
+      "learning_rate": 9.7431714657608e-06,
+      "loss": 0.5417,
+      "step": 5319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4029495344068563,
+      "learning_rate": 9.743046796486699e-06,
+      "loss": 0.2926,
+      "step": 5320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3084022324890412,
+      "learning_rate": 9.742922097759567e-06,
+      "loss": 0.2973,
+      "step": 5321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5733533615148163,
+      "learning_rate": 9.742797369580173e-06,
+      "loss": 0.2464,
+      "step": 5322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4589278007170694,
+      "learning_rate": 9.742672611949294e-06,
+      "loss": 0.3793,
+      "step": 5323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6408833254936184,
+      "learning_rate": 9.742547824867705e-06,
+      "loss": 0.3772,
+      "step": 5324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5100452372671938,
+      "learning_rate": 9.74242300833618e-06,
+      "loss": 0.1663,
+      "step": 5325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.611647390467622,
+      "learning_rate": 9.742298162355495e-06,
+      "loss": 0.4062,
+      "step": 5326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9667867756861757,
+      "learning_rate": 9.742173286926423e-06,
+      "loss": 0.1933,
+      "step": 5327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7651698614573077,
+      "learning_rate": 9.742048382049742e-06,
+      "loss": 0.4663,
+      "step": 5328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2528940345882846,
+      "learning_rate": 9.741923447726227e-06,
+      "loss": 0.1323,
+      "step": 5329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6933106940112739,
+      "learning_rate": 9.741798483956652e-06,
+      "loss": 0.5214,
+      "step": 5330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2532483184096619,
+      "learning_rate": 9.741673490741795e-06,
+      "loss": 0.1752,
+      "step": 5331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5540848335457833,
+      "learning_rate": 9.741548468082432e-06,
+      "loss": 0.3055,
+      "step": 5332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.282944390082316,
+      "learning_rate": 9.741423415979338e-06,
+      "loss": 0.2544,
+      "step": 5333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6986539493968178,
+      "learning_rate": 9.741298334433292e-06,
+      "loss": 0.3327,
+      "step": 5334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6797254247597901,
+      "learning_rate": 9.74117322344507e-06,
+      "loss": 0.3547,
+      "step": 5335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9303041237415783,
+      "learning_rate": 9.741048083015444e-06,
+      "loss": 0.3349,
+      "step": 5336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5101281356365324,
+      "learning_rate": 9.7409229131452e-06,
+      "loss": 0.2252,
+      "step": 5337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4556989254366541,
+      "learning_rate": 9.740797713835109e-06,
+      "loss": 0.1573,
+      "step": 5338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7738517684889696,
+      "learning_rate": 9.74067248508595e-06,
+      "loss": 0.3993,
+      "step": 5339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8239863858297862,
+      "learning_rate": 9.740547226898498e-06,
+      "loss": 0.3792,
+      "step": 5340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9457695359003255,
+      "learning_rate": 9.740421939273537e-06,
+      "loss": 0.5809,
+      "step": 5341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0474942704703962,
+      "learning_rate": 9.740296622211837e-06,
+      "loss": 0.578,
+      "step": 5342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.647364719296847,
+      "learning_rate": 9.740171275714186e-06,
+      "loss": 0.3409,
+      "step": 5343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1944156593200015,
+      "learning_rate": 9.740045899781353e-06,
+      "loss": 0.4465,
+      "step": 5344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7932348324009137,
+      "learning_rate": 9.739920494414122e-06,
+      "loss": 0.4898,
+      "step": 5345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8206468201471293,
+      "learning_rate": 9.73979505961327e-06,
+      "loss": 0.2102,
+      "step": 5346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6068135972775075,
+      "learning_rate": 9.739669595379577e-06,
+      "loss": 0.1838,
+      "step": 5347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8793022210396573,
+      "learning_rate": 9.73954410171382e-06,
+      "loss": 0.2473,
+      "step": 5348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6170506512747633,
+      "learning_rate": 9.73941857861678e-06,
+      "loss": 0.7011,
+      "step": 5349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0336282855401313,
+      "learning_rate": 9.739293026089236e-06,
+      "loss": 0.425,
+      "step": 5350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9984132313391865,
+      "learning_rate": 9.739167444131968e-06,
+      "loss": 0.4858,
+      "step": 5351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8914313724744918,
+      "learning_rate": 9.739041832745754e-06,
+      "loss": 0.4892,
+      "step": 5352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2907172881282685,
+      "learning_rate": 9.738916191931377e-06,
+      "loss": 0.1696,
+      "step": 5353
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.05371386881069,
+      "learning_rate": 9.738790521689615e-06,
+      "loss": 0.6805,
+      "step": 5354
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1786861546507694,
+      "learning_rate": 9.738664822021249e-06,
+      "loss": 0.4881,
+      "step": 5355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2027673623249775,
+      "learning_rate": 9.73853909292706e-06,
+      "loss": 0.2279,
+      "step": 5356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4303644276636363,
+      "learning_rate": 9.738413334407829e-06,
+      "loss": 0.3068,
+      "step": 5357
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9547317304722225,
+      "learning_rate": 9.738287546464335e-06,
+      "loss": 0.3702,
+      "step": 5358
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9561133906772998,
+      "learning_rate": 9.738161729097361e-06,
+      "loss": 0.3958,
+      "step": 5359
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.694810449319169,
+      "learning_rate": 9.738035882307688e-06,
+      "loss": 0.4235,
+      "step": 5360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2920966236325797,
+      "learning_rate": 9.737910006096096e-06,
+      "loss": 0.8166,
+      "step": 5361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.916082686282116,
+      "learning_rate": 9.737784100463369e-06,
+      "loss": 0.4426,
+      "step": 5362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9883918731433015,
+      "learning_rate": 9.737658165410288e-06,
+      "loss": 0.3678,
+      "step": 5363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2887694134010839,
+      "learning_rate": 9.737532200937634e-06,
+      "loss": 0.1998,
+      "step": 5364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6680296187718464,
+      "learning_rate": 9.73740620704619e-06,
+      "loss": 0.4474,
+      "step": 5365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.585657104359641,
+      "learning_rate": 9.737280183736736e-06,
+      "loss": 0.3297,
+      "step": 5366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5250384135372042,
+      "learning_rate": 9.73715413101006e-06,
+      "loss": 0.2844,
+      "step": 5367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.726700303675848,
+      "learning_rate": 9.737028048866942e-06,
+      "loss": 0.1959,
+      "step": 5368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.175183828850543,
+      "learning_rate": 9.736901937308163e-06,
+      "loss": 0.4861,
+      "step": 5369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9990054354767899,
+      "learning_rate": 9.736775796334508e-06,
+      "loss": 0.3657,
+      "step": 5370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6573336043487417,
+      "learning_rate": 9.736649625946758e-06,
+      "loss": 0.2416,
+      "step": 5371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9975245176106875,
+      "learning_rate": 9.736523426145701e-06,
+      "loss": 0.4862,
+      "step": 5372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.272911094618473,
+      "learning_rate": 9.736397196932117e-06,
+      "loss": 0.2403,
+      "step": 5373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3971157340707245,
+      "learning_rate": 9.736270938306792e-06,
+      "loss": 0.1792,
+      "step": 5374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7914440470458308,
+      "learning_rate": 9.736144650270507e-06,
+      "loss": 0.3406,
+      "step": 5375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0409915484709205,
+      "learning_rate": 9.736018332824052e-06,
+      "loss": 0.143,
+      "step": 5376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9576850834428987,
+      "learning_rate": 9.735891985968204e-06,
+      "loss": 0.3636,
+      "step": 5377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8944474073911557,
+      "learning_rate": 9.735765609703752e-06,
+      "loss": 0.2612,
+      "step": 5378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6516402891203916,
+      "learning_rate": 9.735639204031479e-06,
+      "loss": 0.3577,
+      "step": 5379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7187710500641895,
+      "learning_rate": 9.735512768952173e-06,
+      "loss": 0.251,
+      "step": 5380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.23220378915183,
+      "learning_rate": 9.735386304466616e-06,
+      "loss": 0.5461,
+      "step": 5381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6728455843487404,
+      "learning_rate": 9.735259810575595e-06,
+      "loss": 0.2959,
+      "step": 5382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6861277183933872,
+      "learning_rate": 9.735133287279894e-06,
+      "loss": 0.2903,
+      "step": 5383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1926248055780877,
+      "learning_rate": 9.7350067345803e-06,
+      "loss": 0.4938,
+      "step": 5384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.75445794772061,
+      "learning_rate": 9.734880152477597e-06,
+      "loss": 0.4749,
+      "step": 5385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4694322006128728,
+      "learning_rate": 9.734753540972574e-06,
+      "loss": 0.1801,
+      "step": 5386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.88694327461825,
+      "learning_rate": 9.734626900066015e-06,
+      "loss": 0.3453,
+      "step": 5387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6711629750669843,
+      "learning_rate": 9.734500229758708e-06,
+      "loss": 0.2094,
+      "step": 5388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8067191524450734,
+      "learning_rate": 9.734373530051437e-06,
+      "loss": 0.4662,
+      "step": 5389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9763808562756733,
+      "learning_rate": 9.73424680094499e-06,
+      "loss": 0.5295,
+      "step": 5390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6730092158633771,
+      "learning_rate": 9.734120042440157e-06,
+      "loss": 0.3455,
+      "step": 5391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5643613219263293,
+      "learning_rate": 9.733993254537722e-06,
+      "loss": 0.2917,
+      "step": 5392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.996112300838857,
+      "learning_rate": 9.73386643723847e-06,
+      "loss": 0.3288,
+      "step": 5393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9759186259017145,
+      "learning_rate": 9.733739590543195e-06,
+      "loss": 0.3418,
+      "step": 5394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.606992551822603,
+      "learning_rate": 9.733612714452678e-06,
+      "loss": 0.1919,
+      "step": 5395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3268702807097708,
+      "learning_rate": 9.733485808967712e-06,
+      "loss": 0.3322,
+      "step": 5396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.623466001486228,
+      "learning_rate": 9.733358874089081e-06,
+      "loss": 0.3122,
+      "step": 5397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.054685904498623,
+      "learning_rate": 9.733231909817576e-06,
+      "loss": 0.4443,
+      "step": 5398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.021871687231932,
+      "learning_rate": 9.733104916153984e-06,
+      "loss": 0.1955,
+      "step": 5399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8644751556456776,
+      "learning_rate": 9.732977893099094e-06,
+      "loss": 0.3525,
+      "step": 5400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4944715794058963,
+      "learning_rate": 9.732850840653696e-06,
+      "loss": 0.1827,
+      "step": 5401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8144388064834767,
+      "learning_rate": 9.732723758818576e-06,
+      "loss": 0.287,
+      "step": 5402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1185399447884983,
+      "learning_rate": 9.732596647594527e-06,
+      "loss": 0.4795,
+      "step": 5403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0412537358302956,
+      "learning_rate": 9.732469506982337e-06,
+      "loss": 0.5344,
+      "step": 5404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.489033158746593,
+      "learning_rate": 9.732342336982791e-06,
+      "loss": 0.4654,
+      "step": 5405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3085047278798028,
+      "learning_rate": 9.732215137596686e-06,
+      "loss": 0.242,
+      "step": 5406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7921717790215106,
+      "learning_rate": 9.732087908824808e-06,
+      "loss": 0.4868,
+      "step": 5407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6529755477916943,
+      "learning_rate": 9.731960650667947e-06,
+      "loss": 0.4786,
+      "step": 5408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5287760192266338,
+      "learning_rate": 9.731833363126893e-06,
+      "loss": 0.3103,
+      "step": 5409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.984357017150256,
+      "learning_rate": 9.731706046202439e-06,
+      "loss": 0.328,
+      "step": 5410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1005289521372976,
+      "learning_rate": 9.73157869989537e-06,
+      "loss": 0.4979,
+      "step": 5411
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6822865671094325,
+      "learning_rate": 9.731451324206485e-06,
+      "loss": 0.5435,
+      "step": 5412
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5958907671902693,
+      "learning_rate": 9.731323919136567e-06,
+      "loss": 0.4755,
+      "step": 5413
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8187605024104998,
+      "learning_rate": 9.731196484686413e-06,
+      "loss": 0.2195,
+      "step": 5414
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.123088336616928,
+      "learning_rate": 9.73106902085681e-06,
+      "loss": 0.2063,
+      "step": 5415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.006504976404247,
+      "learning_rate": 9.730941527648553e-06,
+      "loss": 0.5265,
+      "step": 5416
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6033622003865071,
+      "learning_rate": 9.730814005062433e-06,
+      "loss": 0.2185,
+      "step": 5417
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1063850900020005,
+      "learning_rate": 9.730686453099238e-06,
+      "loss": 0.4161,
+      "step": 5418
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.733386902031998,
+      "learning_rate": 9.730558871759765e-06,
+      "loss": 0.236,
+      "step": 5419
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9146968257571721,
+      "learning_rate": 9.730431261044804e-06,
+      "loss": 0.5161,
+      "step": 5420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6849262679700705,
+      "learning_rate": 9.730303620955148e-06,
+      "loss": 0.365,
+      "step": 5421
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9618315143310762,
+      "learning_rate": 9.73017595149159e-06,
+      "loss": 0.1314,
+      "step": 5422
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6941591439918235,
+      "learning_rate": 9.730048252654922e-06,
+      "loss": 0.2598,
+      "step": 5423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5581208905947306,
+      "learning_rate": 9.729920524445938e-06,
+      "loss": 0.2584,
+      "step": 5424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5704232575371966,
+      "learning_rate": 9.729792766865429e-06,
+      "loss": 0.3446,
+      "step": 5425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.895012210476974,
+      "learning_rate": 9.729664979914188e-06,
+      "loss": 0.5023,
+      "step": 5426
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4555527585019057,
+      "learning_rate": 9.729537163593013e-06,
+      "loss": 0.2039,
+      "step": 5427
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2985530059957793,
+      "learning_rate": 9.729409317902693e-06,
+      "loss": 0.1777,
+      "step": 5428
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.32659079209252,
+      "learning_rate": 9.729281442844026e-06,
+      "loss": 0.3371,
+      "step": 5429
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0058983434069546,
+      "learning_rate": 9.729153538417802e-06,
+      "loss": 0.3765,
+      "step": 5430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8564737308741097,
+      "learning_rate": 9.72902560462482e-06,
+      "loss": 0.4766,
+      "step": 5431
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9169455629140776,
+      "learning_rate": 9.728897641465868e-06,
+      "loss": 0.1843,
+      "step": 5432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.460120110143627,
+      "learning_rate": 9.728769648941746e-06,
+      "loss": 0.2176,
+      "step": 5433
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6442225431525204,
+      "learning_rate": 9.728641627053245e-06,
+      "loss": 0.4212,
+      "step": 5434
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.217173506549445,
+      "learning_rate": 9.728513575801163e-06,
+      "loss": 0.1857,
+      "step": 5435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9496197870855827,
+      "learning_rate": 9.728385495186297e-06,
+      "loss": 0.1441,
+      "step": 5436
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.924622053348877,
+      "learning_rate": 9.728257385209436e-06,
+      "loss": 0.1969,
+      "step": 5437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9558043717317926,
+      "learning_rate": 9.728129245871381e-06,
+      "loss": 0.2607,
+      "step": 5438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.151282389766585,
+      "learning_rate": 9.728001077172925e-06,
+      "loss": 0.3165,
+      "step": 5439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5909498149130696,
+      "learning_rate": 9.727872879114863e-06,
+      "loss": 0.2782,
+      "step": 5440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8229664091862094,
+      "learning_rate": 9.727744651697994e-06,
+      "loss": 0.3956,
+      "step": 5441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8221428145946799,
+      "learning_rate": 9.727616394923112e-06,
+      "loss": 0.4565,
+      "step": 5442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.497362589812637,
+      "learning_rate": 9.727488108791016e-06,
+      "loss": 0.3976,
+      "step": 5443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9078463491237883,
+      "learning_rate": 9.7273597933025e-06,
+      "loss": 0.545,
+      "step": 5444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5317331058024417,
+      "learning_rate": 9.72723144845836e-06,
+      "loss": 0.1728,
+      "step": 5445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.924877941454358,
+      "learning_rate": 9.727103074259399e-06,
+      "loss": 0.597,
+      "step": 5446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.956301235100853,
+      "learning_rate": 9.726974670706406e-06,
+      "loss": 0.4573,
+      "step": 5447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.828217009633606,
+      "learning_rate": 9.726846237800184e-06,
+      "loss": 0.5088,
+      "step": 5448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9889377941465458,
+      "learning_rate": 9.72671777554153e-06,
+      "loss": 0.474,
+      "step": 5449
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.352969230445607,
+      "learning_rate": 9.726589283931239e-06,
+      "loss": 0.4962,
+      "step": 5450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.471712186741817,
+      "learning_rate": 9.72646076297011e-06,
+      "loss": 0.463,
+      "step": 5451
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9311582466353567,
+      "learning_rate": 9.726332212658942e-06,
+      "loss": 0.356,
+      "step": 5452
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7221115358309078,
+      "learning_rate": 9.726203632998535e-06,
+      "loss": 0.4448,
+      "step": 5453
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3418979578937065,
+      "learning_rate": 9.726075023989682e-06,
+      "loss": 0.3707,
+      "step": 5454
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7935417273691971,
+      "learning_rate": 9.725946385633186e-06,
+      "loss": 0.4883,
+      "step": 5455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7053691183110553,
+      "learning_rate": 9.725817717929845e-06,
+      "loss": 0.3667,
+      "step": 5456
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.828415863829684,
+      "learning_rate": 9.725689020880457e-06,
+      "loss": 0.4604,
+      "step": 5457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6992668671207896,
+      "learning_rate": 9.725560294485823e-06,
+      "loss": 0.3599,
+      "step": 5458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1647860979054998,
+      "learning_rate": 9.72543153874674e-06,
+      "loss": 0.2378,
+      "step": 5459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3390173059245938,
+      "learning_rate": 9.72530275366401e-06,
+      "loss": 0.195,
+      "step": 5460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.039294817906327,
+      "learning_rate": 9.72517393923843e-06,
+      "loss": 0.641,
+      "step": 5461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2065385386656127,
+      "learning_rate": 9.725045095470802e-06,
+      "loss": 0.5083,
+      "step": 5462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6808443211070412,
+      "learning_rate": 9.724916222361925e-06,
+      "loss": 0.4584,
+      "step": 5463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8540763649935634,
+      "learning_rate": 9.7247873199126e-06,
+      "loss": 0.2998,
+      "step": 5464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7382014759917803,
+      "learning_rate": 9.724658388123627e-06,
+      "loss": 0.3477,
+      "step": 5465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0705599951174996,
+      "learning_rate": 9.724529426995808e-06,
+      "loss": 0.1771,
+      "step": 5466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.199939936510576,
+      "learning_rate": 9.724400436529942e-06,
+      "loss": 0.2297,
+      "step": 5467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3555589502977485,
+      "learning_rate": 9.724271416726829e-06,
+      "loss": 0.3338,
+      "step": 5468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9398307816759073,
+      "learning_rate": 9.724142367587273e-06,
+      "loss": 0.5183,
+      "step": 5469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5567758566055432,
+      "learning_rate": 9.724013289112075e-06,
+      "loss": 0.1818,
+      "step": 5470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0245797236260503,
+      "learning_rate": 9.723884181302035e-06,
+      "loss": 0.2328,
+      "step": 5471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.450321312770523,
+      "learning_rate": 9.723755044157954e-06,
+      "loss": 0.2966,
+      "step": 5472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.657122584572619,
+      "learning_rate": 9.723625877680635e-06,
+      "loss": 0.3853,
+      "step": 5473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6655356687384129,
+      "learning_rate": 9.723496681870882e-06,
+      "loss": 0.3891,
+      "step": 5474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0630097680411867,
+      "learning_rate": 9.723367456729497e-06,
+      "loss": 0.2296,
+      "step": 5475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.667925608521481,
+      "learning_rate": 9.72323820225728e-06,
+      "loss": 0.3579,
+      "step": 5476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7836317777652597,
+      "learning_rate": 9.723108918455033e-06,
+      "loss": 0.3638,
+      "step": 5477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8988812088976923,
+      "learning_rate": 9.722979605323562e-06,
+      "loss": 0.3913,
+      "step": 5478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5463107457483074,
+      "learning_rate": 9.72285026286367e-06,
+      "loss": 0.3084,
+      "step": 5479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8575368097567715,
+      "learning_rate": 9.722720891076156e-06,
+      "loss": 0.3181,
+      "step": 5480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.793604926978315,
+      "learning_rate": 9.722591489961829e-06,
+      "loss": 0.2454,
+      "step": 5481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8330110886249318,
+      "learning_rate": 9.722462059521486e-06,
+      "loss": 0.393,
+      "step": 5482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3861262922925117,
+      "learning_rate": 9.722332599755937e-06,
+      "loss": 0.305,
+      "step": 5483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8602577125726252,
+      "learning_rate": 9.722203110665983e-06,
+      "loss": 0.493,
+      "step": 5484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7467164373685364,
+      "learning_rate": 9.722073592252427e-06,
+      "loss": 0.4102,
+      "step": 5485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4767286040716257,
+      "learning_rate": 9.721944044516077e-06,
+      "loss": 0.3365,
+      "step": 5486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3430184990332406,
+      "learning_rate": 9.721814467457733e-06,
+      "loss": 0.142,
+      "step": 5487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.62255702332918,
+      "learning_rate": 9.721684861078202e-06,
+      "loss": 0.4212,
+      "step": 5488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.484886258319779,
+      "learning_rate": 9.72155522537829e-06,
+      "loss": 0.3148,
+      "step": 5489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.334005355669971,
+      "learning_rate": 9.7214255603588e-06,
+      "loss": 0.3613,
+      "step": 5490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3037122719289973,
+      "learning_rate": 9.721295866020537e-06,
+      "loss": 0.3127,
+      "step": 5491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.343681439047901,
+      "learning_rate": 9.721166142364307e-06,
+      "loss": 0.3097,
+      "step": 5492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8047095252010001,
+      "learning_rate": 9.721036389390916e-06,
+      "loss": 0.4386,
+      "step": 5493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.526324280705043,
+      "learning_rate": 9.72090660710117e-06,
+      "loss": 0.2552,
+      "step": 5494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.197575906453385,
+      "learning_rate": 9.720776795495874e-06,
+      "loss": 0.2195,
+      "step": 5495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.292686999298893,
+      "learning_rate": 9.720646954575834e-06,
+      "loss": 0.2196,
+      "step": 5496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.072104560136819,
+      "learning_rate": 9.720517084341857e-06,
+      "loss": 0.4424,
+      "step": 5497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6045908604450871,
+      "learning_rate": 9.72038718479475e-06,
+      "loss": 0.3257,
+      "step": 5498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9591190143538306,
+      "learning_rate": 9.720257255935316e-06,
+      "loss": 0.2319,
+      "step": 5499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5902303136681402,
+      "learning_rate": 9.720127297764367e-06,
+      "loss": 0.3077,
+      "step": 5500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4280861526856803,
+      "learning_rate": 9.719997310282709e-06,
+      "loss": 0.2392,
+      "step": 5501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1004198293303284,
+      "learning_rate": 9.719867293491146e-06,
+      "loss": 0.5944,
+      "step": 5502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6340049856346521,
+      "learning_rate": 9.719737247390486e-06,
+      "loss": 0.3492,
+      "step": 5503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0597527442614747,
+      "learning_rate": 9.71960717198154e-06,
+      "loss": 0.2022,
+      "step": 5504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.403475152289786,
+      "learning_rate": 9.719477067265113e-06,
+      "loss": 0.4832,
+      "step": 5505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3089072364406322,
+      "learning_rate": 9.71934693324201e-06,
+      "loss": 0.1828,
+      "step": 5506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5829693969869747,
+      "learning_rate": 9.719216769913048e-06,
+      "loss": 0.8327,
+      "step": 5507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6226966516775687,
+      "learning_rate": 9.719086577279027e-06,
+      "loss": 0.2993,
+      "step": 5508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4835291899665042,
+      "learning_rate": 9.718956355340758e-06,
+      "loss": 0.1653,
+      "step": 5509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8674868259961066,
+      "learning_rate": 9.71882610409905e-06,
+      "loss": 0.4017,
+      "step": 5510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.921198326481921,
+      "learning_rate": 9.718695823554712e-06,
+      "loss": 0.7008,
+      "step": 5511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.476170590074347,
+      "learning_rate": 9.718565513708552e-06,
+      "loss": 0.3706,
+      "step": 5512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.004991663130054,
+      "learning_rate": 9.71843517456138e-06,
+      "loss": 0.47,
+      "step": 5513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6397638149688587,
+      "learning_rate": 9.718304806114004e-06,
+      "loss": 0.1743,
+      "step": 5514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7139862410447686,
+      "learning_rate": 9.718174408367239e-06,
+      "loss": 0.3259,
+      "step": 5515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.562894492051228,
+      "learning_rate": 9.718043981321888e-06,
+      "loss": 0.278,
+      "step": 5516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5528268921365043,
+      "learning_rate": 9.717913524978762e-06,
+      "loss": 0.4332,
+      "step": 5517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3905930274331078,
+      "learning_rate": 9.717783039338674e-06,
+      "loss": 0.1567,
+      "step": 5518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8741305978102796,
+      "learning_rate": 9.717652524402432e-06,
+      "loss": 0.3904,
+      "step": 5519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8768645750944617,
+      "learning_rate": 9.717521980170848e-06,
+      "loss": 0.3501,
+      "step": 5520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0618333335134276,
+      "learning_rate": 9.717391406644733e-06,
+      "loss": 0.3404,
+      "step": 5521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7317414780808125,
+      "learning_rate": 9.717260803824895e-06,
+      "loss": 0.2666,
+      "step": 5522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4075900554093155,
+      "learning_rate": 9.717130171712148e-06,
+      "loss": 0.1736,
+      "step": 5523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4953715770136664,
+      "learning_rate": 9.7169995103073e-06,
+      "loss": 0.352,
+      "step": 5524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.157494135953088,
+      "learning_rate": 9.716868819611168e-06,
+      "loss": 0.4486,
+      "step": 5525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4540009575877388,
+      "learning_rate": 9.716738099624556e-06,
+      "loss": 0.1918,
+      "step": 5526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.899584498272964,
+      "learning_rate": 9.716607350348282e-06,
+      "loss": 0.3874,
+      "step": 5527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.811201197504454,
+      "learning_rate": 9.716476571783156e-06,
+      "loss": 0.5045,
+      "step": 5528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.48552320142004,
+      "learning_rate": 9.716345763929988e-06,
+      "loss": 0.4478,
+      "step": 5529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1304140943748002,
+      "learning_rate": 9.716214926789592e-06,
+      "loss": 0.2189,
+      "step": 5530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.058523690842987,
+      "learning_rate": 9.716084060362781e-06,
+      "loss": 0.3368,
+      "step": 5531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9131478206640553,
+      "learning_rate": 9.715953164650367e-06,
+      "loss": 0.4487,
+      "step": 5532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.343898116835676,
+      "learning_rate": 9.715822239653162e-06,
+      "loss": 0.6791,
+      "step": 5533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7174549816025804,
+      "learning_rate": 9.71569128537198e-06,
+      "loss": 0.3618,
+      "step": 5534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.128406871244233,
+      "learning_rate": 9.715560301807635e-06,
+      "loss": 0.168,
+      "step": 5535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.07643514777984,
+      "learning_rate": 9.715429288960939e-06,
+      "loss": 0.408,
+      "step": 5536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7036191087941819,
+      "learning_rate": 9.715298246832705e-06,
+      "loss": 0.2951,
+      "step": 5537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.535508836048469,
+      "learning_rate": 9.71516717542375e-06,
+      "loss": 0.2184,
+      "step": 5538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2036555462703513,
+      "learning_rate": 9.715036074734884e-06,
+      "loss": 0.2875,
+      "step": 5539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0064549560151743,
+      "learning_rate": 9.714904944766922e-06,
+      "loss": 0.1868,
+      "step": 5540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0733821208949097,
+      "learning_rate": 9.714773785520681e-06,
+      "loss": 0.4627,
+      "step": 5541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7267492493226306,
+      "learning_rate": 9.714642596996973e-06,
+      "loss": 0.22,
+      "step": 5542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3746684796395627,
+      "learning_rate": 9.714511379196611e-06,
+      "loss": 0.1735,
+      "step": 5543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3983122300669621,
+      "learning_rate": 9.714380132120414e-06,
+      "loss": 0.2233,
+      "step": 5544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6922448130066137,
+      "learning_rate": 9.714248855769195e-06,
+      "loss": 0.2042,
+      "step": 5545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6949303411627066,
+      "learning_rate": 9.71411755014377e-06,
+      "loss": 0.4113,
+      "step": 5546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1242137253685383,
+      "learning_rate": 9.713986215244952e-06,
+      "loss": 0.3319,
+      "step": 5547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4253705332767552,
+      "learning_rate": 9.713854851073558e-06,
+      "loss": 0.3105,
+      "step": 5548
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0141791806334024,
+      "learning_rate": 9.713723457630405e-06,
+      "loss": 0.6778,
+      "step": 5549
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5501540042140747,
+      "learning_rate": 9.713592034916305e-06,
+      "loss": 0.1956,
+      "step": 5550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3950375618201685,
+      "learning_rate": 9.713460582932079e-06,
+      "loss": 0.2684,
+      "step": 5551
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.130545997827871,
+      "learning_rate": 9.713329101678541e-06,
+      "loss": 0.5695,
+      "step": 5552
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.701077980927783,
+      "learning_rate": 9.713197591156507e-06,
+      "loss": 0.4126,
+      "step": 5553
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4984548179890649,
+      "learning_rate": 9.713066051366796e-06,
+      "loss": 0.1856,
+      "step": 5554
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7657640621819508,
+      "learning_rate": 9.712934482310221e-06,
+      "loss": 0.3273,
+      "step": 5555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3824905718517275,
+      "learning_rate": 9.712802883987602e-06,
+      "loss": 0.1859,
+      "step": 5556
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.272138434679505,
+      "learning_rate": 9.712671256399755e-06,
+      "loss": 0.1396,
+      "step": 5557
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1884395591130223,
+      "learning_rate": 9.712539599547497e-06,
+      "loss": 0.3157,
+      "step": 5558
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.211756672427967,
+      "learning_rate": 9.712407913431648e-06,
+      "loss": 0.5369,
+      "step": 5559
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5863923045507007,
+      "learning_rate": 9.712276198053021e-06,
+      "loss": 0.5078,
+      "step": 5560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.443378374803175,
+      "learning_rate": 9.712144453412439e-06,
+      "loss": 0.1626,
+      "step": 5561
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3091146713336739,
+      "learning_rate": 9.712012679510717e-06,
+      "loss": 0.1796,
+      "step": 5562
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0287527248801442,
+      "learning_rate": 9.711880876348674e-06,
+      "loss": 0.3792,
+      "step": 5563
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2591520282526028,
+      "learning_rate": 9.71174904392713e-06,
+      "loss": 0.4644,
+      "step": 5564
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8176894690200578,
+      "learning_rate": 9.711617182246903e-06,
+      "loss": 0.3151,
+      "step": 5565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.98733991362073,
+      "learning_rate": 9.711485291308809e-06,
+      "loss": 0.4148,
+      "step": 5566
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.861213572480636,
+      "learning_rate": 9.711353371113669e-06,
+      "loss": 0.286,
+      "step": 5567
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5064811832841454,
+      "learning_rate": 9.711221421662302e-06,
+      "loss": 0.3039,
+      "step": 5568
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9806268391912125,
+      "learning_rate": 9.71108944295553e-06,
+      "loss": 0.5462,
+      "step": 5569
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4396507665801366,
+      "learning_rate": 9.710957434994168e-06,
+      "loss": 0.292,
+      "step": 5570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.982845775965318,
+      "learning_rate": 9.71082539777904e-06,
+      "loss": 0.4266,
+      "step": 5571
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3696201079386887,
+      "learning_rate": 9.710693331310963e-06,
+      "loss": 0.2138,
+      "step": 5572
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4687388855939725,
+      "learning_rate": 9.710561235590757e-06,
+      "loss": 0.2259,
+      "step": 5573
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5845974912678824,
+      "learning_rate": 9.710429110619246e-06,
+      "loss": 0.2179,
+      "step": 5574
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2383872268380314,
+      "learning_rate": 9.710296956397245e-06,
+      "loss": 0.551,
+      "step": 5575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.52728271484375,
+      "learning_rate": 9.710164772925578e-06,
+      "loss": 0.1557,
+      "step": 5576
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6665487040308666,
+      "learning_rate": 9.710032560205065e-06,
+      "loss": 0.3567,
+      "step": 5577
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.451638235368687,
+      "learning_rate": 9.709900318236528e-06,
+      "loss": 0.3698,
+      "step": 5578
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8322741260054034,
+      "learning_rate": 9.709768047020787e-06,
+      "loss": 0.3804,
+      "step": 5579
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3023339689948126,
+      "learning_rate": 9.709635746558663e-06,
+      "loss": 0.208,
+      "step": 5580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4100469457051583,
+      "learning_rate": 9.70950341685098e-06,
+      "loss": 0.1541,
+      "step": 5581
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4934972905179358,
+      "learning_rate": 9.709371057898557e-06,
+      "loss": 0.2501,
+      "step": 5582
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8887966947129358,
+      "learning_rate": 9.709238669702216e-06,
+      "loss": 0.4334,
+      "step": 5583
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2026167390490274,
+      "learning_rate": 9.709106252262783e-06,
+      "loss": 0.2139,
+      "step": 5584
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.174655568321073,
+      "learning_rate": 9.708973805581075e-06,
+      "loss": 0.2832,
+      "step": 5585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5274591329167997,
+      "learning_rate": 9.708841329657918e-06,
+      "loss": 0.2171,
+      "step": 5586
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1724239208167744,
+      "learning_rate": 9.708708824494132e-06,
+      "loss": 0.1935,
+      "step": 5587
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7364941100348248,
+      "learning_rate": 9.708576290090543e-06,
+      "loss": 0.5332,
+      "step": 5588
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4622217133659483,
+      "learning_rate": 9.70844372644797e-06,
+      "loss": 0.2835,
+      "step": 5589
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.298883792329591,
+      "learning_rate": 9.70831113356724e-06,
+      "loss": 0.5572,
+      "step": 5590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.259449335777477,
+      "learning_rate": 9.708178511449174e-06,
+      "loss": 0.2605,
+      "step": 5591
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7513961736056047,
+      "learning_rate": 9.708045860094597e-06,
+      "loss": 0.2458,
+      "step": 5592
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6770529417055373,
+      "learning_rate": 9.707913179504331e-06,
+      "loss": 0.6368,
+      "step": 5593
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7911844454773584,
+      "learning_rate": 9.707780469679203e-06,
+      "loss": 0.4986,
+      "step": 5594
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6165745029486867,
+      "learning_rate": 9.707647730620034e-06,
+      "loss": 0.2482,
+      "step": 5595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1768870052233322,
+      "learning_rate": 9.707514962327648e-06,
+      "loss": 0.2641,
+      "step": 5596
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6918926640840168,
+      "learning_rate": 9.707382164802873e-06,
+      "loss": 0.2152,
+      "step": 5597
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4793461000438384,
+      "learning_rate": 9.707249338046531e-06,
+      "loss": 0.2549,
+      "step": 5598
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.439368348361981,
+      "learning_rate": 9.707116482059447e-06,
+      "loss": 0.2091,
+      "step": 5599
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.493246057528163,
+      "learning_rate": 9.706983596842446e-06,
+      "loss": 0.175,
+      "step": 5600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.566342694150219,
+      "learning_rate": 9.706850682396353e-06,
+      "loss": 0.7121,
+      "step": 5601
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8036768382865802,
+      "learning_rate": 9.706717738721993e-06,
+      "loss": 0.3946,
+      "step": 5602
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9825968848338764,
+      "learning_rate": 9.706584765820195e-06,
+      "loss": 0.4339,
+      "step": 5603
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6995975642143513,
+      "learning_rate": 9.70645176369178e-06,
+      "loss": 0.5128,
+      "step": 5604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9576712682980502,
+      "learning_rate": 9.706318732337576e-06,
+      "loss": 0.3726,
+      "step": 5605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.949171413542518,
+      "learning_rate": 9.706185671758411e-06,
+      "loss": 0.3915,
+      "step": 5606
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1529662414232744,
+      "learning_rate": 9.706052581955108e-06,
+      "loss": 0.2036,
+      "step": 5607
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6173641642764662,
+      "learning_rate": 9.705919462928493e-06,
+      "loss": 0.4431,
+      "step": 5608
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8071235011426297,
+      "learning_rate": 9.705786314679397e-06,
+      "loss": 0.4897,
+      "step": 5609
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3884320380605502,
+      "learning_rate": 9.705653137208644e-06,
+      "loss": 0.3613,
+      "step": 5610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.001570113420085,
+      "learning_rate": 9.705519930517059e-06,
+      "loss": 0.4038,
+      "step": 5611
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.047384246056515,
+      "learning_rate": 9.705386694605474e-06,
+      "loss": 0.7205,
+      "step": 5612
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3136983606196366,
+      "learning_rate": 9.705253429474713e-06,
+      "loss": 0.1747,
+      "step": 5613
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8689957326378028,
+      "learning_rate": 9.705120135125603e-06,
+      "loss": 0.5936,
+      "step": 5614
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7027103505514893,
+      "learning_rate": 9.704986811558976e-06,
+      "loss": 0.38,
+      "step": 5615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.923461821097863,
+      "learning_rate": 9.704853458775656e-06,
+      "loss": 0.4755,
+      "step": 5616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.91263319094002,
+      "learning_rate": 9.704720076776471e-06,
+      "loss": 0.4474,
+      "step": 5617
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0203873905896335,
+      "learning_rate": 9.70458666556225e-06,
+      "loss": 0.4333,
+      "step": 5618
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9761377488865848,
+      "learning_rate": 9.70445322513382e-06,
+      "loss": 0.3427,
+      "step": 5619
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.567377215818116,
+      "learning_rate": 9.704319755492016e-06,
+      "loss": 0.1311,
+      "step": 5620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5328844796252883,
+      "learning_rate": 9.704186256637659e-06,
+      "loss": 0.3814,
+      "step": 5621
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.834156722968065,
+      "learning_rate": 9.70405272857158e-06,
+      "loss": 0.6273,
+      "step": 5622
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8503663144121564,
+      "learning_rate": 9.703919171294613e-06,
+      "loss": 0.1693,
+      "step": 5623
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8908331259935698,
+      "learning_rate": 9.703785584807582e-06,
+      "loss": 0.3606,
+      "step": 5624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6063435159913149,
+      "learning_rate": 9.703651969111318e-06,
+      "loss": 0.4597,
+      "step": 5625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.34072252085033,
+      "learning_rate": 9.703518324206652e-06,
+      "loss": 0.2178,
+      "step": 5626
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4510019749700933,
+      "learning_rate": 9.703384650094411e-06,
+      "loss": 0.3588,
+      "step": 5627
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.706019341681858,
+      "learning_rate": 9.703250946775426e-06,
+      "loss": 0.2896,
+      "step": 5628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2775169653475236,
+      "learning_rate": 9.703117214250531e-06,
+      "loss": 0.6557,
+      "step": 5629
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2623792918141088,
+      "learning_rate": 9.702983452520554e-06,
+      "loss": 0.1554,
+      "step": 5630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6479506790719805,
+      "learning_rate": 9.702849661586322e-06,
+      "loss": 0.2295,
+      "step": 5631
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6923626625019312,
+      "learning_rate": 9.702715841448671e-06,
+      "loss": 0.1872,
+      "step": 5632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9774039139895914,
+      "learning_rate": 9.702581992108431e-06,
+      "loss": 0.2372,
+      "step": 5633
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.509187412452008,
+      "learning_rate": 9.70244811356643e-06,
+      "loss": 0.4621,
+      "step": 5634
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2110292123089295,
+      "learning_rate": 9.702314205823504e-06,
+      "loss": 0.1764,
+      "step": 5635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4748639018757252,
+      "learning_rate": 9.70218026888048e-06,
+      "loss": 0.6691,
+      "step": 5636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4176317246616632,
+      "learning_rate": 9.702046302738194e-06,
+      "loss": 0.2112,
+      "step": 5637
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.570366877473125,
+      "learning_rate": 9.701912307397473e-06,
+      "loss": 0.343,
+      "step": 5638
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.698798178002939,
+      "learning_rate": 9.701778282859154e-06,
+      "loss": 0.4555,
+      "step": 5639
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6424558823102382,
+      "learning_rate": 9.701644229124065e-06,
+      "loss": 0.2509,
+      "step": 5640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2991046410397469,
+      "learning_rate": 9.701510146193043e-06,
+      "loss": 0.2684,
+      "step": 5641
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.127350170253293,
+      "learning_rate": 9.701376034066916e-06,
+      "loss": 0.2147,
+      "step": 5642
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7603583718886304,
+      "learning_rate": 9.70124189274652e-06,
+      "loss": 0.4116,
+      "step": 5643
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1264671719841632,
+      "learning_rate": 9.701107722232685e-06,
+      "loss": 0.223,
+      "step": 5644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7341361826498505,
+      "learning_rate": 9.700973522526248e-06,
+      "loss": 0.372,
+      "step": 5645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4161855095866567,
+      "learning_rate": 9.700839293628039e-06,
+      "loss": 0.1756,
+      "step": 5646
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.031793521642622,
+      "learning_rate": 9.700705035538894e-06,
+      "loss": 0.6093,
+      "step": 5647
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7789081228361125,
+      "learning_rate": 9.700570748259645e-06,
+      "loss": 0.4701,
+      "step": 5648
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6070007036650575,
+      "learning_rate": 9.700436431791126e-06,
+      "loss": 0.4122,
+      "step": 5649
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2779839366724552,
+      "learning_rate": 9.700302086134175e-06,
+      "loss": 0.3393,
+      "step": 5650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9813876305800209,
+      "learning_rate": 9.70016771128962e-06,
+      "loss": 0.2843,
+      "step": 5651
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2606743672622684,
+      "learning_rate": 9.700033307258297e-06,
+      "loss": 0.4221,
+      "step": 5652
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8966512930709445,
+      "learning_rate": 9.699898874041043e-06,
+      "loss": 0.1887,
+      "step": 5653
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6130049632873196,
+      "learning_rate": 9.699764411638693e-06,
+      "loss": 0.1941,
+      "step": 5654
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.853184579000853,
+      "learning_rate": 9.699629920052079e-06,
+      "loss": 0.381,
+      "step": 5655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.562407266726017,
+      "learning_rate": 9.699495399282038e-06,
+      "loss": 0.4049,
+      "step": 5656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9482591747795794,
+      "learning_rate": 9.699360849329406e-06,
+      "loss": 0.395,
+      "step": 5657
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9401473198428874,
+      "learning_rate": 9.699226270195018e-06,
+      "loss": 0.5383,
+      "step": 5658
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.796289367194615,
+      "learning_rate": 9.699091661879707e-06,
+      "loss": 0.3779,
+      "step": 5659
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9388091390586821,
+      "learning_rate": 9.698957024384312e-06,
+      "loss": 0.3569,
+      "step": 5660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0246580780077097,
+      "learning_rate": 9.69882235770967e-06,
+      "loss": 0.5196,
+      "step": 5661
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3594749315837498,
+      "learning_rate": 9.698687661856614e-06,
+      "loss": 0.5624,
+      "step": 5662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.728890034702552,
+      "learning_rate": 9.698552936825983e-06,
+      "loss": 0.4,
+      "step": 5663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5877835412416883,
+      "learning_rate": 9.69841818261861e-06,
+      "loss": 0.2327,
+      "step": 5664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1119561121599904,
+      "learning_rate": 9.698283399235338e-06,
+      "loss": 0.3355,
+      "step": 5665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5756506406626298,
+      "learning_rate": 9.698148586676997e-06,
+      "loss": 0.3384,
+      "step": 5666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8034836760697115,
+      "learning_rate": 9.69801374494443e-06,
+      "loss": 0.2462,
+      "step": 5667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6292955152990192,
+      "learning_rate": 9.69787887403847e-06,
+      "loss": 0.2524,
+      "step": 5668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5711942623116264,
+      "learning_rate": 9.697743973959956e-06,
+      "loss": 0.3276,
+      "step": 5669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0441862129305113,
+      "learning_rate": 9.697609044709728e-06,
+      "loss": 0.4054,
+      "step": 5670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4908388306836848,
+      "learning_rate": 9.69747408628862e-06,
+      "loss": 0.2267,
+      "step": 5671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.51483447803416,
+      "learning_rate": 9.697339098697473e-06,
+      "loss": 0.3634,
+      "step": 5672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.013591649256329,
+      "learning_rate": 9.697204081937124e-06,
+      "loss": 0.1404,
+      "step": 5673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0005113722790933,
+      "learning_rate": 9.697069036008413e-06,
+      "loss": 0.4714,
+      "step": 5674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6657820489786863,
+      "learning_rate": 9.696933960912176e-06,
+      "loss": 0.4302,
+      "step": 5675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5342322327801992,
+      "learning_rate": 9.696798856649253e-06,
+      "loss": 0.3981,
+      "step": 5676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2127554533195943,
+      "learning_rate": 9.696663723220481e-06,
+      "loss": 0.2849,
+      "step": 5677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4365257675494314,
+      "learning_rate": 9.696528560626704e-06,
+      "loss": 0.1667,
+      "step": 5678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5387199596746657,
+      "learning_rate": 9.696393368868756e-06,
+      "loss": 0.3102,
+      "step": 5679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.973870094679015,
+      "learning_rate": 9.69625814794748e-06,
+      "loss": 0.2943,
+      "step": 5680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.146048920798169,
+      "learning_rate": 9.696122897863713e-06,
+      "loss": 0.1721,
+      "step": 5681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.070136152710191,
+      "learning_rate": 9.695987618618298e-06,
+      "loss": 0.306,
+      "step": 5682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.101001329995012,
+      "learning_rate": 9.695852310212075e-06,
+      "loss": 0.2111,
+      "step": 5683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9611620660443938,
+      "learning_rate": 9.69571697264588e-06,
+      "loss": 0.2986,
+      "step": 5684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7437207655353149,
+      "learning_rate": 9.695581605920557e-06,
+      "loss": 0.4551,
+      "step": 5685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.78087104250497,
+      "learning_rate": 9.695446210036947e-06,
+      "loss": 0.3327,
+      "step": 5686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4699156104133515,
+      "learning_rate": 9.695310784995887e-06,
+      "loss": 0.1547,
+      "step": 5687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5798516411782777,
+      "learning_rate": 9.695175330798223e-06,
+      "loss": 0.1801,
+      "step": 5688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7349125978313968,
+      "learning_rate": 9.695039847444791e-06,
+      "loss": 0.3133,
+      "step": 5689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4534359107114427,
+      "learning_rate": 9.694904334936436e-06,
+      "loss": 0.2376,
+      "step": 5690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7386106778998305,
+      "learning_rate": 9.694768793274e-06,
+      "loss": 0.2119,
+      "step": 5691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6486586517308393,
+      "learning_rate": 9.69463322245832e-06,
+      "loss": 0.3507,
+      "step": 5692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.007350386047915,
+      "learning_rate": 9.694497622490244e-06,
+      "loss": 0.5158,
+      "step": 5693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5938098115543442,
+      "learning_rate": 9.69436199337061e-06,
+      "loss": 0.3515,
+      "step": 5694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4544533149746572,
+      "learning_rate": 9.694226335100261e-06,
+      "loss": 0.2068,
+      "step": 5695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6849338172471338,
+      "learning_rate": 9.694090647680038e-06,
+      "loss": 0.4176,
+      "step": 5696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6071616109253375,
+      "learning_rate": 9.693954931110788e-06,
+      "loss": 0.2657,
+      "step": 5697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.196442901684623,
+      "learning_rate": 9.693819185393348e-06,
+      "loss": 0.2063,
+      "step": 5698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9681697285494797,
+      "learning_rate": 9.693683410528565e-06,
+      "loss": 0.4744,
+      "step": 5699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5851789737010824,
+      "learning_rate": 9.693547606517281e-06,
+      "loss": 0.2815,
+      "step": 5700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5931249215510255,
+      "learning_rate": 9.69341177336034e-06,
+      "loss": 0.3401,
+      "step": 5701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5846643599346093,
+      "learning_rate": 9.693275911058585e-06,
+      "loss": 0.3124,
+      "step": 5702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1607186526878575,
+      "learning_rate": 9.69314001961286e-06,
+      "loss": 0.1938,
+      "step": 5703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8746892532496864,
+      "learning_rate": 9.693004099024006e-06,
+      "loss": 0.4917,
+      "step": 5704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3654582013413405,
+      "learning_rate": 9.69286814929287e-06,
+      "loss": 0.2903,
+      "step": 5705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1814692198544527,
+      "learning_rate": 9.692732170420295e-06,
+      "loss": 0.3817,
+      "step": 5706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9707280871522459,
+      "learning_rate": 9.692596162407128e-06,
+      "loss": 0.2416,
+      "step": 5707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9123544896314584,
+      "learning_rate": 9.692460125254209e-06,
+      "loss": 0.4145,
+      "step": 5708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5282771680305678,
+      "learning_rate": 9.692324058962388e-06,
+      "loss": 0.3483,
+      "step": 5709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3796061598214675,
+      "learning_rate": 9.692187963532505e-06,
+      "loss": 0.1724,
+      "step": 5710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.360457216359507,
+      "learning_rate": 9.692051838965408e-06,
+      "loss": 0.2785,
+      "step": 5711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.695223107445015,
+      "learning_rate": 9.691915685261942e-06,
+      "loss": 0.2759,
+      "step": 5712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9416172212299867,
+      "learning_rate": 9.691779502422951e-06,
+      "loss": 0.5115,
+      "step": 5713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.37870608678965,
+      "learning_rate": 9.691643290449283e-06,
+      "loss": 0.2058,
+      "step": 5714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.004705477906723,
+      "learning_rate": 9.691507049341781e-06,
+      "loss": 0.3185,
+      "step": 5715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7053276547979617,
+      "learning_rate": 9.691370779101294e-06,
+      "loss": 0.4311,
+      "step": 5716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6921572086373993,
+      "learning_rate": 9.691234479728668e-06,
+      "loss": 0.5372,
+      "step": 5717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.79182753469126,
+      "learning_rate": 9.691098151224747e-06,
+      "loss": 0.4091,
+      "step": 5718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9500731531181972,
+      "learning_rate": 9.690961793590377e-06,
+      "loss": 0.2028,
+      "step": 5719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9940597523768521,
+      "learning_rate": 9.690825406826408e-06,
+      "loss": 0.2176,
+      "step": 5720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2575498034130774,
+      "learning_rate": 9.690688990933685e-06,
+      "loss": 0.2748,
+      "step": 5721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2751597711088902,
+      "learning_rate": 9.690552545913057e-06,
+      "loss": 0.3627,
+      "step": 5722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9127161351101487,
+      "learning_rate": 9.690416071765368e-06,
+      "loss": 0.509,
+      "step": 5723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4991312715096563,
+      "learning_rate": 9.690279568491467e-06,
+      "loss": 0.1813,
+      "step": 5724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8814463348211654,
+      "learning_rate": 9.690143036092203e-06,
+      "loss": 0.4036,
+      "step": 5725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8031438174497192,
+      "learning_rate": 9.690006474568423e-06,
+      "loss": 0.3019,
+      "step": 5726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5134662217749226,
+      "learning_rate": 9.689869883920975e-06,
+      "loss": 0.3851,
+      "step": 5727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6566641802508477,
+      "learning_rate": 9.689733264150708e-06,
+      "loss": 0.3316,
+      "step": 5728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6329863925674668,
+      "learning_rate": 9.689596615258465e-06,
+      "loss": 0.3619,
+      "step": 5729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.927314202820908,
+      "learning_rate": 9.689459937245103e-06,
+      "loss": 0.2634,
+      "step": 5730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6101918277206277,
+      "learning_rate": 9.689323230111466e-06,
+      "loss": 0.3088,
+      "step": 5731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.943638728780785,
+      "learning_rate": 9.689186493858401e-06,
+      "loss": 0.3428,
+      "step": 5732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.332814415779528,
+      "learning_rate": 9.689049728486762e-06,
+      "loss": 0.5673,
+      "step": 5733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.120007934285557,
+      "learning_rate": 9.688912933997394e-06,
+      "loss": 0.3052,
+      "step": 5734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4924048694847278,
+      "learning_rate": 9.68877611039115e-06,
+      "loss": 0.2961,
+      "step": 5735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7276402665377228,
+      "learning_rate": 9.688639257668877e-06,
+      "loss": 0.2439,
+      "step": 5736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5308643191083269,
+      "learning_rate": 9.688502375831428e-06,
+      "loss": 0.3347,
+      "step": 5737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4545550539095524,
+      "learning_rate": 9.688365464879648e-06,
+      "loss": 0.3264,
+      "step": 5738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5869628523148804,
+      "learning_rate": 9.688228524814392e-06,
+      "loss": 0.4404,
+      "step": 5739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4742461635100768,
+      "learning_rate": 9.688091555636507e-06,
+      "loss": 0.3019,
+      "step": 5740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9851258013963686,
+      "learning_rate": 9.687954557346844e-06,
+      "loss": 0.2134,
+      "step": 5741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.104163978456127,
+      "learning_rate": 9.687817529946257e-06,
+      "loss": 0.1839,
+      "step": 5742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7139275695075749,
+      "learning_rate": 9.687680473435594e-06,
+      "loss": 0.503,
+      "step": 5743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7992328453588502,
+      "learning_rate": 9.687543387815706e-06,
+      "loss": 0.3742,
+      "step": 5744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9494366724667387,
+      "learning_rate": 9.687406273087445e-06,
+      "loss": 0.4508,
+      "step": 5745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7308399499807428,
+      "learning_rate": 9.687269129251663e-06,
+      "loss": 0.3733,
+      "step": 5746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7663199451744238,
+      "learning_rate": 9.687131956309209e-06,
+      "loss": 0.2613,
+      "step": 5747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6107700838106596,
+      "learning_rate": 9.686994754260938e-06,
+      "loss": 0.2713,
+      "step": 5748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5068233870195615,
+      "learning_rate": 9.6868575231077e-06,
+      "loss": 0.1786,
+      "step": 5749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.052675447788729,
+      "learning_rate": 9.686720262850349e-06,
+      "loss": 0.3328,
+      "step": 5750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.055393511077665,
+      "learning_rate": 9.686582973489735e-06,
+      "loss": 0.4736,
+      "step": 5751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2055611851711627,
+      "learning_rate": 9.686445655026713e-06,
+      "loss": 0.1842,
+      "step": 5752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5388146844790052,
+      "learning_rate": 9.686308307462134e-06,
+      "loss": 0.2345,
+      "step": 5753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5663906558014473,
+      "learning_rate": 9.686170930796852e-06,
+      "loss": 0.2407,
+      "step": 5754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.707597567678896,
+      "learning_rate": 9.68603352503172e-06,
+      "loss": 0.2392,
+      "step": 5755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4460962635169914,
+      "learning_rate": 9.685896090167588e-06,
+      "loss": 0.2504,
+      "step": 5756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2828583599813204,
+      "learning_rate": 9.685758626205315e-06,
+      "loss": 0.3426,
+      "step": 5757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.465720111642149,
+      "learning_rate": 9.685621133145751e-06,
+      "loss": 0.3148,
+      "step": 5758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.007667873977427,
+      "learning_rate": 9.685483610989748e-06,
+      "loss": 0.5932,
+      "step": 5759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8710893220789369,
+      "learning_rate": 9.685346059738165e-06,
+      "loss": 0.3267,
+      "step": 5760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6209460547394101,
+      "learning_rate": 9.685208479391854e-06,
+      "loss": 0.3199,
+      "step": 5761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8635491472466919,
+      "learning_rate": 9.685070869951668e-06,
+      "loss": 0.3832,
+      "step": 5762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5395789054627123,
+      "learning_rate": 9.684933231418462e-06,
+      "loss": 0.4275,
+      "step": 5763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9061887703895743,
+      "learning_rate": 9.684795563793092e-06,
+      "loss": 0.1913,
+      "step": 5764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8901057042479776,
+      "learning_rate": 9.684657867076413e-06,
+      "loss": 0.3788,
+      "step": 5765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2301747924626285,
+      "learning_rate": 9.684520141269278e-06,
+      "loss": 0.2356,
+      "step": 5766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0164632532752265,
+      "learning_rate": 9.684382386372543e-06,
+      "loss": 0.5144,
+      "step": 5767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9155129250541187,
+      "learning_rate": 9.684244602387065e-06,
+      "loss": 0.2811,
+      "step": 5768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4215066453615357,
+      "learning_rate": 9.684106789313699e-06,
+      "loss": 0.2286,
+      "step": 5769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6990771552536603,
+      "learning_rate": 9.683968947153299e-06,
+      "loss": 0.2518,
+      "step": 5770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5128767866372934,
+      "learning_rate": 9.68383107590672e-06,
+      "loss": 0.4032,
+      "step": 5771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0170705919494383,
+      "learning_rate": 9.683693175574825e-06,
+      "loss": 0.2039,
+      "step": 5772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6202455806615992,
+      "learning_rate": 9.683555246158462e-06,
+      "loss": 0.3359,
+      "step": 5773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.861231642280928,
+      "learning_rate": 9.683417287658492e-06,
+      "loss": 0.5447,
+      "step": 5774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.664479277359888,
+      "learning_rate": 9.683279300075772e-06,
+      "loss": 0.2259,
+      "step": 5775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.616748492228258,
+      "learning_rate": 9.683141283411157e-06,
+      "loss": 0.3071,
+      "step": 5776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6850686085441244,
+      "learning_rate": 9.683003237665505e-06,
+      "loss": 0.2873,
+      "step": 5777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.080628235477259,
+      "learning_rate": 9.682865162839671e-06,
+      "loss": 0.5282,
+      "step": 5778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0602840472912063,
+      "learning_rate": 9.682727058934516e-06,
+      "loss": 0.2266,
+      "step": 5779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9336658734058192,
+      "learning_rate": 9.682588925950897e-06,
+      "loss": 0.2809,
+      "step": 5780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7491942406965177,
+      "learning_rate": 9.68245076388967e-06,
+      "loss": 0.3332,
+      "step": 5781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.258804236854023,
+      "learning_rate": 9.682312572751693e-06,
+      "loss": 0.2051,
+      "step": 5782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7182918913420595,
+      "learning_rate": 9.682174352537826e-06,
+      "loss": 0.3356,
+      "step": 5783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1962868472945125,
+      "learning_rate": 9.682036103248924e-06,
+      "loss": 0.51,
+      "step": 5784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9852586997512518,
+      "learning_rate": 9.68189782488585e-06,
+      "loss": 0.2626,
+      "step": 5785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3886700990990124,
+      "learning_rate": 9.681759517449458e-06,
+      "loss": 0.3003,
+      "step": 5786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.016303343809241,
+      "learning_rate": 9.68162118094061e-06,
+      "loss": 0.356,
+      "step": 5787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3878808270019172,
+      "learning_rate": 9.681482815360166e-06,
+      "loss": 0.3253,
+      "step": 5788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3076023744822416,
+      "learning_rate": 9.68134442070898e-06,
+      "loss": 0.1607,
+      "step": 5789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6406686186668737,
+      "learning_rate": 9.681205996987916e-06,
+      "loss": 0.5093,
+      "step": 5790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0349077529257593,
+      "learning_rate": 9.681067544197834e-06,
+      "loss": 0.1443,
+      "step": 5791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6136934109508665,
+      "learning_rate": 9.680929062339592e-06,
+      "loss": 0.4132,
+      "step": 5792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8716400748551933,
+      "learning_rate": 9.680790551414047e-06,
+      "loss": 0.2447,
+      "step": 5793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8280962102204377,
+      "learning_rate": 9.680652011422066e-06,
+      "loss": 0.5354,
+      "step": 5794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6660057259877104,
+      "learning_rate": 9.680513442364504e-06,
+      "loss": 0.2109,
+      "step": 5795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.811035500796633,
+      "learning_rate": 9.680374844242221e-06,
+      "loss": 0.2074,
+      "step": 5796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8592622466320687,
+      "learning_rate": 9.680236217056082e-06,
+      "loss": 0.393,
+      "step": 5797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6633195724166727,
+      "learning_rate": 9.680097560806944e-06,
+      "loss": 0.2025,
+      "step": 5798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0262815416000244,
+      "learning_rate": 9.679958875495669e-06,
+      "loss": 0.6781,
+      "step": 5799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5894625579336576,
+      "learning_rate": 9.679820161123122e-06,
+      "loss": 0.2592,
+      "step": 5800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7014453571760186,
+      "learning_rate": 9.679681417690157e-06,
+      "loss": 0.4168,
+      "step": 5801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.712040809992149,
+      "learning_rate": 9.679542645197641e-06,
+      "loss": 0.2252,
+      "step": 5802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0288446122517225,
+      "learning_rate": 9.679403843646436e-06,
+      "loss": 0.5275,
+      "step": 5803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.553215121575038,
+      "learning_rate": 9.6792650130374e-06,
+      "loss": 0.1448,
+      "step": 5804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1140233638887964,
+      "learning_rate": 9.679126153371398e-06,
+      "loss": 0.3845,
+      "step": 5805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7249265446088498,
+      "learning_rate": 9.678987264649291e-06,
+      "loss": 0.331,
+      "step": 5806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5941111211610035,
+      "learning_rate": 9.678848346871943e-06,
+      "loss": 0.2203,
+      "step": 5807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7657912015756856,
+      "learning_rate": 9.678709400040216e-06,
+      "loss": 0.5039,
+      "step": 5808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.159496793058734,
+      "learning_rate": 9.678570424154971e-06,
+      "loss": 0.1961,
+      "step": 5809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9325862648809429,
+      "learning_rate": 9.678431419217073e-06,
+      "loss": 0.3963,
+      "step": 5810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8372734647060383,
+      "learning_rate": 9.678292385227386e-06,
+      "loss": 0.4329,
+      "step": 5811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.210551526445105,
+      "learning_rate": 9.678153322186772e-06,
+      "loss": 0.342,
+      "step": 5812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.56220403729282,
+      "learning_rate": 9.678014230096093e-06,
+      "loss": 0.2757,
+      "step": 5813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1282238366163684,
+      "learning_rate": 9.677875108956216e-06,
+      "loss": 0.2771,
+      "step": 5814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5536042182183727,
+      "learning_rate": 9.677735958768002e-06,
+      "loss": 0.3746,
+      "step": 5815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4615023128048805,
+      "learning_rate": 9.677596779532317e-06,
+      "loss": 0.2492,
+      "step": 5816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3934071388538662,
+      "learning_rate": 9.677457571250023e-06,
+      "loss": 0.2823,
+      "step": 5817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5236601006663635,
+      "learning_rate": 9.677318333921988e-06,
+      "loss": 0.3781,
+      "step": 5818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9623052340947467,
+      "learning_rate": 9.677179067549074e-06,
+      "loss": 0.3388,
+      "step": 5819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.627622761938046,
+      "learning_rate": 9.677039772132147e-06,
+      "loss": 0.3896,
+      "step": 5820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.472102050384626,
+      "learning_rate": 9.67690044767207e-06,
+      "loss": 0.3923,
+      "step": 5821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4580721626977005,
+      "learning_rate": 9.676761094169711e-06,
+      "loss": 0.2761,
+      "step": 5822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6594107536012996,
+      "learning_rate": 9.676621711625936e-06,
+      "loss": 0.3209,
+      "step": 5823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.516152083780327,
+      "learning_rate": 9.676482300041605e-06,
+      "loss": 0.1942,
+      "step": 5824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.637917291454635,
+      "learning_rate": 9.67634285941759e-06,
+      "loss": 0.3481,
+      "step": 5825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7744476323581653,
+      "learning_rate": 9.676203389754752e-06,
+      "loss": 0.6017,
+      "step": 5826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8324876635631908,
+      "learning_rate": 9.67606389105396e-06,
+      "loss": 0.4816,
+      "step": 5827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1013076656083185,
+      "learning_rate": 9.675924363316077e-06,
+      "loss": 0.7367,
+      "step": 5828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8388681193960688,
+      "learning_rate": 9.675784806541975e-06,
+      "loss": 0.2267,
+      "step": 5829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.046695039888897,
+      "learning_rate": 9.675645220732516e-06,
+      "loss": 0.224,
+      "step": 5830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5343193364583687,
+      "learning_rate": 9.675505605888569e-06,
+      "loss": 0.2321,
+      "step": 5831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8925238090728274,
+      "learning_rate": 9.675365962011e-06,
+      "loss": 0.2309,
+      "step": 5832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.123458681990382,
+      "learning_rate": 9.675226289100676e-06,
+      "loss": 0.4384,
+      "step": 5833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.397367390314606,
+      "learning_rate": 9.675086587158465e-06,
+      "loss": 0.1993,
+      "step": 5834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9884641297179062,
+      "learning_rate": 9.674946856185235e-06,
+      "loss": 0.2298,
+      "step": 5835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.497881709894844,
+      "learning_rate": 9.674807096181852e-06,
+      "loss": 0.377,
+      "step": 5836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.750391592763145,
+      "learning_rate": 9.674667307149185e-06,
+      "loss": 0.3681,
+      "step": 5837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0375055236975563,
+      "learning_rate": 9.674527489088102e-06,
+      "loss": 0.5606,
+      "step": 5838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5630216466839497,
+      "learning_rate": 9.67438764199947e-06,
+      "loss": 0.2277,
+      "step": 5839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8122619357463932,
+      "learning_rate": 9.67424776588416e-06,
+      "loss": 0.3089,
+      "step": 5840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.735427610054484,
+      "learning_rate": 9.674107860743038e-06,
+      "loss": 0.4684,
+      "step": 5841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.397747432949876,
+      "learning_rate": 9.673967926576975e-06,
+      "loss": 0.2008,
+      "step": 5842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6595913094090982,
+      "learning_rate": 9.673827963386837e-06,
+      "loss": 0.1844,
+      "step": 5843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0227139933132203,
+      "learning_rate": 9.673687971173496e-06,
+      "loss": 0.2325,
+      "step": 5844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7861771366882977,
+      "learning_rate": 9.67354794993782e-06,
+      "loss": 0.4315,
+      "step": 5845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6254845593662801,
+      "learning_rate": 9.67340789968068e-06,
+      "loss": 0.4382,
+      "step": 5846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5407221748006508,
+      "learning_rate": 9.673267820402944e-06,
+      "loss": 0.2656,
+      "step": 5847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2786452286272763,
+      "learning_rate": 9.67312771210548e-06,
+      "loss": 0.1684,
+      "step": 5848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4319656024043554,
+      "learning_rate": 9.672987574789162e-06,
+      "loss": 0.3416,
+      "step": 5849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9809223329035945,
+      "learning_rate": 9.67284740845486e-06,
+      "loss": 0.3818,
+      "step": 5850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8099188348054105,
+      "learning_rate": 9.672707213103441e-06,
+      "loss": 0.4418,
+      "step": 5851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5111646597835489,
+      "learning_rate": 9.672566988735777e-06,
+      "loss": 0.4638,
+      "step": 5852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0867661105500974,
+      "learning_rate": 9.672426735352741e-06,
+      "loss": 0.7022,
+      "step": 5853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7401090805896233,
+      "learning_rate": 9.6722864529552e-06,
+      "loss": 0.2647,
+      "step": 5854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8724068193462657,
+      "learning_rate": 9.67214614154403e-06,
+      "loss": 0.3911,
+      "step": 5855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8039372275870065,
+      "learning_rate": 9.672005801120097e-06,
+      "loss": 0.3362,
+      "step": 5856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2916044098773503,
+      "learning_rate": 9.671865431684276e-06,
+      "loss": 0.2615,
+      "step": 5857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7286364669369818,
+      "learning_rate": 9.671725033237438e-06,
+      "loss": 0.4174,
+      "step": 5858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7870764765812095,
+      "learning_rate": 9.671584605780454e-06,
+      "loss": 0.5324,
+      "step": 5859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8188241985267684,
+      "learning_rate": 9.671444149314198e-06,
+      "loss": 0.2342,
+      "step": 5860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7116791324043927,
+      "learning_rate": 9.67130366383954e-06,
+      "loss": 0.4721,
+      "step": 5861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.028451047530926,
+      "learning_rate": 9.671163149357352e-06,
+      "loss": 0.5819,
+      "step": 5862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5022068838638947,
+      "learning_rate": 9.67102260586851e-06,
+      "loss": 0.1984,
+      "step": 5863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8561394091777788,
+      "learning_rate": 9.670882033373882e-06,
+      "loss": 0.4471,
+      "step": 5864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4558018363143161,
+      "learning_rate": 9.670741431874343e-06,
+      "loss": 0.2494,
+      "step": 5865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6748916703482906,
+      "learning_rate": 9.670600801370768e-06,
+      "loss": 0.1761,
+      "step": 5866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5207137946502438,
+      "learning_rate": 9.670460141864028e-06,
+      "loss": 0.1875,
+      "step": 5867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7890914348278082,
+      "learning_rate": 9.670319453355e-06,
+      "loss": 0.3788,
+      "step": 5868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.975177686882912,
+      "learning_rate": 9.670178735844553e-06,
+      "loss": 0.1906,
+      "step": 5869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2588073412483287,
+      "learning_rate": 9.670037989333561e-06,
+      "loss": 0.1935,
+      "step": 5870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9963429121915555,
+      "learning_rate": 9.6698972138229e-06,
+      "loss": 0.3494,
+      "step": 5871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.78860179516411,
+      "learning_rate": 9.669756409313445e-06,
+      "loss": 0.5318,
+      "step": 5872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1100085967010815,
+      "learning_rate": 9.66961557580607e-06,
+      "loss": 0.6902,
+      "step": 5873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8236159108655332,
+      "learning_rate": 9.669474713301647e-06,
+      "loss": 0.2932,
+      "step": 5874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7982374007259831,
+      "learning_rate": 9.669333821801053e-06,
+      "loss": 0.219,
+      "step": 5875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8212929406009735,
+      "learning_rate": 9.669192901305162e-06,
+      "loss": 0.4473,
+      "step": 5876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.329584682761384,
+      "learning_rate": 9.66905195181485e-06,
+      "loss": 0.1428,
+      "step": 5877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.552954306711387,
+      "learning_rate": 9.668910973330991e-06,
+      "loss": 0.1972,
+      "step": 5878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.057640535091624,
+      "learning_rate": 9.668769965854463e-06,
+      "loss": 0.1698,
+      "step": 5879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7164950317007557,
+      "learning_rate": 9.668628929386136e-06,
+      "loss": 0.408,
+      "step": 5880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5317629397799701,
+      "learning_rate": 9.668487863926892e-06,
+      "loss": 0.1786,
+      "step": 5881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.206856845397406,
+      "learning_rate": 9.668346769477605e-06,
+      "loss": 0.196,
+      "step": 5882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6380468123818896,
+      "learning_rate": 9.668205646039151e-06,
+      "loss": 0.1966,
+      "step": 5883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.705380175951394,
+      "learning_rate": 9.668064493612405e-06,
+      "loss": 0.3061,
+      "step": 5884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5248053655109275,
+      "learning_rate": 9.667923312198244e-06,
+      "loss": 0.5367,
+      "step": 5885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8534756667208485,
+      "learning_rate": 9.667782101797544e-06,
+      "loss": 0.2075,
+      "step": 5886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.860761614285967,
+      "learning_rate": 9.667640862411186e-06,
+      "loss": 0.5061,
+      "step": 5887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4439198453657582,
+      "learning_rate": 9.667499594040045e-06,
+      "loss": 0.2127,
+      "step": 5888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7509600590163747,
+      "learning_rate": 9.667358296684992e-06,
+      "loss": 0.3661,
+      "step": 5889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7225584664131992,
+      "learning_rate": 9.667216970346916e-06,
+      "loss": 0.232,
+      "step": 5890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6414613363181472,
+      "learning_rate": 9.667075615026684e-06,
+      "loss": 0.2133,
+      "step": 5891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6551423147474327,
+      "learning_rate": 9.66693423072518e-06,
+      "loss": 0.5126,
+      "step": 5892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6573932498686943,
+      "learning_rate": 9.66679281744328e-06,
+      "loss": 0.3548,
+      "step": 5893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4003561034263694,
+      "learning_rate": 9.666651375181861e-06,
+      "loss": 0.2127,
+      "step": 5894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6836452343496808,
+      "learning_rate": 9.666509903941804e-06,
+      "loss": 0.3271,
+      "step": 5895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8613217704960687,
+      "learning_rate": 9.666368403723987e-06,
+      "loss": 0.4025,
+      "step": 5896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.581121281899562,
+      "learning_rate": 9.666226874529285e-06,
+      "loss": 0.3113,
+      "step": 5897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8212630855558865,
+      "learning_rate": 9.666085316358582e-06,
+      "loss": 0.4773,
+      "step": 5898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8587925904402745,
+      "learning_rate": 9.665943729212754e-06,
+      "loss": 0.321,
+      "step": 5899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.443288731685193,
+      "learning_rate": 9.66580211309268e-06,
+      "loss": 0.2977,
+      "step": 5900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.741915818486569,
+      "learning_rate": 9.665660467999241e-06,
+      "loss": 0.298,
+      "step": 5901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5344356830779302,
+      "learning_rate": 9.665518793933316e-06,
+      "loss": 0.2553,
+      "step": 5902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0463325567958086,
+      "learning_rate": 9.665377090895784e-06,
+      "loss": 0.2948,
+      "step": 5903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8062075564981197,
+      "learning_rate": 9.665235358887527e-06,
+      "loss": 0.3091,
+      "step": 5904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2064849548749685,
+      "learning_rate": 9.665093597909423e-06,
+      "loss": 0.3808,
+      "step": 5905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0563243394533286,
+      "learning_rate": 9.664951807962352e-06,
+      "loss": 0.7722,
+      "step": 5906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9368575015266987,
+      "learning_rate": 9.664809989047197e-06,
+      "loss": 0.419,
+      "step": 5907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5098861921316666,
+      "learning_rate": 9.664668141164836e-06,
+      "loss": 0.2371,
+      "step": 5908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2011747654825804,
+      "learning_rate": 9.664526264316152e-06,
+      "loss": 0.1264,
+      "step": 5909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6112856402720674,
+      "learning_rate": 9.664384358502024e-06,
+      "loss": 0.392,
+      "step": 5910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7313733001043095,
+      "learning_rate": 9.664242423723334e-06,
+      "loss": 0.9233,
+      "step": 5911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9717665299387386,
+      "learning_rate": 9.664100459980964e-06,
+      "loss": 0.6927,
+      "step": 5912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9121154925333264,
+      "learning_rate": 9.663958467275794e-06,
+      "loss": 0.3566,
+      "step": 5913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7519811630131807,
+      "learning_rate": 9.663816445608708e-06,
+      "loss": 0.4457,
+      "step": 5914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7465406652366497,
+      "learning_rate": 9.663674394980587e-06,
+      "loss": 0.2371,
+      "step": 5915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7299766292847039,
+      "learning_rate": 9.663532315392312e-06,
+      "loss": 0.2641,
+      "step": 5916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8386072541327159,
+      "learning_rate": 9.663390206844765e-06,
+      "loss": 0.3272,
+      "step": 5917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9330188937070483,
+      "learning_rate": 9.663248069338833e-06,
+      "loss": 0.4101,
+      "step": 5918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.804129854967241,
+      "learning_rate": 9.663105902875391e-06,
+      "loss": 0.3619,
+      "step": 5919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.665175394870076,
+      "learning_rate": 9.662963707455329e-06,
+      "loss": 0.2859,
+      "step": 5920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.128903236737,
+      "learning_rate": 9.662821483079526e-06,
+      "loss": 0.5608,
+      "step": 5921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4005940873421354,
+      "learning_rate": 9.662679229748867e-06,
+      "loss": 0.2462,
+      "step": 5922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8597623437645403,
+      "learning_rate": 9.662536947464234e-06,
+      "loss": 0.2341,
+      "step": 5923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0623300041597856,
+      "learning_rate": 9.66239463622651e-06,
+      "loss": 0.3944,
+      "step": 5924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.599534444020436,
+      "learning_rate": 9.662252296036581e-06,
+      "loss": 0.3354,
+      "step": 5925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5808664649293502,
+      "learning_rate": 9.66210992689533e-06,
+      "loss": 0.3742,
+      "step": 5926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1238042100040877,
+      "learning_rate": 9.661967528803641e-06,
+      "loss": 0.3381,
+      "step": 5927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2470019507733208,
+      "learning_rate": 9.661825101762397e-06,
+      "loss": 0.2089,
+      "step": 5928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.621441282224918,
+      "learning_rate": 9.661682645772483e-06,
+      "loss": 0.3911,
+      "step": 5929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1871025022487895,
+      "learning_rate": 9.661540160834785e-06,
+      "loss": 0.208,
+      "step": 5930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3918934492481694,
+      "learning_rate": 9.661397646950188e-06,
+      "loss": 0.4742,
+      "step": 5931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6709370031760942,
+      "learning_rate": 9.661255104119574e-06,
+      "loss": 0.3442,
+      "step": 5932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.421238927257051,
+      "learning_rate": 9.66111253234383e-06,
+      "loss": 0.2177,
+      "step": 5933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5317635356267043,
+      "learning_rate": 9.660969931623842e-06,
+      "loss": 0.2736,
+      "step": 5934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0704514716066187,
+      "learning_rate": 9.660827301960494e-06,
+      "loss": 0.1869,
+      "step": 5935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8851276435452198,
+      "learning_rate": 9.660684643354673e-06,
+      "loss": 0.2492,
+      "step": 5936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6938784726534963,
+      "learning_rate": 9.660541955807264e-06,
+      "loss": 0.3386,
+      "step": 5937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8826803679420911,
+      "learning_rate": 9.660399239319153e-06,
+      "loss": 0.4759,
+      "step": 5938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.224298564308251,
+      "learning_rate": 9.66025649389123e-06,
+      "loss": 0.5659,
+      "step": 5939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.598202959269418,
+      "learning_rate": 9.660113719524375e-06,
+      "loss": 0.2387,
+      "step": 5940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5013919086884926,
+      "learning_rate": 9.659970916219478e-06,
+      "loss": 0.2093,
+      "step": 5941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6134311985038365,
+      "learning_rate": 9.659828083977425e-06,
+      "loss": 0.2464,
+      "step": 5942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6241273118483195,
+      "learning_rate": 9.659685222799105e-06,
+      "loss": 0.3361,
+      "step": 5943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.306633356242354,
+      "learning_rate": 9.659542332685402e-06,
+      "loss": 0.157,
+      "step": 5944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.696382005041657,
+      "learning_rate": 9.659399413637205e-06,
+      "loss": 0.3344,
+      "step": 5945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3100949658076864,
+      "learning_rate": 9.659256465655403e-06,
+      "loss": 0.3128,
+      "step": 5946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8616505410669384,
+      "learning_rate": 9.65911348874088e-06,
+      "loss": 0.4466,
+      "step": 5947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2499860127332165,
+      "learning_rate": 9.658970482894527e-06,
+      "loss": 0.2862,
+      "step": 5948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0695909814337625,
+      "learning_rate": 9.65882744811723e-06,
+      "loss": 0.2335,
+      "step": 5949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4987645450654605,
+      "learning_rate": 9.658684384409878e-06,
+      "loss": 0.4107,
+      "step": 5950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9306865453496378,
+      "learning_rate": 9.65854129177336e-06,
+      "loss": 0.2515,
+      "step": 5951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5968974183742883,
+      "learning_rate": 9.658398170208565e-06,
+      "loss": 0.394,
+      "step": 5952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3640400474487673,
+      "learning_rate": 9.658255019716378e-06,
+      "loss": 0.1612,
+      "step": 5953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9665830547095702,
+      "learning_rate": 9.658111840297692e-06,
+      "loss": 0.5252,
+      "step": 5954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5001963180232691,
+      "learning_rate": 9.657968631953396e-06,
+      "loss": 0.1483,
+      "step": 5955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.023608137911245,
+      "learning_rate": 9.657825394684378e-06,
+      "loss": 0.2029,
+      "step": 5956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.218900567976837,
+      "learning_rate": 9.657682128491525e-06,
+      "loss": 0.4123,
+      "step": 5957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5542942351044524,
+      "learning_rate": 9.657538833375733e-06,
+      "loss": 0.5064,
+      "step": 5958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0894557984955995,
+      "learning_rate": 9.657395509337885e-06,
+      "loss": 0.4993,
+      "step": 5959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5654541008961878,
+      "learning_rate": 9.657252156378876e-06,
+      "loss": 0.1802,
+      "step": 5960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8101807484391048,
+      "learning_rate": 9.657108774499593e-06,
+      "loss": 0.317,
+      "step": 5961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.214358463306358,
+      "learning_rate": 9.656965363700928e-06,
+      "loss": 0.5443,
+      "step": 5962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0420937880019663,
+      "learning_rate": 9.65682192398377e-06,
+      "loss": 0.3764,
+      "step": 5963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3343543617326916,
+      "learning_rate": 9.656678455349012e-06,
+      "loss": 0.3949,
+      "step": 5964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5737626595893144,
+      "learning_rate": 9.656534957797543e-06,
+      "loss": 0.3741,
+      "step": 5965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7558542788286267,
+      "learning_rate": 9.656391431330256e-06,
+      "loss": 0.5268,
+      "step": 5966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2033783347659193,
+      "learning_rate": 9.65624787594804e-06,
+      "loss": 0.4143,
+      "step": 5967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0617896756963665,
+      "learning_rate": 9.656104291651787e-06,
+      "loss": 0.5627,
+      "step": 5968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6713695385936203,
+      "learning_rate": 9.65596067844239e-06,
+      "loss": 0.1906,
+      "step": 5969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.99021358310669,
+      "learning_rate": 9.65581703632074e-06,
+      "loss": 0.4514,
+      "step": 5970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7124025852278255,
+      "learning_rate": 9.655673365287728e-06,
+      "loss": 0.3256,
+      "step": 5971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6390841787193753,
+      "learning_rate": 9.655529665344248e-06,
+      "loss": 0.3968,
+      "step": 5972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4535769226427142,
+      "learning_rate": 9.65538593649119e-06,
+      "loss": 0.1702,
+      "step": 5973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6632975826115535,
+      "learning_rate": 9.655242178729451e-06,
+      "loss": 0.255,
+      "step": 5974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4932663148406748,
+      "learning_rate": 9.655098392059918e-06,
+      "loss": 0.3446,
+      "step": 5975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7105324627185714,
+      "learning_rate": 9.654954576483488e-06,
+      "loss": 0.3028,
+      "step": 5976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.035617095257289,
+      "learning_rate": 9.65481073200105e-06,
+      "loss": 0.5251,
+      "step": 5977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2064118554232044,
+      "learning_rate": 9.654666858613502e-06,
+      "loss": 0.545,
+      "step": 5978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7290304096283793,
+      "learning_rate": 9.654522956321735e-06,
+      "loss": 0.2833,
+      "step": 5979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5853038564950563,
+      "learning_rate": 9.654379025126642e-06,
+      "loss": 0.2878,
+      "step": 5980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.913009562192166,
+      "learning_rate": 9.654235065029119e-06,
+      "loss": 0.4456,
+      "step": 5981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6875470671447674,
+      "learning_rate": 9.654091076030057e-06,
+      "loss": 0.3828,
+      "step": 5982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5992985861338282,
+      "learning_rate": 9.653947058130353e-06,
+      "loss": 0.3409,
+      "step": 5983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5229702844413398,
+      "learning_rate": 9.653803011330897e-06,
+      "loss": 0.3387,
+      "step": 5984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1081388626461366,
+      "learning_rate": 9.65365893563259e-06,
+      "loss": 0.2553,
+      "step": 5985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4751317242221103,
+      "learning_rate": 9.653514831036322e-06,
+      "loss": 0.1875,
+      "step": 5986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.175680073792161,
+      "learning_rate": 9.653370697542988e-06,
+      "loss": 0.1626,
+      "step": 5987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.000954929009592,
+      "learning_rate": 9.653226535153485e-06,
+      "loss": 0.1864,
+      "step": 5988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8186315973976184,
+      "learning_rate": 9.653082343868707e-06,
+      "loss": 0.1739,
+      "step": 5989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.012098085878696,
+      "learning_rate": 9.65293812368955e-06,
+      "loss": 0.6141,
+      "step": 5990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9016243633842669,
+      "learning_rate": 9.652793874616908e-06,
+      "loss": 0.3504,
+      "step": 5991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2049858360872776,
+      "learning_rate": 9.652649596651679e-06,
+      "loss": 0.1679,
+      "step": 5992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8736788207709834,
+      "learning_rate": 9.652505289794759e-06,
+      "loss": 0.3969,
+      "step": 5993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6180996188475032,
+      "learning_rate": 9.65236095404704e-06,
+      "loss": 0.1217,
+      "step": 5994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.130108439630104,
+      "learning_rate": 9.652216589409422e-06,
+      "loss": 0.3821,
+      "step": 5995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8234769049965487,
+      "learning_rate": 9.652072195882802e-06,
+      "loss": 0.2992,
+      "step": 5996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8239528709344712,
+      "learning_rate": 9.651927773468075e-06,
+      "loss": 0.2518,
+      "step": 5997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.190494755778928,
+      "learning_rate": 9.651783322166137e-06,
+      "loss": 0.8075,
+      "step": 5998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7457337659091645,
+      "learning_rate": 9.651638841977887e-06,
+      "loss": 0.4196,
+      "step": 5999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3604226829739956,
+      "learning_rate": 9.65149433290422e-06,
+      "loss": 0.2122,
+      "step": 6000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5023514823298005,
+      "learning_rate": 9.651349794946036e-06,
+      "loss": 0.2377,
+      "step": 6001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.608397295638179,
+      "learning_rate": 9.651205228104232e-06,
+      "loss": 0.3061,
+      "step": 6002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7392748918074925,
+      "learning_rate": 9.651060632379705e-06,
+      "loss": 0.3102,
+      "step": 6003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4432505241072633,
+      "learning_rate": 9.650916007773352e-06,
+      "loss": 0.3807,
+      "step": 6004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.406948530069002,
+      "learning_rate": 9.650771354286072e-06,
+      "loss": 0.2676,
+      "step": 6005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.588163840788452,
+      "learning_rate": 9.650626671918763e-06,
+      "loss": 0.5234,
+      "step": 6006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8028342714249799,
+      "learning_rate": 9.650481960672324e-06,
+      "loss": 0.2382,
+      "step": 6007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0587484674384333,
+      "learning_rate": 9.650337220547655e-06,
+      "loss": 0.589,
+      "step": 6008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1641331155045656,
+      "learning_rate": 9.65019245154565e-06,
+      "loss": 0.6436,
+      "step": 6009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.999856817067331,
+      "learning_rate": 9.650047653667212e-06,
+      "loss": 0.3282,
+      "step": 6010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6322767551469708,
+      "learning_rate": 9.64990282691324e-06,
+      "loss": 0.345,
+      "step": 6011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.751100411319461,
+      "learning_rate": 9.649757971284631e-06,
+      "loss": 0.3887,
+      "step": 6012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.68378575949372,
+      "learning_rate": 9.649613086782288e-06,
+      "loss": 0.3128,
+      "step": 6013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5794950312323153,
+      "learning_rate": 9.649468173407108e-06,
+      "loss": 0.2199,
+      "step": 6014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5522708584855005,
+      "learning_rate": 9.649323231159993e-06,
+      "loss": 0.2251,
+      "step": 6015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.643398379476632,
+      "learning_rate": 9.64917826004184e-06,
+      "loss": 0.403,
+      "step": 6016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3540372765298299,
+      "learning_rate": 9.64903326005355e-06,
+      "loss": 0.3023,
+      "step": 6017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.826276251926033,
+      "learning_rate": 9.648888231196026e-06,
+      "loss": 0.3715,
+      "step": 6018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.102282857987647,
+      "learning_rate": 9.648743173470166e-06,
+      "loss": 0.4669,
+      "step": 6019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9802884263365794,
+      "learning_rate": 9.648598086876872e-06,
+      "loss": 0.2081,
+      "step": 6020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3057248765453722,
+      "learning_rate": 9.648452971417047e-06,
+      "loss": 0.197,
+      "step": 6021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8458212960979496,
+      "learning_rate": 9.648307827091586e-06,
+      "loss": 0.2433,
+      "step": 6022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1198204157894367,
+      "learning_rate": 9.648162653901395e-06,
+      "loss": 0.1233,
+      "step": 6023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.824924091661277,
+      "learning_rate": 9.648017451847376e-06,
+      "loss": 0.4608,
+      "step": 6024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.355129102002393,
+      "learning_rate": 9.647872220930428e-06,
+      "loss": 0.3606,
+      "step": 6025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9390959645654224,
+      "learning_rate": 9.647726961151454e-06,
+      "loss": 0.1409,
+      "step": 6026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.519066661133724,
+      "learning_rate": 9.647581672511357e-06,
+      "loss": 0.3506,
+      "step": 6027
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9130395218659613,
+      "learning_rate": 9.647436355011038e-06,
+      "loss": 0.2668,
+      "step": 6028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8801474427078981,
+      "learning_rate": 9.647291008651398e-06,
+      "loss": 0.4129,
+      "step": 6029
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.107783586149648,
+      "learning_rate": 9.647145633433341e-06,
+      "loss": 0.5629,
+      "step": 6030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5669302313694125,
+      "learning_rate": 9.647000229357773e-06,
+      "loss": 0.5103,
+      "step": 6031
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2050471556902227,
+      "learning_rate": 9.646854796425592e-06,
+      "loss": 0.3038,
+      "step": 6032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5651837431800653,
+      "learning_rate": 9.646709334637703e-06,
+      "loss": 0.6806,
+      "step": 6033
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8237442097790935,
+      "learning_rate": 9.64656384399501e-06,
+      "loss": 0.2429,
+      "step": 6034
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5196191191089983,
+      "learning_rate": 9.646418324498414e-06,
+      "loss": 0.2394,
+      "step": 6035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3153732304681238,
+      "learning_rate": 9.646272776148823e-06,
+      "loss": 0.2029,
+      "step": 6036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1754668366447683,
+      "learning_rate": 9.646127198947137e-06,
+      "loss": 0.3606,
+      "step": 6037
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.827814085922688,
+      "learning_rate": 9.64598159289426e-06,
+      "loss": 0.4323,
+      "step": 6038
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.149774693588026,
+      "learning_rate": 9.645835957991098e-06,
+      "loss": 0.532,
+      "step": 6039
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4720549177853093,
+      "learning_rate": 9.645690294238557e-06,
+      "loss": 0.1769,
+      "step": 6040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4169377198848643,
+      "learning_rate": 9.645544601637536e-06,
+      "loss": 0.1739,
+      "step": 6041
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.641442594711368,
+      "learning_rate": 9.645398880188944e-06,
+      "loss": 0.3771,
+      "step": 6042
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7986620351053906,
+      "learning_rate": 9.645253129893685e-06,
+      "loss": 0.4747,
+      "step": 6043
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3809429990491042,
+      "learning_rate": 9.645107350752663e-06,
+      "loss": 0.3179,
+      "step": 6044
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8413832595484059,
+      "learning_rate": 9.644961542766786e-06,
+      "loss": 0.3591,
+      "step": 6045
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.115801485989047,
+      "learning_rate": 9.644815705936956e-06,
+      "loss": 0.2792,
+      "step": 6046
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1972231818512404,
+      "learning_rate": 9.644669840264081e-06,
+      "loss": 0.2553,
+      "step": 6047
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.549833092299364,
+      "learning_rate": 9.644523945749067e-06,
+      "loss": 0.5405,
+      "step": 6048
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4884975055962633,
+      "learning_rate": 9.644378022392817e-06,
+      "loss": 0.3404,
+      "step": 6049
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.884958184243894,
+      "learning_rate": 9.64423207019624e-06,
+      "loss": 0.3714,
+      "step": 6050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.124647633705521,
+      "learning_rate": 9.644086089160241e-06,
+      "loss": 0.3749,
+      "step": 6051
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.798648421416259,
+      "learning_rate": 9.643940079285727e-06,
+      "loss": 0.3655,
+      "step": 6052
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3516537106337347,
+      "learning_rate": 9.643794040573604e-06,
+      "loss": 0.2053,
+      "step": 6053
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.522945830893899,
+      "learning_rate": 9.64364797302478e-06,
+      "loss": 0.2257,
+      "step": 6054
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2063682990113351,
+      "learning_rate": 9.643501876640162e-06,
+      "loss": 0.2615,
+      "step": 6055
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7881807669843854,
+      "learning_rate": 9.643355751420656e-06,
+      "loss": 0.2167,
+      "step": 6056
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6649041621803726,
+      "learning_rate": 9.643209597367171e-06,
+      "loss": 0.4658,
+      "step": 6057
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8136486594238166,
+      "learning_rate": 9.643063414480614e-06,
+      "loss": 0.5221,
+      "step": 6058
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7046311700466872,
+      "learning_rate": 9.64291720276189e-06,
+      "loss": 0.4235,
+      "step": 6059
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.606128809861955,
+      "learning_rate": 9.642770962211911e-06,
+      "loss": 0.4672,
+      "step": 6060
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3357636857314263,
+      "learning_rate": 9.642624692831584e-06,
+      "loss": 0.2596,
+      "step": 6061
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6244252247511133,
+      "learning_rate": 9.642478394621814e-06,
+      "loss": 0.2672,
+      "step": 6062
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7875570804770475,
+      "learning_rate": 9.642332067583515e-06,
+      "loss": 0.2022,
+      "step": 6063
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6199959105385568,
+      "learning_rate": 9.642185711717594e-06,
+      "loss": 0.3365,
+      "step": 6064
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.889879751710149,
+      "learning_rate": 9.642039327024957e-06,
+      "loss": 0.3675,
+      "step": 6065
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7539884297376418,
+      "learning_rate": 9.641892913506514e-06,
+      "loss": 0.3777,
+      "step": 6066
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8542063102520259,
+      "learning_rate": 9.641746471163175e-06,
+      "loss": 0.3934,
+      "step": 6067
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.185124456169567,
+      "learning_rate": 9.64159999999585e-06,
+      "loss": 0.7488,
+      "step": 6068
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.570157859551638,
+      "learning_rate": 9.641453500005448e-06,
+      "loss": 0.3811,
+      "step": 6069
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.687815972169053,
+      "learning_rate": 9.641306971192878e-06,
+      "loss": 0.3708,
+      "step": 6070
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6620401016125879,
+      "learning_rate": 9.641160413559051e-06,
+      "loss": 0.337,
+      "step": 6071
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7180586204597412,
+      "learning_rate": 9.641013827104876e-06,
+      "loss": 0.4798,
+      "step": 6072
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.303594317037356,
+      "learning_rate": 9.640867211831264e-06,
+      "loss": 0.76,
+      "step": 6073
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.906813721867811,
+      "learning_rate": 9.640720567739126e-06,
+      "loss": 0.4843,
+      "step": 6074
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0365811585987301,
+      "learning_rate": 9.640573894829371e-06,
+      "loss": 0.1446,
+      "step": 6075
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2610715259564076,
+      "learning_rate": 9.640427193102911e-06,
+      "loss": 0.582,
+      "step": 6076
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7942485937134545,
+      "learning_rate": 9.640280462560657e-06,
+      "loss": 0.5184,
+      "step": 6077
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0225196840898576,
+      "learning_rate": 9.64013370320352e-06,
+      "loss": 0.5757,
+      "step": 6078
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9936880556486858,
+      "learning_rate": 9.639986915032411e-06,
+      "loss": 0.4496,
+      "step": 6079
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.85040075135475,
+      "learning_rate": 9.639840098048241e-06,
+      "loss": 0.3198,
+      "step": 6080
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.435680955347322,
+      "learning_rate": 9.639693252251922e-06,
+      "loss": 0.264,
+      "step": 6081
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9669886971871602,
+      "learning_rate": 9.639546377644367e-06,
+      "loss": 0.3078,
+      "step": 6082
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2974289951759626,
+      "learning_rate": 9.639399474226488e-06,
+      "loss": 0.1977,
+      "step": 6083
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9084199530364974,
+      "learning_rate": 9.639252541999196e-06,
+      "loss": 0.6086,
+      "step": 6084
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.922011149631367,
+      "learning_rate": 9.639105580963403e-06,
+      "loss": 0.4639,
+      "step": 6085
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.382451743142525,
+      "learning_rate": 9.638958591120022e-06,
+      "loss": 0.239,
+      "step": 6086
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9396640647509875,
+      "learning_rate": 9.638811572469969e-06,
+      "loss": 0.4777,
+      "step": 6087
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.404316614195313,
+      "learning_rate": 9.638664525014153e-06,
+      "loss": 0.2238,
+      "step": 6088
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8248626177171916,
+      "learning_rate": 9.638517448753486e-06,
+      "loss": 0.3845,
+      "step": 6089
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5825169853700793,
+      "learning_rate": 9.638370343688886e-06,
+      "loss": 0.2057,
+      "step": 6090
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4751502140776607,
+      "learning_rate": 9.638223209821264e-06,
+      "loss": 0.5014,
+      "step": 6091
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8340840662957185,
+      "learning_rate": 9.638076047151533e-06,
+      "loss": 0.5688,
+      "step": 6092
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8324961733225151,
+      "learning_rate": 9.637928855680608e-06,
+      "loss": 0.2324,
+      "step": 6093
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1913371661715515,
+      "learning_rate": 9.637781635409403e-06,
+      "loss": 0.3326,
+      "step": 6094
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.065739370806929,
+      "learning_rate": 9.637634386338833e-06,
+      "loss": 0.3495,
+      "step": 6095
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8157902647299684,
+      "learning_rate": 9.63748710846981e-06,
+      "loss": 0.3891,
+      "step": 6096
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1102920728610233,
+      "learning_rate": 9.637339801803248e-06,
+      "loss": 0.449,
+      "step": 6097
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8972872479349172,
+      "learning_rate": 9.637192466340065e-06,
+      "loss": 0.5157,
+      "step": 6098
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5647836400341129,
+      "learning_rate": 9.637045102081174e-06,
+      "loss": 0.3424,
+      "step": 6099
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1309657153370485,
+      "learning_rate": 9.636897709027491e-06,
+      "loss": 0.5924,
+      "step": 6100
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6732004027961886,
+      "learning_rate": 9.636750287179933e-06,
+      "loss": 0.4389,
+      "step": 6101
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.991000779433172,
+      "learning_rate": 9.63660283653941e-06,
+      "loss": 0.2609,
+      "step": 6102
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.323743982643593,
+      "learning_rate": 9.636455357106843e-06,
+      "loss": 0.1514,
+      "step": 6103
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6176170824376093,
+      "learning_rate": 9.636307848883144e-06,
+      "loss": 0.3208,
+      "step": 6104
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8688396303186234,
+      "learning_rate": 9.636160311869232e-06,
+      "loss": 0.1849,
+      "step": 6105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5620152412892363,
+      "learning_rate": 9.63601274606602e-06,
+      "loss": 0.2311,
+      "step": 6106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8101981000956389,
+      "learning_rate": 9.635865151474427e-06,
+      "loss": 0.2505,
+      "step": 6107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.679809064679998,
+      "learning_rate": 9.635717528095369e-06,
+      "loss": 0.4849,
+      "step": 6108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5739792270039423,
+      "learning_rate": 9.635569875929762e-06,
+      "loss": 0.2851,
+      "step": 6109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3616364073902805,
+      "learning_rate": 9.635422194978525e-06,
+      "loss": 0.1684,
+      "step": 6110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2413313291040717,
+      "learning_rate": 9.635274485242571e-06,
+      "loss": 0.484,
+      "step": 6111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.877306219226653,
+      "learning_rate": 9.635126746722821e-06,
+      "loss": 0.4127,
+      "step": 6112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6440308822219276,
+      "learning_rate": 9.63497897942019e-06,
+      "loss": 0.2604,
+      "step": 6113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9375797301469153,
+      "learning_rate": 9.634831183335598e-06,
+      "loss": 0.5376,
+      "step": 6114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.66743188678349,
+      "learning_rate": 9.63468335846996e-06,
+      "loss": 0.2405,
+      "step": 6115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9288939493002644,
+      "learning_rate": 9.634535504824196e-06,
+      "loss": 0.3412,
+      "step": 6116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7822823878911975,
+      "learning_rate": 9.634387622399223e-06,
+      "loss": 0.444,
+      "step": 6117
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6461758081217996,
+      "learning_rate": 9.63423971119596e-06,
+      "loss": 0.4798,
+      "step": 6118
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5244578930102795,
+      "learning_rate": 9.634091771215326e-06,
+      "loss": 0.3862,
+      "step": 6119
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.718701766160926,
+      "learning_rate": 9.633943802458239e-06,
+      "loss": 0.2695,
+      "step": 6120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.367212641617112,
+      "learning_rate": 9.633795804925617e-06,
+      "loss": 0.376,
+      "step": 6121
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.05033305947267,
+      "learning_rate": 9.63364777861838e-06,
+      "loss": 0.4244,
+      "step": 6122
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1883710353181547,
+      "learning_rate": 9.633499723537445e-06,
+      "loss": 0.4192,
+      "step": 6123
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7648363208656606,
+      "learning_rate": 9.633351639683736e-06,
+      "loss": 0.3259,
+      "step": 6124
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7251576279063838,
+      "learning_rate": 9.633203527058168e-06,
+      "loss": 0.274,
+      "step": 6125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8343973192550043,
+      "learning_rate": 9.633055385661664e-06,
+      "loss": 0.3136,
+      "step": 6126
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.311322491519794,
+      "learning_rate": 9.632907215495143e-06,
+      "loss": 0.1483,
+      "step": 6127
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.13369163098697,
+      "learning_rate": 9.632759016559523e-06,
+      "loss": 0.5942,
+      "step": 6128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5259167378122656,
+      "learning_rate": 9.632610788855728e-06,
+      "loss": 0.2244,
+      "step": 6129
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.274004752831917,
+      "learning_rate": 9.632462532384674e-06,
+      "loss": 0.2302,
+      "step": 6130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.029943552230701,
+      "learning_rate": 9.632314247147284e-06,
+      "loss": 0.4341,
+      "step": 6131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8393145243495428,
+      "learning_rate": 9.632165933144482e-06,
+      "loss": 0.4551,
+      "step": 6132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1448742256992386,
+      "learning_rate": 9.632017590377184e-06,
+      "loss": 0.5264,
+      "step": 6133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.880754002424994,
+      "learning_rate": 9.631869218846312e-06,
+      "loss": 0.3098,
+      "step": 6134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.251571028155206,
+      "learning_rate": 9.63172081855279e-06,
+      "loss": 0.2492,
+      "step": 6135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1123801829426125,
+      "learning_rate": 9.631572389497538e-06,
+      "loss": 0.7329,
+      "step": 6136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7817837664129141,
+      "learning_rate": 9.631423931681476e-06,
+      "loss": 0.4275,
+      "step": 6137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7679255570863859,
+      "learning_rate": 9.631275445105528e-06,
+      "loss": 0.5412,
+      "step": 6138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.427376703634631,
+      "learning_rate": 9.631126929770617e-06,
+      "loss": 0.2101,
+      "step": 6139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.089993209485608,
+      "learning_rate": 9.630978385677663e-06,
+      "loss": 0.4276,
+      "step": 6140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8770181444161964,
+      "learning_rate": 9.630829812827587e-06,
+      "loss": 0.3991,
+      "step": 6141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.046937795942494,
+      "learning_rate": 9.630681211221315e-06,
+      "loss": 0.3971,
+      "step": 6142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1740736553828819,
+      "learning_rate": 9.63053258085977e-06,
+      "loss": 0.1713,
+      "step": 6143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7208050536363584,
+      "learning_rate": 9.630383921743873e-06,
+      "loss": 0.3029,
+      "step": 6144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.705570739688168,
+      "learning_rate": 9.630235233874547e-06,
+      "loss": 0.4954,
+      "step": 6145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2693579602685223,
+      "learning_rate": 9.630086517252717e-06,
+      "loss": 0.5096,
+      "step": 6146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1558603519436856,
+      "learning_rate": 9.629937771879304e-06,
+      "loss": 0.1566,
+      "step": 6147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3295214954506596,
+      "learning_rate": 9.629788997755234e-06,
+      "loss": 0.1887,
+      "step": 6148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7153429899183263,
+      "learning_rate": 9.629640194881428e-06,
+      "loss": 0.3999,
+      "step": 6149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3378743832013036,
+      "learning_rate": 9.629491363258816e-06,
+      "loss": 0.2685,
+      "step": 6150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.230499376188335,
+      "learning_rate": 9.629342502888315e-06,
+      "loss": 0.1769,
+      "step": 6151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7448764756690613,
+      "learning_rate": 9.629193613770853e-06,
+      "loss": 0.3971,
+      "step": 6152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.777646711749225,
+      "learning_rate": 9.629044695907354e-06,
+      "loss": 0.1413,
+      "step": 6153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8170256922311536,
+      "learning_rate": 9.628895749298745e-06,
+      "loss": 0.2444,
+      "step": 6154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.609986668151692,
+      "learning_rate": 9.628746773945948e-06,
+      "loss": 0.2676,
+      "step": 6155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4913030294924416,
+      "learning_rate": 9.628597769849889e-06,
+      "loss": 0.1933,
+      "step": 6156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.152025341152399,
+      "learning_rate": 9.628448737011491e-06,
+      "loss": 0.3554,
+      "step": 6157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.148355468571105,
+      "learning_rate": 9.628299675431683e-06,
+      "loss": 0.5453,
+      "step": 6158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1028252915142915,
+      "learning_rate": 9.628150585111388e-06,
+      "loss": 0.1156,
+      "step": 6159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8284058864516015,
+      "learning_rate": 9.628001466051535e-06,
+      "loss": 0.2276,
+      "step": 6160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.055479310493207,
+      "learning_rate": 9.627852318253048e-06,
+      "loss": 0.6173,
+      "step": 6161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.917674916702624,
+      "learning_rate": 9.627703141716853e-06,
+      "loss": 0.4774,
+      "step": 6162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8196223346323548,
+      "learning_rate": 9.627553936443874e-06,
+      "loss": 0.3918,
+      "step": 6163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5185008343043342,
+      "learning_rate": 9.627404702435044e-06,
+      "loss": 0.2442,
+      "step": 6164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.250806143213343,
+      "learning_rate": 9.627255439691283e-06,
+      "loss": 0.1996,
+      "step": 6165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.449826008593588,
+      "learning_rate": 9.627106148213521e-06,
+      "loss": 0.5305,
+      "step": 6166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.905267208055942,
+      "learning_rate": 9.626956828002686e-06,
+      "loss": 0.3573,
+      "step": 6167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.007055058083811,
+      "learning_rate": 9.626807479059704e-06,
+      "loss": 0.367,
+      "step": 6168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6714241877286289,
+      "learning_rate": 9.6266581013855e-06,
+      "loss": 0.2759,
+      "step": 6169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.007107900490007,
+      "learning_rate": 9.626508694981008e-06,
+      "loss": 0.5141,
+      "step": 6170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9719901707200667,
+      "learning_rate": 9.62635925984715e-06,
+      "loss": 0.404,
+      "step": 6171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8431139430752184,
+      "learning_rate": 9.626209795984854e-06,
+      "loss": 0.3821,
+      "step": 6172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.986628630645906,
+      "learning_rate": 9.62606030339505e-06,
+      "loss": 0.4294,
+      "step": 6173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8475159476886274,
+      "learning_rate": 9.625910782078667e-06,
+      "loss": 0.2134,
+      "step": 6174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7700197795999286,
+      "learning_rate": 9.625761232036633e-06,
+      "loss": 0.4127,
+      "step": 6175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5994123679890555,
+      "learning_rate": 9.625611653269876e-06,
+      "loss": 0.3444,
+      "step": 6176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.020061917559932,
+      "learning_rate": 9.625462045779326e-06,
+      "loss": 0.3965,
+      "step": 6177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.549879489336235,
+      "learning_rate": 9.62531240956591e-06,
+      "loss": 0.2951,
+      "step": 6178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7385696129088901,
+      "learning_rate": 9.625162744630559e-06,
+      "loss": 0.3805,
+      "step": 6179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0058404982884865,
+      "learning_rate": 9.625013050974201e-06,
+      "loss": 0.3899,
+      "step": 6180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6965456314499032,
+      "learning_rate": 9.624863328597767e-06,
+      "loss": 0.2652,
+      "step": 6181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5341096567042338,
+      "learning_rate": 9.624713577502185e-06,
+      "loss": 0.363,
+      "step": 6182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6681625914670468,
+      "learning_rate": 9.624563797688387e-06,
+      "loss": 0.5302,
+      "step": 6183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5981834260034493,
+      "learning_rate": 9.624413989157301e-06,
+      "loss": 0.2331,
+      "step": 6184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9091731059163402,
+      "learning_rate": 9.62426415190986e-06,
+      "loss": 0.3352,
+      "step": 6185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6106554372736614,
+      "learning_rate": 9.624114285946993e-06,
+      "loss": 0.3074,
+      "step": 6186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.220940212876992,
+      "learning_rate": 9.623964391269629e-06,
+      "loss": 0.2436,
+      "step": 6187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.005457710678573,
+      "learning_rate": 9.623814467878699e-06,
+      "loss": 0.3793,
+      "step": 6188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.555766428556659,
+      "learning_rate": 9.623664515775137e-06,
+      "loss": 0.3003,
+      "step": 6189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8182089812040565,
+      "learning_rate": 9.623514534959872e-06,
+      "loss": 0.5457,
+      "step": 6190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9515796460573565,
+      "learning_rate": 9.623364525433836e-06,
+      "loss": 0.5316,
+      "step": 6191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8399899904481594,
+      "learning_rate": 9.623214487197959e-06,
+      "loss": 0.4116,
+      "step": 6192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6519498254053444,
+      "learning_rate": 9.623064420253175e-06,
+      "loss": 0.3294,
+      "step": 6193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7561946289999535,
+      "learning_rate": 9.622914324600414e-06,
+      "loss": 0.5135,
+      "step": 6194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.76328199772783,
+      "learning_rate": 9.62276420024061e-06,
+      "loss": 0.5657,
+      "step": 6195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9601345293367765,
+      "learning_rate": 9.622614047174694e-06,
+      "loss": 0.1471,
+      "step": 6196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9109625684477203,
+      "learning_rate": 9.622463865403598e-06,
+      "loss": 0.5211,
+      "step": 6197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5632305273342135,
+      "learning_rate": 9.622313654928253e-06,
+      "loss": 0.2364,
+      "step": 6198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6892664701513609,
+      "learning_rate": 9.622163415749596e-06,
+      "loss": 0.1902,
+      "step": 6199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4341223571115933,
+      "learning_rate": 9.622013147868557e-06,
+      "loss": 0.1127,
+      "step": 6200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9914048321712532,
+      "learning_rate": 9.62186285128607e-06,
+      "loss": 0.4928,
+      "step": 6201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7292496453525168,
+      "learning_rate": 9.621712526003068e-06,
+      "loss": 0.3929,
+      "step": 6202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.719725028822051,
+      "learning_rate": 9.621562172020486e-06,
+      "loss": 0.4681,
+      "step": 6203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8460834212302784,
+      "learning_rate": 9.621411789339257e-06,
+      "loss": 0.4398,
+      "step": 6204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0774407565100141,
+      "learning_rate": 9.621261377960311e-06,
+      "loss": 0.1641,
+      "step": 6205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9618344554254081,
+      "learning_rate": 9.621110937884587e-06,
+      "loss": 0.5943,
+      "step": 6206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8816359789353232,
+      "learning_rate": 9.620960469113017e-06,
+      "loss": 0.4928,
+      "step": 6207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.534168127250397,
+      "learning_rate": 9.620809971646536e-06,
+      "loss": 0.5981,
+      "step": 6208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9783420769059843,
+      "learning_rate": 9.620659445486078e-06,
+      "loss": 0.4741,
+      "step": 6209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9080441973908397,
+      "learning_rate": 9.620508890632578e-06,
+      "loss": 0.4386,
+      "step": 6210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.680409726410263,
+      "learning_rate": 9.62035830708697e-06,
+      "loss": 0.3785,
+      "step": 6211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.061184882875651,
+      "learning_rate": 9.620207694850192e-06,
+      "loss": 0.3548,
+      "step": 6212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5596817050868987,
+      "learning_rate": 9.620057053923177e-06,
+      "loss": 0.3269,
+      "step": 6213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2541585744640846,
+      "learning_rate": 9.61990638430686e-06,
+      "loss": 0.2765,
+      "step": 6214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7058574003448832,
+      "learning_rate": 9.619755686002176e-06,
+      "loss": 0.2213,
+      "step": 6215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.632799042983932,
+      "learning_rate": 9.619604959010065e-06,
+      "loss": 0.1971,
+      "step": 6216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5129167654476416,
+      "learning_rate": 9.619454203331458e-06,
+      "loss": 0.3313,
+      "step": 6217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6622527090952166,
+      "learning_rate": 9.619303418967295e-06,
+      "loss": 0.362,
+      "step": 6218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8265093731231117,
+      "learning_rate": 9.619152605918507e-06,
+      "loss": 0.3264,
+      "step": 6219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.625533650189951,
+      "learning_rate": 9.619001764186037e-06,
+      "loss": 0.3193,
+      "step": 6220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.495066327326907,
+      "learning_rate": 9.618850893770818e-06,
+      "loss": 0.1798,
+      "step": 6221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5429332560066256,
+      "learning_rate": 9.618699994673787e-06,
+      "loss": 0.2084,
+      "step": 6222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.727368984184746,
+      "learning_rate": 9.618549066895884e-06,
+      "loss": 0.252,
+      "step": 6223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5944992108936697,
+      "learning_rate": 9.618398110438042e-06,
+      "loss": 0.4705,
+      "step": 6224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6126782800792685,
+      "learning_rate": 9.6182471253012e-06,
+      "loss": 0.4787,
+      "step": 6225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0205037659925333,
+      "learning_rate": 9.618096111486296e-06,
+      "loss": 0.1857,
+      "step": 6226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6478820775801029,
+      "learning_rate": 9.617945068994267e-06,
+      "loss": 0.323,
+      "step": 6227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.669368290746668,
+      "learning_rate": 9.617793997826054e-06,
+      "loss": 0.26,
+      "step": 6228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9234588501006389,
+      "learning_rate": 9.617642897982591e-06,
+      "loss": 0.3445,
+      "step": 6229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1069701280862327,
+      "learning_rate": 9.617491769464819e-06,
+      "loss": 0.1587,
+      "step": 6230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3622191070693042,
+      "learning_rate": 9.617340612273673e-06,
+      "loss": 0.1914,
+      "step": 6231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3291496403748335,
+      "learning_rate": 9.617189426410096e-06,
+      "loss": 0.1818,
+      "step": 6232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5128946166868613,
+      "learning_rate": 9.617038211875024e-06,
+      "loss": 0.3738,
+      "step": 6233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.175316663413933,
+      "learning_rate": 9.616886968669399e-06,
+      "loss": 0.4595,
+      "step": 6234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6484052419896054,
+      "learning_rate": 9.616735696794156e-06,
+      "loss": 0.2504,
+      "step": 6235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7733737371131926,
+      "learning_rate": 9.616584396250237e-06,
+      "loss": 0.2817,
+      "step": 6236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.697526494688119,
+      "learning_rate": 9.61643306703858e-06,
+      "loss": 0.2722,
+      "step": 6237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2417629351708017,
+      "learning_rate": 9.616281709160127e-06,
+      "loss": 0.5077,
+      "step": 6238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3579512446606024,
+      "learning_rate": 9.616130322615816e-06,
+      "loss": 0.3125,
+      "step": 6239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.003634034642569,
+      "learning_rate": 9.615978907406588e-06,
+      "loss": 0.5268,
+      "step": 6240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.95154503813751,
+      "learning_rate": 9.615827463533382e-06,
+      "loss": 0.3608,
+      "step": 6241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4793617222485826,
+      "learning_rate": 9.61567599099714e-06,
+      "loss": 0.1441,
+      "step": 6242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4462231911463472,
+      "learning_rate": 9.615524489798803e-06,
+      "loss": 0.2125,
+      "step": 6243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4668223484229959,
+      "learning_rate": 9.615372959939308e-06,
+      "loss": 0.3244,
+      "step": 6244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6241906975125933,
+      "learning_rate": 9.615221401419601e-06,
+      "loss": 0.3195,
+      "step": 6245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8336574031409418,
+      "learning_rate": 9.61506981424062e-06,
+      "loss": 0.453,
+      "step": 6246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8251143713608255,
+      "learning_rate": 9.614918198403306e-06,
+      "loss": 0.3392,
+      "step": 6247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0022453020462,
+      "learning_rate": 9.614766553908604e-06,
+      "loss": 0.5121,
+      "step": 6248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7115463344979982,
+      "learning_rate": 9.614614880757452e-06,
+      "loss": 0.2411,
+      "step": 6249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7696202068410798,
+      "learning_rate": 9.614463178950792e-06,
+      "loss": 0.3692,
+      "step": 6250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9661373938238305,
+      "learning_rate": 9.614311448489569e-06,
+      "loss": 0.4907,
+      "step": 6251
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.835543354824187,
+      "learning_rate": 9.614159689374723e-06,
+      "loss": 0.4702,
+      "step": 6252
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6459352555032423,
+      "learning_rate": 9.614007901607196e-06,
+      "loss": 0.2546,
+      "step": 6253
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6374955393825084,
+      "learning_rate": 9.61385608518793e-06,
+      "loss": 0.4027,
+      "step": 6254
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.066634066450222,
+      "learning_rate": 9.613704240117872e-06,
+      "loss": 0.2389,
+      "step": 6255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9395344919767505,
+      "learning_rate": 9.61355236639796e-06,
+      "loss": 0.5155,
+      "step": 6256
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2177796826359253,
+      "learning_rate": 9.61340046402914e-06,
+      "loss": 0.2688,
+      "step": 6257
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7475009854358956,
+      "learning_rate": 9.613248533012353e-06,
+      "loss": 0.3471,
+      "step": 6258
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7238736619257469,
+      "learning_rate": 9.613096573348544e-06,
+      "loss": 0.366,
+      "step": 6259
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.553982665962521,
+      "learning_rate": 9.612944585038658e-06,
+      "loss": 0.3756,
+      "step": 6260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3997110123638603,
+      "learning_rate": 9.612792568083636e-06,
+      "loss": 0.2369,
+      "step": 6261
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.077029254831733,
+      "learning_rate": 9.612640522484423e-06,
+      "loss": 0.5431,
+      "step": 6262
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9068698735076002,
+      "learning_rate": 9.612488448241964e-06,
+      "loss": 0.3438,
+      "step": 6263
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6590972844003724,
+      "learning_rate": 9.612336345357201e-06,
+      "loss": 0.2324,
+      "step": 6264
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.572834944865219,
+      "learning_rate": 9.612184213831081e-06,
+      "loss": 0.3457,
+      "step": 6265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9316566117553733,
+      "learning_rate": 9.612032053664551e-06,
+      "loss": 0.2545,
+      "step": 6266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5393727033507374,
+      "learning_rate": 9.611879864858548e-06,
+      "loss": 0.3355,
+      "step": 6267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9589846260169526,
+      "learning_rate": 9.611727647414026e-06,
+      "loss": 0.4268,
+      "step": 6268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.91917759312128,
+      "learning_rate": 9.611575401331923e-06,
+      "loss": 0.5469,
+      "step": 6269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7417920608950175,
+      "learning_rate": 9.611423126613188e-06,
+      "loss": 0.3743,
+      "step": 6270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3302570438341006,
+      "learning_rate": 9.611270823258767e-06,
+      "loss": 0.1787,
+      "step": 6271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0742577874197354,
+      "learning_rate": 9.611118491269605e-06,
+      "loss": 0.3464,
+      "step": 6272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.272142771586069,
+      "learning_rate": 9.610966130646647e-06,
+      "loss": 0.207,
+      "step": 6273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3997967683214954,
+      "learning_rate": 9.61081374139084e-06,
+      "loss": 0.1829,
+      "step": 6274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9589372326854024,
+      "learning_rate": 9.61066132350313e-06,
+      "loss": 0.562,
+      "step": 6275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3200169830737924,
+      "learning_rate": 9.610508876984463e-06,
+      "loss": 0.1559,
+      "step": 6276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.626119811246087,
+      "learning_rate": 9.610356401835788e-06,
+      "loss": 0.2989,
+      "step": 6277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.661687775068963,
+      "learning_rate": 9.610203898058049e-06,
+      "loss": 0.2244,
+      "step": 6278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1376707529580226,
+      "learning_rate": 9.610051365652196e-06,
+      "loss": 0.2242,
+      "step": 6279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8676047676897236,
+      "learning_rate": 9.60989880461917e-06,
+      "loss": 0.3577,
+      "step": 6280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5862871169909898,
+      "learning_rate": 9.609746214959926e-06,
+      "loss": 0.3676,
+      "step": 6281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4734911714691596,
+      "learning_rate": 9.609593596675408e-06,
+      "loss": 0.3186,
+      "step": 6282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1636448789489118,
+      "learning_rate": 9.609440949766562e-06,
+      "loss": 0.1889,
+      "step": 6283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.660932646571162,
+      "learning_rate": 9.60928827423434e-06,
+      "loss": 0.3942,
+      "step": 6284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2131269754650367,
+      "learning_rate": 9.609135570079689e-06,
+      "loss": 0.801,
+      "step": 6285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0893203527928195,
+      "learning_rate": 9.608982837303554e-06,
+      "loss": 0.21,
+      "step": 6286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5638206289199101,
+      "learning_rate": 9.608830075906885e-06,
+      "loss": 0.3445,
+      "step": 6287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.746794086165982,
+      "learning_rate": 9.608677285890632e-06,
+      "loss": 0.5298,
+      "step": 6288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3222930386784175,
+      "learning_rate": 9.608524467255742e-06,
+      "loss": 0.4639,
+      "step": 6289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.717470898208159,
+      "learning_rate": 9.608371620003167e-06,
+      "loss": 0.4886,
+      "step": 6290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3485686641190162,
+      "learning_rate": 9.608218744133853e-06,
+      "loss": 0.2248,
+      "step": 6291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6472994369644236,
+      "learning_rate": 9.60806583964875e-06,
+      "loss": 0.3535,
+      "step": 6292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.345062466940307,
+      "learning_rate": 9.607912906548809e-06,
+      "loss": 0.2686,
+      "step": 6293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5400415782145946,
+      "learning_rate": 9.607759944834977e-06,
+      "loss": 0.3302,
+      "step": 6294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3069027009214675,
+      "learning_rate": 9.607606954508207e-06,
+      "loss": 0.2171,
+      "step": 6295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1916205955989898,
+      "learning_rate": 9.607453935569446e-06,
+      "loss": 0.2258,
+      "step": 6296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4776968181575914,
+      "learning_rate": 9.607300888019645e-06,
+      "loss": 0.2329,
+      "step": 6297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8393985306793044,
+      "learning_rate": 9.607147811859759e-06,
+      "loss": 0.3842,
+      "step": 6298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8893196628226179,
+      "learning_rate": 9.60699470709073e-06,
+      "loss": 0.3997,
+      "step": 6299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6309697964685328,
+      "learning_rate": 9.606841573713516e-06,
+      "loss": 0.3393,
+      "step": 6300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.996571134750698,
+      "learning_rate": 9.606688411729064e-06,
+      "loss": 0.4517,
+      "step": 6301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0616315590963117,
+      "learning_rate": 9.606535221138326e-06,
+      "loss": 0.2067,
+      "step": 6302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3252540496650238,
+      "learning_rate": 9.606382001942256e-06,
+      "loss": 0.2351,
+      "step": 6303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0073667320847512,
+      "learning_rate": 9.606228754141801e-06,
+      "loss": 0.3652,
+      "step": 6304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7296827589312933,
+      "learning_rate": 9.606075477737915e-06,
+      "loss": 0.2585,
+      "step": 6305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7662501599817277,
+      "learning_rate": 9.60592217273155e-06,
+      "loss": 0.2435,
+      "step": 6306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.80872035561808,
+      "learning_rate": 9.605768839123656e-06,
+      "loss": 0.3821,
+      "step": 6307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1600115884557223,
+      "learning_rate": 9.605615476915188e-06,
+      "loss": 0.4114,
+      "step": 6308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1571928809987306,
+      "learning_rate": 9.605462086107099e-06,
+      "loss": 0.6099,
+      "step": 6309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.061781801543967,
+      "learning_rate": 9.605308666700335e-06,
+      "loss": 0.2263,
+      "step": 6310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7460846698021384,
+      "learning_rate": 9.605155218695856e-06,
+      "loss": 0.2934,
+      "step": 6311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5529468127143504,
+      "learning_rate": 9.605001742094611e-06,
+      "loss": 0.2207,
+      "step": 6312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.211301982434628,
+      "learning_rate": 9.604848236897554e-06,
+      "loss": 0.348,
+      "step": 6313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9519019989897162,
+      "learning_rate": 9.60469470310564e-06,
+      "loss": 0.5719,
+      "step": 6314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.426619462917831,
+      "learning_rate": 9.60454114071982e-06,
+      "loss": 0.4023,
+      "step": 6315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9634244050714456,
+      "learning_rate": 9.604387549741047e-06,
+      "loss": 0.6336,
+      "step": 6316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.770492923966439,
+      "learning_rate": 9.604233930170277e-06,
+      "loss": 0.3821,
+      "step": 6317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.174823652988983,
+      "learning_rate": 9.604080282008463e-06,
+      "loss": 0.5139,
+      "step": 6318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8052093671956067,
+      "learning_rate": 9.60392660525656e-06,
+      "loss": 0.2414,
+      "step": 6319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.683991425051938,
+      "learning_rate": 9.60377289991552e-06,
+      "loss": 0.3026,
+      "step": 6320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6153132754962556,
+      "learning_rate": 9.6036191659863e-06,
+      "loss": 0.3677,
+      "step": 6321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5491663886920393,
+      "learning_rate": 9.603465403469851e-06,
+      "loss": 0.4191,
+      "step": 6322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6903708065891812,
+      "learning_rate": 9.603311612367135e-06,
+      "loss": 0.1979,
+      "step": 6323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.377205978648694,
+      "learning_rate": 9.6031577926791e-06,
+      "loss": 0.816,
+      "step": 6324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6376442449178994,
+      "learning_rate": 9.603003944406702e-06,
+      "loss": 0.2573,
+      "step": 6325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4650591691883572,
+      "learning_rate": 9.602850067550897e-06,
+      "loss": 0.2766,
+      "step": 6326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.145160227999113,
+      "learning_rate": 9.602696162112644e-06,
+      "loss": 0.1618,
+      "step": 6327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3788001881514846,
+      "learning_rate": 9.602542228092896e-06,
+      "loss": 0.3849,
+      "step": 6328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6070191862759728,
+      "learning_rate": 9.602388265492608e-06,
+      "loss": 0.4582,
+      "step": 6329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8358712082927138,
+      "learning_rate": 9.602234274312739e-06,
+      "loss": 0.3683,
+      "step": 6330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7142535238830598,
+      "learning_rate": 9.60208025455424e-06,
+      "loss": 0.3035,
+      "step": 6331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6535931140776239,
+      "learning_rate": 9.601926206218074e-06,
+      "loss": 0.4813,
+      "step": 6332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9119589320724224,
+      "learning_rate": 9.601772129305192e-06,
+      "loss": 0.5682,
+      "step": 6333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2176634350955053,
+      "learning_rate": 9.601618023816555e-06,
+      "loss": 0.5433,
+      "step": 6334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6627055666685855,
+      "learning_rate": 9.601463889753118e-06,
+      "loss": 0.532,
+      "step": 6335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.661888533734738,
+      "learning_rate": 9.601309727115837e-06,
+      "loss": 0.4065,
+      "step": 6336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6989369235413698,
+      "learning_rate": 9.601155535905671e-06,
+      "loss": 0.3779,
+      "step": 6337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1384430836966026,
+      "learning_rate": 9.601001316123576e-06,
+      "loss": 0.4448,
+      "step": 6338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.387425131270654,
+      "learning_rate": 9.600847067770511e-06,
+      "loss": 0.1693,
+      "step": 6339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4744688914498054,
+      "learning_rate": 9.600692790847435e-06,
+      "loss": 0.3508,
+      "step": 6340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.425575423734324,
+      "learning_rate": 9.600538485355304e-06,
+      "loss": 0.1464,
+      "step": 6341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6289224966871882,
+      "learning_rate": 9.600384151295075e-06,
+      "loss": 0.4628,
+      "step": 6342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5461907559226442,
+      "learning_rate": 9.60022978866771e-06,
+      "loss": 0.296,
+      "step": 6343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3313993943032254,
+      "learning_rate": 9.600075397474164e-06,
+      "loss": 0.8338,
+      "step": 6344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.767043134963234,
+      "learning_rate": 9.5999209777154e-06,
+      "loss": 0.4567,
+      "step": 6345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9042996137550416,
+      "learning_rate": 9.599766529392371e-06,
+      "loss": 0.6145,
+      "step": 6346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.316250310768279,
+      "learning_rate": 9.599612052506043e-06,
+      "loss": 0.413,
+      "step": 6347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3934467731012428,
+      "learning_rate": 9.59945754705737e-06,
+      "loss": 0.2317,
+      "step": 6348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4868899955139663,
+      "learning_rate": 9.599303013047312e-06,
+      "loss": 0.2269,
+      "step": 6349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5548459120896303,
+      "learning_rate": 9.59914845047683e-06,
+      "loss": 0.4208,
+      "step": 6350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0326741361333274,
+      "learning_rate": 9.598993859346883e-06,
+      "loss": 0.2743,
+      "step": 6351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.414899292143288,
+      "learning_rate": 9.598839239658433e-06,
+      "loss": 0.4448,
+      "step": 6352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6448246429122109,
+      "learning_rate": 9.598684591412437e-06,
+      "loss": 0.2149,
+      "step": 6353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4881810182075634,
+      "learning_rate": 9.598529914609858e-06,
+      "loss": 0.3702,
+      "step": 6354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1113461365606163,
+      "learning_rate": 9.598375209251654e-06,
+      "loss": 0.1495,
+      "step": 6355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7966900999885278,
+      "learning_rate": 9.598220475338788e-06,
+      "loss": 0.2369,
+      "step": 6356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1089940892783803,
+      "learning_rate": 9.59806571287222e-06,
+      "loss": 0.6335,
+      "step": 6357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6259883158490958,
+      "learning_rate": 9.597910921852912e-06,
+      "loss": 0.4055,
+      "step": 6358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8025770315027185,
+      "learning_rate": 9.597756102281823e-06,
+      "loss": 0.291,
+      "step": 6359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.152833733976525,
+      "learning_rate": 9.597601254159915e-06,
+      "loss": 0.3963,
+      "step": 6360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.43580812166088,
+      "learning_rate": 9.59744637748815e-06,
+      "loss": 0.3153,
+      "step": 6361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5723607727546316,
+      "learning_rate": 9.597291472267491e-06,
+      "loss": 0.3077,
+      "step": 6362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5727422288036557,
+      "learning_rate": 9.5971365384989e-06,
+      "loss": 0.3726,
+      "step": 6363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.480979081620779,
+      "learning_rate": 9.596981576183335e-06,
+      "loss": 0.2294,
+      "step": 6364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6268691788868588,
+      "learning_rate": 9.596826585321763e-06,
+      "loss": 0.3261,
+      "step": 6365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.426633125721228,
+      "learning_rate": 9.596671565915145e-06,
+      "loss": 0.3289,
+      "step": 6366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7572827664187416,
+      "learning_rate": 9.596516517964441e-06,
+      "loss": 0.5088,
+      "step": 6367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8106868405853933,
+      "learning_rate": 9.596361441470618e-06,
+      "loss": 0.435,
+      "step": 6368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.261155716047469,
+      "learning_rate": 9.596206336434635e-06,
+      "loss": 0.5348,
+      "step": 6369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.677015636886413,
+      "learning_rate": 9.59605120285746e-06,
+      "loss": 0.2744,
+      "step": 6370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5430796185027962,
+      "learning_rate": 9.595896040740052e-06,
+      "loss": 0.204,
+      "step": 6371
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5283532209174477,
+      "learning_rate": 9.595740850083375e-06,
+      "loss": 0.2221,
+      "step": 6372
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.588285248156092,
+      "learning_rate": 9.595585630888395e-06,
+      "loss": 0.2096,
+      "step": 6373
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5191223383749688,
+      "learning_rate": 9.595430383156075e-06,
+      "loss": 0.1253,
+      "step": 6374
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1244677171864876,
+      "learning_rate": 9.595275106887378e-06,
+      "loss": 0.1199,
+      "step": 6375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.807827406107149,
+      "learning_rate": 9.595119802083268e-06,
+      "loss": 0.452,
+      "step": 6376
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3389218015406685,
+      "learning_rate": 9.594964468744711e-06,
+      "loss": 0.3446,
+      "step": 6377
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1891184650790143,
+      "learning_rate": 9.594809106872672e-06,
+      "loss": 0.1256,
+      "step": 6378
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9530520654411994,
+      "learning_rate": 9.594653716468114e-06,
+      "loss": 0.5318,
+      "step": 6379
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9020801494720765,
+      "learning_rate": 9.594498297532002e-06,
+      "loss": 0.4216,
+      "step": 6380
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3688475965951767,
+      "learning_rate": 9.594342850065302e-06,
+      "loss": 0.1012,
+      "step": 6381
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2997517334936945,
+      "learning_rate": 9.594187374068979e-06,
+      "loss": 0.2374,
+      "step": 6382
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.662571899841938,
+      "learning_rate": 9.594031869543998e-06,
+      "loss": 0.1759,
+      "step": 6383
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.721684633292258,
+      "learning_rate": 9.593876336491325e-06,
+      "loss": 0.2555,
+      "step": 6384
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.214547694644685,
+      "learning_rate": 9.593720774911926e-06,
+      "loss": 0.137,
+      "step": 6385
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4618109040440592,
+      "learning_rate": 9.593565184806765e-06,
+      "loss": 0.24,
+      "step": 6386
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9363223773193012,
+      "learning_rate": 9.593409566176812e-06,
+      "loss": 0.3864,
+      "step": 6387
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.065577771805467,
+      "learning_rate": 9.593253919023033e-06,
+      "loss": 0.4242,
+      "step": 6388
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3015133255396554,
+      "learning_rate": 9.59309824334639e-06,
+      "loss": 0.3447,
+      "step": 6389
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.244462318039088,
+      "learning_rate": 9.592942539147854e-06,
+      "loss": 0.6859,
+      "step": 6390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3951193747409176,
+      "learning_rate": 9.59278680642839e-06,
+      "loss": 0.1382,
+      "step": 6391
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4149592394503743,
+      "learning_rate": 9.592631045188966e-06,
+      "loss": 0.2423,
+      "step": 6392
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1520196491228032,
+      "learning_rate": 9.59247525543055e-06,
+      "loss": 0.1355,
+      "step": 6393
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4221393894833243,
+      "learning_rate": 9.592319437154106e-06,
+      "loss": 0.2346,
+      "step": 6394
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.71848996752959,
+      "learning_rate": 9.592163590360604e-06,
+      "loss": 0.3291,
+      "step": 6395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3644808979505085,
+      "learning_rate": 9.59200771505101e-06,
+      "loss": 0.7446,
+      "step": 6396
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7037711820629073,
+      "learning_rate": 9.591851811226298e-06,
+      "loss": 0.2133,
+      "step": 6397
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7353424862016222,
+      "learning_rate": 9.591695878887428e-06,
+      "loss": 0.2902,
+      "step": 6398
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2564062669377734,
+      "learning_rate": 9.591539918035374e-06,
+      "loss": 0.4,
+      "step": 6399
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6469775849264923,
+      "learning_rate": 9.5913839286711e-06,
+      "loss": 0.2732,
+      "step": 6400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.854329717267719,
+      "learning_rate": 9.591227910795578e-06,
+      "loss": 0.3765,
+      "step": 6401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.395592664911209,
+      "learning_rate": 9.591071864409777e-06,
+      "loss": 0.1363,
+      "step": 6402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5238789285670638,
+      "learning_rate": 9.590915789514663e-06,
+      "loss": 0.5994,
+      "step": 6403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0671656624728314,
+      "learning_rate": 9.590759686111208e-06,
+      "loss": 0.346,
+      "step": 6404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9163539327267536,
+      "learning_rate": 9.590603554200382e-06,
+      "loss": 0.523,
+      "step": 6405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0579092700271044,
+      "learning_rate": 9.590447393783151e-06,
+      "loss": 0.2197,
+      "step": 6406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9283851019948994,
+      "learning_rate": 9.590291204860487e-06,
+      "loss": 0.2469,
+      "step": 6407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7128891286984806,
+      "learning_rate": 9.590134987433359e-06,
+      "loss": 0.2545,
+      "step": 6408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.638853515451589,
+      "learning_rate": 9.58997874150274e-06,
+      "loss": 0.3043,
+      "step": 6409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.024648749387964,
+      "learning_rate": 9.589822467069594e-06,
+      "loss": 0.233,
+      "step": 6410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5023332134736294,
+      "learning_rate": 9.589666164134898e-06,
+      "loss": 0.1723,
+      "step": 6411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.462358082402399,
+      "learning_rate": 9.58950983269962e-06,
+      "loss": 0.1893,
+      "step": 6412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7602007367150052,
+      "learning_rate": 9.589353472764729e-06,
+      "loss": 0.214,
+      "step": 6413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4030693642135994,
+      "learning_rate": 9.589197084331198e-06,
+      "loss": 0.3802,
+      "step": 6414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8335494824800487,
+      "learning_rate": 9.589040667399999e-06,
+      "loss": 0.2577,
+      "step": 6415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.995696530319073,
+      "learning_rate": 9.5888842219721e-06,
+      "loss": 0.4391,
+      "step": 6416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4731948130835824,
+      "learning_rate": 9.588727748048478e-06,
+      "loss": 0.3482,
+      "step": 6417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3409990648557795,
+      "learning_rate": 9.588571245630099e-06,
+      "loss": 0.248,
+      "step": 6418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7033290675757162,
+      "learning_rate": 9.588414714717937e-06,
+      "loss": 0.3422,
+      "step": 6419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5990634983552399,
+      "learning_rate": 9.588258155312964e-06,
+      "loss": 0.3052,
+      "step": 6420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7567543002335593,
+      "learning_rate": 9.588101567416153e-06,
+      "loss": 0.3963,
+      "step": 6421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5333751156794386,
+      "learning_rate": 9.587944951028473e-06,
+      "loss": 0.3099,
+      "step": 6422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0644417314025545,
+      "learning_rate": 9.587788306150903e-06,
+      "loss": 0.21,
+      "step": 6423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2955932901177616,
+      "learning_rate": 9.587631632784409e-06,
+      "loss": 0.465,
+      "step": 6424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.540687228700582,
+      "learning_rate": 9.587474930929968e-06,
+      "loss": 0.3951,
+      "step": 6425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9782060529727041,
+      "learning_rate": 9.587318200588552e-06,
+      "loss": 0.5655,
+      "step": 6426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6352905925418322,
+      "learning_rate": 9.587161441761133e-06,
+      "loss": 0.3281,
+      "step": 6427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.854551084365604,
+      "learning_rate": 9.587004654448686e-06,
+      "loss": 0.5497,
+      "step": 6428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4916148160290936,
+      "learning_rate": 9.586847838652184e-06,
+      "loss": 0.3701,
+      "step": 6429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9551911450014812,
+      "learning_rate": 9.5866909943726e-06,
+      "loss": 0.5186,
+      "step": 6430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9448275555297823,
+      "learning_rate": 9.58653412161091e-06,
+      "loss": 0.5723,
+      "step": 6431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7707164071319372,
+      "learning_rate": 9.586377220368089e-06,
+      "loss": 0.2553,
+      "step": 6432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2625252466742674,
+      "learning_rate": 9.586220290645107e-06,
+      "loss": 0.6298,
+      "step": 6433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3811586344393003,
+      "learning_rate": 9.58606333244294e-06,
+      "loss": 0.2363,
+      "step": 6434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3247435161673686,
+      "learning_rate": 9.585906345762566e-06,
+      "loss": 0.5966,
+      "step": 6435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6822666648851257,
+      "learning_rate": 9.585749330604955e-06,
+      "loss": 0.276,
+      "step": 6436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5651185029230659,
+      "learning_rate": 9.585592286971086e-06,
+      "loss": 0.3785,
+      "step": 6437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9641753286457433,
+      "learning_rate": 9.58543521486193e-06,
+      "loss": 0.5012,
+      "step": 6438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.546862676841123,
+      "learning_rate": 9.58527811427847e-06,
+      "loss": 0.3878,
+      "step": 6439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.782094420145831,
+      "learning_rate": 9.585120985221672e-06,
+      "loss": 0.1816,
+      "step": 6440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2176856504121896,
+      "learning_rate": 9.584963827692518e-06,
+      "loss": 0.209,
+      "step": 6441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6369169952763205,
+      "learning_rate": 9.584806641691981e-06,
+      "loss": 0.3276,
+      "step": 6442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.580528949415098,
+      "learning_rate": 9.58464942722104e-06,
+      "loss": 0.2307,
+      "step": 6443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1473159151258576,
+      "learning_rate": 9.584492184280667e-06,
+      "loss": 0.6507,
+      "step": 6444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9889360382036467,
+      "learning_rate": 9.584334912871843e-06,
+      "loss": 0.3171,
+      "step": 6445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.616977247646338,
+      "learning_rate": 9.584177612995541e-06,
+      "loss": 0.3145,
+      "step": 6446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6166406020008393,
+      "learning_rate": 9.58402028465274e-06,
+      "loss": 0.3471,
+      "step": 6447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.559749394730507,
+      "learning_rate": 9.583862927844416e-06,
+      "loss": 0.3306,
+      "step": 6448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8469664917308843,
+      "learning_rate": 9.583705542571547e-06,
+      "loss": 0.5144,
+      "step": 6449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.104556924837655,
+      "learning_rate": 9.58354812883511e-06,
+      "loss": 0.2959,
+      "step": 6450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6059690401365352,
+      "learning_rate": 9.583390686636081e-06,
+      "loss": 0.2756,
+      "step": 6451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5620151721263893,
+      "learning_rate": 9.58323321597544e-06,
+      "loss": 0.3707,
+      "step": 6452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6624817750405094,
+      "learning_rate": 9.583075716854163e-06,
+      "loss": 0.476,
+      "step": 6453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.77088824822791,
+      "learning_rate": 9.582918189273229e-06,
+      "loss": 0.3513,
+      "step": 6454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4206537606277148,
+      "learning_rate": 9.582760633233615e-06,
+      "loss": 0.2063,
+      "step": 6455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8399202448337304,
+      "learning_rate": 9.582603048736302e-06,
+      "loss": 0.4063,
+      "step": 6456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3827162967773665,
+      "learning_rate": 9.582445435782266e-06,
+      "loss": 0.1375,
+      "step": 6457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.270494412618494,
+      "learning_rate": 9.582287794372486e-06,
+      "loss": 0.4092,
+      "step": 6458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9047681332951842,
+      "learning_rate": 9.582130124507942e-06,
+      "loss": 0.244,
+      "step": 6459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6736407091584597,
+      "learning_rate": 9.581972426189614e-06,
+      "loss": 0.2262,
+      "step": 6460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.603718079251513,
+      "learning_rate": 9.58181469941848e-06,
+      "loss": 0.2659,
+      "step": 6461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.447083522238455,
+      "learning_rate": 9.581656944195517e-06,
+      "loss": 0.3238,
+      "step": 6462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0360736792597334,
+      "learning_rate": 9.581499160521707e-06,
+      "loss": 0.4053,
+      "step": 6463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5781086647019176,
+      "learning_rate": 9.581341348398032e-06,
+      "loss": 0.2467,
+      "step": 6464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7211968214883229,
+      "learning_rate": 9.581183507825469e-06,
+      "loss": 0.2438,
+      "step": 6465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9293719508595248,
+      "learning_rate": 9.581025638805e-06,
+      "loss": 0.2541,
+      "step": 6466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9837614050637872,
+      "learning_rate": 9.580867741337601e-06,
+      "loss": 0.2109,
+      "step": 6467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8076248477915677,
+      "learning_rate": 9.580709815424259e-06,
+      "loss": 0.4408,
+      "step": 6468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9883900520818312,
+      "learning_rate": 9.580551861065949e-06,
+      "loss": 0.4044,
+      "step": 6469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8789161951511666,
+      "learning_rate": 9.580393878263655e-06,
+      "loss": 0.4735,
+      "step": 6470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3332060125092233,
+      "learning_rate": 9.580235867018357e-06,
+      "loss": 0.2579,
+      "step": 6471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3753483443463745,
+      "learning_rate": 9.580077827331038e-06,
+      "loss": 0.1734,
+      "step": 6472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8578686181937767,
+      "learning_rate": 9.579919759202676e-06,
+      "loss": 0.3845,
+      "step": 6473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.626400249141031,
+      "learning_rate": 9.579761662634255e-06,
+      "loss": 0.1741,
+      "step": 6474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8188996623752052,
+      "learning_rate": 9.579603537626755e-06,
+      "loss": 0.5174,
+      "step": 6475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.493005323735574,
+      "learning_rate": 9.579445384181161e-06,
+      "loss": 0.1288,
+      "step": 6476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8327625519075905,
+      "learning_rate": 9.579287202298454e-06,
+      "loss": 0.4186,
+      "step": 6477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3907348182263761,
+      "learning_rate": 9.579128991979613e-06,
+      "loss": 0.3759,
+      "step": 6478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8732404544191097,
+      "learning_rate": 9.578970753225623e-06,
+      "loss": 0.2146,
+      "step": 6479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5941287974316336,
+      "learning_rate": 9.578812486037468e-06,
+      "loss": 0.3492,
+      "step": 6480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.073963681215231,
+      "learning_rate": 9.578654190416128e-06,
+      "loss": 0.2692,
+      "step": 6481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6057859383570583,
+      "learning_rate": 9.578495866362588e-06,
+      "loss": 0.1729,
+      "step": 6482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.678856266958724,
+      "learning_rate": 9.578337513877829e-06,
+      "loss": 0.3618,
+      "step": 6483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0338660706689864,
+      "learning_rate": 9.578179132962838e-06,
+      "loss": 0.2161,
+      "step": 6484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.54548437051821,
+      "learning_rate": 9.578020723618595e-06,
+      "loss": 0.3918,
+      "step": 6485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6059835332607346,
+      "learning_rate": 9.577862285846084e-06,
+      "loss": 0.4059,
+      "step": 6486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.619195995833527,
+      "learning_rate": 9.577703819646291e-06,
+      "loss": 0.2222,
+      "step": 6487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9088164157092207,
+      "learning_rate": 9.577545325020199e-06,
+      "loss": 0.3343,
+      "step": 6488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1260637134917815,
+      "learning_rate": 9.577386801968791e-06,
+      "loss": 0.366,
+      "step": 6489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4737867625054089,
+      "learning_rate": 9.577228250493054e-06,
+      "loss": 0.4007,
+      "step": 6490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1396240831919053,
+      "learning_rate": 9.577069670593969e-06,
+      "loss": 0.5465,
+      "step": 6491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.638537557141663,
+      "learning_rate": 9.576911062272522e-06,
+      "loss": 0.2521,
+      "step": 6492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.38462531881457,
+      "learning_rate": 9.576752425529701e-06,
+      "loss": 0.1878,
+      "step": 6493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4299682839146783,
+      "learning_rate": 9.576593760366488e-06,
+      "loss": 0.1503,
+      "step": 6494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.046134144900255,
+      "learning_rate": 9.576435066783868e-06,
+      "loss": 0.5069,
+      "step": 6495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5437837259183909,
+      "learning_rate": 9.57627634478283e-06,
+      "loss": 0.2134,
+      "step": 6496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.04811454688796,
+      "learning_rate": 9.576117594364355e-06,
+      "loss": 0.5336,
+      "step": 6497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8084275929397318,
+      "learning_rate": 9.57595881552943e-06,
+      "loss": 0.3772,
+      "step": 6498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5757892007017245,
+      "learning_rate": 9.575800008279043e-06,
+      "loss": 0.471,
+      "step": 6499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.639976909332427,
+      "learning_rate": 9.575641172614179e-06,
+      "loss": 0.2078,
+      "step": 6500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4552015649036931,
+      "learning_rate": 9.575482308535824e-06,
+      "loss": 0.2759,
+      "step": 6501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7337757914109262,
+      "learning_rate": 9.575323416044965e-06,
+      "loss": 0.3222,
+      "step": 6502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4752285396390012,
+      "learning_rate": 9.575164495142589e-06,
+      "loss": 0.2193,
+      "step": 6503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5191773502621917,
+      "learning_rate": 9.57500554582968e-06,
+      "loss": 0.1542,
+      "step": 6504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6985241051064734,
+      "learning_rate": 9.57484656810723e-06,
+      "loss": 0.3571,
+      "step": 6505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3614014820600746,
+      "learning_rate": 9.57468756197622e-06,
+      "loss": 0.2311,
+      "step": 6506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.026441810796513,
+      "learning_rate": 9.574528527437646e-06,
+      "loss": 0.6233,
+      "step": 6507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5728213613470075,
+      "learning_rate": 9.574369464492488e-06,
+      "loss": 0.2131,
+      "step": 6508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7210998597385117,
+      "learning_rate": 9.574210373141735e-06,
+      "loss": 0.3022,
+      "step": 6509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0995946129568264,
+      "learning_rate": 9.574051253386379e-06,
+      "loss": 0.4842,
+      "step": 6510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6240317511701896,
+      "learning_rate": 9.573892105227403e-06,
+      "loss": 0.4743,
+      "step": 6511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8021103036016253,
+      "learning_rate": 9.573732928665797e-06,
+      "loss": 0.4301,
+      "step": 6512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4987375191748709,
+      "learning_rate": 9.573573723702552e-06,
+      "loss": 0.2361,
+      "step": 6513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9662431535992682,
+      "learning_rate": 9.573414490338653e-06,
+      "loss": 0.2361,
+      "step": 6514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5636597978105033,
+      "learning_rate": 9.57325522857509e-06,
+      "loss": 0.4199,
+      "step": 6515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2440721846339855,
+      "learning_rate": 9.573095938412853e-06,
+      "loss": 0.3349,
+      "step": 6516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2683637591334105,
+      "learning_rate": 9.57293661985293e-06,
+      "loss": 0.2075,
+      "step": 6517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.875138222050121,
+      "learning_rate": 9.572777272896312e-06,
+      "loss": 0.3065,
+      "step": 6518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3738122708661573,
+      "learning_rate": 9.572617897543987e-06,
+      "loss": 0.1611,
+      "step": 6519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.993224836934051,
+      "learning_rate": 9.572458493796946e-06,
+      "loss": 0.3272,
+      "step": 6520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3690646180402302,
+      "learning_rate": 9.572299061656176e-06,
+      "loss": 0.2074,
+      "step": 6521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5401410581283193,
+      "learning_rate": 9.57213960112267e-06,
+      "loss": 0.1558,
+      "step": 6522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5453793587410642,
+      "learning_rate": 9.571980112197416e-06,
+      "loss": 0.3817,
+      "step": 6523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8135740000278044,
+      "learning_rate": 9.571820594881407e-06,
+      "loss": 0.3872,
+      "step": 6524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4507554947103005,
+      "learning_rate": 9.57166104917563e-06,
+      "loss": 0.2174,
+      "step": 6525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4351665330330514,
+      "learning_rate": 9.571501475081079e-06,
+      "loss": 0.2226,
+      "step": 6526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7737707799577391,
+      "learning_rate": 9.571341872598743e-06,
+      "loss": 0.315,
+      "step": 6527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.386397482169604,
+      "learning_rate": 9.571182241729615e-06,
+      "loss": 0.707,
+      "step": 6528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9771495972460102,
+      "learning_rate": 9.571022582474684e-06,
+      "loss": 0.4107,
+      "step": 6529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4242272779049328,
+      "learning_rate": 9.570862894834943e-06,
+      "loss": 0.1907,
+      "step": 6530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.059620089266313,
+      "learning_rate": 9.570703178811382e-06,
+      "loss": 0.5908,
+      "step": 6531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.660215302707342,
+      "learning_rate": 9.570543434404996e-06,
+      "loss": 0.2819,
+      "step": 6532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.811696412909864,
+      "learning_rate": 9.570383661616774e-06,
+      "loss": 0.4833,
+      "step": 6533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9016812010855177,
+      "learning_rate": 9.570223860447708e-06,
+      "loss": 0.325,
+      "step": 6534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6581576715177386,
+      "learning_rate": 9.570064030898791e-06,
+      "loss": 0.3565,
+      "step": 6535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5620439817644842,
+      "learning_rate": 9.569904172971016e-06,
+      "loss": 0.4879,
+      "step": 6536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7548586929183696,
+      "learning_rate": 9.569744286665377e-06,
+      "loss": 0.3107,
+      "step": 6537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5450193263495544,
+      "learning_rate": 9.569584371982866e-06,
+      "loss": 0.3518,
+      "step": 6538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.584209642490387,
+      "learning_rate": 9.569424428924472e-06,
+      "loss": 0.3822,
+      "step": 6539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.103480757598782,
+      "learning_rate": 9.569264457491194e-06,
+      "loss": 0.5201,
+      "step": 6540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5589114823740575,
+      "learning_rate": 9.569104457684021e-06,
+      "loss": 0.3477,
+      "step": 6541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.036185899220835,
+      "learning_rate": 9.56894442950395e-06,
+      "loss": 0.41,
+      "step": 6542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8777147409116093,
+      "learning_rate": 9.568784372951973e-06,
+      "loss": 0.3928,
+      "step": 6543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5092705362710057,
+      "learning_rate": 9.568624288029084e-06,
+      "loss": 0.3945,
+      "step": 6544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5730574065415497,
+      "learning_rate": 9.568464174736278e-06,
+      "loss": 0.3817,
+      "step": 6545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7855539129398488,
+      "learning_rate": 9.56830403307455e-06,
+      "loss": 0.4019,
+      "step": 6546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7048692127375407,
+      "learning_rate": 9.56814386304489e-06,
+      "loss": 0.3985,
+      "step": 6547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.59846702485506,
+      "learning_rate": 9.567983664648296e-06,
+      "loss": 0.4144,
+      "step": 6548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.463016036073183,
+      "learning_rate": 9.567823437885762e-06,
+      "loss": 0.2647,
+      "step": 6549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1140901133431904,
+      "learning_rate": 9.567663182758284e-06,
+      "loss": 0.1652,
+      "step": 6550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1530884494641227,
+      "learning_rate": 9.567502899266857e-06,
+      "loss": 0.5549,
+      "step": 6551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8280024283636103,
+      "learning_rate": 9.567342587412474e-06,
+      "loss": 0.3896,
+      "step": 6552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6335167894519347,
+      "learning_rate": 9.567182247196133e-06,
+      "loss": 0.3239,
+      "step": 6553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7484922044256812,
+      "learning_rate": 9.56702187861883e-06,
+      "loss": 0.3435,
+      "step": 6554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4623662329595282,
+      "learning_rate": 9.56686148168156e-06,
+      "loss": 0.2442,
+      "step": 6555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.197020638652505,
+      "learning_rate": 9.566701056385317e-06,
+      "loss": 0.1999,
+      "step": 6556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4701131977479733,
+      "learning_rate": 9.5665406027311e-06,
+      "loss": 0.3001,
+      "step": 6557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.459970728098214,
+      "learning_rate": 9.566380120719905e-06,
+      "loss": 0.2579,
+      "step": 6558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8614958200236158,
+      "learning_rate": 9.566219610352726e-06,
+      "loss": 0.3868,
+      "step": 6559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.601213296176273,
+      "learning_rate": 9.566059071630562e-06,
+      "loss": 0.2865,
+      "step": 6560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.646496640359936,
+      "learning_rate": 9.56589850455441e-06,
+      "loss": 0.1983,
+      "step": 6561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.578322379355534,
+      "learning_rate": 9.565737909125268e-06,
+      "loss": 0.4117,
+      "step": 6562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3524484644666006,
+      "learning_rate": 9.56557728534413e-06,
+      "loss": 0.5108,
+      "step": 6563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9696245219060666,
+      "learning_rate": 9.565416633211996e-06,
+      "loss": 0.4911,
+      "step": 6564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.817030567631916,
+      "learning_rate": 9.565255952729862e-06,
+      "loss": 0.4567,
+      "step": 6565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8443522762733053,
+      "learning_rate": 9.565095243898728e-06,
+      "loss": 0.496,
+      "step": 6566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9412315470419643,
+      "learning_rate": 9.56493450671959e-06,
+      "loss": 0.2753,
+      "step": 6567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9204681178234606,
+      "learning_rate": 9.564773741193448e-06,
+      "loss": 0.4524,
+      "step": 6568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1827600624377705,
+      "learning_rate": 9.564612947321297e-06,
+      "loss": 0.1633,
+      "step": 6569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0312097692173565,
+      "learning_rate": 9.564452125104137e-06,
+      "loss": 0.4046,
+      "step": 6570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.245765463416707,
+      "learning_rate": 9.56429127454297e-06,
+      "loss": 0.1413,
+      "step": 6571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4719837838972603,
+      "learning_rate": 9.56413039563879e-06,
+      "loss": 0.303,
+      "step": 6572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.442099293483902,
+      "learning_rate": 9.5639694883926e-06,
+      "loss": 0.3343,
+      "step": 6573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.721737380854332,
+      "learning_rate": 9.563808552805396e-06,
+      "loss": 0.371,
+      "step": 6574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6922153724653892,
+      "learning_rate": 9.563647588878177e-06,
+      "loss": 0.239,
+      "step": 6575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7805441205401995,
+      "learning_rate": 9.563486596611947e-06,
+      "loss": 0.5683,
+      "step": 6576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1034130755037763,
+      "learning_rate": 9.563325576007702e-06,
+      "loss": 0.5047,
+      "step": 6577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7424136130538972,
+      "learning_rate": 9.563164527066441e-06,
+      "loss": 0.4839,
+      "step": 6578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6289098497163437,
+      "learning_rate": 9.563003449789167e-06,
+      "loss": 0.3393,
+      "step": 6579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5996939947641424,
+      "learning_rate": 9.562842344176882e-06,
+      "loss": 0.162,
+      "step": 6580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.496042683977359,
+      "learning_rate": 9.56268121023058e-06,
+      "loss": 0.2409,
+      "step": 6581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7270794284505686,
+      "learning_rate": 9.562520047951265e-06,
+      "loss": 0.1983,
+      "step": 6582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.647687290427424,
+      "learning_rate": 9.56235885733994e-06,
+      "loss": 0.2151,
+      "step": 6583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.000747176017613,
+      "learning_rate": 9.562197638397604e-06,
+      "loss": 0.4641,
+      "step": 6584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.168330274242976,
+      "learning_rate": 9.562036391125255e-06,
+      "loss": 0.3045,
+      "step": 6585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9640949151376967,
+      "learning_rate": 9.5618751155239e-06,
+      "loss": 0.282,
+      "step": 6586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3545971659908713,
+      "learning_rate": 9.561713811594538e-06,
+      "loss": 0.264,
+      "step": 6587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0862557957719003,
+      "learning_rate": 9.561552479338168e-06,
+      "loss": 0.5311,
+      "step": 6588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7399559140717902,
+      "learning_rate": 9.561391118755796e-06,
+      "loss": 0.3973,
+      "step": 6589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3584396306695368,
+      "learning_rate": 9.561229729848421e-06,
+      "loss": 0.2595,
+      "step": 6590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.698258595664291,
+      "learning_rate": 9.561068312617048e-06,
+      "loss": 0.4234,
+      "step": 6591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2325473931562707,
+      "learning_rate": 9.560906867062676e-06,
+      "loss": 0.1675,
+      "step": 6592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4150587729482502,
+      "learning_rate": 9.56074539318631e-06,
+      "loss": 0.3487,
+      "step": 6593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7451000116716597,
+      "learning_rate": 9.560583890988953e-06,
+      "loss": 0.4207,
+      "step": 6594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4597182068895318,
+      "learning_rate": 9.560422360471606e-06,
+      "loss": 0.1885,
+      "step": 6595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2242974457463471,
+      "learning_rate": 9.560260801635272e-06,
+      "loss": 0.1946,
+      "step": 6596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.059798827139749,
+      "learning_rate": 9.560099214480957e-06,
+      "loss": 0.4637,
+      "step": 6597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4187878090067922,
+      "learning_rate": 9.559937599009663e-06,
+      "loss": 0.3108,
+      "step": 6598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.348811254498028,
+      "learning_rate": 9.559775955222392e-06,
+      "loss": 0.2793,
+      "step": 6599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.891263327936105,
+      "learning_rate": 9.559614283120148e-06,
+      "loss": 0.3816,
+      "step": 6600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8275923971987755,
+      "learning_rate": 9.559452582703938e-06,
+      "loss": 0.431,
+      "step": 6601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6859425636141219,
+      "learning_rate": 9.559290853974763e-06,
+      "loss": 0.3546,
+      "step": 6602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6873682133995729,
+      "learning_rate": 9.559129096933628e-06,
+      "loss": 0.3375,
+      "step": 6603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8079927488129923,
+      "learning_rate": 9.558967311581538e-06,
+      "loss": 0.5696,
+      "step": 6604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1101786580330937,
+      "learning_rate": 9.558805497919498e-06,
+      "loss": 0.1549,
+      "step": 6605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6799933753848026,
+      "learning_rate": 9.558643655948512e-06,
+      "loss": 0.3558,
+      "step": 6606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.460041123041306,
+      "learning_rate": 9.558481785669586e-06,
+      "loss": 0.3454,
+      "step": 6607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4454604730378382,
+      "learning_rate": 9.558319887083723e-06,
+      "loss": 0.2681,
+      "step": 6608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4418493819926983,
+      "learning_rate": 9.55815796019193e-06,
+      "loss": 0.2559,
+      "step": 6609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7189103745285412,
+      "learning_rate": 9.557996004995212e-06,
+      "loss": 0.3516,
+      "step": 6610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3851755418889005,
+      "learning_rate": 9.557834021494576e-06,
+      "loss": 0.1958,
+      "step": 6611
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4634511431486343,
+      "learning_rate": 9.557672009691026e-06,
+      "loss": 0.3595,
+      "step": 6612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5134553896005978,
+      "learning_rate": 9.55750996958557e-06,
+      "loss": 0.22,
+      "step": 6613
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9096962954225467,
+      "learning_rate": 9.557347901179213e-06,
+      "loss": 0.251,
+      "step": 6614
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.372947309419295,
+      "learning_rate": 9.55718580447296e-06,
+      "loss": 0.2555,
+      "step": 6615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8206704416064854,
+      "learning_rate": 9.557023679467819e-06,
+      "loss": 0.2367,
+      "step": 6616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3395220701240607,
+      "learning_rate": 9.556861526164798e-06,
+      "loss": 0.3984,
+      "step": 6617
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6377638186626025,
+      "learning_rate": 9.556699344564904e-06,
+      "loss": 0.3543,
+      "step": 6618
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.767013362632412,
+      "learning_rate": 9.556537134669142e-06,
+      "loss": 0.289,
+      "step": 6619
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4558538174248061,
+      "learning_rate": 9.556374896478519e-06,
+      "loss": 0.3242,
+      "step": 6620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9926246574815072,
+      "learning_rate": 9.556212629994044e-06,
+      "loss": 0.4597,
+      "step": 6621
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7623223213920953,
+      "learning_rate": 9.556050335216724e-06,
+      "loss": 0.3426,
+      "step": 6622
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2280262402212605,
+      "learning_rate": 9.555888012147567e-06,
+      "loss": 0.5089,
+      "step": 6623
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.101334357375272,
+      "learning_rate": 9.555725660787582e-06,
+      "loss": 0.3662,
+      "step": 6624
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.219528679475299,
+      "learning_rate": 9.555563281137776e-06,
+      "loss": 0.7177,
+      "step": 6625
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9955959020270052,
+      "learning_rate": 9.555400873199157e-06,
+      "loss": 0.1824,
+      "step": 6626
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.056965927705154,
+      "learning_rate": 9.555238436972736e-06,
+      "loss": 0.189,
+      "step": 6627
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9402989198022376,
+      "learning_rate": 9.555075972459517e-06,
+      "loss": 0.4361,
+      "step": 6628
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6289123716467313,
+      "learning_rate": 9.554913479660512e-06,
+      "loss": 0.2467,
+      "step": 6629
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3770720816016468,
+      "learning_rate": 9.554750958576731e-06,
+      "loss": 0.2331,
+      "step": 6630
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4026032077626904,
+      "learning_rate": 9.554588409209183e-06,
+      "loss": 0.2107,
+      "step": 6631
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3100421615150124,
+      "learning_rate": 9.554425831558873e-06,
+      "loss": 0.2108,
+      "step": 6632
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5863374736578992,
+      "learning_rate": 9.554263225626815e-06,
+      "loss": 0.4088,
+      "step": 6633
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5674584676481769,
+      "learning_rate": 9.554100591414019e-06,
+      "loss": 0.1892,
+      "step": 6634
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7184393927207688,
+      "learning_rate": 9.55393792892149e-06,
+      "loss": 0.379,
+      "step": 6635
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3407607541119788,
+      "learning_rate": 9.553775238150245e-06,
+      "loss": 0.3222,
+      "step": 6636
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2790467828673924,
+      "learning_rate": 9.55361251910129e-06,
+      "loss": 0.2445,
+      "step": 6637
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7602158710322706,
+      "learning_rate": 9.553449771775637e-06,
+      "loss": 0.536,
+      "step": 6638
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5021230913979264,
+      "learning_rate": 9.553286996174294e-06,
+      "loss": 0.2575,
+      "step": 6639
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0791116567286292,
+      "learning_rate": 9.553124192298274e-06,
+      "loss": 0.1615,
+      "step": 6640
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8744162226898315,
+      "learning_rate": 9.552961360148587e-06,
+      "loss": 0.1748,
+      "step": 6641
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.057908358473792,
+      "learning_rate": 9.552798499726246e-06,
+      "loss": 0.5037,
+      "step": 6642
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9386161926869354,
+      "learning_rate": 9.55263561103226e-06,
+      "loss": 0.1707,
+      "step": 6643
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9171956868140685,
+      "learning_rate": 9.552472694067642e-06,
+      "loss": 0.6789,
+      "step": 6644
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5460960734168239,
+      "learning_rate": 9.552309748833404e-06,
+      "loss": 0.2961,
+      "step": 6645
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1889077300153874,
+      "learning_rate": 9.552146775330557e-06,
+      "loss": 0.1331,
+      "step": 6646
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.515260468602517,
+      "learning_rate": 9.551983773560111e-06,
+      "loss": 0.7288,
+      "step": 6647
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7751429419652027,
+      "learning_rate": 9.551820743523084e-06,
+      "loss": 0.2229,
+      "step": 6648
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2212212828497622,
+      "learning_rate": 9.551657685220482e-06,
+      "loss": 0.2966,
+      "step": 6649
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6322362512636672,
+      "learning_rate": 9.55149459865332e-06,
+      "loss": 0.225,
+      "step": 6650
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8680916875368616,
+      "learning_rate": 9.551331483822612e-06,
+      "loss": 0.4961,
+      "step": 6651
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9610739732647346,
+      "learning_rate": 9.551168340729369e-06,
+      "loss": 0.4002,
+      "step": 6652
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9720661527112437,
+      "learning_rate": 9.551005169374606e-06,
+      "loss": 0.4855,
+      "step": 6653
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.829271726853333,
+      "learning_rate": 9.550841969759335e-06,
+      "loss": 0.3691,
+      "step": 6654
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7357823487107573,
+      "learning_rate": 9.550678741884569e-06,
+      "loss": 0.3728,
+      "step": 6655
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4641985883943829,
+      "learning_rate": 9.550515485751324e-06,
+      "loss": 0.2086,
+      "step": 6656
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5606634301784488,
+      "learning_rate": 9.55035220136061e-06,
+      "loss": 0.2908,
+      "step": 6657
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1216621868502619,
+      "learning_rate": 9.550188888713442e-06,
+      "loss": 0.1616,
+      "step": 6658
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4001329403050067,
+      "learning_rate": 9.550025547810837e-06,
+      "loss": 0.1533,
+      "step": 6659
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5778169567372358,
+      "learning_rate": 9.549862178653807e-06,
+      "loss": 0.3641,
+      "step": 6660
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.284903181360025,
+      "learning_rate": 9.549698781243367e-06,
+      "loss": 0.5342,
+      "step": 6661
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1308565847796417,
+      "learning_rate": 9.549535355580531e-06,
+      "loss": 0.595,
+      "step": 6662
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6020812799352113,
+      "learning_rate": 9.549371901666316e-06,
+      "loss": 0.2917,
+      "step": 6663
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5206500693080167,
+      "learning_rate": 9.549208419501732e-06,
+      "loss": 0.2481,
+      "step": 6664
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.238286066632639,
+      "learning_rate": 9.549044909087799e-06,
+      "loss": 0.2222,
+      "step": 6665
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2918126622363755,
+      "learning_rate": 9.548881370425532e-06,
+      "loss": 0.2227,
+      "step": 6666
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.973589481895482,
+      "learning_rate": 9.548717803515944e-06,
+      "loss": 0.5174,
+      "step": 6667
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.570018921164069,
+      "learning_rate": 9.548554208360052e-06,
+      "loss": 0.2732,
+      "step": 6668
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8954993065511028,
+      "learning_rate": 9.548390584958874e-06,
+      "loss": 0.5452,
+      "step": 6669
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.43765577736884,
+      "learning_rate": 9.548226933313423e-06,
+      "loss": 0.1408,
+      "step": 6670
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8619771678398094,
+      "learning_rate": 9.548063253424716e-06,
+      "loss": 0.2238,
+      "step": 6671
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8830645697184532,
+      "learning_rate": 9.547899545293768e-06,
+      "loss": 0.6348,
+      "step": 6672
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5582918912133512,
+      "learning_rate": 9.547735808921598e-06,
+      "loss": 0.3797,
+      "step": 6673
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4691464061668955,
+      "learning_rate": 9.547572044309223e-06,
+      "loss": 0.1668,
+      "step": 6674
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0648905424051094,
+      "learning_rate": 9.54740825145766e-06,
+      "loss": 0.1725,
+      "step": 6675
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5518434091610367,
+      "learning_rate": 9.547244430367924e-06,
+      "loss": 0.2537,
+      "step": 6676
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8865978368078686,
+      "learning_rate": 9.547080581041033e-06,
+      "loss": 0.228,
+      "step": 6677
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.808204951010355,
+      "learning_rate": 9.546916703478004e-06,
+      "loss": 0.3231,
+      "step": 6678
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.723994749285807,
+      "learning_rate": 9.546752797679856e-06,
+      "loss": 0.3438,
+      "step": 6679
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7242938462632158,
+      "learning_rate": 9.546588863647609e-06,
+      "loss": 0.2516,
+      "step": 6680
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.510298093438576,
+      "learning_rate": 9.546424901382274e-06,
+      "loss": 0.156,
+      "step": 6681
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6764803986298011,
+      "learning_rate": 9.546260910884875e-06,
+      "loss": 0.2997,
+      "step": 6682
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.054205711069767,
+      "learning_rate": 9.54609689215643e-06,
+      "loss": 0.5922,
+      "step": 6683
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7632656389964292,
+      "learning_rate": 9.545932845197954e-06,
+      "loss": 0.4032,
+      "step": 6684
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4615628197041393,
+      "learning_rate": 9.545768770010471e-06,
+      "loss": 0.1967,
+      "step": 6685
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9077996849485697,
+      "learning_rate": 9.545604666594994e-06,
+      "loss": 0.3438,
+      "step": 6686
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7799617266357597,
+      "learning_rate": 9.545440534952546e-06,
+      "loss": 0.4339,
+      "step": 6687
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.618846742397671,
+      "learning_rate": 9.545276375084145e-06,
+      "loss": 0.3602,
+      "step": 6688
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4471903123706764,
+      "learning_rate": 9.545112186990811e-06,
+      "loss": 0.2983,
+      "step": 6689
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.961095868132992,
+      "learning_rate": 9.544947970673561e-06,
+      "loss": 0.3527,
+      "step": 6690
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4526451200354484,
+      "learning_rate": 9.544783726133418e-06,
+      "loss": 0.32,
+      "step": 6691
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6821490326716853,
+      "learning_rate": 9.5446194533714e-06,
+      "loss": 0.4484,
+      "step": 6692
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6569453165450478,
+      "learning_rate": 9.544455152388529e-06,
+      "loss": 0.3243,
+      "step": 6693
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5656150665272641,
+      "learning_rate": 9.544290823185823e-06,
+      "loss": 0.197,
+      "step": 6694
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6632291292021837,
+      "learning_rate": 9.544126465764304e-06,
+      "loss": 0.3768,
+      "step": 6695
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2931502681561653,
+      "learning_rate": 9.543962080124992e-06,
+      "loss": 0.1625,
+      "step": 6696
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3099054736875355,
+      "learning_rate": 9.543797666268908e-06,
+      "loss": 0.1667,
+      "step": 6697
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9762229090561358,
+      "learning_rate": 9.543633224197071e-06,
+      "loss": 0.3293,
+      "step": 6698
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.726253296277865,
+      "learning_rate": 9.543468753910506e-06,
+      "loss": 0.3204,
+      "step": 6699
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6171358118688786,
+      "learning_rate": 9.543304255410232e-06,
+      "loss": 0.3322,
+      "step": 6700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9520424913793994,
+      "learning_rate": 9.543139728697271e-06,
+      "loss": 0.2255,
+      "step": 6701
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3913139929539142,
+      "learning_rate": 9.542975173772644e-06,
+      "loss": 0.4029,
+      "step": 6702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.855782909901644,
+      "learning_rate": 9.542810590637372e-06,
+      "loss": 0.4323,
+      "step": 6703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.916861498819439,
+      "learning_rate": 9.54264597929248e-06,
+      "loss": 0.3068,
+      "step": 6704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3520708533795054,
+      "learning_rate": 9.542481339738988e-06,
+      "loss": 0.1835,
+      "step": 6705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.64247562619346,
+      "learning_rate": 9.542316671977917e-06,
+      "loss": 0.3206,
+      "step": 6706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.028937046285875,
+      "learning_rate": 9.542151976010294e-06,
+      "loss": 0.4639,
+      "step": 6707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6153418148891363,
+      "learning_rate": 9.54198725183714e-06,
+      "loss": 0.2116,
+      "step": 6708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6362182015665234,
+      "learning_rate": 9.541822499459475e-06,
+      "loss": 0.3918,
+      "step": 6709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.187573336325625,
+      "learning_rate": 9.541657718878324e-06,
+      "loss": 0.6164,
+      "step": 6710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7026360361083104,
+      "learning_rate": 9.54149291009471e-06,
+      "loss": 0.4523,
+      "step": 6711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1290213230075596,
+      "learning_rate": 9.541328073109656e-06,
+      "loss": 0.3688,
+      "step": 6712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0766725950139886,
+      "learning_rate": 9.54116320792419e-06,
+      "loss": 0.4577,
+      "step": 6713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9983666966670774,
+      "learning_rate": 9.540998314539327e-06,
+      "loss": 0.3606,
+      "step": 6714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5749990834127516,
+      "learning_rate": 9.5408333929561e-06,
+      "loss": 0.3346,
+      "step": 6715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.747938090940659,
+      "learning_rate": 9.540668443175528e-06,
+      "loss": 0.4582,
+      "step": 6716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.623685167308964,
+      "learning_rate": 9.540503465198636e-06,
+      "loss": 0.2137,
+      "step": 6717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7057739005509499,
+      "learning_rate": 9.54033845902645e-06,
+      "loss": 0.1427,
+      "step": 6718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7132257070726986,
+      "learning_rate": 9.540173424659992e-06,
+      "loss": 0.3704,
+      "step": 6719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0826969549663166,
+      "learning_rate": 9.540008362100288e-06,
+      "loss": 0.3845,
+      "step": 6720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0934888370811695,
+      "learning_rate": 9.539843271348366e-06,
+      "loss": 0.6121,
+      "step": 6721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4558550207166598,
+      "learning_rate": 9.539678152405247e-06,
+      "loss": 0.2674,
+      "step": 6722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8365906547814599,
+      "learning_rate": 9.539513005271956e-06,
+      "loss": 0.49,
+      "step": 6723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.069660796751045,
+      "learning_rate": 9.539347829949523e-06,
+      "loss": 0.4583,
+      "step": 6724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7298491102667415,
+      "learning_rate": 9.53918262643897e-06,
+      "loss": 0.3857,
+      "step": 6725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9749990280668064,
+      "learning_rate": 9.539017394741326e-06,
+      "loss": 0.5664,
+      "step": 6726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3279593097891182,
+      "learning_rate": 9.538852134857613e-06,
+      "loss": 0.3412,
+      "step": 6727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8656052385494457,
+      "learning_rate": 9.538686846788858e-06,
+      "loss": 0.2782,
+      "step": 6728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6031735621081395,
+      "learning_rate": 9.53852153053609e-06,
+      "loss": 0.3617,
+      "step": 6729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8326363530112741,
+      "learning_rate": 9.538356186100335e-06,
+      "loss": 0.2602,
+      "step": 6730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5375996380984296,
+      "learning_rate": 9.538190813482618e-06,
+      "loss": 0.2885,
+      "step": 6731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9270713152811398,
+      "learning_rate": 9.538025412683968e-06,
+      "loss": 0.3908,
+      "step": 6732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8492008873963584,
+      "learning_rate": 9.53785998370541e-06,
+      "loss": 0.469,
+      "step": 6733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6184398713051231,
+      "learning_rate": 9.537694526547971e-06,
+      "loss": 0.2348,
+      "step": 6734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5237172921326172,
+      "learning_rate": 9.53752904121268e-06,
+      "loss": 0.2542,
+      "step": 6735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.523677247084946,
+      "learning_rate": 9.537363527700567e-06,
+      "loss": 0.2682,
+      "step": 6736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0499231351522256,
+      "learning_rate": 9.537197986012656e-06,
+      "loss": 0.2021,
+      "step": 6737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7943190142322507,
+      "learning_rate": 9.537032416149973e-06,
+      "loss": 0.5394,
+      "step": 6738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2896733469494988,
+      "learning_rate": 9.536866818113552e-06,
+      "loss": 0.2216,
+      "step": 6739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8242779462795427,
+      "learning_rate": 9.536701191904419e-06,
+      "loss": 0.2101,
+      "step": 6740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.138175809279386,
+      "learning_rate": 9.5365355375236e-06,
+      "loss": 0.2947,
+      "step": 6741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8550226699274144,
+      "learning_rate": 9.536369854972127e-06,
+      "loss": 0.335,
+      "step": 6742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5461707763895414,
+      "learning_rate": 9.536204144251027e-06,
+      "loss": 0.2428,
+      "step": 6743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0106492875635955,
+      "learning_rate": 9.536038405361329e-06,
+      "loss": 0.6297,
+      "step": 6744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5186897831034674,
+      "learning_rate": 9.535872638304064e-06,
+      "loss": 0.1698,
+      "step": 6745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9854877273242235,
+      "learning_rate": 9.535706843080258e-06,
+      "loss": 0.243,
+      "step": 6746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9980992956360624,
+      "learning_rate": 9.535541019690945e-06,
+      "loss": 0.276,
+      "step": 6747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8171987916984274,
+      "learning_rate": 9.535375168137152e-06,
+      "loss": 0.2103,
+      "step": 6748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9433906859362737,
+      "learning_rate": 9.535209288419909e-06,
+      "loss": 0.4194,
+      "step": 6749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6174455634903413,
+      "learning_rate": 9.535043380540245e-06,
+      "loss": 0.3299,
+      "step": 6750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1533800153274945,
+      "learning_rate": 9.534877444499194e-06,
+      "loss": 0.2314,
+      "step": 6751
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.056731786274344,
+      "learning_rate": 9.53471148029778e-06,
+      "loss": 0.3928,
+      "step": 6752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.259641190567251,
+      "learning_rate": 9.53454548793704e-06,
+      "loss": 0.4079,
+      "step": 6753
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.71979053384831,
+      "learning_rate": 9.534379467418002e-06,
+      "loss": 0.4227,
+      "step": 6754
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6924100632628016,
+      "learning_rate": 9.534213418741698e-06,
+      "loss": 0.3427,
+      "step": 6755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5769100422392044,
+      "learning_rate": 9.534047341909158e-06,
+      "loss": 0.2364,
+      "step": 6756
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6253337586128744,
+      "learning_rate": 9.533881236921414e-06,
+      "loss": 0.1638,
+      "step": 6757
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9029615405624491,
+      "learning_rate": 9.533715103779497e-06,
+      "loss": 0.3982,
+      "step": 6758
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0221512473117795,
+      "learning_rate": 9.533548942484439e-06,
+      "loss": 0.5419,
+      "step": 6759
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6900863431001496,
+      "learning_rate": 9.53338275303727e-06,
+      "loss": 0.3052,
+      "step": 6760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0790319704079083,
+      "learning_rate": 9.533216535439026e-06,
+      "loss": 0.6369,
+      "step": 6761
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8297755093993862,
+      "learning_rate": 9.533050289690735e-06,
+      "loss": 0.518,
+      "step": 6762
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6264991539044997,
+      "learning_rate": 9.532884015793432e-06,
+      "loss": 0.2471,
+      "step": 6763
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6189013949776956,
+      "learning_rate": 9.532717713748148e-06,
+      "loss": 0.4072,
+      "step": 6764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7056223578658,
+      "learning_rate": 9.532551383555917e-06,
+      "loss": 0.2753,
+      "step": 6765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7613299209270852,
+      "learning_rate": 9.53238502521777e-06,
+      "loss": 0.3757,
+      "step": 6766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.905244417424139,
+      "learning_rate": 9.532218638734743e-06,
+      "loss": 0.221,
+      "step": 6767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2716283453020165,
+      "learning_rate": 9.532052224107866e-06,
+      "loss": 0.2388,
+      "step": 6768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6541119216520692,
+      "learning_rate": 9.531885781338173e-06,
+      "loss": 0.5519,
+      "step": 6769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6574949879987668,
+      "learning_rate": 9.5317193104267e-06,
+      "loss": 0.2951,
+      "step": 6770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9172242384159863,
+      "learning_rate": 9.531552811374479e-06,
+      "loss": 0.3701,
+      "step": 6771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0650317518919707,
+      "learning_rate": 9.531386284182544e-06,
+      "loss": 0.5561,
+      "step": 6772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7707850833004106,
+      "learning_rate": 9.531219728851929e-06,
+      "loss": 0.2676,
+      "step": 6773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3525196721308628,
+      "learning_rate": 9.531053145383667e-06,
+      "loss": 0.1904,
+      "step": 6774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9285450074490174,
+      "learning_rate": 9.530886533778797e-06,
+      "loss": 0.3685,
+      "step": 6775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0765074744924048,
+      "learning_rate": 9.530719894038346e-06,
+      "loss": 0.1528,
+      "step": 6776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0950715706422207,
+      "learning_rate": 9.530553226163356e-06,
+      "loss": 0.1613,
+      "step": 6777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9348814712614337,
+      "learning_rate": 9.53038653015486e-06,
+      "loss": 0.3265,
+      "step": 6778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4492173259142023,
+      "learning_rate": 9.530219806013892e-06,
+      "loss": 0.1851,
+      "step": 6779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.397220122446438,
+      "learning_rate": 9.530053053741486e-06,
+      "loss": 0.6404,
+      "step": 6780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4858683828771508,
+      "learning_rate": 9.52988627333868e-06,
+      "loss": 0.3368,
+      "step": 6781
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1605063184077666,
+      "learning_rate": 9.529719464806509e-06,
+      "loss": 0.7106,
+      "step": 6782
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1446264138152324,
+      "learning_rate": 9.529552628146008e-06,
+      "loss": 0.2158,
+      "step": 6783
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8949052618458773,
+      "learning_rate": 9.529385763358214e-06,
+      "loss": 0.2851,
+      "step": 6784
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.657408232838419,
+      "learning_rate": 9.529218870444163e-06,
+      "loss": 0.2486,
+      "step": 6785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6303589584205131,
+      "learning_rate": 9.529051949404891e-06,
+      "loss": 0.3327,
+      "step": 6786
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6406120345193718,
+      "learning_rate": 9.528885000241435e-06,
+      "loss": 0.4007,
+      "step": 6787
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.279880978556634,
+      "learning_rate": 9.52871802295483e-06,
+      "loss": 0.1919,
+      "step": 6788
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6914393573718791,
+      "learning_rate": 9.528551017546118e-06,
+      "loss": 0.299,
+      "step": 6789
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1971915948900445,
+      "learning_rate": 9.528383984016329e-06,
+      "loss": 0.5337,
+      "step": 6790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3997669697918687,
+      "learning_rate": 9.528216922366504e-06,
+      "loss": 0.6261,
+      "step": 6791
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3483546292524085,
+      "learning_rate": 9.528049832597681e-06,
+      "loss": 0.4174,
+      "step": 6792
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3732346825398203,
+      "learning_rate": 9.527882714710895e-06,
+      "loss": 0.3649,
+      "step": 6793
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7998055671776678,
+      "learning_rate": 9.527715568707187e-06,
+      "loss": 0.3558,
+      "step": 6794
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3544623076671163,
+      "learning_rate": 9.527548394587593e-06,
+      "loss": 0.2261,
+      "step": 6795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6268394712154692,
+      "learning_rate": 9.527381192353152e-06,
+      "loss": 0.2161,
+      "step": 6796
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6201615571792212,
+      "learning_rate": 9.5272139620049e-06,
+      "loss": 0.3565,
+      "step": 6797
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9728279832171653,
+      "learning_rate": 9.527046703543877e-06,
+      "loss": 0.4366,
+      "step": 6798
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5913090221576467,
+      "learning_rate": 9.526879416971123e-06,
+      "loss": 0.4647,
+      "step": 6799
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7547124186620662,
+      "learning_rate": 9.526712102287674e-06,
+      "loss": 0.3824,
+      "step": 6800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0494485939262845,
+      "learning_rate": 9.526544759494572e-06,
+      "loss": 0.4248,
+      "step": 6801
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4074395393832142,
+      "learning_rate": 9.526377388592854e-06,
+      "loss": 0.3,
+      "step": 6802
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8177314661643538,
+      "learning_rate": 9.526209989583561e-06,
+      "loss": 0.3532,
+      "step": 6803
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4672219591288664,
+      "learning_rate": 9.52604256246773e-06,
+      "loss": 0.1861,
+      "step": 6804
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8346572696270065,
+      "learning_rate": 9.525875107246402e-06,
+      "loss": 0.2047,
+      "step": 6805
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.73115094259544,
+      "learning_rate": 9.525707623920618e-06,
+      "loss": 0.4022,
+      "step": 6806
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6722580340604554,
+      "learning_rate": 9.525540112491418e-06,
+      "loss": 0.2453,
+      "step": 6807
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.785188030045556,
+      "learning_rate": 9.525372572959842e-06,
+      "loss": 0.6448,
+      "step": 6808
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0426399457389335,
+      "learning_rate": 9.525205005326928e-06,
+      "loss": 0.1966,
+      "step": 6809
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0422665791571144,
+      "learning_rate": 9.525037409593718e-06,
+      "loss": 0.5095,
+      "step": 6810
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.970776120848103,
+      "learning_rate": 9.524869785761253e-06,
+      "loss": 0.3395,
+      "step": 6811
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.255524415631892,
+      "learning_rate": 9.524702133830573e-06,
+      "loss": 0.2573,
+      "step": 6812
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6335450130947795,
+      "learning_rate": 9.52453445380272e-06,
+      "loss": 0.3815,
+      "step": 6813
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9441151668452223,
+      "learning_rate": 9.524366745678737e-06,
+      "loss": 0.7519,
+      "step": 6814
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2919743085764437,
+      "learning_rate": 9.524199009459662e-06,
+      "loss": 0.7967,
+      "step": 6815
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5341495041090787,
+      "learning_rate": 9.52403124514654e-06,
+      "loss": 0.215,
+      "step": 6816
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2793071913301843,
+      "learning_rate": 9.523863452740409e-06,
+      "loss": 0.4702,
+      "step": 6817
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5806925443107487,
+      "learning_rate": 9.523695632242313e-06,
+      "loss": 0.4669,
+      "step": 6818
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6340234933721685,
+      "learning_rate": 9.523527783653295e-06,
+      "loss": 0.3568,
+      "step": 6819
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6166881075851403,
+      "learning_rate": 9.523359906974396e-06,
+      "loss": 0.3751,
+      "step": 6820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.377800102007228,
+      "learning_rate": 9.523192002206658e-06,
+      "loss": 0.2284,
+      "step": 6821
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8105686272854546,
+      "learning_rate": 9.523024069351125e-06,
+      "loss": 0.5073,
+      "step": 6822
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0878800658627203,
+      "learning_rate": 9.52285610840884e-06,
+      "loss": 0.5786,
+      "step": 6823
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2264017429516854,
+      "learning_rate": 9.522688119380842e-06,
+      "loss": 0.176,
+      "step": 6824
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4940601152657533,
+      "learning_rate": 9.522520102268182e-06,
+      "loss": 0.2217,
+      "step": 6825
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9924690286460656,
+      "learning_rate": 9.522352057071895e-06,
+      "loss": 0.1412,
+      "step": 6826
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1253288305040998,
+      "learning_rate": 9.52218398379303e-06,
+      "loss": 0.1537,
+      "step": 6827
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1162539320637133,
+      "learning_rate": 9.52201588243263e-06,
+      "loss": 0.542,
+      "step": 6828
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2743012915515963,
+      "learning_rate": 9.521847752991735e-06,
+      "loss": 0.2381,
+      "step": 6829
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3553138872556865,
+      "learning_rate": 9.521679595471394e-06,
+      "loss": 0.2315,
+      "step": 6830
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0994468945466256,
+      "learning_rate": 9.521511409872649e-06,
+      "loss": 0.5879,
+      "step": 6831
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.067947114269954,
+      "learning_rate": 9.521343196196541e-06,
+      "loss": 0.3971,
+      "step": 6832
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6975998302516766,
+      "learning_rate": 9.521174954444121e-06,
+      "loss": 0.3921,
+      "step": 6833
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5125369930472379,
+      "learning_rate": 9.52100668461643e-06,
+      "loss": 0.3053,
+      "step": 6834
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.368191886817051,
+      "learning_rate": 9.520838386714514e-06,
+      "loss": 0.4265,
+      "step": 6835
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8194424305747765,
+      "learning_rate": 9.52067006073942e-06,
+      "loss": 0.2979,
+      "step": 6836
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7097502730623562,
+      "learning_rate": 9.520501706692187e-06,
+      "loss": 0.2127,
+      "step": 6837
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5521272047987544,
+      "learning_rate": 9.520333324573865e-06,
+      "loss": 0.1887,
+      "step": 6838
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0791537759844378,
+      "learning_rate": 9.520164914385501e-06,
+      "loss": 0.1938,
+      "step": 6839
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.831132526014805,
+      "learning_rate": 9.519996476128137e-06,
+      "loss": 0.3082,
+      "step": 6840
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8249550789373414,
+      "learning_rate": 9.519828009802822e-06,
+      "loss": 0.4018,
+      "step": 6841
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4682475914892352,
+      "learning_rate": 9.5196595154106e-06,
+      "loss": 0.3968,
+      "step": 6842
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6362025919662322,
+      "learning_rate": 9.519490992952518e-06,
+      "loss": 0.2467,
+      "step": 6843
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.128660429314984,
+      "learning_rate": 9.519322442429624e-06,
+      "loss": 0.5395,
+      "step": 6844
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.307277430037643,
+      "learning_rate": 9.519153863842963e-06,
+      "loss": 0.5497,
+      "step": 6845
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.651671701301946,
+      "learning_rate": 9.518985257193581e-06,
+      "loss": 0.2905,
+      "step": 6846
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8482254595152208,
+      "learning_rate": 9.518816622482527e-06,
+      "loss": 0.2215,
+      "step": 6847
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0306272358997983,
+      "learning_rate": 9.518647959710848e-06,
+      "loss": 0.7481,
+      "step": 6848
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.073706335507117,
+      "learning_rate": 9.518479268879587e-06,
+      "loss": 0.2755,
+      "step": 6849
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0620256366036993,
+      "learning_rate": 9.5183105499898e-06,
+      "loss": 0.1722,
+      "step": 6850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6275072470736551,
+      "learning_rate": 9.518141803042528e-06,
+      "loss": 0.2918,
+      "step": 6851
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0546291219736443,
+      "learning_rate": 9.51797302803882e-06,
+      "loss": 0.4376,
+      "step": 6852
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.377235013472668,
+      "learning_rate": 9.517804224979725e-06,
+      "loss": 0.2192,
+      "step": 6853
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.981182609278413,
+      "learning_rate": 9.517635393866291e-06,
+      "loss": 0.3649,
+      "step": 6854
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8870037665217085,
+      "learning_rate": 9.517466534699567e-06,
+      "loss": 0.4479,
+      "step": 6855
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.933798230080796,
+      "learning_rate": 9.5172976474806e-06,
+      "loss": 0.3569,
+      "step": 6856
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.070301826017682,
+      "learning_rate": 9.517128732210444e-06,
+      "loss": 0.2971,
+      "step": 6857
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3997293637516233,
+      "learning_rate": 9.51695978889014e-06,
+      "loss": 0.1994,
+      "step": 6858
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.075838824856647,
+      "learning_rate": 9.51679081752074e-06,
+      "loss": 0.5684,
+      "step": 6859
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4639325881202663,
+      "learning_rate": 9.516621818103295e-06,
+      "loss": 0.3437,
+      "step": 6860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.775697368985256,
+      "learning_rate": 9.516452790638852e-06,
+      "loss": 0.2279,
+      "step": 6861
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0713579741025647,
+      "learning_rate": 9.516283735128464e-06,
+      "loss": 0.4783,
+      "step": 6862
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3781747147292875,
+      "learning_rate": 9.516114651573178e-06,
+      "loss": 0.2469,
+      "step": 6863
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4815346069180428,
+      "learning_rate": 9.515945539974047e-06,
+      "loss": 0.1857,
+      "step": 6864
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.531115022051012,
+      "learning_rate": 9.515776400332117e-06,
+      "loss": 0.2811,
+      "step": 6865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6523694426430624,
+      "learning_rate": 9.515607232648441e-06,
+      "loss": 0.5533,
+      "step": 6866
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.56742609504999,
+      "learning_rate": 9.515438036924068e-06,
+      "loss": 0.1813,
+      "step": 6867
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3027181163728314,
+      "learning_rate": 9.515268813160048e-06,
+      "loss": 0.4342,
+      "step": 6868
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8694969563062795,
+      "learning_rate": 9.515099561357435e-06,
+      "loss": 0.5763,
+      "step": 6869
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.598755064561829,
+      "learning_rate": 9.514930281517278e-06,
+      "loss": 0.3784,
+      "step": 6870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8729731253861417,
+      "learning_rate": 9.51476097364063e-06,
+      "loss": 0.4905,
+      "step": 6871
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0454319606483016,
+      "learning_rate": 9.514591637728539e-06,
+      "loss": 0.5597,
+      "step": 6872
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6453726330932619,
+      "learning_rate": 9.514422273782058e-06,
+      "loss": 0.1971,
+      "step": 6873
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7129465070110415,
+      "learning_rate": 9.51425288180224e-06,
+      "loss": 0.1904,
+      "step": 6874
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0547465761924073,
+      "learning_rate": 9.514083461790136e-06,
+      "loss": 0.4374,
+      "step": 6875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8741372507748661,
+      "learning_rate": 9.513914013746798e-06,
+      "loss": 0.5359,
+      "step": 6876
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6132447005373594,
+      "learning_rate": 9.513744537673278e-06,
+      "loss": 0.2191,
+      "step": 6877
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.058593666757079,
+      "learning_rate": 9.513575033570628e-06,
+      "loss": 0.638,
+      "step": 6878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9693887522080185,
+      "learning_rate": 9.513405501439901e-06,
+      "loss": 0.5933,
+      "step": 6879
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.794429474910239,
+      "learning_rate": 9.51323594128215e-06,
+      "loss": 0.5218,
+      "step": 6880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7312559528799345,
+      "learning_rate": 9.51306635309843e-06,
+      "loss": 0.3147,
+      "step": 6881
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3678907710953374,
+      "learning_rate": 9.51289673688979e-06,
+      "loss": 0.1691,
+      "step": 6882
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7076620424800493,
+      "learning_rate": 9.512727092657285e-06,
+      "loss": 0.3683,
+      "step": 6883
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8420285254194666,
+      "learning_rate": 9.512557420401969e-06,
+      "loss": 0.4003,
+      "step": 6884
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.690038987378721,
+      "learning_rate": 9.512387720124896e-06,
+      "loss": 0.2317,
+      "step": 6885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9468794567002037,
+      "learning_rate": 9.512217991827117e-06,
+      "loss": 0.3979,
+      "step": 6886
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7891412784273224,
+      "learning_rate": 9.51204823550969e-06,
+      "loss": 0.3929,
+      "step": 6887
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9138233346797773,
+      "learning_rate": 9.511878451173666e-06,
+      "loss": 0.4085,
+      "step": 6888
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9039511890140288,
+      "learning_rate": 9.5117086388201e-06,
+      "loss": 0.4001,
+      "step": 6889
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7366213985310492,
+      "learning_rate": 9.511538798450049e-06,
+      "loss": 0.2186,
+      "step": 6890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2607087796466148,
+      "learning_rate": 9.511368930064566e-06,
+      "loss": 0.1818,
+      "step": 6891
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4509042783774317,
+      "learning_rate": 9.511199033664703e-06,
+      "loss": 0.5052,
+      "step": 6892
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.845844792171873,
+      "learning_rate": 9.51102910925152e-06,
+      "loss": 0.38,
+      "step": 6893
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7874678205381223,
+      "learning_rate": 9.510859156826067e-06,
+      "loss": 0.2691,
+      "step": 6894
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9715266704462906,
+      "learning_rate": 9.510689176389402e-06,
+      "loss": 0.5069,
+      "step": 6895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5172533328410533,
+      "learning_rate": 9.510519167942583e-06,
+      "loss": 0.2075,
+      "step": 6896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3403019066154278,
+      "learning_rate": 9.510349131486663e-06,
+      "loss": 0.2039,
+      "step": 6897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5569954088510263,
+      "learning_rate": 9.510179067022695e-06,
+      "loss": 0.2052,
+      "step": 6898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2070747236173083,
+      "learning_rate": 9.510008974551742e-06,
+      "loss": 0.3662,
+      "step": 6899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.821594908042014,
+      "learning_rate": 9.509838854074855e-06,
+      "loss": 0.3417,
+      "step": 6900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.789114751427574,
+      "learning_rate": 9.50966870559309e-06,
+      "loss": 0.4452,
+      "step": 6901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0917314158352291,
+      "learning_rate": 9.509498529107508e-06,
+      "loss": 0.2065,
+      "step": 6902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6097563551015897,
+      "learning_rate": 9.509328324619162e-06,
+      "loss": 0.3378,
+      "step": 6903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2361709357699697,
+      "learning_rate": 9.509158092129111e-06,
+      "loss": 0.1983,
+      "step": 6904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.709207358935652,
+      "learning_rate": 9.508987831638409e-06,
+      "loss": 0.3149,
+      "step": 6905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9055956983956535,
+      "learning_rate": 9.508817543148118e-06,
+      "loss": 0.3534,
+      "step": 6906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9631892873809276,
+      "learning_rate": 9.50864722665929e-06,
+      "loss": 0.3569,
+      "step": 6907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.863515123448724,
+      "learning_rate": 9.508476882172988e-06,
+      "loss": 0.5866,
+      "step": 6908
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.837814260552323,
+      "learning_rate": 9.508306509690268e-06,
+      "loss": 0.2959,
+      "step": 6909
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5985465821701388,
+      "learning_rate": 9.508136109212184e-06,
+      "loss": 0.2501,
+      "step": 6910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4779279516025137,
+      "learning_rate": 9.5079656807398e-06,
+      "loss": 0.3356,
+      "step": 6911
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8242650158914868,
+      "learning_rate": 9.507795224274172e-06,
+      "loss": 0.2382,
+      "step": 6912
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.18755987970957,
+      "learning_rate": 9.507624739816356e-06,
+      "loss": 0.4901,
+      "step": 6913
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7716770172519556,
+      "learning_rate": 9.507454227367413e-06,
+      "loss": 0.4261,
+      "step": 6914
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1702126154561796,
+      "learning_rate": 9.507283686928404e-06,
+      "loss": 0.2275,
+      "step": 6915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4461811399733637,
+      "learning_rate": 9.507113118500385e-06,
+      "loss": 0.8447,
+      "step": 6916
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5714633434840695,
+      "learning_rate": 9.506942522084416e-06,
+      "loss": 0.3086,
+      "step": 6917
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4599013504575504,
+      "learning_rate": 9.506771897681554e-06,
+      "loss": 0.3091,
+      "step": 6918
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5961043158175403,
+      "learning_rate": 9.506601245292863e-06,
+      "loss": 0.2529,
+      "step": 6919
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6130455574610034,
+      "learning_rate": 9.5064305649194e-06,
+      "loss": 0.2053,
+      "step": 6920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8092196475259934,
+      "learning_rate": 9.506259856562226e-06,
+      "loss": 0.5423,
+      "step": 6921
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7450321392389678,
+      "learning_rate": 9.506089120222399e-06,
+      "loss": 0.3969,
+      "step": 6922
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1947721572090768,
+      "learning_rate": 9.505918355900982e-06,
+      "loss": 0.1931,
+      "step": 6923
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4673629714719132,
+      "learning_rate": 9.505747563599033e-06,
+      "loss": 0.2376,
+      "step": 6924
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4933266810002201,
+      "learning_rate": 9.505576743317615e-06,
+      "loss": 0.3424,
+      "step": 6925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.075493726479546,
+      "learning_rate": 9.505405895057787e-06,
+      "loss": 0.1902,
+      "step": 6926
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9055924219441411,
+      "learning_rate": 9.505235018820609e-06,
+      "loss": 0.52,
+      "step": 6927
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9012396974930277,
+      "learning_rate": 9.505064114607144e-06,
+      "loss": 0.4264,
+      "step": 6928
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1138699514296888,
+      "learning_rate": 9.504893182418452e-06,
+      "loss": 0.1207,
+      "step": 6929
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6849355152494403,
+      "learning_rate": 9.504722222255595e-06,
+      "loss": 0.5169,
+      "step": 6930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7338451100123684,
+      "learning_rate": 9.504551234119637e-06,
+      "loss": 0.4854,
+      "step": 6931
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0314211516533907,
+      "learning_rate": 9.504380218011636e-06,
+      "loss": 0.6792,
+      "step": 6932
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.317434487970324,
+      "learning_rate": 9.504209173932655e-06,
+      "loss": 0.2741,
+      "step": 6933
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9284893130678118,
+      "learning_rate": 9.504038101883759e-06,
+      "loss": 0.5412,
+      "step": 6934
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2432928220918207,
+      "learning_rate": 9.503867001866005e-06,
+      "loss": 0.2354,
+      "step": 6935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.589200076990314,
+      "learning_rate": 9.50369587388046e-06,
+      "loss": 0.2533,
+      "step": 6936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.189255660422965,
+      "learning_rate": 9.503524717928182e-06,
+      "loss": 0.3988,
+      "step": 6937
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1174937621444734,
+      "learning_rate": 9.50335353401024e-06,
+      "loss": 0.53,
+      "step": 6938
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0881435144785954,
+      "learning_rate": 9.503182322127695e-06,
+      "loss": 0.3442,
+      "step": 6939
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4321104543377081,
+      "learning_rate": 9.503011082281608e-06,
+      "loss": 0.2671,
+      "step": 6940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5403908235298576,
+      "learning_rate": 9.502839814473041e-06,
+      "loss": 0.3492,
+      "step": 6941
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.213911418700109,
+      "learning_rate": 9.502668518703062e-06,
+      "loss": 0.3371,
+      "step": 6942
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2724311573352591,
+      "learning_rate": 9.502497194972733e-06,
+      "loss": 0.2577,
+      "step": 6943
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0387134521443608,
+      "learning_rate": 9.502325843283118e-06,
+      "loss": 0.1631,
+      "step": 6944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6137098638956622,
+      "learning_rate": 9.502154463635279e-06,
+      "loss": 0.289,
+      "step": 6945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4736224478122655,
+      "learning_rate": 9.501983056030282e-06,
+      "loss": 0.2143,
+      "step": 6946
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6730792240860384,
+      "learning_rate": 9.501811620469192e-06,
+      "loss": 0.3498,
+      "step": 6947
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4796181074384593,
+      "learning_rate": 9.501640156953072e-06,
+      "loss": 0.3641,
+      "step": 6948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8657362528603021,
+      "learning_rate": 9.501468665482988e-06,
+      "loss": 0.1806,
+      "step": 6949
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9004581749134235,
+      "learning_rate": 9.501297146060005e-06,
+      "loss": 0.3787,
+      "step": 6950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3321794010791321,
+      "learning_rate": 9.501125598685186e-06,
+      "loss": 0.2433,
+      "step": 6951
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7586283422361466,
+      "learning_rate": 9.500954023359598e-06,
+      "loss": 0.1943,
+      "step": 6952
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0252648648365272,
+      "learning_rate": 9.500782420084308e-06,
+      "loss": 0.6544,
+      "step": 6953
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5911167707471114,
+      "learning_rate": 9.500610788860379e-06,
+      "loss": 0.2145,
+      "step": 6954
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6788262177319424,
+      "learning_rate": 9.500439129688877e-06,
+      "loss": 0.4311,
+      "step": 6955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1421067198277055,
+      "learning_rate": 9.500267442570869e-06,
+      "loss": 0.4537,
+      "step": 6956
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6607441750287641,
+      "learning_rate": 9.50009572750742e-06,
+      "loss": 0.1616,
+      "step": 6957
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.944218860513705,
+      "learning_rate": 9.499923984499598e-06,
+      "loss": 0.6636,
+      "step": 6958
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7047705793546937,
+      "learning_rate": 9.499752213548469e-06,
+      "loss": 0.2788,
+      "step": 6959
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8228351933573963,
+      "learning_rate": 9.499580414655096e-06,
+      "loss": 0.5466,
+      "step": 6960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.831198081950316,
+      "learning_rate": 9.499408587820551e-06,
+      "loss": 0.2896,
+      "step": 6961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6440121540459414,
+      "learning_rate": 9.4992367330459e-06,
+      "loss": 0.3637,
+      "step": 6962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7947338906558967,
+      "learning_rate": 9.499064850332207e-06,
+      "loss": 0.278,
+      "step": 6963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.571533370243933,
+      "learning_rate": 9.498892939680543e-06,
+      "loss": 0.2832,
+      "step": 6964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.142171402111823,
+      "learning_rate": 9.498721001091975e-06,
+      "loss": 0.53,
+      "step": 6965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5091508166253456,
+      "learning_rate": 9.498549034567568e-06,
+      "loss": 0.2602,
+      "step": 6966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2816948163076174,
+      "learning_rate": 9.498377040108392e-06,
+      "loss": 0.4493,
+      "step": 6967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9473334083505964,
+      "learning_rate": 9.498205017715513e-06,
+      "loss": 0.5369,
+      "step": 6968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.181021519707061,
+      "learning_rate": 9.498032967390002e-06,
+      "loss": 0.4329,
+      "step": 6969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4211155623931995,
+      "learning_rate": 9.497860889132926e-06,
+      "loss": 0.3361,
+      "step": 6970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9455882876590673,
+      "learning_rate": 9.497688782945355e-06,
+      "loss": 0.5529,
+      "step": 6971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9273935343262327,
+      "learning_rate": 9.497516648828356e-06,
+      "loss": 0.4005,
+      "step": 6972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.565731415573885,
+      "learning_rate": 9.497344486782998e-06,
+      "loss": 0.2428,
+      "step": 6973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.739463662244747,
+      "learning_rate": 9.49717229681035e-06,
+      "loss": 0.2576,
+      "step": 6974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.794518443067303,
+      "learning_rate": 9.49700007891148e-06,
+      "loss": 0.3816,
+      "step": 6975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1024341512782834,
+      "learning_rate": 9.496827833087464e-06,
+      "loss": 0.3211,
+      "step": 6976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4173259828320417,
+      "learning_rate": 9.496655559339362e-06,
+      "loss": 0.2375,
+      "step": 6977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2500676568773463,
+      "learning_rate": 9.496483257668252e-06,
+      "loss": 0.8433,
+      "step": 6978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7463140498042795,
+      "learning_rate": 9.4963109280752e-06,
+      "loss": 0.3451,
+      "step": 6979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.321600796168781,
+      "learning_rate": 9.496138570561272e-06,
+      "loss": 0.3593,
+      "step": 6980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6132279264860307,
+      "learning_rate": 9.495966185127547e-06,
+      "loss": 0.3303,
+      "step": 6981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5812906533318998,
+      "learning_rate": 9.495793771775092e-06,
+      "loss": 0.4134,
+      "step": 6982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5023936256941373,
+      "learning_rate": 9.495621330504975e-06,
+      "loss": 0.3259,
+      "step": 6983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7594821968419139,
+      "learning_rate": 9.49544886131827e-06,
+      "loss": 0.2721,
+      "step": 6984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6163762978958691,
+      "learning_rate": 9.495276364216046e-06,
+      "loss": 0.3767,
+      "step": 6985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9508761161741854,
+      "learning_rate": 9.495103839199376e-06,
+      "loss": 0.4179,
+      "step": 6986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6187236902499615,
+      "learning_rate": 9.494931286269329e-06,
+      "loss": 0.1969,
+      "step": 6987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6377202977803471,
+      "learning_rate": 9.494758705426978e-06,
+      "loss": 0.2584,
+      "step": 6988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.493531388382521,
+      "learning_rate": 9.494586096673396e-06,
+      "loss": 0.3895,
+      "step": 6989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6714187070709179,
+      "learning_rate": 9.49441346000965e-06,
+      "loss": 0.323,
+      "step": 6990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5616316957622576,
+      "learning_rate": 9.494240795436817e-06,
+      "loss": 0.478,
+      "step": 6991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4813476927620375,
+      "learning_rate": 9.494068102955969e-06,
+      "loss": 0.6352,
+      "step": 6992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.138364593371376,
+      "learning_rate": 9.493895382568174e-06,
+      "loss": 0.2079,
+      "step": 6993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6265324143660593,
+      "learning_rate": 9.493722634274509e-06,
+      "loss": 0.2408,
+      "step": 6994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4304950965859309,
+      "learning_rate": 9.493549858076047e-06,
+      "loss": 0.3471,
+      "step": 6995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6785227796061628,
+      "learning_rate": 9.493377053973858e-06,
+      "loss": 0.2748,
+      "step": 6996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9627643449691798,
+      "learning_rate": 9.493204221969014e-06,
+      "loss": 0.2845,
+      "step": 6997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.541872806288508,
+      "learning_rate": 9.493031362062592e-06,
+      "loss": 0.3791,
+      "step": 6998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9147201031249586,
+      "learning_rate": 9.492858474255663e-06,
+      "loss": 0.7147,
+      "step": 6999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.384619133414513,
+      "learning_rate": 9.492685558549304e-06,
+      "loss": 0.5384,
+      "step": 7000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6927566104508995,
+      "learning_rate": 9.492512614944584e-06,
+      "loss": 0.2345,
+      "step": 7001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.287592497242715,
+      "learning_rate": 9.492339643442579e-06,
+      "loss": 0.2878,
+      "step": 7002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5020850189396169,
+      "learning_rate": 9.492166644044363e-06,
+      "loss": 0.2149,
+      "step": 7003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3186678659705962,
+      "learning_rate": 9.491993616751012e-06,
+      "loss": 0.228,
+      "step": 7004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7050144056930978,
+      "learning_rate": 9.491820561563597e-06,
+      "loss": 0.4143,
+      "step": 7005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.296815598945901,
+      "learning_rate": 9.491647478483195e-06,
+      "loss": 0.293,
+      "step": 7006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7509358917806288,
+      "learning_rate": 9.491474367510882e-06,
+      "loss": 0.4708,
+      "step": 7007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.265685374385832,
+      "learning_rate": 9.49130122864773e-06,
+      "loss": 0.1589,
+      "step": 7008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.849631868284064,
+      "learning_rate": 9.491128061894815e-06,
+      "loss": 0.4096,
+      "step": 7009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.606047788092792,
+      "learning_rate": 9.490954867253213e-06,
+      "loss": 0.477,
+      "step": 7010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.071892218201254,
+      "learning_rate": 9.490781644724e-06,
+      "loss": 0.1538,
+      "step": 7011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3717267139048073,
+      "learning_rate": 9.490608394308249e-06,
+      "loss": 0.1657,
+      "step": 7012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0816726193529969,
+      "learning_rate": 9.49043511600704e-06,
+      "loss": 0.2725,
+      "step": 7013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0070245714473716,
+      "learning_rate": 9.490261809821446e-06,
+      "loss": 0.4936,
+      "step": 7014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5052064823570441,
+      "learning_rate": 9.490088475752543e-06,
+      "loss": 0.1855,
+      "step": 7015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.405639478821771,
+      "learning_rate": 9.48991511380141e-06,
+      "loss": 0.2431,
+      "step": 7016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3925290520409424,
+      "learning_rate": 9.489741723969122e-06,
+      "loss": 0.2678,
+      "step": 7017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8375060786905755,
+      "learning_rate": 9.489568306256753e-06,
+      "loss": 0.2713,
+      "step": 7018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4166484102891337,
+      "learning_rate": 9.489394860665385e-06,
+      "loss": 0.195,
+      "step": 7019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6566433012224964,
+      "learning_rate": 9.489221387196092e-06,
+      "loss": 0.2454,
+      "step": 7020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6609396668040513,
+      "learning_rate": 9.489047885849952e-06,
+      "loss": 0.2894,
+      "step": 7021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.741635085310853,
+      "learning_rate": 9.488874356628039e-06,
+      "loss": 0.2149,
+      "step": 7022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5116820228881545,
+      "learning_rate": 9.488700799531437e-06,
+      "loss": 0.2629,
+      "step": 7023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8345399334371155,
+      "learning_rate": 9.48852721456122e-06,
+      "loss": 0.2379,
+      "step": 7024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.293037335541452,
+      "learning_rate": 9.488353601718465e-06,
+      "loss": 0.5682,
+      "step": 7025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4363386967663911,
+      "learning_rate": 9.488179961004252e-06,
+      "loss": 0.1724,
+      "step": 7026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9042657429033234,
+      "learning_rate": 9.488006292419659e-06,
+      "loss": 0.5512,
+      "step": 7027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7813634334384372,
+      "learning_rate": 9.487832595965765e-06,
+      "loss": 0.3674,
+      "step": 7028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53700203381415,
+      "learning_rate": 9.487658871643646e-06,
+      "loss": 0.2984,
+      "step": 7029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1187117744755994,
+      "learning_rate": 9.487485119454382e-06,
+      "loss": 0.4363,
+      "step": 7030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6444352636165127,
+      "learning_rate": 9.487311339399055e-06,
+      "loss": 0.1931,
+      "step": 7031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7665712766228565,
+      "learning_rate": 9.487137531478741e-06,
+      "loss": 0.5471,
+      "step": 7032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4200737857495225,
+      "learning_rate": 9.486963695694517e-06,
+      "loss": 0.342,
+      "step": 7033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8004419540865038,
+      "learning_rate": 9.486789832047468e-06,
+      "loss": 0.5352,
+      "step": 7034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2228727918280031,
+      "learning_rate": 9.48661594053867e-06,
+      "loss": 0.1963,
+      "step": 7035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8783831393585368,
+      "learning_rate": 9.486442021169204e-06,
+      "loss": 0.5324,
+      "step": 7036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6651884644005197,
+      "learning_rate": 9.48626807394015e-06,
+      "loss": 0.2268,
+      "step": 7037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.814080597378983,
+      "learning_rate": 9.48609409885259e-06,
+      "loss": 0.3155,
+      "step": 7038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4520206759951706,
+      "learning_rate": 9.485920095907599e-06,
+      "loss": 0.2328,
+      "step": 7039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9146316382824846,
+      "learning_rate": 9.485746065106263e-06,
+      "loss": 0.4873,
+      "step": 7040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6057394152250415,
+      "learning_rate": 9.485572006449659e-06,
+      "loss": 0.4394,
+      "step": 7041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.305787767528899,
+      "learning_rate": 9.48539791993887e-06,
+      "loss": 0.2235,
+      "step": 7042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.657457599708904,
+      "learning_rate": 9.485223805574977e-06,
+      "loss": 0.5219,
+      "step": 7043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3617227275219077,
+      "learning_rate": 9.485049663359059e-06,
+      "loss": 0.2885,
+      "step": 7044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8064866490088485,
+      "learning_rate": 9.4848754932922e-06,
+      "loss": 0.3179,
+      "step": 7045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3671309759172117,
+      "learning_rate": 9.48470129537548e-06,
+      "loss": 0.3722,
+      "step": 7046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6603438574308491,
+      "learning_rate": 9.484527069609983e-06,
+      "loss": 0.3442,
+      "step": 7047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8182096409429924,
+      "learning_rate": 9.484352815996788e-06,
+      "loss": 0.3788,
+      "step": 7048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9575466687935006,
+      "learning_rate": 9.484178534536978e-06,
+      "loss": 0.2348,
+      "step": 7049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2237816125457193,
+      "learning_rate": 9.484004225231636e-06,
+      "loss": 0.5849,
+      "step": 7050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7633439664288746,
+      "learning_rate": 9.483829888081843e-06,
+      "loss": 0.2541,
+      "step": 7051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0858676770129096,
+      "learning_rate": 9.483655523088682e-06,
+      "loss": 0.4282,
+      "step": 7052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0956270287626504,
+      "learning_rate": 9.483481130253238e-06,
+      "loss": 0.3828,
+      "step": 7053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7484013906539297,
+      "learning_rate": 9.48330670957659e-06,
+      "loss": 0.3269,
+      "step": 7054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0902630715065826,
+      "learning_rate": 9.483132261059825e-06,
+      "loss": 0.5413,
+      "step": 7055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2121441129920476,
+      "learning_rate": 9.482957784704024e-06,
+      "loss": 0.2351,
+      "step": 7056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.869171899675393,
+      "learning_rate": 9.48278328051027e-06,
+      "loss": 0.539,
+      "step": 7057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6885189560730391,
+      "learning_rate": 9.482608748479649e-06,
+      "loss": 0.332,
+      "step": 7058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4802460943390214,
+      "learning_rate": 9.482434188613243e-06,
+      "loss": 0.3189,
+      "step": 7059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6903646645031467,
+      "learning_rate": 9.482259600912137e-06,
+      "loss": 0.3305,
+      "step": 7060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.713610405446637,
+      "learning_rate": 9.482084985377414e-06,
+      "loss": 0.4979,
+      "step": 7061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3917122425436006,
+      "learning_rate": 9.48191034201016e-06,
+      "loss": 0.2433,
+      "step": 7062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.603344562015807,
+      "learning_rate": 9.481735670811456e-06,
+      "loss": 0.1761,
+      "step": 7063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8216037427258354,
+      "learning_rate": 9.481560971782389e-06,
+      "loss": 0.4601,
+      "step": 7064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.053098183577248,
+      "learning_rate": 9.481386244924044e-06,
+      "loss": 0.1874,
+      "step": 7065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5724097570851798,
+      "learning_rate": 9.481211490237508e-06,
+      "loss": 0.3815,
+      "step": 7066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.644678385624319,
+      "learning_rate": 9.481036707723862e-06,
+      "loss": 0.3947,
+      "step": 7067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8214651232345658,
+      "learning_rate": 9.480861897384193e-06,
+      "loss": 0.3268,
+      "step": 7068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4659080739596178,
+      "learning_rate": 9.480687059219589e-06,
+      "loss": 0.229,
+      "step": 7069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1293731737366937,
+      "learning_rate": 9.480512193231132e-06,
+      "loss": 0.3706,
+      "step": 7070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1812990940684656,
+      "learning_rate": 9.48033729941991e-06,
+      "loss": 0.356,
+      "step": 7071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.285371802833489,
+      "learning_rate": 9.480162377787008e-06,
+      "loss": 0.379,
+      "step": 7072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8048851454040358,
+      "learning_rate": 9.479987428333513e-06,
+      "loss": 0.4722,
+      "step": 7073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.240114869020699,
+      "learning_rate": 9.479812451060511e-06,
+      "loss": 0.4444,
+      "step": 7074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.423864198026271,
+      "learning_rate": 9.479637445969087e-06,
+      "loss": 0.2457,
+      "step": 7075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4648427543636755,
+      "learning_rate": 9.479462413060332e-06,
+      "loss": 0.3119,
+      "step": 7076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4721513687134165,
+      "learning_rate": 9.479287352335328e-06,
+      "loss": 0.2164,
+      "step": 7077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.483259474553206,
+      "learning_rate": 9.479112263795167e-06,
+      "loss": 0.3356,
+      "step": 7078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.004698613434074,
+      "learning_rate": 9.478937147440932e-06,
+      "loss": 0.2898,
+      "step": 7079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3421037175959558,
+      "learning_rate": 9.478762003273712e-06,
+      "loss": 0.1509,
+      "step": 7080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6101571729723096,
+      "learning_rate": 9.478586831294597e-06,
+      "loss": 0.2061,
+      "step": 7081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4436605033548444,
+      "learning_rate": 9.478411631504669e-06,
+      "loss": 0.3947,
+      "step": 7082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.491674599701904,
+      "learning_rate": 9.47823640390502e-06,
+      "loss": 0.3269,
+      "step": 7083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7400997443734312,
+      "learning_rate": 9.478061148496737e-06,
+      "loss": 0.5216,
+      "step": 7084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3880663489154783,
+      "learning_rate": 9.477885865280912e-06,
+      "loss": 0.3183,
+      "step": 7085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5817994689694048,
+      "learning_rate": 9.477710554258628e-06,
+      "loss": 0.3154,
+      "step": 7086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4492894384207655,
+      "learning_rate": 9.477535215430974e-06,
+      "loss": 0.1753,
+      "step": 7087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6469242259600227,
+      "learning_rate": 9.477359848799044e-06,
+      "loss": 0.2585,
+      "step": 7088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53924495080309,
+      "learning_rate": 9.477184454363922e-06,
+      "loss": 0.377,
+      "step": 7089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.089791010253543,
+      "learning_rate": 9.4770090321267e-06,
+      "loss": 0.5709,
+      "step": 7090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.650413251191287,
+      "learning_rate": 9.476833582088465e-06,
+      "loss": 0.342,
+      "step": 7091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4601646663052996,
+      "learning_rate": 9.476658104250307e-06,
+      "loss": 0.5493,
+      "step": 7092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2784050779765663,
+      "learning_rate": 9.476482598613318e-06,
+      "loss": 0.5354,
+      "step": 7093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7487919311833926,
+      "learning_rate": 9.476307065178584e-06,
+      "loss": 0.2384,
+      "step": 7094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5381600776847633,
+      "learning_rate": 9.476131503947199e-06,
+      "loss": 0.2614,
+      "step": 7095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.058307519228323,
+      "learning_rate": 9.47595591492025e-06,
+      "loss": 0.3954,
+      "step": 7096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.201776480716154,
+      "learning_rate": 9.47578029809883e-06,
+      "loss": 0.6013,
+      "step": 7097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3253868093978145,
+      "learning_rate": 9.475604653484027e-06,
+      "loss": 0.4095,
+      "step": 7098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8559864024804138,
+      "learning_rate": 9.475428981076933e-06,
+      "loss": 0.3843,
+      "step": 7099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8188837895228516,
+      "learning_rate": 9.47525328087864e-06,
+      "loss": 0.3892,
+      "step": 7100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8588631931316286,
+      "learning_rate": 9.475077552890237e-06,
+      "loss": 0.2177,
+      "step": 7101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0329902163676845,
+      "learning_rate": 9.474901797112817e-06,
+      "loss": 0.3346,
+      "step": 7102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0851275446699997,
+      "learning_rate": 9.474726013547469e-06,
+      "loss": 0.33,
+      "step": 7103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0910708701797427,
+      "learning_rate": 9.474550202195288e-06,
+      "loss": 0.6018,
+      "step": 7104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.508314373440441,
+      "learning_rate": 9.474374363057362e-06,
+      "loss": 0.1612,
+      "step": 7105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8299713653235425,
+      "learning_rate": 9.474198496134784e-06,
+      "loss": 0.3771,
+      "step": 7106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3491354660941959,
+      "learning_rate": 9.474022601428647e-06,
+      "loss": 0.2624,
+      "step": 7107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.520929507149707,
+      "learning_rate": 9.473846678940045e-06,
+      "loss": 0.4698,
+      "step": 7108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.341692100516311,
+      "learning_rate": 9.473670728670068e-06,
+      "loss": 0.5512,
+      "step": 7109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.063774610912173,
+      "learning_rate": 9.473494750619809e-06,
+      "loss": 0.3784,
+      "step": 7110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.212942386450276,
+      "learning_rate": 9.47331874479036e-06,
+      "loss": 0.1971,
+      "step": 7111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7852261652704908,
+      "learning_rate": 9.473142711182814e-06,
+      "loss": 0.3547,
+      "step": 7112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6958804958949234,
+      "learning_rate": 9.472966649798267e-06,
+      "loss": 0.5083,
+      "step": 7113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0471209822335372,
+      "learning_rate": 9.472790560637809e-06,
+      "loss": 0.4595,
+      "step": 7114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.610534620550676,
+      "learning_rate": 9.472614443702534e-06,
+      "loss": 0.2655,
+      "step": 7115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0109952366298036,
+      "learning_rate": 9.472438298993538e-06,
+      "loss": 0.3687,
+      "step": 7116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8417858566407668,
+      "learning_rate": 9.472262126511913e-06,
+      "loss": 0.3296,
+      "step": 7117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7091409665981985,
+      "learning_rate": 9.472085926258752e-06,
+      "loss": 0.3741,
+      "step": 7118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.730579340395455,
+      "learning_rate": 9.471909698235152e-06,
+      "loss": 0.1866,
+      "step": 7119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8850789259392264,
+      "learning_rate": 9.471733442442203e-06,
+      "loss": 0.485,
+      "step": 7120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2438755291941586,
+      "learning_rate": 9.471557158881004e-06,
+      "loss": 0.1754,
+      "step": 7121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8200257734488556,
+      "learning_rate": 9.471380847552647e-06,
+      "loss": 0.1941,
+      "step": 7122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4240658218915534,
+      "learning_rate": 9.471204508458228e-06,
+      "loss": 0.1699,
+      "step": 7123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.816305902231834,
+      "learning_rate": 9.471028141598841e-06,
+      "loss": 0.3796,
+      "step": 7124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9368625253581537,
+      "learning_rate": 9.470851746975582e-06,
+      "loss": 0.4189,
+      "step": 7125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9528986742494945,
+      "learning_rate": 9.470675324589546e-06,
+      "loss": 0.2271,
+      "step": 7126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9280604364043485,
+      "learning_rate": 9.470498874441832e-06,
+      "loss": 0.5238,
+      "step": 7127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4189416736994567,
+      "learning_rate": 9.47032239653353e-06,
+      "loss": 0.3069,
+      "step": 7128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4221278335240117,
+      "learning_rate": 9.470145890865736e-06,
+      "loss": 0.1992,
+      "step": 7129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7319330640802253,
+      "learning_rate": 9.46996935743955e-06,
+      "loss": 0.2188,
+      "step": 7130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.971249409205627,
+      "learning_rate": 9.469792796256068e-06,
+      "loss": 0.4296,
+      "step": 7131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6861439307173363,
+      "learning_rate": 9.469616207316383e-06,
+      "loss": 0.2742,
+      "step": 7132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2187197710046513,
+      "learning_rate": 9.469439590621594e-06,
+      "loss": 0.386,
+      "step": 7133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.456021111867451,
+      "learning_rate": 9.469262946172799e-06,
+      "loss": 0.658,
+      "step": 7134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.844378226951845,
+      "learning_rate": 9.469086273971091e-06,
+      "loss": 0.4794,
+      "step": 7135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9612224483910083,
+      "learning_rate": 9.46890957401757e-06,
+      "loss": 0.2927,
+      "step": 7136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6301202412550377,
+      "learning_rate": 9.468732846313334e-06,
+      "loss": 0.3292,
+      "step": 7137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7544659133688305,
+      "learning_rate": 9.468556090859477e-06,
+      "loss": 0.2627,
+      "step": 7138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7350340591922953,
+      "learning_rate": 9.4683793076571e-06,
+      "loss": 0.3868,
+      "step": 7139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5564617140321402,
+      "learning_rate": 9.468202496707299e-06,
+      "loss": 0.2886,
+      "step": 7140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3683397805142,
+      "learning_rate": 9.46802565801117e-06,
+      "loss": 0.2025,
+      "step": 7141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4020455409934693,
+      "learning_rate": 9.467848791569817e-06,
+      "loss": 0.3169,
+      "step": 7142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.603430760840984,
+      "learning_rate": 9.467671897384333e-06,
+      "loss": 0.3785,
+      "step": 7143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2664797781819126,
+      "learning_rate": 9.467494975455818e-06,
+      "loss": 0.2219,
+      "step": 7144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7107481307348833,
+      "learning_rate": 9.467318025785371e-06,
+      "loss": 0.2809,
+      "step": 7145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.23697566689753,
+      "learning_rate": 9.46714104837409e-06,
+      "loss": 0.539,
+      "step": 7146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6663530931431365,
+      "learning_rate": 9.466964043223077e-06,
+      "loss": 0.2517,
+      "step": 7147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7009645986141742,
+      "learning_rate": 9.466787010333428e-06,
+      "loss": 0.4739,
+      "step": 7148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1074344804332,
+      "learning_rate": 9.46660994970624e-06,
+      "loss": 0.4216,
+      "step": 7149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4308688228794095,
+      "learning_rate": 9.466432861342619e-06,
+      "loss": 0.2333,
+      "step": 7150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.267182403521919,
+      "learning_rate": 9.46625574524366e-06,
+      "loss": 0.5398,
+      "step": 7151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.08424552775591,
+      "learning_rate": 9.466078601410464e-06,
+      "loss": 0.4317,
+      "step": 7152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8251227481269832,
+      "learning_rate": 9.465901429844131e-06,
+      "loss": 0.3747,
+      "step": 7153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.986433812183274,
+      "learning_rate": 9.465724230545764e-06,
+      "loss": 0.3843,
+      "step": 7154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5272559387057534,
+      "learning_rate": 9.465547003516457e-06,
+      "loss": 0.1605,
+      "step": 7155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.782945511960431,
+      "learning_rate": 9.465369748757316e-06,
+      "loss": 0.4051,
+      "step": 7156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0329625906281814,
+      "learning_rate": 9.465192466269439e-06,
+      "loss": 0.3744,
+      "step": 7157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3829181329068676,
+      "learning_rate": 9.465015156053928e-06,
+      "loss": 0.4121,
+      "step": 7158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4727371886699345,
+      "learning_rate": 9.464837818111886e-06,
+      "loss": 0.3203,
+      "step": 7159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.214917872969779,
+      "learning_rate": 9.464660452444409e-06,
+      "loss": 0.2027,
+      "step": 7160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.829743481943415,
+      "learning_rate": 9.464483059052602e-06,
+      "loss": 0.2176,
+      "step": 7161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.013774528809851,
+      "learning_rate": 9.464305637937567e-06,
+      "loss": 0.6362,
+      "step": 7162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9877470906528643,
+      "learning_rate": 9.464128189100405e-06,
+      "loss": 0.3777,
+      "step": 7163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.865262252040075,
+      "learning_rate": 9.463950712542216e-06,
+      "loss": 0.2036,
+      "step": 7164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6907003176159807,
+      "learning_rate": 9.463773208264104e-06,
+      "loss": 0.2465,
+      "step": 7165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0065285328899813,
+      "learning_rate": 9.463595676267171e-06,
+      "loss": 0.5378,
+      "step": 7166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.336475370000694,
+      "learning_rate": 9.463418116552522e-06,
+      "loss": 0.2323,
+      "step": 7167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8828353247307679,
+      "learning_rate": 9.463240529121256e-06,
+      "loss": 0.562,
+      "step": 7168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6677445293359876,
+      "learning_rate": 9.463062913974475e-06,
+      "loss": 0.2049,
+      "step": 7169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7986992490579492,
+      "learning_rate": 9.462885271113284e-06,
+      "loss": 0.48,
+      "step": 7170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5690534678102146,
+      "learning_rate": 9.462707600538788e-06,
+      "loss": 0.3752,
+      "step": 7171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4164572928016606,
+      "learning_rate": 9.462529902252086e-06,
+      "loss": 0.3474,
+      "step": 7172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7232959018201657,
+      "learning_rate": 9.462352176254283e-06,
+      "loss": 0.4615,
+      "step": 7173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7601725566123907,
+      "learning_rate": 9.462174422546487e-06,
+      "loss": 0.2217,
+      "step": 7174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.958128071656848,
+      "learning_rate": 9.461996641129794e-06,
+      "loss": 0.2389,
+      "step": 7175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.654856632160194,
+      "learning_rate": 9.461818832005315e-06,
+      "loss": 0.2861,
+      "step": 7176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.94459842115277,
+      "learning_rate": 9.461640995174148e-06,
+      "loss": 0.47,
+      "step": 7177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4829411583486871,
+      "learning_rate": 9.461463130637401e-06,
+      "loss": 0.3571,
+      "step": 7178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3480085276821696,
+      "learning_rate": 9.461285238396179e-06,
+      "loss": 0.5613,
+      "step": 7179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1536766433656025,
+      "learning_rate": 9.461107318451585e-06,
+      "loss": 0.2074,
+      "step": 7180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8735732411959414,
+      "learning_rate": 9.460929370804726e-06,
+      "loss": 0.5674,
+      "step": 7181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6758037601195097,
+      "learning_rate": 9.460751395456705e-06,
+      "loss": 0.3702,
+      "step": 7182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8189291793679592,
+      "learning_rate": 9.460573392408628e-06,
+      "loss": 0.4033,
+      "step": 7183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3702890950338777,
+      "learning_rate": 9.460395361661598e-06,
+      "loss": 0.2168,
+      "step": 7184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6094395036640845,
+      "learning_rate": 9.460217303216724e-06,
+      "loss": 0.2451,
+      "step": 7185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4984132301346036,
+      "learning_rate": 9.46003921707511e-06,
+      "loss": 0.222,
+      "step": 7186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.07675885809881,
+      "learning_rate": 9.459861103237862e-06,
+      "loss": 0.1935,
+      "step": 7187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3600589001001455,
+      "learning_rate": 9.459682961706085e-06,
+      "loss": 0.4179,
+      "step": 7188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9815162694868573,
+      "learning_rate": 9.459504792480887e-06,
+      "loss": 0.2279,
+      "step": 7189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9368346094620421,
+      "learning_rate": 9.459326595563374e-06,
+      "loss": 0.5122,
+      "step": 7190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5468320316364377,
+      "learning_rate": 9.459148370954652e-06,
+      "loss": 0.4105,
+      "step": 7191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2443339109366853,
+      "learning_rate": 9.458970118655828e-06,
+      "loss": 0.2118,
+      "step": 7192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8347779142615077,
+      "learning_rate": 9.45879183866801e-06,
+      "loss": 0.4345,
+      "step": 7193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.83743373692984,
+      "learning_rate": 9.458613530992302e-06,
+      "loss": 0.4266,
+      "step": 7194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7868334364351814,
+      "learning_rate": 9.458435195629815e-06,
+      "loss": 0.2515,
+      "step": 7195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8951517170560854,
+      "learning_rate": 9.458256832581654e-06,
+      "loss": 0.4115,
+      "step": 7196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.711557158517016,
+      "learning_rate": 9.458078441848926e-06,
+      "loss": 0.3005,
+      "step": 7197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8852340155764733,
+      "learning_rate": 9.457900023432741e-06,
+      "loss": 0.1789,
+      "step": 7198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.160518098569536,
+      "learning_rate": 9.457721577334206e-06,
+      "loss": 0.3416,
+      "step": 7199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.696392422932708,
+      "learning_rate": 9.45754310355443e-06,
+      "loss": 0.4258,
+      "step": 7200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3735654104123582,
+      "learning_rate": 9.457364602094517e-06,
+      "loss": 0.2393,
+      "step": 7201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6028297899830537,
+      "learning_rate": 9.457186072955581e-06,
+      "loss": 0.2573,
+      "step": 7202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5336027341395855,
+      "learning_rate": 9.457007516138729e-06,
+      "loss": 0.1998,
+      "step": 7203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.378553217106689,
+      "learning_rate": 9.456828931645068e-06,
+      "loss": 0.1968,
+      "step": 7204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.604133983894267,
+      "learning_rate": 9.456650319475707e-06,
+      "loss": 0.4421,
+      "step": 7205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3547855354257834,
+      "learning_rate": 9.456471679631758e-06,
+      "loss": 0.4384,
+      "step": 7206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8120601445077062,
+      "learning_rate": 9.456293012114328e-06,
+      "loss": 0.3661,
+      "step": 7207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7838584535551885,
+      "learning_rate": 9.456114316924527e-06,
+      "loss": 0.4639,
+      "step": 7208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8378821462989006,
+      "learning_rate": 9.455935594063465e-06,
+      "loss": 0.2572,
+      "step": 7209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5281047766562306,
+      "learning_rate": 9.455756843532251e-06,
+      "loss": 0.1407,
+      "step": 7210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8938062480089233,
+      "learning_rate": 9.455578065331995e-06,
+      "loss": 0.5616,
+      "step": 7211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.420942188404645,
+      "learning_rate": 9.455399259463808e-06,
+      "loss": 0.546,
+      "step": 7212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.501143565966908,
+      "learning_rate": 9.4552204259288e-06,
+      "loss": 0.3516,
+      "step": 7213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.964116819953709,
+      "learning_rate": 9.455041564728081e-06,
+      "loss": 0.1908,
+      "step": 7214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8269994961471547,
+      "learning_rate": 9.454862675862763e-06,
+      "loss": 0.3101,
+      "step": 7215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.546362690909063,
+      "learning_rate": 9.454683759333954e-06,
+      "loss": 0.2042,
+      "step": 7216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3526602735500326,
+      "learning_rate": 9.45450481514277e-06,
+      "loss": 0.2514,
+      "step": 7217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5231923566332826,
+      "learning_rate": 9.454325843290317e-06,
+      "loss": 0.2285,
+      "step": 7218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0936960343262758,
+      "learning_rate": 9.454146843777709e-06,
+      "loss": 0.1395,
+      "step": 7219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.931287623319836,
+      "learning_rate": 9.453967816606058e-06,
+      "loss": 0.2565,
+      "step": 7220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9708457721239712,
+      "learning_rate": 9.453788761776475e-06,
+      "loss": 0.5594,
+      "step": 7221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.635962208995745,
+      "learning_rate": 9.45360967929007e-06,
+      "loss": 0.4545,
+      "step": 7222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5993319935809762,
+      "learning_rate": 9.453430569147957e-06,
+      "loss": 0.2906,
+      "step": 7223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1499679376947762,
+      "learning_rate": 9.45325143135125e-06,
+      "loss": 0.2413,
+      "step": 7224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5485771668590942,
+      "learning_rate": 9.453072265901058e-06,
+      "loss": 0.3204,
+      "step": 7225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.356113137753688,
+      "learning_rate": 9.452893072798494e-06,
+      "loss": 0.3055,
+      "step": 7226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5683010750809998,
+      "learning_rate": 9.452713852044673e-06,
+      "loss": 0.3426,
+      "step": 7227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.169513694903705,
+      "learning_rate": 9.452534603640706e-06,
+      "loss": 0.3643,
+      "step": 7228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2639200594776372,
+      "learning_rate": 9.452355327587705e-06,
+      "loss": 0.1638,
+      "step": 7229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8194403953685148,
+      "learning_rate": 9.452176023886788e-06,
+      "loss": 0.498,
+      "step": 7230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8279145294965237,
+      "learning_rate": 9.451996692539065e-06,
+      "loss": 0.1156,
+      "step": 7231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9005020751891617,
+      "learning_rate": 9.451817333545647e-06,
+      "loss": 0.4091,
+      "step": 7232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3017332084717295,
+      "learning_rate": 9.451637946907655e-06,
+      "loss": 0.124,
+      "step": 7233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4818472244759988,
+      "learning_rate": 9.451458532626195e-06,
+      "loss": 0.3705,
+      "step": 7234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2658402854404356,
+      "learning_rate": 9.451279090702387e-06,
+      "loss": 0.2342,
+      "step": 7235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2778896483513036,
+      "learning_rate": 9.451099621137343e-06,
+      "loss": 0.3135,
+      "step": 7236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.741199095424192,
+      "learning_rate": 9.450920123932177e-06,
+      "loss": 0.1991,
+      "step": 7237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9292443612199066,
+      "learning_rate": 9.450740599088003e-06,
+      "loss": 0.5537,
+      "step": 7238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4715068135931733,
+      "learning_rate": 9.450561046605939e-06,
+      "loss": 0.2749,
+      "step": 7239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9422549612738331,
+      "learning_rate": 9.450381466487095e-06,
+      "loss": 0.4431,
+      "step": 7240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4373957114122051,
+      "learning_rate": 9.450201858732592e-06,
+      "loss": 0.3641,
+      "step": 7241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.356553440714027,
+      "learning_rate": 9.450022223343541e-06,
+      "loss": 0.2018,
+      "step": 7242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.575469639369385,
+      "learning_rate": 9.44984256032106e-06,
+      "loss": 0.2395,
+      "step": 7243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1682495072860273,
+      "learning_rate": 9.449662869666264e-06,
+      "loss": 0.8263,
+      "step": 7244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7373635988073401,
+      "learning_rate": 9.449483151380265e-06,
+      "loss": 0.4902,
+      "step": 7245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5196093610285315,
+      "learning_rate": 9.449303405464185e-06,
+      "loss": 0.319,
+      "step": 7246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7264196220243975,
+      "learning_rate": 9.449123631919137e-06,
+      "loss": 0.13,
+      "step": 7247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5263043547503143,
+      "learning_rate": 9.448943830746238e-06,
+      "loss": 0.2349,
+      "step": 7248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.749596583325382,
+      "learning_rate": 9.448764001946604e-06,
+      "loss": 0.4058,
+      "step": 7249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1589405577844216,
+      "learning_rate": 9.448584145521354e-06,
+      "loss": 0.5255,
+      "step": 7250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4410075719230397,
+      "learning_rate": 9.4484042614716e-06,
+      "loss": 0.2648,
+      "step": 7251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.924457354200445,
+      "learning_rate": 9.448224349798464e-06,
+      "loss": 0.4157,
+      "step": 7252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5701996207589175,
+      "learning_rate": 9.44804441050306e-06,
+      "loss": 0.4034,
+      "step": 7253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8896566839309274,
+      "learning_rate": 9.447864443586507e-06,
+      "loss": 0.408,
+      "step": 7254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3203887926910474,
+      "learning_rate": 9.447684449049923e-06,
+      "loss": 0.6302,
+      "step": 7255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.827379612695112,
+      "learning_rate": 9.447504426894424e-06,
+      "loss": 0.44,
+      "step": 7256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4547982017033072,
+      "learning_rate": 9.447324377121127e-06,
+      "loss": 0.3851,
+      "step": 7257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7141232030337938,
+      "learning_rate": 9.447144299731155e-06,
+      "loss": 0.3501,
+      "step": 7258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2936754063838585,
+      "learning_rate": 9.44696419472562e-06,
+      "loss": 0.3735,
+      "step": 7259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.319535019464343,
+      "learning_rate": 9.446784062105647e-06,
+      "loss": 0.2268,
+      "step": 7260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5942970720699365,
+      "learning_rate": 9.446603901872349e-06,
+      "loss": 0.28,
+      "step": 7261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3915165764807875,
+      "learning_rate": 9.446423714026846e-06,
+      "loss": 0.2512,
+      "step": 7262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9802175095385857,
+      "learning_rate": 9.446243498570258e-06,
+      "loss": 0.2423,
+      "step": 7263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3961023781909527,
+      "learning_rate": 9.446063255503704e-06,
+      "loss": 0.1469,
+      "step": 7264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.864431541984501,
+      "learning_rate": 9.445882984828302e-06,
+      "loss": 0.5006,
+      "step": 7265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.547117462325845,
+      "learning_rate": 9.445702686545175e-06,
+      "loss": 0.2322,
+      "step": 7266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6932790312576929,
+      "learning_rate": 9.445522360655438e-06,
+      "loss": 0.406,
+      "step": 7267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.550273322033482,
+      "learning_rate": 9.445342007160213e-06,
+      "loss": 0.3667,
+      "step": 7268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8797613367242874,
+      "learning_rate": 9.44516162606062e-06,
+      "loss": 0.2275,
+      "step": 7269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.320266344036763,
+      "learning_rate": 9.44498121735778e-06,
+      "loss": 0.1799,
+      "step": 7270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5006008534234434,
+      "learning_rate": 9.444800781052809e-06,
+      "loss": 0.2675,
+      "step": 7271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9208369514309893,
+      "learning_rate": 9.444620317146832e-06,
+      "loss": 0.3095,
+      "step": 7272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7350662761171833,
+      "learning_rate": 9.444439825640972e-06,
+      "loss": 0.3808,
+      "step": 7273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6460716345649793,
+      "learning_rate": 9.444259306536341e-06,
+      "loss": 0.3966,
+      "step": 7274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7523152934964175,
+      "learning_rate": 9.444078759834068e-06,
+      "loss": 0.3165,
+      "step": 7275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.27942006486251,
+      "learning_rate": 9.44389818553527e-06,
+      "loss": 0.2415,
+      "step": 7276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.631628770992954,
+      "learning_rate": 9.44371758364107e-06,
+      "loss": 0.4048,
+      "step": 7277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4697279920602735,
+      "learning_rate": 9.443536954152588e-06,
+      "loss": 0.2887,
+      "step": 7278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5821292187532845,
+      "learning_rate": 9.443356297070948e-06,
+      "loss": 0.2514,
+      "step": 7279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0421848981529886,
+      "learning_rate": 9.443175612397268e-06,
+      "loss": 0.2386,
+      "step": 7280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7910124800559815,
+      "learning_rate": 9.442994900132676e-06,
+      "loss": 0.5646,
+      "step": 7281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8512972911394032,
+      "learning_rate": 9.442814160278289e-06,
+      "loss": 0.5612,
+      "step": 7282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0165979528510274,
+      "learning_rate": 9.44263339283523e-06,
+      "loss": 0.6819,
+      "step": 7283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6397023240414064,
+      "learning_rate": 9.442452597804623e-06,
+      "loss": 0.4494,
+      "step": 7284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3400599486464286,
+      "learning_rate": 9.44227177518759e-06,
+      "loss": 0.6913,
+      "step": 7285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9189764665787703,
+      "learning_rate": 9.442090924985255e-06,
+      "loss": 0.3101,
+      "step": 7286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.641935821669274,
+      "learning_rate": 9.44191004719874e-06,
+      "loss": 0.1751,
+      "step": 7287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6573915236492593,
+      "learning_rate": 9.441729141829168e-06,
+      "loss": 0.4116,
+      "step": 7288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4189470137541957,
+      "learning_rate": 9.441548208877663e-06,
+      "loss": 0.3465,
+      "step": 7289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7156786873760543,
+      "learning_rate": 9.441367248345348e-06,
+      "loss": 0.3024,
+      "step": 7290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.022397433121571,
+      "learning_rate": 9.441186260233347e-06,
+      "loss": 0.2838,
+      "step": 7291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2164260386033035,
+      "learning_rate": 9.441005244542783e-06,
+      "loss": 0.6525,
+      "step": 7292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5531894556770571,
+      "learning_rate": 9.440824201274781e-06,
+      "loss": 0.2167,
+      "step": 7293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.033509684820775,
+      "learning_rate": 9.440643130430467e-06,
+      "loss": 0.5716,
+      "step": 7294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6806670947493618,
+      "learning_rate": 9.440462032010962e-06,
+      "loss": 0.3713,
+      "step": 7295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9837965798092376,
+      "learning_rate": 9.440280906017391e-06,
+      "loss": 0.3588,
+      "step": 7296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3519681310826634,
+      "learning_rate": 9.440099752450882e-06,
+      "loss": 0.1535,
+      "step": 7297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9684108638823512,
+      "learning_rate": 9.439918571312556e-06,
+      "loss": 0.2753,
+      "step": 7298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3893741784713245,
+      "learning_rate": 9.439737362603541e-06,
+      "loss": 0.2358,
+      "step": 7299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0678290244623034,
+      "learning_rate": 9.43955612632496e-06,
+      "loss": 0.3919,
+      "step": 7300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2648795619958295,
+      "learning_rate": 9.43937486247794e-06,
+      "loss": 0.3205,
+      "step": 7301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3614024247372205,
+      "learning_rate": 9.439193571063606e-06,
+      "loss": 0.1754,
+      "step": 7302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5024169681856345,
+      "learning_rate": 9.439012252083083e-06,
+      "loss": 0.3499,
+      "step": 7303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.595518026740676,
+      "learning_rate": 9.4388309055375e-06,
+      "loss": 0.4763,
+      "step": 7304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.641978148609893,
+      "learning_rate": 9.43864953142798e-06,
+      "loss": 0.3177,
+      "step": 7305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7691356225089352,
+      "learning_rate": 9.43846812975565e-06,
+      "loss": 0.3371,
+      "step": 7306
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1269848296798035,
+      "learning_rate": 9.438286700521637e-06,
+      "loss": 0.1677,
+      "step": 7307
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9288272754976803,
+      "learning_rate": 9.438105243727067e-06,
+      "loss": 0.3897,
+      "step": 7308
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2973405795108575,
+      "learning_rate": 9.437923759373067e-06,
+      "loss": 0.2164,
+      "step": 7309
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8373360075723828,
+      "learning_rate": 9.437742247460765e-06,
+      "loss": 0.277,
+      "step": 7310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3454119051300846,
+      "learning_rate": 9.437560707991284e-06,
+      "loss": 0.4062,
+      "step": 7311
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8652730308641399,
+      "learning_rate": 9.437379140965757e-06,
+      "loss": 0.2653,
+      "step": 7312
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2265113021586003,
+      "learning_rate": 9.437197546385311e-06,
+      "loss": 0.2106,
+      "step": 7313
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4305347787815323,
+      "learning_rate": 9.437015924251068e-06,
+      "loss": 0.2975,
+      "step": 7314
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9979774075212626,
+      "learning_rate": 9.436834274564162e-06,
+      "loss": 0.2568,
+      "step": 7315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7405823489190912,
+      "learning_rate": 9.436652597325718e-06,
+      "loss": 0.3647,
+      "step": 7316
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4456005479548095,
+      "learning_rate": 9.436470892536861e-06,
+      "loss": 0.3373,
+      "step": 7317
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6004301086082222,
+      "learning_rate": 9.436289160198728e-06,
+      "loss": 0.2328,
+      "step": 7318
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3068294417231676,
+      "learning_rate": 9.43610740031244e-06,
+      "loss": 0.1979,
+      "step": 7319
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1428574872868125,
+      "learning_rate": 9.43592561287913e-06,
+      "loss": 0.4777,
+      "step": 7320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.870323858967565,
+      "learning_rate": 9.435743797899923e-06,
+      "loss": 0.6427,
+      "step": 7321
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.70488516817991,
+      "learning_rate": 9.435561955375949e-06,
+      "loss": 0.3263,
+      "step": 7322
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7302940616310762,
+      "learning_rate": 9.43538008530834e-06,
+      "loss": 0.3035,
+      "step": 7323
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.663380832477279,
+      "learning_rate": 9.435198187698224e-06,
+      "loss": 0.4777,
+      "step": 7324
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0050210551824827,
+      "learning_rate": 9.43501626254673e-06,
+      "loss": 0.2042,
+      "step": 7325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8511526402610954,
+      "learning_rate": 9.434834309854987e-06,
+      "loss": 0.3492,
+      "step": 7326
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3341564577900384,
+      "learning_rate": 9.434652329624125e-06,
+      "loss": 0.6209,
+      "step": 7327
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.880502486708827,
+      "learning_rate": 9.434470321855277e-06,
+      "loss": 0.4103,
+      "step": 7328
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3965511919589386,
+      "learning_rate": 9.434288286549568e-06,
+      "loss": 0.1616,
+      "step": 7329
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1383539721269584,
+      "learning_rate": 9.434106223708135e-06,
+      "loss": 0.446,
+      "step": 7330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3972911771592662,
+      "learning_rate": 9.433924133332104e-06,
+      "loss": 0.2568,
+      "step": 7331
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5553249527130135,
+      "learning_rate": 9.433742015422604e-06,
+      "loss": 0.2797,
+      "step": 7332
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.272077687941052,
+      "learning_rate": 9.43355986998077e-06,
+      "loss": 0.222,
+      "step": 7333
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6038253893864443,
+      "learning_rate": 9.433377697007733e-06,
+      "loss": 0.1879,
+      "step": 7334
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.272961684130472,
+      "learning_rate": 9.433195496504621e-06,
+      "loss": 0.196,
+      "step": 7335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.571273168242813,
+      "learning_rate": 9.433013268472568e-06,
+      "loss": 0.279,
+      "step": 7336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9404689744960798,
+      "learning_rate": 9.432831012912706e-06,
+      "loss": 0.1176,
+      "step": 7337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5204282309813795,
+      "learning_rate": 9.432648729826164e-06,
+      "loss": 0.2457,
+      "step": 7338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.104471625207696,
+      "learning_rate": 9.432466419214075e-06,
+      "loss": 0.6336,
+      "step": 7339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.701541116017887,
+      "learning_rate": 9.432284081077574e-06,
+      "loss": 0.343,
+      "step": 7340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2915456278359094,
+      "learning_rate": 9.43210171541779e-06,
+      "loss": 0.2208,
+      "step": 7341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4683652282500759,
+      "learning_rate": 9.431919322235857e-06,
+      "loss": 0.3304,
+      "step": 7342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4649246663950968,
+      "learning_rate": 9.431736901532904e-06,
+      "loss": 0.2252,
+      "step": 7343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2337524206930746,
+      "learning_rate": 9.43155445331007e-06,
+      "loss": 0.5632,
+      "step": 7344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8802664149886186,
+      "learning_rate": 9.431371977568483e-06,
+      "loss": 0.3029,
+      "step": 7345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.002211394979381,
+      "learning_rate": 9.431189474309277e-06,
+      "loss": 0.4258,
+      "step": 7346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4083865095243662,
+      "learning_rate": 9.43100694353359e-06,
+      "loss": 0.3482,
+      "step": 7347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.376322513947171,
+      "learning_rate": 9.430824385242548e-06,
+      "loss": 0.1538,
+      "step": 7348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5620952940389927,
+      "learning_rate": 9.43064179943729e-06,
+      "loss": 0.3937,
+      "step": 7349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5039470462715263,
+      "learning_rate": 9.430459186118948e-06,
+      "loss": 0.1747,
+      "step": 7350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3760600365902866,
+      "learning_rate": 9.430276545288655e-06,
+      "loss": 0.1886,
+      "step": 7351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4428083945140138,
+      "learning_rate": 9.430093876947546e-06,
+      "loss": 0.3225,
+      "step": 7352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1759939250890252,
+      "learning_rate": 9.429911181096756e-06,
+      "loss": 0.202,
+      "step": 7353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8961307537741723,
+      "learning_rate": 9.42972845773742e-06,
+      "loss": 0.2173,
+      "step": 7354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.44788002378308,
+      "learning_rate": 9.42954570687067e-06,
+      "loss": 0.371,
+      "step": 7355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8675087369129733,
+      "learning_rate": 9.429362928497643e-06,
+      "loss": 0.3357,
+      "step": 7356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8261462724900308,
+      "learning_rate": 9.429180122619474e-06,
+      "loss": 0.3651,
+      "step": 7357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.595054901016757,
+      "learning_rate": 9.428997289237297e-06,
+      "loss": 0.3319,
+      "step": 7358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0985197850708706,
+      "learning_rate": 9.42881442835225e-06,
+      "loss": 0.2053,
+      "step": 7359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.395774530988365,
+      "learning_rate": 9.428631539965463e-06,
+      "loss": 0.8792,
+      "step": 7360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0140128563858393,
+      "learning_rate": 9.428448624078077e-06,
+      "loss": 0.2628,
+      "step": 7361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5048332924093784,
+      "learning_rate": 9.428265680691228e-06,
+      "loss": 0.2972,
+      "step": 7362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4942195597224288,
+      "learning_rate": 9.428082709806049e-06,
+      "loss": 0.294,
+      "step": 7363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0420722890984746,
+      "learning_rate": 9.427899711423676e-06,
+      "loss": 0.3776,
+      "step": 7364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6527069269838948,
+      "learning_rate": 9.427716685545247e-06,
+      "loss": 0.4955,
+      "step": 7365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3675001393212192,
+      "learning_rate": 9.4275336321719e-06,
+      "loss": 0.204,
+      "step": 7366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3072771309681543,
+      "learning_rate": 9.427350551304767e-06,
+      "loss": 0.3273,
+      "step": 7367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2293318325002187,
+      "learning_rate": 9.427167442944988e-06,
+      "loss": 0.248,
+      "step": 7368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.794470535209473,
+      "learning_rate": 9.426984307093703e-06,
+      "loss": 0.2761,
+      "step": 7369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8150779953925649,
+      "learning_rate": 9.426801143752044e-06,
+      "loss": 0.5218,
+      "step": 7370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7428009467095116,
+      "learning_rate": 9.426617952921151e-06,
+      "loss": 0.4749,
+      "step": 7371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.187585178487982,
+      "learning_rate": 9.426434734602161e-06,
+      "loss": 0.5828,
+      "step": 7372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9216506524678714,
+      "learning_rate": 9.426251488796211e-06,
+      "loss": 0.4278,
+      "step": 7373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3143059653821698,
+      "learning_rate": 9.42606821550444e-06,
+      "loss": 0.2009,
+      "step": 7374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7567253544489974,
+      "learning_rate": 9.425884914727987e-06,
+      "loss": 0.3238,
+      "step": 7375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.051797596810374,
+      "learning_rate": 9.425701586467987e-06,
+      "loss": 0.4309,
+      "step": 7376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6014085951568438,
+      "learning_rate": 9.425518230725582e-06,
+      "loss": 0.373,
+      "step": 7377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7611523244182405,
+      "learning_rate": 9.425334847501908e-06,
+      "loss": 0.5063,
+      "step": 7378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6868138292458903,
+      "learning_rate": 9.425151436798106e-06,
+      "loss": 0.2248,
+      "step": 7379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.169692431662272,
+      "learning_rate": 9.424967998615312e-06,
+      "loss": 0.8446,
+      "step": 7380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7713430783617257,
+      "learning_rate": 9.42478453295467e-06,
+      "loss": 0.3912,
+      "step": 7381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.648058395725226,
+      "learning_rate": 9.424601039817312e-06,
+      "loss": 0.2963,
+      "step": 7382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3393516012004512,
+      "learning_rate": 9.424417519204384e-06,
+      "loss": 0.1811,
+      "step": 7383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.061698068606835,
+      "learning_rate": 9.424233971117024e-06,
+      "loss": 0.2552,
+      "step": 7384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1592884079482046,
+      "learning_rate": 9.424050395556368e-06,
+      "loss": 0.408,
+      "step": 7385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3484985953245177,
+      "learning_rate": 9.423866792523562e-06,
+      "loss": 0.1513,
+      "step": 7386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9436861386495072,
+      "learning_rate": 9.423683162019742e-06,
+      "loss": 0.4465,
+      "step": 7387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6961877690511855,
+      "learning_rate": 9.423499504046047e-06,
+      "loss": 0.3721,
+      "step": 7388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.63951429280679,
+      "learning_rate": 9.423315818603622e-06,
+      "loss": 0.4731,
+      "step": 7389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.348293325905993,
+      "learning_rate": 9.423132105693603e-06,
+      "loss": 0.3549,
+      "step": 7390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7642521627017478,
+      "learning_rate": 9.422948365317137e-06,
+      "loss": 0.19,
+      "step": 7391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3085945100924434,
+      "learning_rate": 9.422764597475359e-06,
+      "loss": 0.2178,
+      "step": 7392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6880237552445296,
+      "learning_rate": 9.422580802169412e-06,
+      "loss": 0.3345,
+      "step": 7393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8147992685739966,
+      "learning_rate": 9.422396979400438e-06,
+      "loss": 0.3948,
+      "step": 7394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4134902507503453,
+      "learning_rate": 9.42221312916958e-06,
+      "loss": 0.2047,
+      "step": 7395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3572525687257522,
+      "learning_rate": 9.422029251477975e-06,
+      "loss": 0.1932,
+      "step": 7396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4724172555071038,
+      "learning_rate": 9.42184534632677e-06,
+      "loss": 0.2815,
+      "step": 7397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9068532950554808,
+      "learning_rate": 9.421661413717104e-06,
+      "loss": 0.5915,
+      "step": 7398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7400231656003373,
+      "learning_rate": 9.421477453650118e-06,
+      "loss": 0.2914,
+      "step": 7399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7719275562625905,
+      "learning_rate": 9.421293466126959e-06,
+      "loss": 0.3337,
+      "step": 7400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3451027936014353,
+      "learning_rate": 9.421109451148765e-06,
+      "loss": 0.3216,
+      "step": 7401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5661551543012218,
+      "learning_rate": 9.42092540871668e-06,
+      "loss": 0.3699,
+      "step": 7402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3105776651008865,
+      "learning_rate": 9.420741338831847e-06,
+      "loss": 0.2536,
+      "step": 7403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0499306504219375,
+      "learning_rate": 9.420557241495412e-06,
+      "loss": 0.4292,
+      "step": 7404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.20068602001291,
+      "learning_rate": 9.420373116708513e-06,
+      "loss": 0.1958,
+      "step": 7405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2607794179228204,
+      "learning_rate": 9.420188964472296e-06,
+      "loss": 0.3566,
+      "step": 7406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8979591583313584,
+      "learning_rate": 9.420004784787903e-06,
+      "loss": 0.2103,
+      "step": 7407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6346351156699883,
+      "learning_rate": 9.419820577656482e-06,
+      "loss": 0.3828,
+      "step": 7408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3447685063180175,
+      "learning_rate": 9.419636343079173e-06,
+      "loss": 0.6023,
+      "step": 7409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.709547566938408,
+      "learning_rate": 9.419452081057118e-06,
+      "loss": 0.5145,
+      "step": 7410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1699927345176633,
+      "learning_rate": 9.419267791591469e-06,
+      "loss": 0.1873,
+      "step": 7411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8996269165647395,
+      "learning_rate": 9.419083474683362e-06,
+      "loss": 0.534,
+      "step": 7412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5375209097487113,
+      "learning_rate": 9.418899130333947e-06,
+      "loss": 0.1568,
+      "step": 7413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.496291389060194,
+      "learning_rate": 9.418714758544366e-06,
+      "loss": 0.296,
+      "step": 7414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3599250228132935,
+      "learning_rate": 9.418530359315766e-06,
+      "loss": 0.6015,
+      "step": 7415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.597658115376547,
+      "learning_rate": 9.418345932649288e-06,
+      "loss": 0.2994,
+      "step": 7416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4059100733627892,
+      "learning_rate": 9.418161478546082e-06,
+      "loss": 0.2191,
+      "step": 7417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4513267842576283,
+      "learning_rate": 9.417976997007291e-06,
+      "loss": 0.2297,
+      "step": 7418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4775142377645034,
+      "learning_rate": 9.417792488034061e-06,
+      "loss": 0.2709,
+      "step": 7419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9603966692844736,
+      "learning_rate": 9.417607951627539e-06,
+      "loss": 0.3893,
+      "step": 7420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6810025412377476,
+      "learning_rate": 9.417423387788868e-06,
+      "loss": 0.5278,
+      "step": 7421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6990527071888986,
+      "learning_rate": 9.417238796519196e-06,
+      "loss": 0.2418,
+      "step": 7422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7514665793460005,
+      "learning_rate": 9.41705417781967e-06,
+      "loss": 0.5055,
+      "step": 7423
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0093527556088393,
+      "learning_rate": 9.416869531691436e-06,
+      "loss": 0.2161,
+      "step": 7424
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6310697636839981,
+      "learning_rate": 9.416684858135639e-06,
+      "loss": 0.3548,
+      "step": 7425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3982588611392552,
+      "learning_rate": 9.416500157153426e-06,
+      "loss": 0.4384,
+      "step": 7426
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1489153276673953,
+      "learning_rate": 9.416315428745946e-06,
+      "loss": 0.1957,
+      "step": 7427
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5733699924264746,
+      "learning_rate": 9.416130672914346e-06,
+      "loss": 0.4988,
+      "step": 7428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6152211734181496,
+      "learning_rate": 9.41594588965977e-06,
+      "loss": 0.3505,
+      "step": 7429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7439794220281788,
+      "learning_rate": 9.415761078983368e-06,
+      "loss": 0.2969,
+      "step": 7430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7398986449743536,
+      "learning_rate": 9.415576240886288e-06,
+      "loss": 0.5447,
+      "step": 7431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5851230937955176,
+      "learning_rate": 9.415391375369678e-06,
+      "loss": 0.2008,
+      "step": 7432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2023986568394092,
+      "learning_rate": 9.415206482434685e-06,
+      "loss": 0.1901,
+      "step": 7433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4204831280000048,
+      "learning_rate": 9.415021562082456e-06,
+      "loss": 0.2562,
+      "step": 7434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4106026415812867,
+      "learning_rate": 9.41483661431414e-06,
+      "loss": 0.3563,
+      "step": 7435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.610806162130227,
+      "learning_rate": 9.414651639130888e-06,
+      "loss": 0.2346,
+      "step": 7436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.91841224468613,
+      "learning_rate": 9.414466636533845e-06,
+      "loss": 0.4089,
+      "step": 7437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6286332063965177,
+      "learning_rate": 9.414281606524161e-06,
+      "loss": 0.4157,
+      "step": 7438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5600299384838947,
+      "learning_rate": 9.414096549102987e-06,
+      "loss": 0.2583,
+      "step": 7439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.056098899246379,
+      "learning_rate": 9.413911464271469e-06,
+      "loss": 0.6034,
+      "step": 7440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3189988019669734,
+      "learning_rate": 9.41372635203076e-06,
+      "loss": 0.1733,
+      "step": 7441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5376826142007607,
+      "learning_rate": 9.413541212382005e-06,
+      "loss": 0.3182,
+      "step": 7442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4945743280395059,
+      "learning_rate": 9.413356045326358e-06,
+      "loss": 0.3287,
+      "step": 7443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9959497272864815,
+      "learning_rate": 9.413170850864966e-06,
+      "loss": 0.1984,
+      "step": 7444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5485524753646844,
+      "learning_rate": 9.412985628998979e-06,
+      "loss": 0.2489,
+      "step": 7445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9909602648676425,
+      "learning_rate": 9.41280037972955e-06,
+      "loss": 0.2301,
+      "step": 7446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.23542561365665,
+      "learning_rate": 9.412615103057826e-06,
+      "loss": 0.3371,
+      "step": 7447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.583201655969096,
+      "learning_rate": 9.412429798984958e-06,
+      "loss": 0.2908,
+      "step": 7448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.564410062618734,
+      "learning_rate": 9.4122444675121e-06,
+      "loss": 0.3762,
+      "step": 7449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2334479226147272,
+      "learning_rate": 9.412059108640399e-06,
+      "loss": 0.2219,
+      "step": 7450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4444890679410056,
+      "learning_rate": 9.411873722371009e-06,
+      "loss": 0.1959,
+      "step": 7451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2865099965724376,
+      "learning_rate": 9.411688308705078e-06,
+      "loss": 0.2528,
+      "step": 7452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6178313738064825,
+      "learning_rate": 9.411502867643761e-06,
+      "loss": 0.187,
+      "step": 7453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.327284763711705,
+      "learning_rate": 9.411317399188206e-06,
+      "loss": 0.3734,
+      "step": 7454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7310489172819408,
+      "learning_rate": 9.411131903339566e-06,
+      "loss": 0.5481,
+      "step": 7455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.832798445404618,
+      "learning_rate": 9.410946380098995e-06,
+      "loss": 0.5324,
+      "step": 7456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.672622475690671,
+      "learning_rate": 9.410760829467643e-06,
+      "loss": 0.3406,
+      "step": 7457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5872923391356342,
+      "learning_rate": 9.410575251446661e-06,
+      "loss": 0.3279,
+      "step": 7458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5596402617779648,
+      "learning_rate": 9.410389646037203e-06,
+      "loss": 0.2303,
+      "step": 7459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6574811048710885,
+      "learning_rate": 9.410204013240422e-06,
+      "loss": 0.2622,
+      "step": 7460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3163549147198237,
+      "learning_rate": 9.41001835305747e-06,
+      "loss": 0.185,
+      "step": 7461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.276152552792219,
+      "learning_rate": 9.4098326654895e-06,
+      "loss": 0.6475,
+      "step": 7462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9476578766012305,
+      "learning_rate": 9.409646950537666e-06,
+      "loss": 0.2537,
+      "step": 7463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.734584565861689,
+      "learning_rate": 9.409461208203121e-06,
+      "loss": 0.3357,
+      "step": 7464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.106533913479944,
+      "learning_rate": 9.409275438487015e-06,
+      "loss": 0.266,
+      "step": 7465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5258883385938935,
+      "learning_rate": 9.409089641390508e-06,
+      "loss": 0.3488,
+      "step": 7466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.949316423679006,
+      "learning_rate": 9.408903816914747e-06,
+      "loss": 0.2139,
+      "step": 7467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0009179244006576,
+      "learning_rate": 9.40871796506089e-06,
+      "loss": 0.3636,
+      "step": 7468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0380845008190667,
+      "learning_rate": 9.408532085830091e-06,
+      "loss": 0.4442,
+      "step": 7469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2147783632866056,
+      "learning_rate": 9.408346179223502e-06,
+      "loss": 0.1971,
+      "step": 7470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.557056199221782,
+      "learning_rate": 9.408160245242279e-06,
+      "loss": 0.2435,
+      "step": 7471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8546265989812065,
+      "learning_rate": 9.407974283887577e-06,
+      "loss": 0.2726,
+      "step": 7472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8081795978313477,
+      "learning_rate": 9.40778829516055e-06,
+      "loss": 0.4257,
+      "step": 7473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3651205746723316,
+      "learning_rate": 9.407602279062353e-06,
+      "loss": 0.2345,
+      "step": 7474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4014059932001197,
+      "learning_rate": 9.407416235594142e-06,
+      "loss": 0.3016,
+      "step": 7475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.518362669748589,
+      "learning_rate": 9.40723016475707e-06,
+      "loss": 0.3271,
+      "step": 7476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5731557565409717,
+      "learning_rate": 9.407044066552295e-06,
+      "loss": 0.3233,
+      "step": 7477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.704072159565408,
+      "learning_rate": 9.406857940980972e-06,
+      "loss": 0.3249,
+      "step": 7478
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5416206965702288,
+      "learning_rate": 9.406671788044256e-06,
+      "loss": 0.3324,
+      "step": 7479
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5028852340935492,
+      "learning_rate": 9.406485607743303e-06,
+      "loss": 0.3049,
+      "step": 7480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6889489805342872,
+      "learning_rate": 9.406299400079269e-06,
+      "loss": 0.491,
+      "step": 7481
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8634022107638002,
+      "learning_rate": 9.406113165053313e-06,
+      "loss": 0.4204,
+      "step": 7482
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.975594471994578,
+      "learning_rate": 9.405926902666587e-06,
+      "loss": 0.4755,
+      "step": 7483
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5663211732815046,
+      "learning_rate": 9.405740612920249e-06,
+      "loss": 0.3183,
+      "step": 7484
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2869249423017706,
+      "learning_rate": 9.405554295815458e-06,
+      "loss": 0.1985,
+      "step": 7485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6003678006198925,
+      "learning_rate": 9.40536795135337e-06,
+      "loss": 0.201,
+      "step": 7486
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7510705083354912,
+      "learning_rate": 9.40518157953514e-06,
+      "loss": 0.4909,
+      "step": 7487
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5899090624616306,
+      "learning_rate": 9.40499518036193e-06,
+      "loss": 0.2235,
+      "step": 7488
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.649426082479203,
+      "learning_rate": 9.404808753834893e-06,
+      "loss": 0.3681,
+      "step": 7489
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.104495085251798,
+      "learning_rate": 9.404622299955188e-06,
+      "loss": 0.4194,
+      "step": 7490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4848799047571328,
+      "learning_rate": 9.404435818723975e-06,
+      "loss": 0.2948,
+      "step": 7491
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9250904453691917,
+      "learning_rate": 9.404249310142407e-06,
+      "loss": 0.2373,
+      "step": 7492
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6888196827116795,
+      "learning_rate": 9.404062774211649e-06,
+      "loss": 0.4971,
+      "step": 7493
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.358424620441153,
+      "learning_rate": 9.403876210932852e-06,
+      "loss": 0.1443,
+      "step": 7494
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.252200037843399,
+      "learning_rate": 9.403689620307179e-06,
+      "loss": 0.3864,
+      "step": 7495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9846494965420285,
+      "learning_rate": 9.40350300233579e-06,
+      "loss": 0.4069,
+      "step": 7496
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.801217679988582,
+      "learning_rate": 9.40331635701984e-06,
+      "loss": 0.2195,
+      "step": 7497
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5744584495042893,
+      "learning_rate": 9.403129684360487e-06,
+      "loss": 0.3645,
+      "step": 7498
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6348942736796166,
+      "learning_rate": 9.402942984358897e-06,
+      "loss": 0.3677,
+      "step": 7499
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.255173768064218,
+      "learning_rate": 9.402756257016224e-06,
+      "loss": 0.8155,
+      "step": 7500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2274855190061948,
+      "learning_rate": 9.402569502333629e-06,
+      "loss": 0.1431,
+      "step": 7501
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6897654647082794,
+      "learning_rate": 9.40238272031227e-06,
+      "loss": 0.351,
+      "step": 7502
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3837779207762337,
+      "learning_rate": 9.402195910953309e-06,
+      "loss": 0.2554,
+      "step": 7503
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9573472047782758,
+      "learning_rate": 9.402009074257905e-06,
+      "loss": 0.4489,
+      "step": 7504
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.016547834590432,
+      "learning_rate": 9.401822210227219e-06,
+      "loss": 0.3488,
+      "step": 7505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4296417098711902,
+      "learning_rate": 9.40163531886241e-06,
+      "loss": 0.3005,
+      "step": 7506
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6800730151959293,
+      "learning_rate": 9.401448400164642e-06,
+      "loss": 0.4021,
+      "step": 7507
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9659741829897126,
+      "learning_rate": 9.40126145413507e-06,
+      "loss": 0.6955,
+      "step": 7508
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.360864949071795,
+      "learning_rate": 9.401074480774858e-06,
+      "loss": 0.325,
+      "step": 7509
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4470889631035628,
+      "learning_rate": 9.40088748008517e-06,
+      "loss": 0.3228,
+      "step": 7510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9815298262310138,
+      "learning_rate": 9.400700452067162e-06,
+      "loss": 0.167,
+      "step": 7511
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6448926370015653,
+      "learning_rate": 9.400513396721999e-06,
+      "loss": 0.4074,
+      "step": 7512
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.858521133197247,
+      "learning_rate": 9.40032631405084e-06,
+      "loss": 0.3852,
+      "step": 7513
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3702063534870328,
+      "learning_rate": 9.400139204054848e-06,
+      "loss": 0.2313,
+      "step": 7514
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9227422757765482,
+      "learning_rate": 9.399952066735187e-06,
+      "loss": 0.1909,
+      "step": 7515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7930026747209689,
+      "learning_rate": 9.399764902093015e-06,
+      "loss": 0.2714,
+      "step": 7516
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3322975194602145,
+      "learning_rate": 9.399577710129497e-06,
+      "loss": 0.1643,
+      "step": 7517
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0937409436328944,
+      "learning_rate": 9.399390490845794e-06,
+      "loss": 0.2566,
+      "step": 7518
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.828332986982319,
+      "learning_rate": 9.39920324424307e-06,
+      "loss": 0.3647,
+      "step": 7519
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3778530588883522,
+      "learning_rate": 9.399015970322486e-06,
+      "loss": 0.2526,
+      "step": 7520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3422177583350143,
+      "learning_rate": 9.398828669085207e-06,
+      "loss": 0.2457,
+      "step": 7521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.406025968031271,
+      "learning_rate": 9.398641340532394e-06,
+      "loss": 0.1727,
+      "step": 7522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1206653401185935,
+      "learning_rate": 9.398453984665213e-06,
+      "loss": 0.4604,
+      "step": 7523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2273408566619692,
+      "learning_rate": 9.398266601484823e-06,
+      "loss": 0.2231,
+      "step": 7524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5741609132048011,
+      "learning_rate": 9.39807919099239e-06,
+      "loss": 0.2487,
+      "step": 7525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7309823533372939,
+      "learning_rate": 9.397891753189082e-06,
+      "loss": 0.3111,
+      "step": 7526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.987612928625057,
+      "learning_rate": 9.397704288076055e-06,
+      "loss": 0.5204,
+      "step": 7527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0746362551062805,
+      "learning_rate": 9.397516795654478e-06,
+      "loss": 0.1622,
+      "step": 7528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7945746292431346,
+      "learning_rate": 9.397329275925516e-06,
+      "loss": 0.4674,
+      "step": 7529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7113107356912258,
+      "learning_rate": 9.39714172889033e-06,
+      "loss": 0.3809,
+      "step": 7530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4583981982065293,
+      "learning_rate": 9.396954154550086e-06,
+      "loss": 0.177,
+      "step": 7531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3859443436446126,
+      "learning_rate": 9.396766552905952e-06,
+      "loss": 0.2763,
+      "step": 7532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3263819778094097,
+      "learning_rate": 9.396578923959088e-06,
+      "loss": 0.2341,
+      "step": 7533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.479508025333249,
+      "learning_rate": 9.396391267710662e-06,
+      "loss": 0.356,
+      "step": 7534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.940702715759615,
+      "learning_rate": 9.396203584161839e-06,
+      "loss": 0.331,
+      "step": 7535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.834810704551957,
+      "learning_rate": 9.396015873313781e-06,
+      "loss": 0.3502,
+      "step": 7536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5735949162908904,
+      "learning_rate": 9.39582813516766e-06,
+      "loss": 0.3743,
+      "step": 7537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8815853781545486,
+      "learning_rate": 9.395640369724636e-06,
+      "loss": 0.4496,
+      "step": 7538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6242454122489847,
+      "learning_rate": 9.39545257698588e-06,
+      "loss": 0.253,
+      "step": 7539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.514381857457774,
+      "learning_rate": 9.395264756952555e-06,
+      "loss": 0.3566,
+      "step": 7540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7896919392576824,
+      "learning_rate": 9.395076909625826e-06,
+      "loss": 0.4818,
+      "step": 7541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.98091280772413,
+      "learning_rate": 9.394889035006864e-06,
+      "loss": 0.5999,
+      "step": 7542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5845318329139226,
+      "learning_rate": 9.394701133096831e-06,
+      "loss": 0.3004,
+      "step": 7543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8089048600024684,
+      "learning_rate": 9.394513203896897e-06,
+      "loss": 0.2351,
+      "step": 7544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8467883033219477,
+      "learning_rate": 9.394325247408226e-06,
+      "loss": 0.571,
+      "step": 7545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8485833497440434,
+      "learning_rate": 9.394137263631991e-06,
+      "loss": 0.4458,
+      "step": 7546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7318862680573157,
+      "learning_rate": 9.393949252569352e-06,
+      "loss": 0.2254,
+      "step": 7547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6759763350673869,
+      "learning_rate": 9.39376121422148e-06,
+      "loss": 0.3581,
+      "step": 7548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.491775246039112,
+      "learning_rate": 9.393573148589544e-06,
+      "loss": 0.3016,
+      "step": 7549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6176666572698077,
+      "learning_rate": 9.39338505567471e-06,
+      "loss": 0.3265,
+      "step": 7550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3779872474030146,
+      "learning_rate": 9.393196935478143e-06,
+      "loss": 0.4862,
+      "step": 7551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9667286969957167,
+      "learning_rate": 9.393008788001017e-06,
+      "loss": 0.3013,
+      "step": 7552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4543291891126444,
+      "learning_rate": 9.392820613244499e-06,
+      "loss": 0.2753,
+      "step": 7553
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6097702309557145,
+      "learning_rate": 9.392632411209754e-06,
+      "loss": 0.3463,
+      "step": 7554
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7308128307629973,
+      "learning_rate": 9.392444181897956e-06,
+      "loss": 0.413,
+      "step": 7555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3617175761538955,
+      "learning_rate": 9.392255925310269e-06,
+      "loss": 0.157,
+      "step": 7556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.640504959801196,
+      "learning_rate": 9.392067641447864e-06,
+      "loss": 0.3421,
+      "step": 7557
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6603519257061674,
+      "learning_rate": 9.39187933031191e-06,
+      "loss": 0.3279,
+      "step": 7558
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3948431547754847,
+      "learning_rate": 9.391690991903577e-06,
+      "loss": 0.3655,
+      "step": 7559
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.238052875791593,
+      "learning_rate": 9.391502626224033e-06,
+      "loss": 0.2721,
+      "step": 7560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.285511745612558,
+      "learning_rate": 9.391314233274452e-06,
+      "loss": 0.3988,
+      "step": 7561
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8716089817478325,
+      "learning_rate": 9.391125813055997e-06,
+      "loss": 0.2896,
+      "step": 7562
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5330227885521417,
+      "learning_rate": 9.390937365569841e-06,
+      "loss": 0.3294,
+      "step": 7563
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8806622699994833,
+      "learning_rate": 9.390748890817159e-06,
+      "loss": 0.2413,
+      "step": 7564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4888911211676532,
+      "learning_rate": 9.390560388799114e-06,
+      "loss": 0.1413,
+      "step": 7565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1839593809823303,
+      "learning_rate": 9.39037185951688e-06,
+      "loss": 0.3939,
+      "step": 7566
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1769591827945685,
+      "learning_rate": 9.390183302971629e-06,
+      "loss": 0.1267,
+      "step": 7567
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3203436948946266,
+      "learning_rate": 9.389994719164528e-06,
+      "loss": 0.4869,
+      "step": 7568
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2946654536243698,
+      "learning_rate": 9.389806108096753e-06,
+      "loss": 0.1863,
+      "step": 7569
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0664763087202864,
+      "learning_rate": 9.38961746976947e-06,
+      "loss": 0.4167,
+      "step": 7570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.263406256156836,
+      "learning_rate": 9.389428804183855e-06,
+      "loss": 0.1935,
+      "step": 7571
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4816939070554989,
+      "learning_rate": 9.389240111341076e-06,
+      "loss": 0.2737,
+      "step": 7572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8213878124148486,
+      "learning_rate": 9.389051391242308e-06,
+      "loss": 0.4843,
+      "step": 7573
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1988016108982773,
+      "learning_rate": 9.38886264388872e-06,
+      "loss": 0.4212,
+      "step": 7574
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.793484287485858,
+      "learning_rate": 9.388673869281486e-06,
+      "loss": 0.3709,
+      "step": 7575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.103215386048558,
+      "learning_rate": 9.388485067421777e-06,
+      "loss": 0.113,
+      "step": 7576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.518140573245031,
+      "learning_rate": 9.388296238310765e-06,
+      "loss": 0.3086,
+      "step": 7577
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3918749309870802,
+      "learning_rate": 9.388107381949624e-06,
+      "loss": 0.1578,
+      "step": 7578
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9690047667428433,
+      "learning_rate": 9.387918498339528e-06,
+      "loss": 0.192,
+      "step": 7579
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1371316520372228,
+      "learning_rate": 9.387729587481648e-06,
+      "loss": 0.1836,
+      "step": 7580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4200243106187347,
+      "learning_rate": 9.387540649377156e-06,
+      "loss": 0.3016,
+      "step": 7581
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4197918020329046,
+      "learning_rate": 9.387351684027227e-06,
+      "loss": 0.3011,
+      "step": 7582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.118665784383052,
+      "learning_rate": 9.387162691433034e-06,
+      "loss": 0.4668,
+      "step": 7583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2941222203077554,
+      "learning_rate": 9.386973671595749e-06,
+      "loss": 0.236,
+      "step": 7584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3519616695201273,
+      "learning_rate": 9.38678462451655e-06,
+      "loss": 0.2323,
+      "step": 7585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.596749437774325,
+      "learning_rate": 9.386595550196606e-06,
+      "loss": 0.359,
+      "step": 7586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6519586010524419,
+      "learning_rate": 9.386406448637093e-06,
+      "loss": 0.4854,
+      "step": 7587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9478036104804526,
+      "learning_rate": 9.386217319839188e-06,
+      "loss": 0.5551,
+      "step": 7588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.382981854375511,
+      "learning_rate": 9.386028163804062e-06,
+      "loss": 0.2629,
+      "step": 7589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1747909554608844,
+      "learning_rate": 9.385838980532888e-06,
+      "loss": 0.3364,
+      "step": 7590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5599462597062015,
+      "learning_rate": 9.385649770026847e-06,
+      "loss": 0.2012,
+      "step": 7591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8884637869886955,
+      "learning_rate": 9.38546053228711e-06,
+      "loss": 0.4036,
+      "step": 7592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1029230214534436,
+      "learning_rate": 9.38527126731485e-06,
+      "loss": 0.1679,
+      "step": 7593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.719342892429161,
+      "learning_rate": 9.385081975111248e-06,
+      "loss": 0.353,
+      "step": 7594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7867287694319003,
+      "learning_rate": 9.384892655677473e-06,
+      "loss": 0.5062,
+      "step": 7595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9909933138729818,
+      "learning_rate": 9.384703309014705e-06,
+      "loss": 0.5391,
+      "step": 7596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0109412014681904,
+      "learning_rate": 9.38451393512412e-06,
+      "loss": 0.552,
+      "step": 7597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3931932429349136,
+      "learning_rate": 9.384324534006892e-06,
+      "loss": 0.3648,
+      "step": 7598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8162615400526614,
+      "learning_rate": 9.384135105664195e-06,
+      "loss": 0.3551,
+      "step": 7599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6121544840190625,
+      "learning_rate": 9.383945650097211e-06,
+      "loss": 0.4193,
+      "step": 7600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7056698116623632,
+      "learning_rate": 9.383756167307113e-06,
+      "loss": 0.3731,
+      "step": 7601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7876771577329802,
+      "learning_rate": 9.383566657295078e-06,
+      "loss": 0.553,
+      "step": 7602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0310198648921127,
+      "learning_rate": 9.383377120062283e-06,
+      "loss": 0.2725,
+      "step": 7603
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7585660802629373,
+      "learning_rate": 9.383187555609904e-06,
+      "loss": 0.3747,
+      "step": 7604
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.324046868291843,
+      "learning_rate": 9.38299796393912e-06,
+      "loss": 0.2881,
+      "step": 7605
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8876524312664591,
+      "learning_rate": 9.382808345051107e-06,
+      "loss": 0.1185,
+      "step": 7606
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1697010156603254,
+      "learning_rate": 9.382618698947044e-06,
+      "loss": 0.242,
+      "step": 7607
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7727231338815013,
+      "learning_rate": 9.382429025628106e-06,
+      "loss": 0.4597,
+      "step": 7608
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9607568776339805,
+      "learning_rate": 9.382239325095472e-06,
+      "loss": 0.3076,
+      "step": 7609
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5644616020820679,
+      "learning_rate": 9.38204959735032e-06,
+      "loss": 0.2484,
+      "step": 7610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7060769379359786,
+      "learning_rate": 9.38185984239383e-06,
+      "loss": 0.3977,
+      "step": 7611
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2915361771761282,
+      "learning_rate": 9.381670060227178e-06,
+      "loss": 0.1805,
+      "step": 7612
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6451194418065533,
+      "learning_rate": 9.381480250851545e-06,
+      "loss": 0.4377,
+      "step": 7613
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5339607043634165,
+      "learning_rate": 9.381290414268104e-06,
+      "loss": 0.3749,
+      "step": 7614
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1252150628042232,
+      "learning_rate": 9.381100550478041e-06,
+      "loss": 0.1861,
+      "step": 7615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7747734642199917,
+      "learning_rate": 9.380910659482532e-06,
+      "loss": 0.2711,
+      "step": 7616
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2014982865491397,
+      "learning_rate": 9.380720741282754e-06,
+      "loss": 0.244,
+      "step": 7617
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4992687310439714,
+      "learning_rate": 9.38053079587989e-06,
+      "loss": 0.2125,
+      "step": 7618
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4839744203037313,
+      "learning_rate": 9.380340823275116e-06,
+      "loss": 0.2771,
+      "step": 7619
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.760254470253694,
+      "learning_rate": 9.380150823469614e-06,
+      "loss": 0.2085,
+      "step": 7620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.606601083734043,
+      "learning_rate": 9.379960796464564e-06,
+      "loss": 0.2087,
+      "step": 7621
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4894477097807777,
+      "learning_rate": 9.379770742261146e-06,
+      "loss": 0.3069,
+      "step": 7622
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4355441138831333,
+      "learning_rate": 9.379580660860538e-06,
+      "loss": 0.3273,
+      "step": 7623
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9555587444560134,
+      "learning_rate": 9.379390552263923e-06,
+      "loss": 0.3908,
+      "step": 7624
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8379041669227094,
+      "learning_rate": 9.37920041647248e-06,
+      "loss": 0.3377,
+      "step": 7625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5547283229548112,
+      "learning_rate": 9.379010253487388e-06,
+      "loss": 0.3063,
+      "step": 7626
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.007106163227537,
+      "learning_rate": 9.378820063309833e-06,
+      "loss": 0.5549,
+      "step": 7627
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4111067495612148,
+      "learning_rate": 9.378629845940993e-06,
+      "loss": 0.2328,
+      "step": 7628
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6357645744955331,
+      "learning_rate": 9.378439601382047e-06,
+      "loss": 0.1927,
+      "step": 7629
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0371899611716713,
+      "learning_rate": 9.37824932963418e-06,
+      "loss": 0.5135,
+      "step": 7630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.330301845741755,
+      "learning_rate": 9.378059030698572e-06,
+      "loss": 0.1947,
+      "step": 7631
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.576719777340621,
+      "learning_rate": 9.377868704576403e-06,
+      "loss": 0.3156,
+      "step": 7632
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6697495399471036,
+      "learning_rate": 9.37767835126886e-06,
+      "loss": 0.3276,
+      "step": 7633
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3812695114022036,
+      "learning_rate": 9.377487970777118e-06,
+      "loss": 0.1455,
+      "step": 7634
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1499772227664602,
+      "learning_rate": 9.377297563102366e-06,
+      "loss": 0.1312,
+      "step": 7635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5581764221772298,
+      "learning_rate": 9.377107128245782e-06,
+      "loss": 0.2675,
+      "step": 7636
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.626406125440546,
+      "learning_rate": 9.376916666208549e-06,
+      "loss": 0.5261,
+      "step": 7637
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6168465231815978,
+      "learning_rate": 9.376726176991851e-06,
+      "loss": 0.3952,
+      "step": 7638
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8166880132703513,
+      "learning_rate": 9.376535660596872e-06,
+      "loss": 0.2251,
+      "step": 7639
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.346386437026263,
+      "learning_rate": 9.376345117024791e-06,
+      "loss": 0.1941,
+      "step": 7640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8949989095586595,
+      "learning_rate": 9.376154546276795e-06,
+      "loss": 0.5444,
+      "step": 7641
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8114292192249104,
+      "learning_rate": 9.375963948354068e-06,
+      "loss": 0.2219,
+      "step": 7642
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0011598178247874,
+      "learning_rate": 9.375773323257789e-06,
+      "loss": 0.3129,
+      "step": 7643
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.778728665610168,
+      "learning_rate": 9.375582670989144e-06,
+      "loss": 0.226,
+      "step": 7644
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3240529001516832,
+      "learning_rate": 9.37539199154932e-06,
+      "loss": 0.2374,
+      "step": 7645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4829343078475385,
+      "learning_rate": 9.375201284939497e-06,
+      "loss": 0.3138,
+      "step": 7646
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7598664977771028,
+      "learning_rate": 9.37501055116086e-06,
+      "loss": 0.4988,
+      "step": 7647
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8057108780696063,
+      "learning_rate": 9.374819790214596e-06,
+      "loss": 0.2192,
+      "step": 7648
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4153950666591746,
+      "learning_rate": 9.374629002101884e-06,
+      "loss": 0.4649,
+      "step": 7649
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6983610509741947,
+      "learning_rate": 9.374438186823915e-06,
+      "loss": 0.2454,
+      "step": 7650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8574579518181005,
+      "learning_rate": 9.374247344381871e-06,
+      "loss": 0.5156,
+      "step": 7651
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4204037122950726,
+      "learning_rate": 9.374056474776938e-06,
+      "loss": 0.3747,
+      "step": 7652
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4830253940656926,
+      "learning_rate": 9.373865578010299e-06,
+      "loss": 0.2056,
+      "step": 7653
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1285447924664969,
+      "learning_rate": 9.373674654083142e-06,
+      "loss": 0.2314,
+      "step": 7654
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1676126985162771,
+      "learning_rate": 9.373483702996653e-06,
+      "loss": 0.1752,
+      "step": 7655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5374572844268997,
+      "learning_rate": 9.373292724752014e-06,
+      "loss": 0.5659,
+      "step": 7656
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6314840091375558,
+      "learning_rate": 9.373101719350413e-06,
+      "loss": 0.2455,
+      "step": 7657
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.064103219323881,
+      "learning_rate": 9.372910686793037e-06,
+      "loss": 0.2973,
+      "step": 7658
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4971733555476958,
+      "learning_rate": 9.372719627081072e-06,
+      "loss": 0.3256,
+      "step": 7659
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3389323473989114,
+      "learning_rate": 9.372528540215705e-06,
+      "loss": 0.2553,
+      "step": 7660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7068020581646768,
+      "learning_rate": 9.372337426198119e-06,
+      "loss": 0.4087,
+      "step": 7661
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.681813191314275,
+      "learning_rate": 9.372146285029504e-06,
+      "loss": 0.4583,
+      "step": 7662
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3848191992076901,
+      "learning_rate": 9.371955116711048e-06,
+      "loss": 0.2884,
+      "step": 7663
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.710134313345962,
+      "learning_rate": 9.371763921243936e-06,
+      "loss": 0.6304,
+      "step": 7664
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.383250476960659,
+      "learning_rate": 9.371572698629355e-06,
+      "loss": 0.3325,
+      "step": 7665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.055829896023414,
+      "learning_rate": 9.371381448868494e-06,
+      "loss": 0.3251,
+      "step": 7666
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6464631338321842,
+      "learning_rate": 9.371190171962539e-06,
+      "loss": 0.3421,
+      "step": 7667
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7134855829112554,
+      "learning_rate": 9.370998867912679e-06,
+      "loss": 0.3495,
+      "step": 7668
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6070283011639828,
+      "learning_rate": 9.370807536720102e-06,
+      "loss": 0.5096,
+      "step": 7669
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0542637070706546,
+      "learning_rate": 9.370616178385994e-06,
+      "loss": 0.2218,
+      "step": 7670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5795299772211597,
+      "learning_rate": 9.370424792911545e-06,
+      "loss": 0.5117,
+      "step": 7671
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3543823846757401,
+      "learning_rate": 9.370233380297945e-06,
+      "loss": 0.187,
+      "step": 7672
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.106125621282593,
+      "learning_rate": 9.37004194054638e-06,
+      "loss": 0.2757,
+      "step": 7673
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4529110830525813,
+      "learning_rate": 9.36985047365804e-06,
+      "loss": 0.2687,
+      "step": 7674
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6257496695542355,
+      "learning_rate": 9.369658979634113e-06,
+      "loss": 0.3781,
+      "step": 7675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.047733680049974,
+      "learning_rate": 9.369467458475789e-06,
+      "loss": 0.6485,
+      "step": 7676
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4117619085654878,
+      "learning_rate": 9.369275910184256e-06,
+      "loss": 0.1885,
+      "step": 7677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7141544329557972,
+      "learning_rate": 9.369084334760707e-06,
+      "loss": 0.1954,
+      "step": 7678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.48313569485023,
+      "learning_rate": 9.36889273220633e-06,
+      "loss": 0.2788,
+      "step": 7679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5001424297124875,
+      "learning_rate": 9.36870110252231e-06,
+      "loss": 0.1913,
+      "step": 7680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9411932772788687,
+      "learning_rate": 9.368509445709843e-06,
+      "loss": 0.3956,
+      "step": 7681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3158483879646679,
+      "learning_rate": 9.368317761770117e-06,
+      "loss": 0.2614,
+      "step": 7682
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.767821325779436,
+      "learning_rate": 9.368126050704324e-06,
+      "loss": 0.3578,
+      "step": 7683
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1131494396373744,
+      "learning_rate": 9.367934312513651e-06,
+      "loss": 0.3525,
+      "step": 7684
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2133993166275703,
+      "learning_rate": 9.367742547199293e-06,
+      "loss": 0.2048,
+      "step": 7685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3399927540216678,
+      "learning_rate": 9.367550754762436e-06,
+      "loss": 0.3522,
+      "step": 7686
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9371931348331946,
+      "learning_rate": 9.367358935204274e-06,
+      "loss": 0.4226,
+      "step": 7687
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.576830400705117,
+      "learning_rate": 9.367167088525997e-06,
+      "loss": 0.3711,
+      "step": 7688
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6413034262535815,
+      "learning_rate": 9.366975214728797e-06,
+      "loss": 0.4653,
+      "step": 7689
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6771597894178003,
+      "learning_rate": 9.366783313813865e-06,
+      "loss": 0.5056,
+      "step": 7690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.812010847053945,
+      "learning_rate": 9.366591385782394e-06,
+      "loss": 0.4714,
+      "step": 7691
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3329440147028564,
+      "learning_rate": 9.366399430635573e-06,
+      "loss": 0.3004,
+      "step": 7692
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7854803446390488,
+      "learning_rate": 9.366207448374598e-06,
+      "loss": 0.2422,
+      "step": 7693
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9259870778984267,
+      "learning_rate": 9.366015439000657e-06,
+      "loss": 0.4177,
+      "step": 7694
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1083940346248076,
+      "learning_rate": 9.365823402514945e-06,
+      "loss": 0.2176,
+      "step": 7695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9040648196732333,
+      "learning_rate": 9.365631338918654e-06,
+      "loss": 0.184,
+      "step": 7696
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.888291412345109,
+      "learning_rate": 9.365439248212975e-06,
+      "loss": 0.5716,
+      "step": 7697
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8193525435405984,
+      "learning_rate": 9.365247130399104e-06,
+      "loss": 0.3892,
+      "step": 7698
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9508513911985468,
+      "learning_rate": 9.365054985478233e-06,
+      "loss": 0.5544,
+      "step": 7699
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0009879637393855,
+      "learning_rate": 9.364862813451552e-06,
+      "loss": 0.4179,
+      "step": 7700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.978612059552391,
+      "learning_rate": 9.364670614320257e-06,
+      "loss": 0.3655,
+      "step": 7701
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4482685325937517,
+      "learning_rate": 9.36447838808554e-06,
+      "loss": 0.3489,
+      "step": 7702
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2745051973788337,
+      "learning_rate": 9.3642861347486e-06,
+      "loss": 0.4377,
+      "step": 7703
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8050176723043063,
+      "learning_rate": 9.364093854310624e-06,
+      "loss": 0.4697,
+      "step": 7704
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3840945356920205,
+      "learning_rate": 9.36390154677281e-06,
+      "loss": 0.335,
+      "step": 7705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5580751971415627,
+      "learning_rate": 9.36370921213635e-06,
+      "loss": 0.3069,
+      "step": 7706
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7366824560924785,
+      "learning_rate": 9.363516850402439e-06,
+      "loss": 0.2288,
+      "step": 7707
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.498108476493761,
+      "learning_rate": 9.363324461572271e-06,
+      "loss": 0.1958,
+      "step": 7708
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9467616084189807,
+      "learning_rate": 9.363132045647044e-06,
+      "loss": 0.2045,
+      "step": 7709
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.936908543201802,
+      "learning_rate": 9.362939602627948e-06,
+      "loss": 0.4409,
+      "step": 7710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.469494912277758,
+      "learning_rate": 9.362747132516181e-06,
+      "loss": 0.1899,
+      "step": 7711
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.562205432305123,
+      "learning_rate": 9.362554635312939e-06,
+      "loss": 0.4033,
+      "step": 7712
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5857416064856504,
+      "learning_rate": 9.362362111019414e-06,
+      "loss": 0.1681,
+      "step": 7713
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.223626329012982,
+      "learning_rate": 9.362169559636805e-06,
+      "loss": 0.4565,
+      "step": 7714
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5393415575657912,
+      "learning_rate": 9.361976981166304e-06,
+      "loss": 0.4354,
+      "step": 7715
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7426519544602026,
+      "learning_rate": 9.36178437560911e-06,
+      "loss": 0.4081,
+      "step": 7716
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9347694714248225,
+      "learning_rate": 9.361591742966419e-06,
+      "loss": 0.3894,
+      "step": 7717
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4660927330675433,
+      "learning_rate": 9.361399083239424e-06,
+      "loss": 0.1472,
+      "step": 7718
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1840851594189132,
+      "learning_rate": 9.361206396429325e-06,
+      "loss": 0.6662,
+      "step": 7719
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6269798688762678,
+      "learning_rate": 9.361013682537317e-06,
+      "loss": 0.3222,
+      "step": 7720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0215654458269072,
+      "learning_rate": 9.360820941564597e-06,
+      "loss": 0.3131,
+      "step": 7721
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.780304609769693,
+      "learning_rate": 9.360628173512361e-06,
+      "loss": 0.2755,
+      "step": 7722
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7867915522309417,
+      "learning_rate": 9.360435378381806e-06,
+      "loss": 0.2508,
+      "step": 7723
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6498425403153032,
+      "learning_rate": 9.360242556174131e-06,
+      "loss": 0.2362,
+      "step": 7724
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2642133789233645,
+      "learning_rate": 9.360049706890532e-06,
+      "loss": 0.1727,
+      "step": 7725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8818712622153422,
+      "learning_rate": 9.359856830532208e-06,
+      "loss": 0.2414,
+      "step": 7726
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6521371217361116,
+      "learning_rate": 9.359663927100354e-06,
+      "loss": 0.2467,
+      "step": 7727
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8106062137628605,
+      "learning_rate": 9.359470996596169e-06,
+      "loss": 0.2012,
+      "step": 7728
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9991569418556017,
+      "learning_rate": 9.359278039020853e-06,
+      "loss": 0.4276,
+      "step": 7729
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7594823916302949,
+      "learning_rate": 9.359085054375602e-06,
+      "loss": 0.4976,
+      "step": 7730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2530717791459618,
+      "learning_rate": 9.358892042661617e-06,
+      "loss": 0.2263,
+      "step": 7731
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8186413518662263,
+      "learning_rate": 9.358699003880092e-06,
+      "loss": 0.215,
+      "step": 7732
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5066565346860676,
+      "learning_rate": 9.358505938032227e-06,
+      "loss": 0.5023,
+      "step": 7733
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6647301485210235,
+      "learning_rate": 9.358312845119225e-06,
+      "loss": 0.2774,
+      "step": 7734
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0295310939268225,
+      "learning_rate": 9.35811972514228e-06,
+      "loss": 0.161,
+      "step": 7735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.212257293615508,
+      "learning_rate": 9.357926578102597e-06,
+      "loss": 0.1493,
+      "step": 7736
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4512071066131182,
+      "learning_rate": 9.357733404001369e-06,
+      "loss": 0.218,
+      "step": 7737
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1340315622066575,
+      "learning_rate": 9.357540202839799e-06,
+      "loss": 0.3601,
+      "step": 7738
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2276012204622704,
+      "learning_rate": 9.357346974619086e-06,
+      "loss": 0.2263,
+      "step": 7739
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.460387171413195,
+      "learning_rate": 9.357153719340429e-06,
+      "loss": 0.3226,
+      "step": 7740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.569863010514534,
+      "learning_rate": 9.356960437005032e-06,
+      "loss": 0.336,
+      "step": 7741
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.205721938991786,
+      "learning_rate": 9.35676712761409e-06,
+      "loss": 0.1714,
+      "step": 7742
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7936623764982886,
+      "learning_rate": 9.356573791168806e-06,
+      "loss": 0.2258,
+      "step": 7743
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.459085609059396,
+      "learning_rate": 9.35638042767038e-06,
+      "loss": 0.3789,
+      "step": 7744
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8236439675444156,
+      "learning_rate": 9.356187037120014e-06,
+      "loss": 0.1738,
+      "step": 7745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9165879521498457,
+      "learning_rate": 9.355993619518909e-06,
+      "loss": 0.266,
+      "step": 7746
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6024140768924757,
+      "learning_rate": 9.35580017486826e-06,
+      "loss": 0.3994,
+      "step": 7747
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.696667610846524,
+      "learning_rate": 9.355606703169279e-06,
+      "loss": 0.2499,
+      "step": 7748
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8254265350460206,
+      "learning_rate": 9.35541320442316e-06,
+      "loss": 0.4502,
+      "step": 7749
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7006430433522504,
+      "learning_rate": 9.355219678631105e-06,
+      "loss": 0.2456,
+      "step": 7750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8529146625098212,
+      "learning_rate": 9.355026125794317e-06,
+      "loss": 0.1971,
+      "step": 7751
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.215947370516073,
+      "learning_rate": 9.354832545914e-06,
+      "loss": 0.1168,
+      "step": 7752
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0982421146388048,
+      "learning_rate": 9.354638938991351e-06,
+      "loss": 0.289,
+      "step": 7753
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5363646500254131,
+      "learning_rate": 9.354445305027577e-06,
+      "loss": 0.363,
+      "step": 7754
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7675566909080584,
+      "learning_rate": 9.35425164402388e-06,
+      "loss": 0.2949,
+      "step": 7755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8366035916968142,
+      "learning_rate": 9.354057955981459e-06,
+      "loss": 0.5914,
+      "step": 7756
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8762380645087682,
+      "learning_rate": 9.35386424090152e-06,
+      "loss": 0.4578,
+      "step": 7757
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4578313599375388,
+      "learning_rate": 9.353670498785264e-06,
+      "loss": 0.3492,
+      "step": 7758
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.649791866237768,
+      "learning_rate": 9.353476729633895e-06,
+      "loss": 0.32,
+      "step": 7759
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.020797151965921,
+      "learning_rate": 9.353282933448617e-06,
+      "loss": 0.5891,
+      "step": 7760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5234131542118923,
+      "learning_rate": 9.35308911023063e-06,
+      "loss": 0.2466,
+      "step": 7761
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1108167025492721,
+      "learning_rate": 9.352895259981143e-06,
+      "loss": 0.1729,
+      "step": 7762
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.104939771173409,
+      "learning_rate": 9.352701382701356e-06,
+      "loss": 0.495,
+      "step": 7763
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5463085052397312,
+      "learning_rate": 9.352507478392477e-06,
+      "loss": 0.3857,
+      "step": 7764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5576262853470344,
+      "learning_rate": 9.352313547055703e-06,
+      "loss": 0.2337,
+      "step": 7765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.52735042814468,
+      "learning_rate": 9.352119588692244e-06,
+      "loss": 0.3833,
+      "step": 7766
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1916947500158692,
+      "learning_rate": 9.351925603303302e-06,
+      "loss": 0.1846,
+      "step": 7767
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9310872179086491,
+      "learning_rate": 9.351731590890081e-06,
+      "loss": 0.6213,
+      "step": 7768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9448884632054788,
+      "learning_rate": 9.35153755145379e-06,
+      "loss": 0.5718,
+      "step": 7769
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.697267014356951,
+      "learning_rate": 9.35134348499563e-06,
+      "loss": 0.2095,
+      "step": 7770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.157732100661819,
+      "learning_rate": 9.351149391516806e-06,
+      "loss": 0.1459,
+      "step": 7771
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5986555117683279,
+      "learning_rate": 9.350955271018525e-06,
+      "loss": 0.1754,
+      "step": 7772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.033603204956406,
+      "learning_rate": 9.350761123501992e-06,
+      "loss": 0.1529,
+      "step": 7773
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.203195490877588,
+      "learning_rate": 9.350566948968411e-06,
+      "loss": 0.2201,
+      "step": 7774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.095431658831424,
+      "learning_rate": 9.35037274741899e-06,
+      "loss": 0.2104,
+      "step": 7775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5167906889493372,
+      "learning_rate": 9.350178518854935e-06,
+      "loss": 0.2509,
+      "step": 7776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8767619596338834,
+      "learning_rate": 9.34998426327745e-06,
+      "loss": 0.4266,
+      "step": 7777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4344855245173647,
+      "learning_rate": 9.349789980687744e-06,
+      "loss": 0.2977,
+      "step": 7778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4086443637020858,
+      "learning_rate": 9.34959567108702e-06,
+      "loss": 0.1993,
+      "step": 7779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.533708521972058,
+      "learning_rate": 9.349401334476489e-06,
+      "loss": 0.3341,
+      "step": 7780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5280069584491451,
+      "learning_rate": 9.349206970857353e-06,
+      "loss": 0.1543,
+      "step": 7781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.57144461573078,
+      "learning_rate": 9.349012580230822e-06,
+      "loss": 0.4405,
+      "step": 7782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5028380724848518,
+      "learning_rate": 9.348818162598102e-06,
+      "loss": 0.3086,
+      "step": 7783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8297663802776536,
+      "learning_rate": 9.348623717960402e-06,
+      "loss": 0.3612,
+      "step": 7784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.022101871882007,
+      "learning_rate": 9.348429246318927e-06,
+      "loss": 0.2682,
+      "step": 7785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9558962235466117,
+      "learning_rate": 9.348234747674886e-06,
+      "loss": 0.5902,
+      "step": 7786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2285308382096054,
+      "learning_rate": 9.348040222029486e-06,
+      "loss": 0.2196,
+      "step": 7787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.394836675498425,
+      "learning_rate": 9.347845669383936e-06,
+      "loss": 0.1784,
+      "step": 7788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.649345700308015,
+      "learning_rate": 9.347651089739442e-06,
+      "loss": 0.2223,
+      "step": 7789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8665651816606377,
+      "learning_rate": 9.347456483097216e-06,
+      "loss": 0.382,
+      "step": 7790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3523097133016921,
+      "learning_rate": 9.347261849458464e-06,
+      "loss": 0.147,
+      "step": 7791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6214265389858078,
+      "learning_rate": 9.347067188824395e-06,
+      "loss": 0.4126,
+      "step": 7792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.624517987132842,
+      "learning_rate": 9.346872501196216e-06,
+      "loss": 0.3242,
+      "step": 7793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.83661317160849,
+      "learning_rate": 9.34667778657514e-06,
+      "loss": 0.3361,
+      "step": 7794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8779229507682524,
+      "learning_rate": 9.346483044962372e-06,
+      "loss": 0.3644,
+      "step": 7795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5462393538394559,
+      "learning_rate": 9.346288276359123e-06,
+      "loss": 0.1965,
+      "step": 7796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9621486455234062,
+      "learning_rate": 9.346093480766605e-06,
+      "loss": 0.3114,
+      "step": 7797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7209602990795396,
+      "learning_rate": 9.345898658186023e-06,
+      "loss": 0.2019,
+      "step": 7798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4487756941280225,
+      "learning_rate": 9.34570380861859e-06,
+      "loss": 0.1303,
+      "step": 7799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7528943199053826,
+      "learning_rate": 9.345508932065513e-06,
+      "loss": 0.4871,
+      "step": 7800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0033515028465474,
+      "learning_rate": 9.345314028528005e-06,
+      "loss": 0.2539,
+      "step": 7801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5540723709945115,
+      "learning_rate": 9.345119098007277e-06,
+      "loss": 0.3606,
+      "step": 7802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.47939024218341,
+      "learning_rate": 9.344924140504536e-06,
+      "loss": 0.428,
+      "step": 7803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.162222954886244,
+      "learning_rate": 9.344729156020995e-06,
+      "loss": 0.4382,
+      "step": 7804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5757683588426432,
+      "learning_rate": 9.344534144557865e-06,
+      "loss": 0.3147,
+      "step": 7805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9927965925367506,
+      "learning_rate": 9.344339106116355e-06,
+      "loss": 0.296,
+      "step": 7806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9030650805920282,
+      "learning_rate": 9.34414404069768e-06,
+      "loss": 0.4222,
+      "step": 7807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.588942677402,
+      "learning_rate": 9.343948948303046e-06,
+      "loss": 0.4791,
+      "step": 7808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7099454586040859,
+      "learning_rate": 9.343753828933668e-06,
+      "loss": 0.3202,
+      "step": 7809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2572344301625609,
+      "learning_rate": 9.343558682590757e-06,
+      "loss": 0.1916,
+      "step": 7810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5628624066639731,
+      "learning_rate": 9.343363509275522e-06,
+      "loss": 0.2452,
+      "step": 7811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8547983906093548,
+      "learning_rate": 9.34316830898918e-06,
+      "loss": 0.5649,
+      "step": 7812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.535948599836381,
+      "learning_rate": 9.342973081732943e-06,
+      "loss": 0.3192,
+      "step": 7813
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.098034801459068,
+      "learning_rate": 9.342777827508018e-06,
+      "loss": 0.4071,
+      "step": 7814
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.81741372337886,
+      "learning_rate": 9.342582546315619e-06,
+      "loss": 0.356,
+      "step": 7815
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.603208786650625,
+      "learning_rate": 9.342387238156962e-06,
+      "loss": 0.3319,
+      "step": 7816
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1514735808105705,
+      "learning_rate": 9.34219190303326e-06,
+      "loss": 0.5712,
+      "step": 7817
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4306309875206331,
+      "learning_rate": 9.341996540945722e-06,
+      "loss": 0.3005,
+      "step": 7818
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5297991961709447,
+      "learning_rate": 9.341801151895563e-06,
+      "loss": 0.2517,
+      "step": 7819
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6071233159250986,
+      "learning_rate": 9.341605735883996e-06,
+      "loss": 0.2235,
+      "step": 7820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2715904628133405,
+      "learning_rate": 9.341410292912236e-06,
+      "loss": 0.3875,
+      "step": 7821
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9320834889350849,
+      "learning_rate": 9.341214822981494e-06,
+      "loss": 0.3155,
+      "step": 7822
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6073818703772185,
+      "learning_rate": 9.341019326092986e-06,
+      "loss": 0.3285,
+      "step": 7823
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5283264452533953,
+      "learning_rate": 9.340823802247925e-06,
+      "loss": 0.2632,
+      "step": 7824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8060814853122746,
+      "learning_rate": 9.340628251447527e-06,
+      "loss": 0.4511,
+      "step": 7825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.404089720067573,
+      "learning_rate": 9.340432673693003e-06,
+      "loss": 0.2192,
+      "step": 7826
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8337382434787945,
+      "learning_rate": 9.340237068985569e-06,
+      "loss": 0.3523,
+      "step": 7827
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8299814705581565,
+      "learning_rate": 9.340041437326442e-06,
+      "loss": 0.426,
+      "step": 7828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7232593682954667,
+      "learning_rate": 9.339845778716834e-06,
+      "loss": 0.5503,
+      "step": 7829
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0453974071623517,
+      "learning_rate": 9.339650093157959e-06,
+      "loss": 0.234,
+      "step": 7830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1914794008279865,
+      "learning_rate": 9.339454380651037e-06,
+      "loss": 0.2212,
+      "step": 7831
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9378897444129608,
+      "learning_rate": 9.339258641197277e-06,
+      "loss": 0.3045,
+      "step": 7832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.508375413603602,
+      "learning_rate": 9.339062874797899e-06,
+      "loss": 0.3457,
+      "step": 7833
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5433983506533857,
+      "learning_rate": 9.338867081454116e-06,
+      "loss": 0.3282,
+      "step": 7834
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7432008199796634,
+      "learning_rate": 9.338671261167147e-06,
+      "loss": 0.4877,
+      "step": 7835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6449479619414609,
+      "learning_rate": 9.338475413938207e-06,
+      "loss": 0.3871,
+      "step": 7836
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.204044163826549,
+      "learning_rate": 9.33827953976851e-06,
+      "loss": 0.2382,
+      "step": 7837
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6603535344204459,
+      "learning_rate": 9.338083638659273e-06,
+      "loss": 0.4245,
+      "step": 7838
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4485218041179504,
+      "learning_rate": 9.337887710611715e-06,
+      "loss": 0.2057,
+      "step": 7839
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0097552670878276,
+      "learning_rate": 9.33769175562705e-06,
+      "loss": 0.4227,
+      "step": 7840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.171483309848975,
+      "learning_rate": 9.337495773706494e-06,
+      "loss": 0.3937,
+      "step": 7841
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7619942980368504,
+      "learning_rate": 9.337299764851266e-06,
+      "loss": 0.268,
+      "step": 7842
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.60466752513462,
+      "learning_rate": 9.337103729062585e-06,
+      "loss": 0.2715,
+      "step": 7843
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6423592291182267,
+      "learning_rate": 9.336907666341664e-06,
+      "loss": 0.2551,
+      "step": 7844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.972780918676453,
+      "learning_rate": 9.336711576689722e-06,
+      "loss": 0.5145,
+      "step": 7845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.191094760355145,
+      "learning_rate": 9.33651546010798e-06,
+      "loss": 0.2036,
+      "step": 7846
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6912434507302747,
+      "learning_rate": 9.336319316597652e-06,
+      "loss": 0.188,
+      "step": 7847
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4795469519422213,
+      "learning_rate": 9.336123146159956e-06,
+      "loss": 0.3052,
+      "step": 7848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3474637465925305,
+      "learning_rate": 9.335926948796112e-06,
+      "loss": 0.2052,
+      "step": 7849
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6958967687503486,
+      "learning_rate": 9.335730724507337e-06,
+      "loss": 0.4762,
+      "step": 7850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8329565277810742,
+      "learning_rate": 9.335534473294852e-06,
+      "loss": 0.4961,
+      "step": 7851
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7950539815311495,
+      "learning_rate": 9.335338195159872e-06,
+      "loss": 0.4785,
+      "step": 7852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6850672776632132,
+      "learning_rate": 9.335141890103619e-06,
+      "loss": 0.3904,
+      "step": 7853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.425054558328594,
+      "learning_rate": 9.334945558127308e-06,
+      "loss": 0.2319,
+      "step": 7854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2396651876762501,
+      "learning_rate": 9.334749199232162e-06,
+      "loss": 0.2443,
+      "step": 7855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6835734649374348,
+      "learning_rate": 9.3345528134194e-06,
+      "loss": 0.2278,
+      "step": 7856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.853231438415306,
+      "learning_rate": 9.33435640069024e-06,
+      "loss": 0.461,
+      "step": 7857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.083214024862704,
+      "learning_rate": 9.334159961045902e-06,
+      "loss": 0.3796,
+      "step": 7858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1507478048354844,
+      "learning_rate": 9.333963494487607e-06,
+      "loss": 0.8365,
+      "step": 7859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.476519066181886,
+      "learning_rate": 9.333767001016573e-06,
+      "loss": 0.1871,
+      "step": 7860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.598630416952213,
+      "learning_rate": 9.333570480634022e-06,
+      "loss": 0.3776,
+      "step": 7861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.213990456113996,
+      "learning_rate": 9.333373933341174e-06,
+      "loss": 0.1603,
+      "step": 7862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6068798926719268,
+      "learning_rate": 9.333177359139247e-06,
+      "loss": 0.3279,
+      "step": 7863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2717055173842329,
+      "learning_rate": 9.332980758029467e-06,
+      "loss": 0.134,
+      "step": 7864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4890351413543144,
+      "learning_rate": 9.332784130013048e-06,
+      "loss": 0.337,
+      "step": 7865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5084770645459635,
+      "learning_rate": 9.332587475091218e-06,
+      "loss": 0.2746,
+      "step": 7866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3374234292804736,
+      "learning_rate": 9.332390793265193e-06,
+      "loss": 0.5039,
+      "step": 7867
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2111582708612825,
+      "learning_rate": 9.332194084536198e-06,
+      "loss": 0.2003,
+      "step": 7868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.272099602454468,
+      "learning_rate": 9.331997348905451e-06,
+      "loss": 0.2709,
+      "step": 7869
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8520718730969579,
+      "learning_rate": 9.331800586374175e-06,
+      "loss": 0.209,
+      "step": 7870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.628667701891145,
+      "learning_rate": 9.331603796943592e-06,
+      "loss": 0.2219,
+      "step": 7871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6661556985516734,
+      "learning_rate": 9.331406980614925e-06,
+      "loss": 0.4509,
+      "step": 7872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5735054151089112,
+      "learning_rate": 9.331210137389395e-06,
+      "loss": 0.3453,
+      "step": 7873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6947566892402133,
+      "learning_rate": 9.331013267268226e-06,
+      "loss": 0.2951,
+      "step": 7874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7352131502319204,
+      "learning_rate": 9.330816370252637e-06,
+      "loss": 0.2376,
+      "step": 7875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8394474242505423,
+      "learning_rate": 9.330619446343854e-06,
+      "loss": 0.305,
+      "step": 7876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.42170148658617,
+      "learning_rate": 9.330422495543098e-06,
+      "loss": 0.3037,
+      "step": 7877
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8839068180693053,
+      "learning_rate": 9.330225517851594e-06,
+      "loss": 0.7716,
+      "step": 7878
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1015529809615665,
+      "learning_rate": 9.330028513270564e-06,
+      "loss": 0.6101,
+      "step": 7879
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4045345280386548,
+      "learning_rate": 9.32983148180123e-06,
+      "loss": 0.1845,
+      "step": 7880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.279802022314513,
+      "learning_rate": 9.329634423444819e-06,
+      "loss": 0.2506,
+      "step": 7881
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1419138387465106,
+      "learning_rate": 9.329437338202552e-06,
+      "loss": 0.2646,
+      "step": 7882
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8950907362839784,
+      "learning_rate": 9.329240226075651e-06,
+      "loss": 0.6007,
+      "step": 7883
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.950877899690586,
+      "learning_rate": 9.329043087065344e-06,
+      "loss": 0.4436,
+      "step": 7884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6569582576403319,
+      "learning_rate": 9.328845921172853e-06,
+      "loss": 0.4119,
+      "step": 7885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5697288519811885,
+      "learning_rate": 9.328648728399404e-06,
+      "loss": 0.3959,
+      "step": 7886
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9603046216643738,
+      "learning_rate": 9.32845150874622e-06,
+      "loss": 0.4011,
+      "step": 7887
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.136077423716949,
+      "learning_rate": 9.328254262214525e-06,
+      "loss": 0.2212,
+      "step": 7888
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.491591639159179,
+      "learning_rate": 9.328056988805544e-06,
+      "loss": 0.2083,
+      "step": 7889
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3843702039570969,
+      "learning_rate": 9.327859688520505e-06,
+      "loss": 0.1703,
+      "step": 7890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5179463974371554,
+      "learning_rate": 9.32766236136063e-06,
+      "loss": 0.3675,
+      "step": 7891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.489449798216107,
+      "learning_rate": 9.327465007327146e-06,
+      "loss": 0.3056,
+      "step": 7892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2522634465308533,
+      "learning_rate": 9.327267626421277e-06,
+      "loss": 0.2123,
+      "step": 7893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7233135759858504,
+      "learning_rate": 9.327070218644251e-06,
+      "loss": 0.5174,
+      "step": 7894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1284279082127524,
+      "learning_rate": 9.326872783997291e-06,
+      "loss": 0.1823,
+      "step": 7895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6328215028432487,
+      "learning_rate": 9.326675322481626e-06,
+      "loss": 0.2721,
+      "step": 7896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1202770930168247,
+      "learning_rate": 9.32647783409848e-06,
+      "loss": 0.533,
+      "step": 7897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5499699377694014,
+      "learning_rate": 9.32628031884908e-06,
+      "loss": 0.3293,
+      "step": 7898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4340741134908148,
+      "learning_rate": 9.326082776734653e-06,
+      "loss": 0.4369,
+      "step": 7899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4319362387222663,
+      "learning_rate": 9.325885207756425e-06,
+      "loss": 0.2051,
+      "step": 7900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.082115143021536,
+      "learning_rate": 9.325687611915621e-06,
+      "loss": 0.7241,
+      "step": 7901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1875002219488537,
+      "learning_rate": 9.325489989213473e-06,
+      "loss": 0.1312,
+      "step": 7902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2989640458770884,
+      "learning_rate": 9.325292339651205e-06,
+      "loss": 0.1652,
+      "step": 7903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4274784046542164,
+      "learning_rate": 9.325094663230044e-06,
+      "loss": 0.1889,
+      "step": 7904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9709525125396605,
+      "learning_rate": 9.324896959951218e-06,
+      "loss": 0.7586,
+      "step": 7905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0549414590168418,
+      "learning_rate": 9.324699229815954e-06,
+      "loss": 0.4659,
+      "step": 7906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1788908175868773,
+      "learning_rate": 9.324501472825482e-06,
+      "loss": 0.4843,
+      "step": 7907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6940716364180868,
+      "learning_rate": 9.324303688981027e-06,
+      "loss": 0.4335,
+      "step": 7908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7403750421633226,
+      "learning_rate": 9.32410587828382e-06,
+      "loss": 0.3483,
+      "step": 7909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2536334152429431,
+      "learning_rate": 9.32390804073509e-06,
+      "loss": 0.2557,
+      "step": 7910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.511992198155568,
+      "learning_rate": 9.32371017633606e-06,
+      "loss": 0.3362,
+      "step": 7911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2112727972486805,
+      "learning_rate": 9.323512285087965e-06,
+      "loss": 0.2275,
+      "step": 7912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8320249219837992,
+      "learning_rate": 9.323314366992031e-06,
+      "loss": 0.3391,
+      "step": 7913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2548686704857115,
+      "learning_rate": 9.323116422049488e-06,
+      "loss": 0.5698,
+      "step": 7914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.756980436276404,
+      "learning_rate": 9.322918450261562e-06,
+      "loss": 0.2801,
+      "step": 7915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.286353539112414,
+      "learning_rate": 9.322720451629488e-06,
+      "loss": 0.2483,
+      "step": 7916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8354357987603593,
+      "learning_rate": 9.32252242615449e-06,
+      "loss": 0.4274,
+      "step": 7917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4947864632621168,
+      "learning_rate": 9.322324373837801e-06,
+      "loss": 0.4492,
+      "step": 7918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6214832849023326,
+      "learning_rate": 9.32212629468065e-06,
+      "loss": 0.3496,
+      "step": 7919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0940451093770487,
+      "learning_rate": 9.321928188684266e-06,
+      "loss": 0.3824,
+      "step": 7920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5837628236667707,
+      "learning_rate": 9.321730055849881e-06,
+      "loss": 0.2416,
+      "step": 7921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.759925863039046,
+      "learning_rate": 9.321531896178724e-06,
+      "loss": 0.201,
+      "step": 7922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4878166586201444,
+      "learning_rate": 9.321333709672025e-06,
+      "loss": 0.185,
+      "step": 7923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9269770803456912,
+      "learning_rate": 9.321135496331018e-06,
+      "loss": 0.6111,
+      "step": 7924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6747389799969248,
+      "learning_rate": 9.32093725615693e-06,
+      "loss": 0.3558,
+      "step": 7925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3178925902912633,
+      "learning_rate": 9.320738989150994e-06,
+      "loss": 0.1865,
+      "step": 7926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5838230958763826,
+      "learning_rate": 9.32054069531444e-06,
+      "loss": 0.4668,
+      "step": 7927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6150454741680658,
+      "learning_rate": 9.320342374648501e-06,
+      "loss": 0.3137,
+      "step": 7928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8786507668750758,
+      "learning_rate": 9.320144027154407e-06,
+      "loss": 0.3078,
+      "step": 7929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.751997654751891,
+      "learning_rate": 9.31994565283339e-06,
+      "loss": 0.5482,
+      "step": 7930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.922208501210821,
+      "learning_rate": 9.319747251686682e-06,
+      "loss": 0.4279,
+      "step": 7931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.508654072966495,
+      "learning_rate": 9.319548823715517e-06,
+      "loss": 0.2155,
+      "step": 7932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.809684589234114,
+      "learning_rate": 9.319350368921123e-06,
+      "loss": 0.4692,
+      "step": 7933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.082287942546843,
+      "learning_rate": 9.319151887304735e-06,
+      "loss": 0.5259,
+      "step": 7934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4349547040624033,
+      "learning_rate": 9.318953378867586e-06,
+      "loss": 0.3281,
+      "step": 7935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6998407453210103,
+      "learning_rate": 9.318754843610906e-06,
+      "loss": 0.435,
+      "step": 7936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6418138919286975,
+      "learning_rate": 9.318556281535932e-06,
+      "loss": 0.2958,
+      "step": 7937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.980477779468434,
+      "learning_rate": 9.318357692643894e-06,
+      "loss": 0.2001,
+      "step": 7938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6107774406156419,
+      "learning_rate": 9.318159076936026e-06,
+      "loss": 0.3699,
+      "step": 7939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8498310773963016,
+      "learning_rate": 9.317960434413561e-06,
+      "loss": 0.3553,
+      "step": 7940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.468671801697097,
+      "learning_rate": 9.317761765077733e-06,
+      "loss": 0.3595,
+      "step": 7941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2466696224867664,
+      "learning_rate": 9.317563068929775e-06,
+      "loss": 0.4163,
+      "step": 7942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8254881226460924,
+      "learning_rate": 9.317364345970922e-06,
+      "loss": 0.3638,
+      "step": 7943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4291284359215237,
+      "learning_rate": 9.317165596202406e-06,
+      "loss": 0.201,
+      "step": 7944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.852901666573514,
+      "learning_rate": 9.316966819625464e-06,
+      "loss": 0.4047,
+      "step": 7945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1289923029315205,
+      "learning_rate": 9.31676801624133e-06,
+      "loss": 0.3899,
+      "step": 7946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5688278150684019,
+      "learning_rate": 9.316569186051234e-06,
+      "loss": 0.1851,
+      "step": 7947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8729030088479033,
+      "learning_rate": 9.316370329056415e-06,
+      "loss": 0.6415,
+      "step": 7948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3565057462298302,
+      "learning_rate": 9.316171445258108e-06,
+      "loss": 0.1989,
+      "step": 7949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7951387641713215,
+      "learning_rate": 9.315972534657547e-06,
+      "loss": 0.2503,
+      "step": 7950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.82820771511133,
+      "learning_rate": 9.315773597255966e-06,
+      "loss": 0.5194,
+      "step": 7951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.698289103899154,
+      "learning_rate": 9.315574633054603e-06,
+      "loss": 0.3328,
+      "step": 7952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3607143526281524,
+      "learning_rate": 9.31537564205469e-06,
+      "loss": 0.2049,
+      "step": 7953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5007719844867697,
+      "learning_rate": 9.315176624257466e-06,
+      "loss": 0.3061,
+      "step": 7954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1007154594154898,
+      "learning_rate": 9.314977579664165e-06,
+      "loss": 0.5927,
+      "step": 7955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5663754181907363,
+      "learning_rate": 9.314778508276024e-06,
+      "loss": 0.3558,
+      "step": 7956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7280335114976122,
+      "learning_rate": 9.314579410094278e-06,
+      "loss": 0.3866,
+      "step": 7957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.641277766929808,
+      "learning_rate": 9.314380285120164e-06,
+      "loss": 0.4009,
+      "step": 7958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6908694598217933,
+      "learning_rate": 9.314181133354918e-06,
+      "loss": 0.484,
+      "step": 7959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8815040592671384,
+      "learning_rate": 9.313981954799777e-06,
+      "loss": 0.4701,
+      "step": 7960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0479639665555487,
+      "learning_rate": 9.313782749455981e-06,
+      "loss": 0.3685,
+      "step": 7961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6089208202771115,
+      "learning_rate": 9.313583517324763e-06,
+      "loss": 0.1943,
+      "step": 7962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2325013606620594,
+      "learning_rate": 9.31338425840736e-06,
+      "loss": 0.1671,
+      "step": 7963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0482028119669398,
+      "learning_rate": 9.31318497270501e-06,
+      "loss": 0.4191,
+      "step": 7964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1179306677897274,
+      "learning_rate": 9.312985660218951e-06,
+      "loss": 0.1662,
+      "step": 7965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7876778610396273,
+      "learning_rate": 9.312786320950421e-06,
+      "loss": 0.2954,
+      "step": 7966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.054597294208866,
+      "learning_rate": 9.31258695490066e-06,
+      "loss": 0.4729,
+      "step": 7967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.082794593453611,
+      "learning_rate": 9.312387562070901e-06,
+      "loss": 0.4142,
+      "step": 7968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.34182737140898,
+      "learning_rate": 9.312188142462386e-06,
+      "loss": 0.3014,
+      "step": 7969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9940921691626663,
+      "learning_rate": 9.311988696076352e-06,
+      "loss": 0.4611,
+      "step": 7970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5043821547908243,
+      "learning_rate": 9.311789222914038e-06,
+      "loss": 0.3413,
+      "step": 7971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9989154639293008,
+      "learning_rate": 9.311589722976682e-06,
+      "loss": 0.6378,
+      "step": 7972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6777226688567255,
+      "learning_rate": 9.311390196265523e-06,
+      "loss": 0.3672,
+      "step": 7973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1563051048982746,
+      "learning_rate": 9.3111906427818e-06,
+      "loss": 0.3062,
+      "step": 7974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5303913453231035,
+      "learning_rate": 9.310991062526752e-06,
+      "loss": 0.3065,
+      "step": 7975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9676577815671863,
+      "learning_rate": 9.31079145550162e-06,
+      "loss": 0.1996,
+      "step": 7976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.945767613682284,
+      "learning_rate": 9.310591821707642e-06,
+      "loss": 0.5535,
+      "step": 7977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6836318345862604,
+      "learning_rate": 9.310392161146056e-06,
+      "loss": 0.2771,
+      "step": 7978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8169769170013423,
+      "learning_rate": 9.310192473818105e-06,
+      "loss": 0.3521,
+      "step": 7979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.438111268566083,
+      "learning_rate": 9.309992759725028e-06,
+      "loss": 0.2445,
+      "step": 7980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2942312104692524,
+      "learning_rate": 9.309793018868065e-06,
+      "loss": 0.2708,
+      "step": 7981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0431846519157006,
+      "learning_rate": 9.309593251248455e-06,
+      "loss": 0.6549,
+      "step": 7982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4421987629220234,
+      "learning_rate": 9.309393456867442e-06,
+      "loss": 0.2448,
+      "step": 7983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6247709598885642,
+      "learning_rate": 9.309193635726262e-06,
+      "loss": 0.5285,
+      "step": 7984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6504478986384141,
+      "learning_rate": 9.30899378782616e-06,
+      "loss": 0.1865,
+      "step": 7985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5762405049675603,
+      "learning_rate": 9.308793913168372e-06,
+      "loss": 0.3151,
+      "step": 7986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9702769594670204,
+      "learning_rate": 9.308594011754147e-06,
+      "loss": 0.5621,
+      "step": 7987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5520789820548215,
+      "learning_rate": 9.308394083584717e-06,
+      "loss": 0.2464,
+      "step": 7988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.487841543688547,
+      "learning_rate": 9.308194128661332e-06,
+      "loss": 0.2888,
+      "step": 7989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9020787314923018,
+      "learning_rate": 9.307994146985227e-06,
+      "loss": 0.4469,
+      "step": 7990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6388187228981173,
+      "learning_rate": 9.307794138557649e-06,
+      "loss": 0.2378,
+      "step": 7991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6398078223005184,
+      "learning_rate": 9.307594103379838e-06,
+      "loss": 0.4567,
+      "step": 7992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7421604975502318,
+      "learning_rate": 9.307394041453034e-06,
+      "loss": 0.2193,
+      "step": 7993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7254167440353456,
+      "learning_rate": 9.307193952778482e-06,
+      "loss": 0.5403,
+      "step": 7994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.542812498848714,
+      "learning_rate": 9.306993837357424e-06,
+      "loss": 0.2806,
+      "step": 7995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6660881787265787,
+      "learning_rate": 9.3067936951911e-06,
+      "loss": 0.553,
+      "step": 7996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8565594408812376,
+      "learning_rate": 9.306593526280758e-06,
+      "loss": 0.2912,
+      "step": 7997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5539090181942654,
+      "learning_rate": 9.306393330627638e-06,
+      "loss": 0.2514,
+      "step": 7998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7696213808445376,
+      "learning_rate": 9.306193108232982e-06,
+      "loss": 0.4071,
+      "step": 7999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.485993822676974,
+      "learning_rate": 9.305992859098037e-06,
+      "loss": 0.1767,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.919662784080114,
+      "learning_rate": 9.305792583224041e-06,
+      "loss": 0.2707,
+      "step": 8001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.765806555908004,
+      "learning_rate": 9.305592280612243e-06,
+      "loss": 0.4609,
+      "step": 8002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9062070998070322,
+      "learning_rate": 9.305391951263885e-06,
+      "loss": 0.5605,
+      "step": 8003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.216938189693877,
+      "learning_rate": 9.305191595180208e-06,
+      "loss": 0.1367,
+      "step": 8004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5528662719741733,
+      "learning_rate": 9.304991212362461e-06,
+      "loss": 0.4142,
+      "step": 8005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4897106670317115,
+      "learning_rate": 9.304790802811885e-06,
+      "loss": 0.2033,
+      "step": 8006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.412725763677662,
+      "learning_rate": 9.304590366529727e-06,
+      "loss": 0.4035,
+      "step": 8007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5919256989006947,
+      "learning_rate": 9.30438990351723e-06,
+      "loss": 0.2291,
+      "step": 8008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.942459925774545,
+      "learning_rate": 9.304189413775637e-06,
+      "loss": 0.3855,
+      "step": 8009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7654800017916883,
+      "learning_rate": 9.303988897306197e-06,
+      "loss": 0.3295,
+      "step": 8010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7927197748903725,
+      "learning_rate": 9.303788354110153e-06,
+      "loss": 0.2279,
+      "step": 8011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0558629897903247,
+      "learning_rate": 9.30358778418875e-06,
+      "loss": 0.3052,
+      "step": 8012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5419797474055539,
+      "learning_rate": 9.303387187543234e-06,
+      "loss": 0.2356,
+      "step": 8013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.540716352520596,
+      "learning_rate": 9.303186564174852e-06,
+      "loss": 0.1944,
+      "step": 8014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8356352618186829,
+      "learning_rate": 9.302985914084847e-06,
+      "loss": 0.2185,
+      "step": 8015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5218525447105713,
+      "learning_rate": 9.302785237274468e-06,
+      "loss": 0.7539,
+      "step": 8016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3603021786633611,
+      "learning_rate": 9.302584533744958e-06,
+      "loss": 0.1581,
+      "step": 8017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.852892341772208,
+      "learning_rate": 9.302383803497566e-06,
+      "loss": 0.5553,
+      "step": 8018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7216382524821823,
+      "learning_rate": 9.302183046533538e-06,
+      "loss": 0.4556,
+      "step": 8019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3638086255319504,
+      "learning_rate": 9.30198226285412e-06,
+      "loss": 0.1686,
+      "step": 8020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8274618111480652,
+      "learning_rate": 9.301781452460556e-06,
+      "loss": 0.3445,
+      "step": 8021
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4762552082880769,
+      "learning_rate": 9.301580615354098e-06,
+      "loss": 0.372,
+      "step": 8022
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.795232335036397,
+      "learning_rate": 9.301379751535991e-06,
+      "loss": 0.4427,
+      "step": 8023
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7210955524132114,
+      "learning_rate": 9.301178861007483e-06,
+      "loss": 0.2078,
+      "step": 8024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.117483850268916,
+      "learning_rate": 9.30097794376982e-06,
+      "loss": 0.2376,
+      "step": 8025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6023528870201642,
+      "learning_rate": 9.30077699982425e-06,
+      "loss": 0.2536,
+      "step": 8026
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8509952062408719,
+      "learning_rate": 9.300576029172021e-06,
+      "loss": 0.4617,
+      "step": 8027
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.244906679197752,
+      "learning_rate": 9.300375031814383e-06,
+      "loss": 0.2956,
+      "step": 8028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3584564287105547,
+      "learning_rate": 9.300174007752581e-06,
+      "loss": 0.1969,
+      "step": 8029
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4502911776828988,
+      "learning_rate": 9.299972956987864e-06,
+      "loss": 0.2802,
+      "step": 8030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.256027340931707,
+      "learning_rate": 9.299771879521481e-06,
+      "loss": 0.328,
+      "step": 8031
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4805114442049594,
+      "learning_rate": 9.299570775354683e-06,
+      "loss": 0.602,
+      "step": 8032
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7601883959385105,
+      "learning_rate": 9.299369644488716e-06,
+      "loss": 0.3473,
+      "step": 8033
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.059902546016078,
+      "learning_rate": 9.29916848692483e-06,
+      "loss": 0.4743,
+      "step": 8034
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.435967296876467,
+      "learning_rate": 9.29896730266427e-06,
+      "loss": 0.2855,
+      "step": 8035
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.645667831835638,
+      "learning_rate": 9.298766091708293e-06,
+      "loss": 0.442,
+      "step": 8036
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3766635400882452,
+      "learning_rate": 9.298564854058143e-06,
+      "loss": 0.3658,
+      "step": 8037
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1160459643367227,
+      "learning_rate": 9.298363589715071e-06,
+      "loss": 0.5372,
+      "step": 8038
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.53438822751258,
+      "learning_rate": 9.298162298680327e-06,
+      "loss": 0.2549,
+      "step": 8039
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5164519314171803,
+      "learning_rate": 9.29796098095516e-06,
+      "loss": 0.2375,
+      "step": 8040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1280498216880948,
+      "learning_rate": 9.297759636540822e-06,
+      "loss": 0.5474,
+      "step": 8041
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5607892398475058,
+      "learning_rate": 9.297558265438563e-06,
+      "loss": 0.3094,
+      "step": 8042
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1263576435063791,
+      "learning_rate": 9.297356867649629e-06,
+      "loss": 0.2068,
+      "step": 8043
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3438267879946262,
+      "learning_rate": 9.297155443175278e-06,
+      "loss": 0.2505,
+      "step": 8044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7487455495528224,
+      "learning_rate": 9.296953992016757e-06,
+      "loss": 0.3247,
+      "step": 8045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6382646748109504,
+      "learning_rate": 9.296752514175315e-06,
+      "loss": 0.249,
+      "step": 8046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2654620142177042,
+      "learning_rate": 9.296551009652207e-06,
+      "loss": 0.1121,
+      "step": 8047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4395121014960202,
+      "learning_rate": 9.296349478448681e-06,
+      "loss": 0.2195,
+      "step": 8048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.299896769694837,
+      "learning_rate": 9.29614792056599e-06,
+      "loss": 0.5097,
+      "step": 8049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2884365664642607,
+      "learning_rate": 9.295946336005388e-06,
+      "loss": 0.3919,
+      "step": 8050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.466509516590262,
+      "learning_rate": 9.295744724768122e-06,
+      "loss": 0.2798,
+      "step": 8051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.023695837194971,
+      "learning_rate": 9.295543086855448e-06,
+      "loss": 0.558,
+      "step": 8052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2654524244155274,
+      "learning_rate": 9.295341422268614e-06,
+      "loss": 0.4282,
+      "step": 8053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0717178037684283,
+      "learning_rate": 9.295139731008876e-06,
+      "loss": 0.2334,
+      "step": 8054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.719330772214714,
+      "learning_rate": 9.294938013077486e-06,
+      "loss": 0.3608,
+      "step": 8055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5503851675416636,
+      "learning_rate": 9.294736268475695e-06,
+      "loss": 0.3288,
+      "step": 8056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1883415068147882,
+      "learning_rate": 9.294534497204757e-06,
+      "loss": 0.4884,
+      "step": 8057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7924623085791689,
+      "learning_rate": 9.294332699265925e-06,
+      "loss": 0.3951,
+      "step": 8058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.28268740725934,
+      "learning_rate": 9.29413087466045e-06,
+      "loss": 0.1629,
+      "step": 8059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3621901050269987,
+      "learning_rate": 9.293929023389587e-06,
+      "loss": 0.6809,
+      "step": 8060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7598177386663585,
+      "learning_rate": 9.29372714545459e-06,
+      "loss": 0.3452,
+      "step": 8061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.322130898753636,
+      "learning_rate": 9.29352524085671e-06,
+      "loss": 0.3443,
+      "step": 8062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4914559523392097,
+      "learning_rate": 9.293323309597206e-06,
+      "loss": 0.2751,
+      "step": 8063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4102490492320325,
+      "learning_rate": 9.293121351677327e-06,
+      "loss": 0.2611,
+      "step": 8064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9264675149837684,
+      "learning_rate": 9.292919367098329e-06,
+      "loss": 0.2867,
+      "step": 8065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6387555461341592,
+      "learning_rate": 9.292717355861465e-06,
+      "loss": 0.2949,
+      "step": 8066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8126805478639303,
+      "learning_rate": 9.292515317967992e-06,
+      "loss": 0.2737,
+      "step": 8067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4079339078254611,
+      "learning_rate": 9.292313253419162e-06,
+      "loss": 0.1479,
+      "step": 8068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8810404752729732,
+      "learning_rate": 9.29211116221623e-06,
+      "loss": 0.6137,
+      "step": 8069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6760155902352816,
+      "learning_rate": 9.291909044360455e-06,
+      "loss": 0.8638,
+      "step": 8070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.62698854909298,
+      "learning_rate": 9.291706899853086e-06,
+      "loss": 0.3512,
+      "step": 8071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4928507304591658,
+      "learning_rate": 9.291504728695382e-06,
+      "loss": 0.3622,
+      "step": 8072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9977379911584756,
+      "learning_rate": 9.291302530888597e-06,
+      "loss": 0.4034,
+      "step": 8073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9102590158820392,
+      "learning_rate": 9.291100306433987e-06,
+      "loss": 0.4464,
+      "step": 8074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.011773006855657,
+      "learning_rate": 9.290898055332809e-06,
+      "loss": 0.5562,
+      "step": 8075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.055966541208676,
+      "learning_rate": 9.290695777586315e-06,
+      "loss": 0.3422,
+      "step": 8076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.246895169899269,
+      "learning_rate": 9.290493473195767e-06,
+      "loss": 0.2016,
+      "step": 8077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.426704259516226,
+      "learning_rate": 9.290291142162417e-06,
+      "loss": 0.4734,
+      "step": 8078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8102454435578021,
+      "learning_rate": 9.290088784487523e-06,
+      "loss": 0.3693,
+      "step": 8079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.881440917494749,
+      "learning_rate": 9.289886400172341e-06,
+      "loss": 0.4806,
+      "step": 8080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2522595130351923,
+      "learning_rate": 9.289683989218128e-06,
+      "loss": 0.2694,
+      "step": 8081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7279303858477386,
+      "learning_rate": 9.28948155162614e-06,
+      "loss": 0.3728,
+      "step": 8082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6614069277983428,
+      "learning_rate": 9.289279087397636e-06,
+      "loss": 0.216,
+      "step": 8083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4392626415083152,
+      "learning_rate": 9.289076596533873e-06,
+      "loss": 0.1683,
+      "step": 8084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.035437269815857,
+      "learning_rate": 9.288874079036105e-06,
+      "loss": 0.3784,
+      "step": 8085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3511338463775202,
+      "learning_rate": 9.288671534905593e-06,
+      "loss": 0.2055,
+      "step": 8086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1898842016345794,
+      "learning_rate": 9.288468964143596e-06,
+      "loss": 0.2349,
+      "step": 8087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.494832878314494,
+      "learning_rate": 9.288266366751365e-06,
+      "loss": 0.2114,
+      "step": 8088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7007549839330853,
+      "learning_rate": 9.288063742730168e-06,
+      "loss": 0.2902,
+      "step": 8089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4585863052938253,
+      "learning_rate": 9.287861092081256e-06,
+      "loss": 0.2769,
+      "step": 8090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0386515285052822,
+      "learning_rate": 9.28765841480589e-06,
+      "loss": 0.1455,
+      "step": 8091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1633466748572943,
+      "learning_rate": 9.287455710905328e-06,
+      "loss": 0.3055,
+      "step": 8092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3323118076586615,
+      "learning_rate": 9.287252980380828e-06,
+      "loss": 0.2294,
+      "step": 8093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2398064708693886,
+      "learning_rate": 9.28705022323365e-06,
+      "loss": 0.2282,
+      "step": 8094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1904919462841765,
+      "learning_rate": 9.286847439465053e-06,
+      "loss": 0.3783,
+      "step": 8095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4016114211501873,
+      "learning_rate": 9.286644629076295e-06,
+      "loss": 0.167,
+      "step": 8096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.443731809683148,
+      "learning_rate": 9.286441792068639e-06,
+      "loss": 0.3374,
+      "step": 8097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0646677871012464,
+      "learning_rate": 9.28623892844334e-06,
+      "loss": 0.2374,
+      "step": 8098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.142053468602765,
+      "learning_rate": 9.286036038201661e-06,
+      "loss": 0.5562,
+      "step": 8099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6027003991032978,
+      "learning_rate": 9.28583312134486e-06,
+      "loss": 0.2119,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0169380812389233,
+      "learning_rate": 9.285630177874197e-06,
+      "loss": 0.2876,
+      "step": 8101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5146471757717523,
+      "learning_rate": 9.285427207790934e-06,
+      "loss": 0.349,
+      "step": 8102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9522971267903364,
+      "learning_rate": 9.285224211096329e-06,
+      "loss": 0.3681,
+      "step": 8103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6829355727929278,
+      "learning_rate": 9.285021187791645e-06,
+      "loss": 0.4873,
+      "step": 8104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9587379825469264,
+      "learning_rate": 9.284818137878142e-06,
+      "loss": 0.4135,
+      "step": 8105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2324933879193258,
+      "learning_rate": 9.284615061357078e-06,
+      "loss": 0.2726,
+      "step": 8106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8269211379491694,
+      "learning_rate": 9.284411958229717e-06,
+      "loss": 0.3619,
+      "step": 8107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6977094853459263,
+      "learning_rate": 9.284208828497322e-06,
+      "loss": 0.3406,
+      "step": 8108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5123359938047327,
+      "learning_rate": 9.284005672161152e-06,
+      "loss": 0.3668,
+      "step": 8109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7038959871041632,
+      "learning_rate": 9.283802489222467e-06,
+      "loss": 0.3721,
+      "step": 8110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1877282748284939,
+      "learning_rate": 9.28359927968253e-06,
+      "loss": 0.2954,
+      "step": 8111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5508135941156853,
+      "learning_rate": 9.283396043542603e-06,
+      "loss": 0.2322,
+      "step": 8112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3692699941280932,
+      "learning_rate": 9.283192780803952e-06,
+      "loss": 0.2642,
+      "step": 8113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1961490097586187,
+      "learning_rate": 9.28298949146783e-06,
+      "loss": 0.267,
+      "step": 8114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2658231736080054,
+      "learning_rate": 9.282786175535508e-06,
+      "loss": 0.3528,
+      "step": 8115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.291991348439331,
+      "learning_rate": 9.282582833008246e-06,
+      "loss": 0.3393,
+      "step": 8116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7521857252906285,
+      "learning_rate": 9.282379463887306e-06,
+      "loss": 0.4682,
+      "step": 8117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3855136752822614,
+      "learning_rate": 9.28217606817395e-06,
+      "loss": 0.1541,
+      "step": 8118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7706754450440696,
+      "learning_rate": 9.281972645869441e-06,
+      "loss": 0.4031,
+      "step": 8119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6426076807000674,
+      "learning_rate": 9.281769196975044e-06,
+      "loss": 0.4979,
+      "step": 8120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.452838505425823,
+      "learning_rate": 9.281565721492023e-06,
+      "loss": 0.3263,
+      "step": 8121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.095954943883359,
+      "learning_rate": 9.281362219421637e-06,
+      "loss": 0.2331,
+      "step": 8122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4603727110856468,
+      "learning_rate": 9.281158690765155e-06,
+      "loss": 0.2254,
+      "step": 8123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6993601992258254,
+      "learning_rate": 9.280955135523838e-06,
+      "loss": 0.3718,
+      "step": 8124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5248372372162662,
+      "learning_rate": 9.28075155369895e-06,
+      "loss": 0.3638,
+      "step": 8125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3587907091739135,
+      "learning_rate": 9.280547945291758e-06,
+      "loss": 0.2215,
+      "step": 8126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.717065115360452,
+      "learning_rate": 9.280344310303522e-06,
+      "loss": 0.3602,
+      "step": 8127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7442025390434834,
+      "learning_rate": 9.28014064873551e-06,
+      "loss": 0.5032,
+      "step": 8128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4570812237225865,
+      "learning_rate": 9.279936960588984e-06,
+      "loss": 0.3097,
+      "step": 8129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7709504177617952,
+      "learning_rate": 9.279733245865212e-06,
+      "loss": 0.29,
+      "step": 8130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7385469340760205,
+      "learning_rate": 9.279529504565457e-06,
+      "loss": 0.6176,
+      "step": 8131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4736303957646943,
+      "learning_rate": 9.279325736690983e-06,
+      "loss": 0.2089,
+      "step": 8132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9067007610620028,
+      "learning_rate": 9.279121942243057e-06,
+      "loss": 0.4229,
+      "step": 8133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.822013451155867,
+      "learning_rate": 9.278918121222947e-06,
+      "loss": 0.2077,
+      "step": 8134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1564880702717268,
+      "learning_rate": 9.278714273631913e-06,
+      "loss": 0.1611,
+      "step": 8135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7575461800478425,
+      "learning_rate": 9.278510399471223e-06,
+      "loss": 0.2589,
+      "step": 8136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.653419874678091,
+      "learning_rate": 9.278306498742146e-06,
+      "loss": 0.3471,
+      "step": 8137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9369132553242,
+      "learning_rate": 9.278102571445945e-06,
+      "loss": 0.5553,
+      "step": 8138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3703645943664413,
+      "learning_rate": 9.277898617583886e-06,
+      "loss": 0.2558,
+      "step": 8139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3576576714206965,
+      "learning_rate": 9.277694637157239e-06,
+      "loss": 0.1909,
+      "step": 8140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9447628646547188,
+      "learning_rate": 9.27749063016727e-06,
+      "loss": 0.2293,
+      "step": 8141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2925788847689617,
+      "learning_rate": 9.277286596615241e-06,
+      "loss": 0.3525,
+      "step": 8142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5157521663978561,
+      "learning_rate": 9.277082536502423e-06,
+      "loss": 0.2385,
+      "step": 8143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.302834700323264,
+      "learning_rate": 9.276878449830083e-06,
+      "loss": 0.2386,
+      "step": 8144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5165193318060204,
+      "learning_rate": 9.276674336599486e-06,
+      "loss": 0.3013,
+      "step": 8145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5802913941506598,
+      "learning_rate": 9.276470196811905e-06,
+      "loss": 0.4207,
+      "step": 8146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7438471890098712,
+      "learning_rate": 9.276266030468602e-06,
+      "loss": 0.4632,
+      "step": 8147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.143528492511071,
+      "learning_rate": 9.276061837570846e-06,
+      "loss": 0.1671,
+      "step": 8148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.734889847573229,
+      "learning_rate": 9.275857618119906e-06,
+      "loss": 0.2441,
+      "step": 8149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4979319244016347,
+      "learning_rate": 9.275653372117049e-06,
+      "loss": 0.3083,
+      "step": 8150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.07733203164447,
+      "learning_rate": 9.275449099563544e-06,
+      "loss": 0.4079,
+      "step": 8151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5770793090645716,
+      "learning_rate": 9.275244800460663e-06,
+      "loss": 0.2169,
+      "step": 8152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4716988241448759,
+      "learning_rate": 9.275040474809668e-06,
+      "loss": 0.2602,
+      "step": 8153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6806191832398136,
+      "learning_rate": 9.274836122611832e-06,
+      "loss": 0.4953,
+      "step": 8154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8725402917634255,
+      "learning_rate": 9.274631743868423e-06,
+      "loss": 0.1908,
+      "step": 8155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4404281642065238,
+      "learning_rate": 9.27442733858071e-06,
+      "loss": 0.311,
+      "step": 8156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9706304062280078,
+      "learning_rate": 9.274222906749963e-06,
+      "loss": 0.3525,
+      "step": 8157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5940623444805955,
+      "learning_rate": 9.27401844837745e-06,
+      "loss": 0.4088,
+      "step": 8158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2551590789797233,
+      "learning_rate": 9.273813963464444e-06,
+      "loss": 0.2218,
+      "step": 8159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6265349016587412,
+      "learning_rate": 9.27360945201221e-06,
+      "loss": 0.4227,
+      "step": 8160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9572974478920633,
+      "learning_rate": 9.273404914022021e-06,
+      "loss": 0.2589,
+      "step": 8161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4866713893180388,
+      "learning_rate": 9.273200349495147e-06,
+      "loss": 0.3198,
+      "step": 8162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.078253354862699,
+      "learning_rate": 9.272995758432856e-06,
+      "loss": 0.2929,
+      "step": 8163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9764086565450443,
+      "learning_rate": 9.272791140836421e-06,
+      "loss": 0.2368,
+      "step": 8164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1438142879259603,
+      "learning_rate": 9.272586496707113e-06,
+      "loss": 0.2253,
+      "step": 8165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.750161867021174,
+      "learning_rate": 9.272381826046203e-06,
+      "loss": 0.233,
+      "step": 8166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.797744801680755,
+      "learning_rate": 9.272177128854958e-06,
+      "loss": 0.5419,
+      "step": 8167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6372337686037577,
+      "learning_rate": 9.271972405134651e-06,
+      "loss": 0.5668,
+      "step": 8168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3101399865957966,
+      "learning_rate": 9.271767654886555e-06,
+      "loss": 0.2166,
+      "step": 8169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3984338184926253,
+      "learning_rate": 9.271562878111942e-06,
+      "loss": 0.3098,
+      "step": 8170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.590106394112613,
+      "learning_rate": 9.27135807481208e-06,
+      "loss": 0.3124,
+      "step": 8171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3399297558731382,
+      "learning_rate": 9.271153244988242e-06,
+      "loss": 0.3513,
+      "step": 8172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8875612901058698,
+      "learning_rate": 9.270948388641702e-06,
+      "loss": 0.418,
+      "step": 8173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5960692565801247,
+      "learning_rate": 9.270743505773732e-06,
+      "loss": 0.3321,
+      "step": 8174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2557546880301924,
+      "learning_rate": 9.270538596385601e-06,
+      "loss": 0.6698,
+      "step": 8175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9725656091997883,
+      "learning_rate": 9.270333660478586e-06,
+      "loss": 0.6141,
+      "step": 8176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.163929074439942,
+      "learning_rate": 9.270128698053954e-06,
+      "loss": 0.4102,
+      "step": 8177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4745362371029103,
+      "learning_rate": 9.269923709112982e-06,
+      "loss": 0.3052,
+      "step": 8178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.190151495408909,
+      "learning_rate": 9.269718693656944e-06,
+      "loss": 0.2584,
+      "step": 8179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1784661899373048,
+      "learning_rate": 9.26951365168711e-06,
+      "loss": 0.197,
+      "step": 8180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6300040554803588,
+      "learning_rate": 9.269308583204752e-06,
+      "loss": 0.2554,
+      "step": 8181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.823314569279808,
+      "learning_rate": 9.269103488211147e-06,
+      "loss": 0.5425,
+      "step": 8182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.774837898857976,
+      "learning_rate": 9.268898366707568e-06,
+      "loss": 0.546,
+      "step": 8183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7649353714489748,
+      "learning_rate": 9.268693218695287e-06,
+      "loss": 0.2458,
+      "step": 8184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6682944192033407,
+      "learning_rate": 9.268488044175577e-06,
+      "loss": 0.3648,
+      "step": 8185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0578861106098074,
+      "learning_rate": 9.268282843149716e-06,
+      "loss": 0.4847,
+      "step": 8186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8786668398412595,
+      "learning_rate": 9.268077615618976e-06,
+      "loss": 0.2001,
+      "step": 8187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4887733772205636,
+      "learning_rate": 9.267872361584631e-06,
+      "loss": 0.2432,
+      "step": 8188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.450135062150989,
+      "learning_rate": 9.267667081047957e-06,
+      "loss": 0.3707,
+      "step": 8189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4717225496952018,
+      "learning_rate": 9.267461774010227e-06,
+      "loss": 0.2762,
+      "step": 8190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5625478594126014,
+      "learning_rate": 9.267256440472718e-06,
+      "loss": 0.1949,
+      "step": 8191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8865092222391233,
+      "learning_rate": 9.267051080436702e-06,
+      "loss": 0.5206,
+      "step": 8192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9254678119055666,
+      "learning_rate": 9.266845693903457e-06,
+      "loss": 0.498,
+      "step": 8193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2069410642245628,
+      "learning_rate": 9.266640280874257e-06,
+      "loss": 0.1732,
+      "step": 8194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.424174330788212,
+      "learning_rate": 9.266434841350379e-06,
+      "loss": 0.2378,
+      "step": 8195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7995885221392238,
+      "learning_rate": 9.266229375333098e-06,
+      "loss": 0.4137,
+      "step": 8196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0180475920347734,
+      "learning_rate": 9.266023882823688e-06,
+      "loss": 0.5705,
+      "step": 8197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.892259655623283,
+      "learning_rate": 9.265818363823427e-06,
+      "loss": 0.5294,
+      "step": 8198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8376078656708634,
+      "learning_rate": 9.265612818333592e-06,
+      "loss": 0.5131,
+      "step": 8199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0943058756341295,
+      "learning_rate": 9.265407246355457e-06,
+      "loss": 0.6603,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5760691773567805,
+      "learning_rate": 9.2652016478903e-06,
+      "loss": 0.3083,
+      "step": 8201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8571261703054625,
+      "learning_rate": 9.264996022939398e-06,
+      "loss": 0.4042,
+      "step": 8202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9833191950194597,
+      "learning_rate": 9.264790371504028e-06,
+      "loss": 0.2138,
+      "step": 8203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1246368083806377,
+      "learning_rate": 9.264584693585465e-06,
+      "loss": 0.5123,
+      "step": 8204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3944521128214655,
+      "learning_rate": 9.264378989184988e-06,
+      "loss": 0.1761,
+      "step": 8205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.501189395974837,
+      "learning_rate": 9.264173258303876e-06,
+      "loss": 0.6407,
+      "step": 8206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.671153357326834,
+      "learning_rate": 9.263967500943402e-06,
+      "loss": 0.235,
+      "step": 8207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9502871467389207,
+      "learning_rate": 9.263761717104848e-06,
+      "loss": 0.3195,
+      "step": 8208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2815295923375896,
+      "learning_rate": 9.26355590678949e-06,
+      "loss": 0.5127,
+      "step": 8209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5741364620842269,
+      "learning_rate": 9.263350069998605e-06,
+      "loss": 0.3968,
+      "step": 8210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9374997942678283,
+      "learning_rate": 9.263144206733473e-06,
+      "loss": 0.3595,
+      "step": 8211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8250157562592904,
+      "learning_rate": 9.262938316995372e-06,
+      "loss": 0.5628,
+      "step": 8212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0885862256922905,
+      "learning_rate": 9.262732400785578e-06,
+      "loss": 0.5822,
+      "step": 8213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.820065295333893,
+      "learning_rate": 9.262526458105375e-06,
+      "loss": 0.3686,
+      "step": 8214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6571232140258019,
+      "learning_rate": 9.262320488956036e-06,
+      "loss": 0.4345,
+      "step": 8215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6743403767059957,
+      "learning_rate": 9.262114493338843e-06,
+      "loss": 0.2212,
+      "step": 8216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3750318725856565,
+      "learning_rate": 9.261908471255076e-06,
+      "loss": 0.4855,
+      "step": 8217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7771746953275216,
+      "learning_rate": 9.261702422706014e-06,
+      "loss": 0.1812,
+      "step": 8218
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7634342240487846,
+      "learning_rate": 9.261496347692935e-06,
+      "loss": 0.5131,
+      "step": 8219
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.791330137270394,
+      "learning_rate": 9.26129024621712e-06,
+      "loss": 0.3633,
+      "step": 8220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8284551412043986,
+      "learning_rate": 9.261084118279846e-06,
+      "loss": 0.3791,
+      "step": 8221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.820812373637183,
+      "learning_rate": 9.260877963882398e-06,
+      "loss": 0.25,
+      "step": 8222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8060650501467534,
+      "learning_rate": 9.260671783026052e-06,
+      "loss": 0.1695,
+      "step": 8223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8786297434100265,
+      "learning_rate": 9.26046557571209e-06,
+      "loss": 0.4474,
+      "step": 8224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.560745205170377,
+      "learning_rate": 9.260259341941793e-06,
+      "loss": 0.4114,
+      "step": 8225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0582610916544044,
+      "learning_rate": 9.26005308171644e-06,
+      "loss": 0.5736,
+      "step": 8226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4594144276718843,
+      "learning_rate": 9.259846795037313e-06,
+      "loss": 0.3044,
+      "step": 8227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.775315395393824,
+      "learning_rate": 9.259640481905692e-06,
+      "loss": 0.2418,
+      "step": 8228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3947492637437462,
+      "learning_rate": 9.259434142322861e-06,
+      "loss": 0.2224,
+      "step": 8229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7010331124649138,
+      "learning_rate": 9.259227776290095e-06,
+      "loss": 0.4404,
+      "step": 8230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.710813612951705,
+      "learning_rate": 9.259021383808683e-06,
+      "loss": 0.2995,
+      "step": 8231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7429214586538904,
+      "learning_rate": 9.258814964879904e-06,
+      "loss": 0.2725,
+      "step": 8232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9352909846675221,
+      "learning_rate": 9.258608519505037e-06,
+      "loss": 0.4839,
+      "step": 8233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8309569431919162,
+      "learning_rate": 9.258402047685368e-06,
+      "loss": 0.4288,
+      "step": 8234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.18634718763456,
+      "learning_rate": 9.258195549422175e-06,
+      "loss": 0.5015,
+      "step": 8235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5724562085191556,
+      "learning_rate": 9.257989024716745e-06,
+      "loss": 0.369,
+      "step": 8236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1819824291870151,
+      "learning_rate": 9.257782473570356e-06,
+      "loss": 0.1428,
+      "step": 8237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4328815356734235,
+      "learning_rate": 9.257575895984292e-06,
+      "loss": 0.2818,
+      "step": 8238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5874716850068449,
+      "learning_rate": 9.257369291959838e-06,
+      "loss": 0.4261,
+      "step": 8239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.578085083267818,
+      "learning_rate": 9.257162661498274e-06,
+      "loss": 0.2251,
+      "step": 8240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.124955229398188,
+      "learning_rate": 9.256956004600886e-06,
+      "loss": 0.2551,
+      "step": 8241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3627567703321728,
+      "learning_rate": 9.256749321268954e-06,
+      "loss": 0.2364,
+      "step": 8242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.748190440244283,
+      "learning_rate": 9.256542611503764e-06,
+      "loss": 0.4448,
+      "step": 8243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.440910339658442,
+      "learning_rate": 9.256335875306601e-06,
+      "loss": 0.2324,
+      "step": 8244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5776671463048975,
+      "learning_rate": 9.256129112678744e-06,
+      "loss": 0.2221,
+      "step": 8245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6646270734469433,
+      "learning_rate": 9.25592232362148e-06,
+      "loss": 0.2699,
+      "step": 8246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5737736264334645,
+      "learning_rate": 9.255715508136091e-06,
+      "loss": 0.4759,
+      "step": 8247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9919717335261002,
+      "learning_rate": 9.255508666223866e-06,
+      "loss": 0.2343,
+      "step": 8248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5120835763498872,
+      "learning_rate": 9.255301797886085e-06,
+      "loss": 0.3634,
+      "step": 8249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5771752747888972,
+      "learning_rate": 9.255094903124034e-06,
+      "loss": 0.1373,
+      "step": 8250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.395237452677015,
+      "learning_rate": 9.254887981938997e-06,
+      "loss": 0.4135,
+      "step": 8251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.425442650186094,
+      "learning_rate": 9.254681034332259e-06,
+      "loss": 0.2942,
+      "step": 8252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.532413765416407,
+      "learning_rate": 9.254474060305107e-06,
+      "loss": 0.237,
+      "step": 8253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7088623003275347,
+      "learning_rate": 9.254267059858824e-06,
+      "loss": 0.3775,
+      "step": 8254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5648141508688302,
+      "learning_rate": 9.254060032994699e-06,
+      "loss": 0.4636,
+      "step": 8255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.06616605433525,
+      "learning_rate": 9.253852979714011e-06,
+      "loss": 0.3322,
+      "step": 8256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6171473622524104,
+      "learning_rate": 9.253645900018052e-06,
+      "loss": 0.1837,
+      "step": 8257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5382174384199863,
+      "learning_rate": 9.253438793908106e-06,
+      "loss": 0.4042,
+      "step": 8258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1226975038523714,
+      "learning_rate": 9.253231661385456e-06,
+      "loss": 0.1728,
+      "step": 8259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9774558494862158,
+      "learning_rate": 9.253024502451393e-06,
+      "loss": 0.5214,
+      "step": 8260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6096545087810132,
+      "learning_rate": 9.252817317107201e-06,
+      "loss": 0.3981,
+      "step": 8261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.774216615212826,
+      "learning_rate": 9.252610105354167e-06,
+      "loss": 0.3711,
+      "step": 8262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5779496463760372,
+      "learning_rate": 9.252402867193577e-06,
+      "loss": 0.3039,
+      "step": 8263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.888995149830104,
+      "learning_rate": 9.252195602626719e-06,
+      "loss": 0.5153,
+      "step": 8264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7657942015670207,
+      "learning_rate": 9.251988311654881e-06,
+      "loss": 0.2269,
+      "step": 8265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.24712100482334,
+      "learning_rate": 9.251780994279348e-06,
+      "loss": 0.1716,
+      "step": 8266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7963288099033548,
+      "learning_rate": 9.251573650501406e-06,
+      "loss": 0.5025,
+      "step": 8267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5381901781392786,
+      "learning_rate": 9.251366280322347e-06,
+      "loss": 0.3209,
+      "step": 8268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6255500302742707,
+      "learning_rate": 9.251158883743454e-06,
+      "loss": 0.2209,
+      "step": 8269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5145944060594696,
+      "learning_rate": 9.250951460766019e-06,
+      "loss": 0.3616,
+      "step": 8270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1351474846228338,
+      "learning_rate": 9.250744011391329e-06,
+      "loss": 0.4251,
+      "step": 8271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6317574660805587,
+      "learning_rate": 9.25053653562067e-06,
+      "loss": 0.2921,
+      "step": 8272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8732194393408637,
+      "learning_rate": 9.250329033455333e-06,
+      "loss": 0.3856,
+      "step": 8273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9758600637841885,
+      "learning_rate": 9.250121504896605e-06,
+      "loss": 0.5064,
+      "step": 8274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7395332472813814,
+      "learning_rate": 9.249913949945775e-06,
+      "loss": 0.2203,
+      "step": 8275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.813860366297482,
+      "learning_rate": 9.249706368604131e-06,
+      "loss": 0.5408,
+      "step": 8276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1085974493630364,
+      "learning_rate": 9.249498760872964e-06,
+      "loss": 0.1555,
+      "step": 8277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.677074917827445,
+      "learning_rate": 9.249291126753563e-06,
+      "loss": 0.1791,
+      "step": 8278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9446775403596042,
+      "learning_rate": 9.249083466247215e-06,
+      "loss": 0.5114,
+      "step": 8279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2888839410166146,
+      "learning_rate": 9.248875779355212e-06,
+      "loss": 0.2266,
+      "step": 8280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6211871016866721,
+      "learning_rate": 9.248668066078844e-06,
+      "loss": 0.2001,
+      "step": 8281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0550423696222406,
+      "learning_rate": 9.248460326419398e-06,
+      "loss": 0.2082,
+      "step": 8282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6248843537040933,
+      "learning_rate": 9.248252560378165e-06,
+      "loss": 0.3358,
+      "step": 8283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9019365924767263,
+      "learning_rate": 9.248044767956437e-06,
+      "loss": 0.1336,
+      "step": 8284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9687916660441105,
+      "learning_rate": 9.247836949155502e-06,
+      "loss": 0.4314,
+      "step": 8285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9668732489083214,
+      "learning_rate": 9.247629103976652e-06,
+      "loss": 0.3425,
+      "step": 8286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2933231953876323,
+      "learning_rate": 9.247421232421178e-06,
+      "loss": 0.4095,
+      "step": 8287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.110391588197777,
+      "learning_rate": 9.247213334490367e-06,
+      "loss": 0.2266,
+      "step": 8288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.926427069387591,
+      "learning_rate": 9.247005410185517e-06,
+      "loss": 0.2916,
+      "step": 8289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0534738847344673,
+      "learning_rate": 9.246797459507914e-06,
+      "loss": 0.2406,
+      "step": 8290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.490420506103774,
+      "learning_rate": 9.24658948245885e-06,
+      "loss": 0.2414,
+      "step": 8291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.647920134111109,
+      "learning_rate": 9.246381479039619e-06,
+      "loss": 0.4095,
+      "step": 8292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2411384831492491,
+      "learning_rate": 9.246173449251507e-06,
+      "loss": 0.2825,
+      "step": 8293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5389986388032841,
+      "learning_rate": 9.245965393095812e-06,
+      "loss": 0.3589,
+      "step": 8294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2300093837851112,
+      "learning_rate": 9.245757310573823e-06,
+      "loss": 0.1972,
+      "step": 8295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.62074529510676,
+      "learning_rate": 9.245549201686833e-06,
+      "loss": 0.1395,
+      "step": 8296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5913671720947618,
+      "learning_rate": 9.245341066436135e-06,
+      "loss": 0.2242,
+      "step": 8297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0518218315990873,
+      "learning_rate": 9.245132904823018e-06,
+      "loss": 0.3856,
+      "step": 8298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.52088634162346,
+      "learning_rate": 9.244924716848778e-06,
+      "loss": 0.3748,
+      "step": 8299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5515919111139942,
+      "learning_rate": 9.244716502514705e-06,
+      "loss": 0.1905,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4980882457799978,
+      "learning_rate": 9.244508261822097e-06,
+      "loss": 0.2791,
+      "step": 8301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4330378615138586,
+      "learning_rate": 9.244299994772242e-06,
+      "loss": 0.5769,
+      "step": 8302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3802340862263853,
+      "learning_rate": 9.244091701366435e-06,
+      "loss": 0.31,
+      "step": 8303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9892270849417497,
+      "learning_rate": 9.24388338160597e-06,
+      "loss": 0.5233,
+      "step": 8304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6275051366560898,
+      "learning_rate": 9.243675035492138e-06,
+      "loss": 0.467,
+      "step": 8305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2087214681686078,
+      "learning_rate": 9.24346666302624e-06,
+      "loss": 0.1832,
+      "step": 8306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.539831651884926,
+      "learning_rate": 9.24325826420956e-06,
+      "loss": 0.4103,
+      "step": 8307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.290711704729227,
+      "learning_rate": 9.2430498390434e-06,
+      "loss": 0.2089,
+      "step": 8308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3668876462178274,
+      "learning_rate": 9.242841387529048e-06,
+      "loss": 0.1779,
+      "step": 8309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.638405783944903,
+      "learning_rate": 9.242632909667806e-06,
+      "loss": 0.3223,
+      "step": 8310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.095564536341047,
+      "learning_rate": 9.24242440546096e-06,
+      "loss": 0.3513,
+      "step": 8311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.824984775603819,
+      "learning_rate": 9.242215874909812e-06,
+      "loss": 0.3471,
+      "step": 8312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.551051059908591,
+      "learning_rate": 9.242007318015653e-06,
+      "loss": 0.39,
+      "step": 8313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9510706702515659,
+      "learning_rate": 9.241798734779779e-06,
+      "loss": 0.3822,
+      "step": 8314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9164504006078522,
+      "learning_rate": 9.241590125203486e-06,
+      "loss": 0.4684,
+      "step": 8315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8828966519635348,
+      "learning_rate": 9.241381489288068e-06,
+      "loss": 0.1958,
+      "step": 8316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7923847361006966,
+      "learning_rate": 9.241172827034822e-06,
+      "loss": 0.4728,
+      "step": 8317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6052741051066928,
+      "learning_rate": 9.240964138445042e-06,
+      "loss": 0.4198,
+      "step": 8318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4727731906186567,
+      "learning_rate": 9.240755423520024e-06,
+      "loss": 0.3385,
+      "step": 8319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6412485027814834,
+      "learning_rate": 9.240546682261066e-06,
+      "loss": 0.3334,
+      "step": 8320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4260397506902551,
+      "learning_rate": 9.240337914669463e-06,
+      "loss": 0.5456,
+      "step": 8321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3080578119326178,
+      "learning_rate": 9.240129120746513e-06,
+      "loss": 0.3216,
+      "step": 8322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.522256930509723,
+      "learning_rate": 9.239920300493509e-06,
+      "loss": 0.3406,
+      "step": 8323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7978433321837475,
+      "learning_rate": 9.23971145391175e-06,
+      "loss": 0.4855,
+      "step": 8324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.002504412737669,
+      "learning_rate": 9.239502581002535e-06,
+      "loss": 0.1264,
+      "step": 8325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1939177198075004,
+      "learning_rate": 9.239293681767157e-06,
+      "loss": 0.2274,
+      "step": 8326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6486357144873784,
+      "learning_rate": 9.239084756206914e-06,
+      "loss": 0.3364,
+      "step": 8327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7046552747692776,
+      "learning_rate": 9.238875804323107e-06,
+      "loss": 0.5844,
+      "step": 8328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8778416776144002,
+      "learning_rate": 9.23866682611703e-06,
+      "loss": 0.3984,
+      "step": 8329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2343469839900443,
+      "learning_rate": 9.238457821589979e-06,
+      "loss": 0.2696,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9861874525882304,
+      "learning_rate": 9.238248790743258e-06,
+      "loss": 0.5714,
+      "step": 8331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8896037133417074,
+      "learning_rate": 9.23803973357816e-06,
+      "loss": 0.439,
+      "step": 8332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9161585872094702,
+      "learning_rate": 9.237830650095983e-06,
+      "loss": 0.4132,
+      "step": 8333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9678603491509454,
+      "learning_rate": 9.23762154029803e-06,
+      "loss": 0.2993,
+      "step": 8334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6934889201558367,
+      "learning_rate": 9.237412404185595e-06,
+      "loss": 0.3499,
+      "step": 8335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8881591822918264,
+      "learning_rate": 9.237203241759978e-06,
+      "loss": 0.3664,
+      "step": 8336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2930389195097727,
+      "learning_rate": 9.236994053022477e-06,
+      "loss": 0.2023,
+      "step": 8337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6780383863775632,
+      "learning_rate": 9.236784837974393e-06,
+      "loss": 0.4891,
+      "step": 8338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6151570517264995,
+      "learning_rate": 9.236575596617025e-06,
+      "loss": 0.3132,
+      "step": 8339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6605016382750488,
+      "learning_rate": 9.236366328951671e-06,
+      "loss": 0.2803,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6531775537404867,
+      "learning_rate": 9.236157034979632e-06,
+      "loss": 0.2411,
+      "step": 8341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9984499393341437,
+      "learning_rate": 9.235947714702206e-06,
+      "loss": 0.3045,
+      "step": 8342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.504637379470114,
+      "learning_rate": 9.235738368120693e-06,
+      "loss": 0.2932,
+      "step": 8343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0326629199530637,
+      "learning_rate": 9.235528995236394e-06,
+      "loss": 0.6483,
+      "step": 8344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7268430861590183,
+      "learning_rate": 9.235319596050607e-06,
+      "loss": 0.4776,
+      "step": 8345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5319755012994207,
+      "learning_rate": 9.235110170564636e-06,
+      "loss": 0.3879,
+      "step": 8346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.778534802442983,
+      "learning_rate": 9.234900718779779e-06,
+      "loss": 0.713,
+      "step": 8347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9485717953529582,
+      "learning_rate": 9.234691240697336e-06,
+      "loss": 0.442,
+      "step": 8348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.796120953944978,
+      "learning_rate": 9.23448173631861e-06,
+      "loss": 0.4846,
+      "step": 8349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9119690248377483,
+      "learning_rate": 9.234272205644901e-06,
+      "loss": 0.4553,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.803329041124305,
+      "learning_rate": 9.23406264867751e-06,
+      "loss": 0.2881,
+      "step": 8351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6832457501389515,
+      "learning_rate": 9.233853065417736e-06,
+      "loss": 0.26,
+      "step": 8352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.963202705056876,
+      "learning_rate": 9.233643455866886e-06,
+      "loss": 0.3748,
+      "step": 8353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5321023378895773,
+      "learning_rate": 9.233433820026256e-06,
+      "loss": 0.2307,
+      "step": 8354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0035979236368027,
+      "learning_rate": 9.233224157897152e-06,
+      "loss": 0.6756,
+      "step": 8355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4711253255493209,
+      "learning_rate": 9.233014469480874e-06,
+      "loss": 0.1526,
+      "step": 8356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6059412794538195,
+      "learning_rate": 9.232804754778723e-06,
+      "loss": 0.4044,
+      "step": 8357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.201867786463871,
+      "learning_rate": 9.232595013792004e-06,
+      "loss": 0.5114,
+      "step": 8358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8457199862885683,
+      "learning_rate": 9.232385246522016e-06,
+      "loss": 0.5976,
+      "step": 8359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4890362233888639,
+      "learning_rate": 9.232175452970062e-06,
+      "loss": 0.3665,
+      "step": 8360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7606540537276254,
+      "learning_rate": 9.23196563313745e-06,
+      "loss": 0.1948,
+      "step": 8361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6713320483294187,
+      "learning_rate": 9.231755787025476e-06,
+      "loss": 0.3728,
+      "step": 8362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.780740945421351,
+      "learning_rate": 9.231545914635449e-06,
+      "loss": 0.6518,
+      "step": 8363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9866897081359312,
+      "learning_rate": 9.231336015968668e-06,
+      "loss": 0.4007,
+      "step": 8364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5324583103225187,
+      "learning_rate": 9.23112609102644e-06,
+      "loss": 0.353,
+      "step": 8365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2510450122578356,
+      "learning_rate": 9.230916139810063e-06,
+      "loss": 0.1631,
+      "step": 8366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5808990265228546,
+      "learning_rate": 9.230706162320846e-06,
+      "loss": 0.4466,
+      "step": 8367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9501408095138622,
+      "learning_rate": 9.23049615856009e-06,
+      "loss": 0.2461,
+      "step": 8368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.521037264981563,
+      "learning_rate": 9.230286128529102e-06,
+      "loss": 0.2774,
+      "step": 8369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4460858982147116,
+      "learning_rate": 9.230076072229184e-06,
+      "loss": 0.3564,
+      "step": 8370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5300594611121563,
+      "learning_rate": 9.229865989661639e-06,
+      "loss": 0.4556,
+      "step": 8371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0081950613623123,
+      "learning_rate": 9.229655880827776e-06,
+      "loss": 0.5249,
+      "step": 8372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3807385592774903,
+      "learning_rate": 9.229445745728896e-06,
+      "loss": 0.2225,
+      "step": 8373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5678119369081547,
+      "learning_rate": 9.229235584366304e-06,
+      "loss": 0.3717,
+      "step": 8374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5155180224782088,
+      "learning_rate": 9.229025396741307e-06,
+      "loss": 0.3519,
+      "step": 8375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5556378742511368,
+      "learning_rate": 9.22881518285521e-06,
+      "loss": 0.2378,
+      "step": 8376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8045916882797683,
+      "learning_rate": 9.228604942709316e-06,
+      "loss": 0.2624,
+      "step": 8377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8663263370081071,
+      "learning_rate": 9.228394676304934e-06,
+      "loss": 0.5029,
+      "step": 8378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.558494256364686,
+      "learning_rate": 9.228184383643365e-06,
+      "loss": 0.2751,
+      "step": 8379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5850832545853497,
+      "learning_rate": 9.22797406472592e-06,
+      "loss": 0.2996,
+      "step": 8380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9958901215898291,
+      "learning_rate": 9.227763719553903e-06,
+      "loss": 0.229,
+      "step": 8381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6995229251507227,
+      "learning_rate": 9.22755334812862e-06,
+      "loss": 0.1593,
+      "step": 8382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.343929465917968,
+      "learning_rate": 9.227342950451376e-06,
+      "loss": 0.1578,
+      "step": 8383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.003760152179465,
+      "learning_rate": 9.227132526523479e-06,
+      "loss": 0.3054,
+      "step": 8384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7007190964591212,
+      "learning_rate": 9.226922076346237e-06,
+      "loss": 0.4907,
+      "step": 8385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.590613621631688,
+      "learning_rate": 9.226711599920952e-06,
+      "loss": 0.3743,
+      "step": 8386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9857037368403063,
+      "learning_rate": 9.226501097248937e-06,
+      "loss": 0.6079,
+      "step": 8387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7094733252656136,
+      "learning_rate": 9.226290568331495e-06,
+      "loss": 0.2172,
+      "step": 8388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.834840999010959,
+      "learning_rate": 9.226080013169936e-06,
+      "loss": 0.3311,
+      "step": 8389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5087496409775276,
+      "learning_rate": 9.225869431765564e-06,
+      "loss": 0.429,
+      "step": 8390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5103637427902123,
+      "learning_rate": 9.225658824119692e-06,
+      "loss": 0.2657,
+      "step": 8391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6305307525485446,
+      "learning_rate": 9.225448190233623e-06,
+      "loss": 0.4094,
+      "step": 8392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1592966645759428,
+      "learning_rate": 9.225237530108666e-06,
+      "loss": 0.1672,
+      "step": 8393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5367540869852374,
+      "learning_rate": 9.225026843746133e-06,
+      "loss": 0.5174,
+      "step": 8394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.482592498750627,
+      "learning_rate": 9.224816131147326e-06,
+      "loss": 0.1839,
+      "step": 8395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7720176583843454,
+      "learning_rate": 9.224605392313557e-06,
+      "loss": 0.425,
+      "step": 8396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4103587129393407,
+      "learning_rate": 9.224394627246135e-06,
+      "loss": 0.1796,
+      "step": 8397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6191332670810767,
+      "learning_rate": 9.22418383594637e-06,
+      "loss": 0.2678,
+      "step": 8398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9882715432462716,
+      "learning_rate": 9.223973018415567e-06,
+      "loss": 0.3528,
+      "step": 8399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.374267210972618,
+      "learning_rate": 9.223762174655035e-06,
+      "loss": 0.3048,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5870521297547926,
+      "learning_rate": 9.223551304666088e-06,
+      "loss": 0.2261,
+      "step": 8401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3059045649227685,
+      "learning_rate": 9.223340408450032e-06,
+      "loss": 0.2924,
+      "step": 8402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5171038550548492,
+      "learning_rate": 9.223129486008178e-06,
+      "loss": 0.3517,
+      "step": 8403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9820088224002943,
+      "learning_rate": 9.222918537341835e-06,
+      "loss": 0.2039,
+      "step": 8404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6048947767796156,
+      "learning_rate": 9.222707562452313e-06,
+      "loss": 0.3261,
+      "step": 8405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.35315722444764,
+      "learning_rate": 9.222496561340924e-06,
+      "loss": 0.2259,
+      "step": 8406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8721773199582425,
+      "learning_rate": 9.222285534008974e-06,
+      "loss": 0.4086,
+      "step": 8407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6590931915004683,
+      "learning_rate": 9.222074480457777e-06,
+      "loss": 0.3293,
+      "step": 8408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5649723091585477,
+      "learning_rate": 9.221863400688644e-06,
+      "loss": 0.2542,
+      "step": 8409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4300595700846834,
+      "learning_rate": 9.221652294702882e-06,
+      "loss": 0.2889,
+      "step": 8410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4574077480268466,
+      "learning_rate": 9.221441162501806e-06,
+      "loss": 0.2053,
+      "step": 8411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2023685603962275,
+      "learning_rate": 9.221230004086723e-06,
+      "loss": 0.2195,
+      "step": 8412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.267965673010273,
+      "learning_rate": 9.22101881945895e-06,
+      "loss": 0.4088,
+      "step": 8413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1283476000304926,
+      "learning_rate": 9.22080760861979e-06,
+      "loss": 0.443,
+      "step": 8414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.391143337901153,
+      "learning_rate": 9.220596371570562e-06,
+      "loss": 0.2106,
+      "step": 8415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.686076913024333,
+      "learning_rate": 9.220385108312576e-06,
+      "loss": 0.3466,
+      "step": 8416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.98097051851431,
+      "learning_rate": 9.220173818847143e-06,
+      "loss": 0.4032,
+      "step": 8417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.731166014544989,
+      "learning_rate": 9.219962503175575e-06,
+      "loss": 0.5997,
+      "step": 8418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9496854823426564,
+      "learning_rate": 9.219751161299183e-06,
+      "loss": 0.5227,
+      "step": 8419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7086784263318853,
+      "learning_rate": 9.219539793219283e-06,
+      "loss": 0.4623,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.628903755991689,
+      "learning_rate": 9.219328398937183e-06,
+      "loss": 0.3066,
+      "step": 8421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8052358619900744,
+      "learning_rate": 9.219116978454198e-06,
+      "loss": 0.5549,
+      "step": 8422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4599316928958208,
+      "learning_rate": 9.218905531771643e-06,
+      "loss": 0.1673,
+      "step": 8423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.07499892382099,
+      "learning_rate": 9.218694058890827e-06,
+      "loss": 0.4725,
+      "step": 8424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5593943511549597,
+      "learning_rate": 9.218482559813066e-06,
+      "loss": 0.2337,
+      "step": 8425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.476815751978294,
+      "learning_rate": 9.218271034539673e-06,
+      "loss": 0.2335,
+      "step": 8426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.80337491543096,
+      "learning_rate": 9.21805948307196e-06,
+      "loss": 0.4771,
+      "step": 8427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.803031223278276,
+      "learning_rate": 9.217847905411241e-06,
+      "loss": 0.5147,
+      "step": 8428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.612113040137001,
+      "learning_rate": 9.21763630155883e-06,
+      "loss": 0.29,
+      "step": 8429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.796159593554186,
+      "learning_rate": 9.217424671516043e-06,
+      "loss": 0.507,
+      "step": 8430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4250938535200541,
+      "learning_rate": 9.217213015284192e-06,
+      "loss": 0.2663,
+      "step": 8431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8334392513398772,
+      "learning_rate": 9.217001332864593e-06,
+      "loss": 0.4257,
+      "step": 8432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8459926479065074,
+      "learning_rate": 9.216789624258557e-06,
+      "loss": 0.3697,
+      "step": 8433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.821473794931358,
+      "learning_rate": 9.216577889467402e-06,
+      "loss": 0.1937,
+      "step": 8434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4995302716532526,
+      "learning_rate": 9.216366128492442e-06,
+      "loss": 0.4491,
+      "step": 8435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6475538601812112,
+      "learning_rate": 9.216154341334993e-06,
+      "loss": 0.2015,
+      "step": 8436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5322513566159746,
+      "learning_rate": 9.215942527996367e-06,
+      "loss": 0.3394,
+      "step": 8437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6599394982153317,
+      "learning_rate": 9.215730688477884e-06,
+      "loss": 0.4862,
+      "step": 8438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0021023578381647,
+      "learning_rate": 9.215518822780855e-06,
+      "loss": 0.5328,
+      "step": 8439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6644389528155965,
+      "learning_rate": 9.215306930906596e-06,
+      "loss": 0.3188,
+      "step": 8440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.200162241382838,
+      "learning_rate": 9.215095012856426e-06,
+      "loss": 0.4823,
+      "step": 8441
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3996351164639325,
+      "learning_rate": 9.21488306863166e-06,
+      "loss": 0.1904,
+      "step": 8442
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7338840288008694,
+      "learning_rate": 9.214671098233612e-06,
+      "loss": 0.5207,
+      "step": 8443
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4239572746405007,
+      "learning_rate": 9.2144591016636e-06,
+      "loss": 0.3214,
+      "step": 8444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1322804021297075,
+      "learning_rate": 9.21424707892294e-06,
+      "loss": 0.4621,
+      "step": 8445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8620266239191703,
+      "learning_rate": 9.214035030012946e-06,
+      "loss": 0.3689,
+      "step": 8446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3642550054042064,
+      "learning_rate": 9.213822954934941e-06,
+      "loss": 0.2369,
+      "step": 8447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.901634769586618,
+      "learning_rate": 9.213610853690237e-06,
+      "loss": 0.2623,
+      "step": 8448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6194664198269764,
+      "learning_rate": 9.213398726280152e-06,
+      "loss": 0.3394,
+      "step": 8449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.112069947121831,
+      "learning_rate": 9.213186572706003e-06,
+      "loss": 0.3458,
+      "step": 8450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4707891684764487,
+      "learning_rate": 9.21297439296911e-06,
+      "loss": 0.3202,
+      "step": 8451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5599164441317408,
+      "learning_rate": 9.212762187070788e-06,
+      "loss": 0.2789,
+      "step": 8452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1085770299267583,
+      "learning_rate": 9.212549955012353e-06,
+      "loss": 0.1189,
+      "step": 8453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1067485696759878,
+      "learning_rate": 9.212337696795128e-06,
+      "loss": 0.203,
+      "step": 8454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.494477551936818,
+      "learning_rate": 9.212125412420428e-06,
+      "loss": 0.3624,
+      "step": 8455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.450618630735885,
+      "learning_rate": 9.21191310188957e-06,
+      "loss": 0.188,
+      "step": 8456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2667089434378913,
+      "learning_rate": 9.211700765203874e-06,
+      "loss": 0.613,
+      "step": 8457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5509361662946382,
+      "learning_rate": 9.211488402364662e-06,
+      "loss": 0.2632,
+      "step": 8458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4308493432378828,
+      "learning_rate": 9.211276013373245e-06,
+      "loss": 0.1945,
+      "step": 8459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3143063643262085,
+      "learning_rate": 9.211063598230947e-06,
+      "loss": 0.2473,
+      "step": 8460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0547954327913875,
+      "learning_rate": 9.210851156939087e-06,
+      "loss": 0.3426,
+      "step": 8461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4287839546240588,
+      "learning_rate": 9.210638689498983e-06,
+      "loss": 0.3305,
+      "step": 8462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.275041576348806,
+      "learning_rate": 9.210426195911957e-06,
+      "loss": 0.2175,
+      "step": 8463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6507277736857335,
+      "learning_rate": 9.210213676179323e-06,
+      "loss": 0.4642,
+      "step": 8464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7357225874773627,
+      "learning_rate": 9.210001130302405e-06,
+      "loss": 0.3681,
+      "step": 8465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3324926386908533,
+      "learning_rate": 9.209788558282522e-06,
+      "loss": 0.1983,
+      "step": 8466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5471714222233874,
+      "learning_rate": 9.209575960120993e-06,
+      "loss": 0.2024,
+      "step": 8467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8648567017072315,
+      "learning_rate": 9.209363335819138e-06,
+      "loss": 0.309,
+      "step": 8468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2127981319574068,
+      "learning_rate": 9.209150685378281e-06,
+      "loss": 0.1642,
+      "step": 8469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.101861877459994,
+      "learning_rate": 9.208938008799739e-06,
+      "loss": 0.4211,
+      "step": 8470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7448974977015503,
+      "learning_rate": 9.208725306084834e-06,
+      "loss": 0.1746,
+      "step": 8471
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5967955381028651,
+      "learning_rate": 9.208512577234884e-06,
+      "loss": 0.2121,
+      "step": 8472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.325110174830227,
+      "learning_rate": 9.208299822251215e-06,
+      "loss": 0.2131,
+      "step": 8473
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.915109923907989,
+      "learning_rate": 9.208087041135142e-06,
+      "loss": 0.4759,
+      "step": 8474
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.895750263161775,
+      "learning_rate": 9.207874233887993e-06,
+      "loss": 0.5757,
+      "step": 8475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6516152392423549,
+      "learning_rate": 9.207661400511086e-06,
+      "loss": 0.31,
+      "step": 8476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5898703542446975,
+      "learning_rate": 9.207448541005741e-06,
+      "loss": 0.5098,
+      "step": 8477
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7388007127193597,
+      "learning_rate": 9.207235655373284e-06,
+      "loss": 0.243,
+      "step": 8478
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7746244693141873,
+      "learning_rate": 9.207022743615032e-06,
+      "loss": 0.4113,
+      "step": 8479
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4044005444584085,
+      "learning_rate": 9.206809805732311e-06,
+      "loss": 0.2684,
+      "step": 8480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.708840975603439,
+      "learning_rate": 9.206596841726442e-06,
+      "loss": 0.4451,
+      "step": 8481
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.486081106585152,
+      "learning_rate": 9.206383851598747e-06,
+      "loss": 0.2654,
+      "step": 8482
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9030272726514177,
+      "learning_rate": 9.20617083535055e-06,
+      "loss": 0.3769,
+      "step": 8483
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6657255083359261,
+      "learning_rate": 9.205957792983172e-06,
+      "loss": 0.3062,
+      "step": 8484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.589583445897313,
+      "learning_rate": 9.205744724497938e-06,
+      "loss": 0.3947,
+      "step": 8485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7115356061799507,
+      "learning_rate": 9.205531629896168e-06,
+      "loss": 0.4269,
+      "step": 8486
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7897303357151058,
+      "learning_rate": 9.205318509179188e-06,
+      "loss": 0.2069,
+      "step": 8487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4305002320521758,
+      "learning_rate": 9.205105362348321e-06,
+      "loss": 0.3149,
+      "step": 8488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6770157279630225,
+      "learning_rate": 9.20489218940489e-06,
+      "loss": 0.388,
+      "step": 8489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6521542826034017,
+      "learning_rate": 9.20467899035022e-06,
+      "loss": 0.4664,
+      "step": 8490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.371454217998057,
+      "learning_rate": 9.204465765185631e-06,
+      "loss": 0.1995,
+      "step": 8491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2021133589223862,
+      "learning_rate": 9.204252513912453e-06,
+      "loss": 0.4066,
+      "step": 8492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.172716999816387,
+      "learning_rate": 9.204039236532006e-06,
+      "loss": 0.5733,
+      "step": 8493
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6671375419117471,
+      "learning_rate": 9.203825933045614e-06,
+      "loss": 0.3696,
+      "step": 8494
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.773399781149488,
+      "learning_rate": 9.203612603454605e-06,
+      "loss": 0.456,
+      "step": 8495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8960561534044675,
+      "learning_rate": 9.203399247760301e-06,
+      "loss": 0.5502,
+      "step": 8496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0273823339359147,
+      "learning_rate": 9.203185865964028e-06,
+      "loss": 0.3838,
+      "step": 8497
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9065786101906896,
+      "learning_rate": 9.20297245806711e-06,
+      "loss": 0.3335,
+      "step": 8498
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3982724926207863,
+      "learning_rate": 9.202759024070873e-06,
+      "loss": 0.26,
+      "step": 8499
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6577208978151947,
+      "learning_rate": 9.202545563976644e-06,
+      "loss": 0.4007,
+      "step": 8500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6953552961441314,
+      "learning_rate": 9.202332077785745e-06,
+      "loss": 0.4349,
+      "step": 8501
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.873783496895426,
+      "learning_rate": 9.202118565499503e-06,
+      "loss": 0.2438,
+      "step": 8502
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3831326442753762,
+      "learning_rate": 9.201905027119245e-06,
+      "loss": 0.314,
+      "step": 8503
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.569706363317816,
+      "learning_rate": 9.201691462646296e-06,
+      "loss": 0.4066,
+      "step": 8504
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9153132886693585,
+      "learning_rate": 9.201477872081985e-06,
+      "loss": 0.5246,
+      "step": 8505
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7544722334155851,
+      "learning_rate": 9.201264255427633e-06,
+      "loss": 0.2264,
+      "step": 8506
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2140460492520628,
+      "learning_rate": 9.201050612684571e-06,
+      "loss": 0.19,
+      "step": 8507
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6311051965803904,
+      "learning_rate": 9.200836943854124e-06,
+      "loss": 0.4875,
+      "step": 8508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9710329382818443,
+      "learning_rate": 9.200623248937619e-06,
+      "loss": 0.3774,
+      "step": 8509
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.897433026982669,
+      "learning_rate": 9.200409527936382e-06,
+      "loss": 0.4011,
+      "step": 8510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3853100888648413,
+      "learning_rate": 9.200195780851742e-06,
+      "loss": 0.2942,
+      "step": 8511
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8080796815010411,
+      "learning_rate": 9.199982007685024e-06,
+      "loss": 0.3691,
+      "step": 8512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5687157557367017,
+      "learning_rate": 9.199768208437558e-06,
+      "loss": 0.2015,
+      "step": 8513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4943430445567765,
+      "learning_rate": 9.19955438311067e-06,
+      "loss": 0.2881,
+      "step": 8514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5700293423255403,
+      "learning_rate": 9.199340531705688e-06,
+      "loss": 0.1952,
+      "step": 8515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9405912918505013,
+      "learning_rate": 9.19912665422394e-06,
+      "loss": 0.2904,
+      "step": 8516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.238315429877628,
+      "learning_rate": 9.198912750666755e-06,
+      "loss": 0.3514,
+      "step": 8517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.02034310060849,
+      "learning_rate": 9.198698821035458e-06,
+      "loss": 0.2056,
+      "step": 8518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5566404311540165,
+      "learning_rate": 9.198484865331383e-06,
+      "loss": 0.2765,
+      "step": 8519
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7080385849864976,
+      "learning_rate": 9.198270883555855e-06,
+      "loss": 0.3865,
+      "step": 8520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8508911444781144,
+      "learning_rate": 9.198056875710204e-06,
+      "loss": 0.4456,
+      "step": 8521
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4675487008697583,
+      "learning_rate": 9.197842841795758e-06,
+      "loss": 0.543,
+      "step": 8522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3604785657256737,
+      "learning_rate": 9.197628781813846e-06,
+      "loss": 0.228,
+      "step": 8523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2042357496840101,
+      "learning_rate": 9.197414695765798e-06,
+      "loss": 0.1907,
+      "step": 8524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9449163860880267,
+      "learning_rate": 9.197200583652944e-06,
+      "loss": 0.539,
+      "step": 8525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8346155764335217,
+      "learning_rate": 9.19698644547661e-06,
+      "loss": 0.4389,
+      "step": 8526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1030026534830852,
+      "learning_rate": 9.196772281238132e-06,
+      "loss": 0.1875,
+      "step": 8527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2303025383179005,
+      "learning_rate": 9.196558090938832e-06,
+      "loss": 0.1964,
+      "step": 8528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8256086653958443,
+      "learning_rate": 9.196343874580047e-06,
+      "loss": 0.3799,
+      "step": 8529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8766309955937934,
+      "learning_rate": 9.196129632163105e-06,
+      "loss": 0.2262,
+      "step": 8530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5388440398261736,
+      "learning_rate": 9.195915363689336e-06,
+      "loss": 0.3376,
+      "step": 8531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7677937117979632,
+      "learning_rate": 9.195701069160071e-06,
+      "loss": 0.3546,
+      "step": 8532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.677301082237191,
+      "learning_rate": 9.195486748576638e-06,
+      "loss": 0.3184,
+      "step": 8533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7098120118564482,
+      "learning_rate": 9.195272401940371e-06,
+      "loss": 0.5796,
+      "step": 8534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4148653317326811,
+      "learning_rate": 9.195058029252601e-06,
+      "loss": 0.3081,
+      "step": 8535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0316324897396238,
+      "learning_rate": 9.194843630514658e-06,
+      "loss": 0.196,
+      "step": 8536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.469595848176581,
+      "learning_rate": 9.194629205727875e-06,
+      "loss": 0.3639,
+      "step": 8537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.690382738083985,
+      "learning_rate": 9.19441475489358e-06,
+      "loss": 0.3474,
+      "step": 8538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0424704883175653,
+      "learning_rate": 9.194200278013107e-06,
+      "loss": 0.3357,
+      "step": 8539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3346029944474733,
+      "learning_rate": 9.19398577508779e-06,
+      "loss": 0.2616,
+      "step": 8540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.831518870199991,
+      "learning_rate": 9.193771246118957e-06,
+      "loss": 0.5531,
+      "step": 8541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4539918028690073,
+      "learning_rate": 9.193556691107942e-06,
+      "loss": 0.1327,
+      "step": 8542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.751047825466795,
+      "learning_rate": 9.193342110056079e-06,
+      "loss": 0.3592,
+      "step": 8543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8995503040866437,
+      "learning_rate": 9.193127502964695e-06,
+      "loss": 0.7047,
+      "step": 8544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.770883241585931,
+      "learning_rate": 9.192912869835131e-06,
+      "loss": 0.2884,
+      "step": 8545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3005390965679464,
+      "learning_rate": 9.192698210668712e-06,
+      "loss": 0.168,
+      "step": 8546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1572877779193418,
+      "learning_rate": 9.192483525466776e-06,
+      "loss": 0.1508,
+      "step": 8547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3867335708255986,
+      "learning_rate": 9.192268814230654e-06,
+      "loss": 0.337,
+      "step": 8548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7399315575931509,
+      "learning_rate": 9.192054076961678e-06,
+      "loss": 0.5425,
+      "step": 8549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6629347276070237,
+      "learning_rate": 9.191839313661187e-06,
+      "loss": 0.2313,
+      "step": 8550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5167077758697005,
+      "learning_rate": 9.191624524330507e-06,
+      "loss": 0.4057,
+      "step": 8551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2253406752742046,
+      "learning_rate": 9.191409708970976e-06,
+      "loss": 0.2628,
+      "step": 8552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.737305427280931,
+      "learning_rate": 9.191194867583928e-06,
+      "loss": 0.3932,
+      "step": 8553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8789128483836839,
+      "learning_rate": 9.190980000170697e-06,
+      "loss": 0.4059,
+      "step": 8554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5263647517540573,
+      "learning_rate": 9.190765106732615e-06,
+      "loss": 0.3757,
+      "step": 8555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.129119447026386,
+      "learning_rate": 9.19055018727102e-06,
+      "loss": 0.2456,
+      "step": 8556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5496258166714814,
+      "learning_rate": 9.190335241787245e-06,
+      "loss": 0.3121,
+      "step": 8557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7379760731470175,
+      "learning_rate": 9.190120270282625e-06,
+      "loss": 0.4143,
+      "step": 8558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7384981534410215,
+      "learning_rate": 9.189905272758494e-06,
+      "loss": 0.2989,
+      "step": 8559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3693436726657775,
+      "learning_rate": 9.189690249216187e-06,
+      "loss": 0.494,
+      "step": 8560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7114320588915044,
+      "learning_rate": 9.18947519965704e-06,
+      "loss": 0.4269,
+      "step": 8561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2117019265281863,
+      "learning_rate": 9.18926012408239e-06,
+      "loss": 0.4119,
+      "step": 8562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.252169463857028,
+      "learning_rate": 9.189045022493567e-06,
+      "loss": 0.1744,
+      "step": 8563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6486414290565548,
+      "learning_rate": 9.188829894891913e-06,
+      "loss": 0.1958,
+      "step": 8564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4721394271993902,
+      "learning_rate": 9.188614741278762e-06,
+      "loss": 0.3464,
+      "step": 8565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6539111312556325,
+      "learning_rate": 9.188399561655447e-06,
+      "loss": 0.4788,
+      "step": 8566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3721218737488168,
+      "learning_rate": 9.188184356023309e-06,
+      "loss": 0.2933,
+      "step": 8567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4997485367280894,
+      "learning_rate": 9.18796912438368e-06,
+      "loss": 0.3534,
+      "step": 8568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3308208663224015,
+      "learning_rate": 9.1877538667379e-06,
+      "loss": 0.2496,
+      "step": 8569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9216047694020297,
+      "learning_rate": 9.187538583087305e-06,
+      "loss": 0.4957,
+      "step": 8570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.200470851621783,
+      "learning_rate": 9.18732327343323e-06,
+      "loss": 0.2101,
+      "step": 8571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.670652459486559,
+      "learning_rate": 9.187107937777013e-06,
+      "loss": 0.227,
+      "step": 8572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0469385475748174,
+      "learning_rate": 9.186892576119993e-06,
+      "loss": 0.1483,
+      "step": 8573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9048216815527812,
+      "learning_rate": 9.186677188463503e-06,
+      "loss": 0.4934,
+      "step": 8574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4611965347588989,
+      "learning_rate": 9.186461774808883e-06,
+      "loss": 0.2032,
+      "step": 8575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5316231915105154,
+      "learning_rate": 9.186246335157472e-06,
+      "loss": 0.373,
+      "step": 8576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8761566409291337,
+      "learning_rate": 9.186030869510607e-06,
+      "loss": 0.3495,
+      "step": 8577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5922266935666325,
+      "learning_rate": 9.185815377869626e-06,
+      "loss": 0.4388,
+      "step": 8578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.294127053133459,
+      "learning_rate": 9.185599860235866e-06,
+      "loss": 0.6797,
+      "step": 8579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9059600384807676,
+      "learning_rate": 9.185384316610666e-06,
+      "loss": 0.2012,
+      "step": 8580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.012037155730442,
+      "learning_rate": 9.185168746995365e-06,
+      "loss": 0.5936,
+      "step": 8581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3933933635167033,
+      "learning_rate": 9.1849531513913e-06,
+      "loss": 0.2837,
+      "step": 8582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.109574986620319,
+      "learning_rate": 9.184737529799811e-06,
+      "loss": 0.4377,
+      "step": 8583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8671893375180193,
+      "learning_rate": 9.184521882222237e-06,
+      "loss": 0.4858,
+      "step": 8584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.870753820835215,
+      "learning_rate": 9.184306208659918e-06,
+      "loss": 0.6446,
+      "step": 8585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.381065024206542,
+      "learning_rate": 9.184090509114191e-06,
+      "loss": 0.1954,
+      "step": 8586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9833654381102996,
+      "learning_rate": 9.183874783586397e-06,
+      "loss": 0.1511,
+      "step": 8587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9095288964636217,
+      "learning_rate": 9.183659032077876e-06,
+      "loss": 0.4957,
+      "step": 8588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6248149399652307,
+      "learning_rate": 9.183443254589967e-06,
+      "loss": 0.3713,
+      "step": 8589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.826930744163741,
+      "learning_rate": 9.183227451124011e-06,
+      "loss": 0.4585,
+      "step": 8590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8709271462317911,
+      "learning_rate": 9.183011621681345e-06,
+      "loss": 0.3145,
+      "step": 8591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.044451042703966,
+      "learning_rate": 9.182795766263312e-06,
+      "loss": 0.5292,
+      "step": 8592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4445089142487348,
+      "learning_rate": 9.182579884871252e-06,
+      "loss": 0.3202,
+      "step": 8593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8655065707376266,
+      "learning_rate": 9.182363977506505e-06,
+      "loss": 0.3404,
+      "step": 8594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3294032256936996,
+      "learning_rate": 9.182148044170411e-06,
+      "loss": 0.3773,
+      "step": 8595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.605314354073982,
+      "learning_rate": 9.181932084864314e-06,
+      "loss": 0.3635,
+      "step": 8596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2108669075847571,
+      "learning_rate": 9.18171609958955e-06,
+      "loss": 0.3087,
+      "step": 8597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3890171936164304,
+      "learning_rate": 9.181500088347466e-06,
+      "loss": 0.3174,
+      "step": 8598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2459730191474443,
+      "learning_rate": 9.1812840511394e-06,
+      "loss": 0.1743,
+      "step": 8599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8863680555075384,
+      "learning_rate": 9.181067987966693e-06,
+      "loss": 0.2423,
+      "step": 8600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.360988263925439,
+      "learning_rate": 9.180851898830689e-06,
+      "loss": 0.2157,
+      "step": 8601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5817330986042142,
+      "learning_rate": 9.180635783732727e-06,
+      "loss": 0.2876,
+      "step": 8602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9357256378553316,
+      "learning_rate": 9.180419642674153e-06,
+      "loss": 0.4448,
+      "step": 8603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1704061457979436,
+      "learning_rate": 9.180203475656304e-06,
+      "loss": 0.3781,
+      "step": 8604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8869353598492535,
+      "learning_rate": 9.179987282680525e-06,
+      "loss": 0.369,
+      "step": 8605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.438043274014117,
+      "learning_rate": 9.17977106374816e-06,
+      "loss": 0.3361,
+      "step": 8606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0136502802024119,
+      "learning_rate": 9.17955481886055e-06,
+      "loss": 0.1814,
+      "step": 8607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0236260904333583,
+      "learning_rate": 9.179338548019039e-06,
+      "loss": 0.6213,
+      "step": 8608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4572480477637229,
+      "learning_rate": 9.17912225122497e-06,
+      "loss": 0.4381,
+      "step": 8609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.505812274379315,
+      "learning_rate": 9.178905928479683e-06,
+      "loss": 0.2877,
+      "step": 8610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7053703241259697,
+      "learning_rate": 9.178689579784524e-06,
+      "loss": 0.4226,
+      "step": 8611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0917647123291074,
+      "learning_rate": 9.178473205140836e-06,
+      "loss": 0.1645,
+      "step": 8612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2779184850790302,
+      "learning_rate": 9.178256804549964e-06,
+      "loss": 0.3292,
+      "step": 8613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6931967211684822,
+      "learning_rate": 9.178040378013249e-06,
+      "loss": 0.4798,
+      "step": 8614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2393851894878267,
+      "learning_rate": 9.177823925532037e-06,
+      "loss": 0.3312,
+      "step": 8615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9872754368984282,
+      "learning_rate": 9.17760744710767e-06,
+      "loss": 0.4695,
+      "step": 8616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8486669711560704,
+      "learning_rate": 9.177390942741496e-06,
+      "loss": 0.5111,
+      "step": 8617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.560204175863416,
+      "learning_rate": 9.177174412434855e-06,
+      "loss": 0.3894,
+      "step": 8618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9030634166982547,
+      "learning_rate": 9.176957856189095e-06,
+      "loss": 0.472,
+      "step": 8619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1342131768549657,
+      "learning_rate": 9.176741274005558e-06,
+      "loss": 0.1251,
+      "step": 8620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.946564381836628,
+      "learning_rate": 9.176524665885593e-06,
+      "loss": 0.3321,
+      "step": 8621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8743719559098138,
+      "learning_rate": 9.17630803183054e-06,
+      "loss": 0.4717,
+      "step": 8622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.043496772990375,
+      "learning_rate": 9.176091371841749e-06,
+      "loss": 0.2128,
+      "step": 8623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8714347401221412,
+      "learning_rate": 9.175874685920562e-06,
+      "loss": 0.3246,
+      "step": 8624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5227953560058063,
+      "learning_rate": 9.175657974068328e-06,
+      "loss": 0.3274,
+      "step": 8625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1620305502702184,
+      "learning_rate": 9.175441236286387e-06,
+      "loss": 0.4582,
+      "step": 8626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3062585277712495,
+      "learning_rate": 9.17522447257609e-06,
+      "loss": 0.1981,
+      "step": 8627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4790421912602196,
+      "learning_rate": 9.175007682938781e-06,
+      "loss": 0.2764,
+      "step": 8628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2157005018758964,
+      "learning_rate": 9.174790867375807e-06,
+      "loss": 0.4792,
+      "step": 8629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2687451995561483,
+      "learning_rate": 9.174574025888514e-06,
+      "loss": 0.192,
+      "step": 8630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8302799607771976,
+      "learning_rate": 9.174357158478249e-06,
+      "loss": 0.1624,
+      "step": 8631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.483906405921776,
+      "learning_rate": 9.174140265146356e-06,
+      "loss": 0.3617,
+      "step": 8632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9525670799663704,
+      "learning_rate": 9.173923345894186e-06,
+      "loss": 0.2718,
+      "step": 8633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8172883424798116,
+      "learning_rate": 9.173706400723084e-06,
+      "loss": 0.5032,
+      "step": 8634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7742997270696983,
+      "learning_rate": 9.173489429634397e-06,
+      "loss": 0.4742,
+      "step": 8635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7466149373696611,
+      "learning_rate": 9.173272432629472e-06,
+      "loss": 0.4626,
+      "step": 8636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6696615451206935,
+      "learning_rate": 9.173055409709658e-06,
+      "loss": 0.1665,
+      "step": 8637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6185556190094925,
+      "learning_rate": 9.172838360876299e-06,
+      "loss": 0.3074,
+      "step": 8638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6286154013940908,
+      "learning_rate": 9.172621286130747e-06,
+      "loss": 0.3569,
+      "step": 8639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.610454810483664,
+      "learning_rate": 9.17240418547435e-06,
+      "loss": 0.3101,
+      "step": 8640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7040406136967214,
+      "learning_rate": 9.172187058908452e-06,
+      "loss": 0.4529,
+      "step": 8641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6636032222887656,
+      "learning_rate": 9.171969906434405e-06,
+      "loss": 0.3513,
+      "step": 8642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.900040558018218,
+      "learning_rate": 9.171752728053558e-06,
+      "loss": 0.4577,
+      "step": 8643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.352076920020538,
+      "learning_rate": 9.171535523767257e-06,
+      "loss": 0.3087,
+      "step": 8644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6500783500709162,
+      "learning_rate": 9.171318293576851e-06,
+      "loss": 0.4524,
+      "step": 8645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.610322710268034,
+      "learning_rate": 9.17110103748369e-06,
+      "loss": 0.3866,
+      "step": 8646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.745050705474308,
+      "learning_rate": 9.170883755489124e-06,
+      "loss": 0.1901,
+      "step": 8647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6873010897849199,
+      "learning_rate": 9.170666447594499e-06,
+      "loss": 0.562,
+      "step": 8648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6575701104018148,
+      "learning_rate": 9.170449113801167e-06,
+      "loss": 0.3519,
+      "step": 8649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7011594892329052,
+      "learning_rate": 9.170231754110477e-06,
+      "loss": 0.3579,
+      "step": 8650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.793315644043662,
+      "learning_rate": 9.170014368523781e-06,
+      "loss": 0.2096,
+      "step": 8651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.395721121652241,
+      "learning_rate": 9.169796957042424e-06,
+      "loss": 0.2145,
+      "step": 8652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7318506414042694,
+      "learning_rate": 9.16957951966776e-06,
+      "loss": 0.5877,
+      "step": 8653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9001617450633086,
+      "learning_rate": 9.169362056401139e-06,
+      "loss": 0.3402,
+      "step": 8654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6334861775595173,
+      "learning_rate": 9.169144567243908e-06,
+      "loss": 0.417,
+      "step": 8655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3363612239187534,
+      "learning_rate": 9.168927052197421e-06,
+      "loss": 0.1285,
+      "step": 8656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5423311871934915,
+      "learning_rate": 9.168709511263028e-06,
+      "loss": 0.277,
+      "step": 8657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5305748248959457,
+      "learning_rate": 9.16849194444208e-06,
+      "loss": 0.2826,
+      "step": 8658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2747832605062395,
+      "learning_rate": 9.168274351735926e-06,
+      "loss": 0.6842,
+      "step": 8659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8885541660272531,
+      "learning_rate": 9.168056733145921e-06,
+      "loss": 0.4467,
+      "step": 8660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6281247539849562,
+      "learning_rate": 9.167839088673412e-06,
+      "loss": 0.3902,
+      "step": 8661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8825939257368254,
+      "learning_rate": 9.167621418319753e-06,
+      "loss": 0.2884,
+      "step": 8662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2262201097072505,
+      "learning_rate": 9.167403722086295e-06,
+      "loss": 0.1975,
+      "step": 8663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6864563354855775,
+      "learning_rate": 9.16718599997439e-06,
+      "loss": 0.269,
+      "step": 8664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.244919103685783,
+      "learning_rate": 9.166968251985393e-06,
+      "loss": 0.1797,
+      "step": 8665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.281655991592067,
+      "learning_rate": 9.16675047812065e-06,
+      "loss": 0.287,
+      "step": 8666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4849226393120416,
+      "learning_rate": 9.166532678381518e-06,
+      "loss": 0.3411,
+      "step": 8667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8636089092836507,
+      "learning_rate": 9.166314852769348e-06,
+      "loss": 0.5982,
+      "step": 8668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8971825089606564,
+      "learning_rate": 9.166097001285491e-06,
+      "loss": 0.3825,
+      "step": 8669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.408105196041052,
+      "learning_rate": 9.165879123931305e-06,
+      "loss": 0.2325,
+      "step": 8670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8058573329021996,
+      "learning_rate": 9.165661220708137e-06,
+      "loss": 0.5572,
+      "step": 8671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5974856626008274,
+      "learning_rate": 9.165443291617343e-06,
+      "loss": 0.2254,
+      "step": 8672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.717254922040453,
+      "learning_rate": 9.165225336660275e-06,
+      "loss": 0.2214,
+      "step": 8673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8772675273723218,
+      "learning_rate": 9.165007355838289e-06,
+      "loss": 0.4516,
+      "step": 8674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5098222534396657,
+      "learning_rate": 9.164789349152736e-06,
+      "loss": 0.3287,
+      "step": 8675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6837353514049689,
+      "learning_rate": 9.16457131660497e-06,
+      "loss": 0.2734,
+      "step": 8676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.104545638400351,
+      "learning_rate": 9.164353258196349e-06,
+      "loss": 0.3342,
+      "step": 8677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3458461214890955,
+      "learning_rate": 9.164135173928219e-06,
+      "loss": 0.3576,
+      "step": 8678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8926872403052986,
+      "learning_rate": 9.163917063801942e-06,
+      "loss": 0.4716,
+      "step": 8679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5627345931435057,
+      "learning_rate": 9.163698927818867e-06,
+      "loss": 0.3892,
+      "step": 8680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.897667897590331,
+      "learning_rate": 9.163480765980353e-06,
+      "loss": 0.543,
+      "step": 8681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7208697003502345,
+      "learning_rate": 9.163262578287752e-06,
+      "loss": 0.2617,
+      "step": 8682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.042761557962042,
+      "learning_rate": 9.16304436474242e-06,
+      "loss": 0.161,
+      "step": 8683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.023248376440059,
+      "learning_rate": 9.16282612534571e-06,
+      "loss": 0.3118,
+      "step": 8684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4510564539973552,
+      "learning_rate": 9.16260786009898e-06,
+      "loss": 0.3561,
+      "step": 8685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.606998577907766,
+      "learning_rate": 9.162389569003583e-06,
+      "loss": 0.2874,
+      "step": 8686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3239926346289215,
+      "learning_rate": 9.162171252060877e-06,
+      "loss": 0.2085,
+      "step": 8687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.17351954789999,
+      "learning_rate": 9.161952909272215e-06,
+      "loss": 0.55,
+      "step": 8688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8747782695318576,
+      "learning_rate": 9.161734540638955e-06,
+      "loss": 0.5278,
+      "step": 8689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9063715895975635,
+      "learning_rate": 9.161516146162453e-06,
+      "loss": 0.4325,
+      "step": 8690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.005461262360918,
+      "learning_rate": 9.161297725844062e-06,
+      "loss": 0.4275,
+      "step": 8691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7593211164065972,
+      "learning_rate": 9.161079279685143e-06,
+      "loss": 0.3546,
+      "step": 8692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7260511642276097,
+      "learning_rate": 9.160860807687048e-06,
+      "loss": 0.3936,
+      "step": 8693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8251464494315197,
+      "learning_rate": 9.160642309851137e-06,
+      "loss": 0.3838,
+      "step": 8694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9420301022205095,
+      "learning_rate": 9.160423786178767e-06,
+      "loss": 0.4025,
+      "step": 8695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3723438519713755,
+      "learning_rate": 9.160205236671293e-06,
+      "loss": 0.2075,
+      "step": 8696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0564534083085753,
+      "learning_rate": 9.159986661330071e-06,
+      "loss": 0.2356,
+      "step": 8697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8639910963248987,
+      "learning_rate": 9.159768060156462e-06,
+      "loss": 0.4901,
+      "step": 8698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0391540872167395,
+      "learning_rate": 9.15954943315182e-06,
+      "loss": 0.4315,
+      "step": 8699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4075689805881297,
+      "learning_rate": 9.159330780317506e-06,
+      "loss": 0.285,
+      "step": 8700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7095950381664469,
+      "learning_rate": 9.159112101654875e-06,
+      "loss": 0.2743,
+      "step": 8701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9812421285611297,
+      "learning_rate": 9.158893397165286e-06,
+      "loss": 0.5386,
+      "step": 8702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9582396353379512,
+      "learning_rate": 9.158674666850096e-06,
+      "loss": 0.2805,
+      "step": 8703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4760671941604526,
+      "learning_rate": 9.158455910710666e-06,
+      "loss": 0.4128,
+      "step": 8704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.414501384850834,
+      "learning_rate": 9.15823712874835e-06,
+      "loss": 0.446,
+      "step": 8705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.48705643072824,
+      "learning_rate": 9.158018320964511e-06,
+      "loss": 0.4638,
+      "step": 8706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2047727049420476,
+      "learning_rate": 9.157799487360506e-06,
+      "loss": 0.2065,
+      "step": 8707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4181110400012524,
+      "learning_rate": 9.157580627937693e-06,
+      "loss": 0.3202,
+      "step": 8708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6037033287149378,
+      "learning_rate": 9.157361742697433e-06,
+      "loss": 0.3494,
+      "step": 8709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.645693516652281,
+      "learning_rate": 9.157142831641082e-06,
+      "loss": 0.2198,
+      "step": 8710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5406449383379806,
+      "learning_rate": 9.156923894770003e-06,
+      "loss": 0.1799,
+      "step": 8711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9225424057397238,
+      "learning_rate": 9.156704932085554e-06,
+      "loss": 0.5418,
+      "step": 8712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6860526354933478,
+      "learning_rate": 9.156485943589094e-06,
+      "loss": 0.2736,
+      "step": 8713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.968132535290701,
+      "learning_rate": 9.156266929281985e-06,
+      "loss": 0.6727,
+      "step": 8714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9631216530041125,
+      "learning_rate": 9.156047889165584e-06,
+      "loss": 0.4018,
+      "step": 8715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8093753818814626,
+      "learning_rate": 9.155828823241253e-06,
+      "loss": 0.3008,
+      "step": 8716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7418733600113248,
+      "learning_rate": 9.155609731510352e-06,
+      "loss": 0.3233,
+      "step": 8717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6090050545271353,
+      "learning_rate": 9.155390613974242e-06,
+      "loss": 0.3402,
+      "step": 8718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8072633402104217,
+      "learning_rate": 9.155171470634285e-06,
+      "loss": 0.4962,
+      "step": 8719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8930819129259553,
+      "learning_rate": 9.154952301491837e-06,
+      "loss": 0.3496,
+      "step": 8720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9918935558901896,
+      "learning_rate": 9.154733106548263e-06,
+      "loss": 0.4631,
+      "step": 8721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7850563472114056,
+      "learning_rate": 9.154513885804926e-06,
+      "loss": 0.2249,
+      "step": 8722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2905886920920457,
+      "learning_rate": 9.154294639263182e-06,
+      "loss": 0.3473,
+      "step": 8723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2975194993452472,
+      "learning_rate": 9.154075366924396e-06,
+      "loss": 0.4077,
+      "step": 8724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2458846774838677,
+      "learning_rate": 9.153856068789926e-06,
+      "loss": 0.1349,
+      "step": 8725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.052560702743526,
+      "learning_rate": 9.153636744861138e-06,
+      "loss": 0.6045,
+      "step": 8726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6354426847097183,
+      "learning_rate": 9.153417395139392e-06,
+      "loss": 0.3146,
+      "step": 8727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.405034708261067,
+      "learning_rate": 9.153198019626052e-06,
+      "loss": 0.1838,
+      "step": 8728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4479528023535286,
+      "learning_rate": 9.152978618322476e-06,
+      "loss": 0.2685,
+      "step": 8729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.963900216316632,
+      "learning_rate": 9.152759191230032e-06,
+      "loss": 0.4086,
+      "step": 8730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7009654965568004,
+      "learning_rate": 9.152539738350078e-06,
+      "loss": 0.298,
+      "step": 8731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4109739273539665,
+      "learning_rate": 9.15232025968398e-06,
+      "loss": 0.3985,
+      "step": 8732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.54750368077791,
+      "learning_rate": 9.152100755233097e-06,
+      "loss": 0.3362,
+      "step": 8733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7495116633873817,
+      "learning_rate": 9.151881224998796e-06,
+      "loss": 0.4867,
+      "step": 8734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8438759857711966,
+      "learning_rate": 9.151661668982438e-06,
+      "loss": 0.4105,
+      "step": 8735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.564159651529531,
+      "learning_rate": 9.151442087185388e-06,
+      "loss": 0.314,
+      "step": 8736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4217282575823302,
+      "learning_rate": 9.151222479609008e-06,
+      "loss": 0.3098,
+      "step": 8737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.030388172425153,
+      "learning_rate": 9.151002846254661e-06,
+      "loss": 0.4322,
+      "step": 8738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9208454809370172,
+      "learning_rate": 9.150783187123712e-06,
+      "loss": 0.4972,
+      "step": 8739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4720372782530027,
+      "learning_rate": 9.150563502217526e-06,
+      "loss": 0.2575,
+      "step": 8740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.748208499874102,
+      "learning_rate": 9.150343791537468e-06,
+      "loss": 0.3886,
+      "step": 8741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6782539371419862,
+      "learning_rate": 9.150124055084899e-06,
+      "loss": 0.4244,
+      "step": 8742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3678817730087292,
+      "learning_rate": 9.149904292861185e-06,
+      "loss": 0.2341,
+      "step": 8743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0394326016647044,
+      "learning_rate": 9.14968450486769e-06,
+      "loss": 0.3533,
+      "step": 8744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.66585863252482,
+      "learning_rate": 9.149464691105782e-06,
+      "loss": 0.2796,
+      "step": 8745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4207620745373506,
+      "learning_rate": 9.149244851576822e-06,
+      "loss": 0.3138,
+      "step": 8746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9643317744347706,
+      "learning_rate": 9.149024986282178e-06,
+      "loss": 0.4851,
+      "step": 8747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7028284930309305,
+      "learning_rate": 9.148805095223212e-06,
+      "loss": 0.2238,
+      "step": 8748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9250291009350562,
+      "learning_rate": 9.148585178401293e-06,
+      "loss": 0.1421,
+      "step": 8749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2582519262029654,
+      "learning_rate": 9.148365235817787e-06,
+      "loss": 0.7388,
+      "step": 8750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6810411077965626,
+      "learning_rate": 9.148145267474054e-06,
+      "loss": 0.3203,
+      "step": 8751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.27071715999763,
+      "learning_rate": 9.147925273371467e-06,
+      "loss": 0.7124,
+      "step": 8752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6881730913947832,
+      "learning_rate": 9.147705253511387e-06,
+      "loss": 0.2421,
+      "step": 8753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4280388738300684,
+      "learning_rate": 9.147485207895183e-06,
+      "loss": 0.3509,
+      "step": 8754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1781091394292786,
+      "learning_rate": 9.147265136524222e-06,
+      "loss": 0.2153,
+      "step": 8755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.052953938829175,
+      "learning_rate": 9.147045039399868e-06,
+      "loss": 0.4789,
+      "step": 8756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0060533115938237,
+      "learning_rate": 9.146824916523489e-06,
+      "loss": 0.2035,
+      "step": 8757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3384013405457114,
+      "learning_rate": 9.146604767896452e-06,
+      "loss": 0.1502,
+      "step": 8758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7178323463457956,
+      "learning_rate": 9.146384593520124e-06,
+      "loss": 0.2688,
+      "step": 8759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1191664593084334,
+      "learning_rate": 9.146164393395872e-06,
+      "loss": 0.7103,
+      "step": 8760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1247302613441734,
+      "learning_rate": 9.145944167525065e-06,
+      "loss": 0.5192,
+      "step": 8761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6778284763684632,
+      "learning_rate": 9.145723915909068e-06,
+      "loss": 0.3042,
+      "step": 8762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0250641548327593,
+      "learning_rate": 9.145503638549249e-06,
+      "loss": 0.2271,
+      "step": 8763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4436763636664136,
+      "learning_rate": 9.145283335446977e-06,
+      "loss": 0.226,
+      "step": 8764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.740381390904602,
+      "learning_rate": 9.14506300660362e-06,
+      "loss": 0.1763,
+      "step": 8765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2587011408675435,
+      "learning_rate": 9.144842652020548e-06,
+      "loss": 0.1551,
+      "step": 8766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4847253335826585,
+      "learning_rate": 9.144622271699124e-06,
+      "loss": 0.3515,
+      "step": 8767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7006587887273963,
+      "learning_rate": 9.144401865640722e-06,
+      "loss": 0.4885,
+      "step": 8768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0217132576335266,
+      "learning_rate": 9.144181433846707e-06,
+      "loss": 0.1825,
+      "step": 8769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.137799716593955,
+      "learning_rate": 9.14396097631845e-06,
+      "loss": 0.1674,
+      "step": 8770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8101651757439654,
+      "learning_rate": 9.143740493057317e-06,
+      "loss": 0.5457,
+      "step": 8771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.635147211985755,
+      "learning_rate": 9.143519984064684e-06,
+      "loss": 0.3536,
+      "step": 8772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2898484059555337,
+      "learning_rate": 9.143299449341912e-06,
+      "loss": 0.1982,
+      "step": 8773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7982471430778482,
+      "learning_rate": 9.143078888890375e-06,
+      "loss": 0.3513,
+      "step": 8774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3392067999524149,
+      "learning_rate": 9.142858302711442e-06,
+      "loss": 0.2791,
+      "step": 8775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7940392550072386,
+      "learning_rate": 9.142637690806483e-06,
+      "loss": 0.3777,
+      "step": 8776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.458213745618066,
+      "learning_rate": 9.142417053176866e-06,
+      "loss": 0.2337,
+      "step": 8777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.802045721606062,
+      "learning_rate": 9.142196389823964e-06,
+      "loss": 0.2707,
+      "step": 8778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6813800113602728,
+      "learning_rate": 9.141975700749146e-06,
+      "loss": 0.1793,
+      "step": 8779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3386075820600043,
+      "learning_rate": 9.141754985953781e-06,
+      "loss": 0.1662,
+      "step": 8780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4971697837097502,
+      "learning_rate": 9.14153424543924e-06,
+      "loss": 0.2613,
+      "step": 8781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.368388082389116,
+      "learning_rate": 9.141313479206897e-06,
+      "loss": 0.5328,
+      "step": 8782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.526229778994701,
+      "learning_rate": 9.14109268725812e-06,
+      "loss": 0.1898,
+      "step": 8783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0946658038951114,
+      "learning_rate": 9.14087186959428e-06,
+      "loss": 0.5714,
+      "step": 8784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1039191879883212,
+      "learning_rate": 9.140651026216747e-06,
+      "loss": 0.3263,
+      "step": 8785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8710950640393067,
+      "learning_rate": 9.140430157126895e-06,
+      "loss": 0.3658,
+      "step": 8786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5419989157512393,
+      "learning_rate": 9.140209262326096e-06,
+      "loss": 0.3823,
+      "step": 8787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.150797200031216,
+      "learning_rate": 9.13998834181572e-06,
+      "loss": 0.429,
+      "step": 8788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3183657328134457,
+      "learning_rate": 9.13976739559714e-06,
+      "loss": 0.188,
+      "step": 8789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9757188572201239,
+      "learning_rate": 9.139546423671725e-06,
+      "loss": 0.5108,
+      "step": 8790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5224553916198846,
+      "learning_rate": 9.139325426040851e-06,
+      "loss": 0.2904,
+      "step": 8791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6331757684372201,
+      "learning_rate": 9.139104402705887e-06,
+      "loss": 0.4053,
+      "step": 8792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9393296293679751,
+      "learning_rate": 9.138883353668209e-06,
+      "loss": 0.2409,
+      "step": 8793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.431873172540718,
+      "learning_rate": 9.138662278929187e-06,
+      "loss": 0.3615,
+      "step": 8794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5488222297662095,
+      "learning_rate": 9.138441178490195e-06,
+      "loss": 0.1691,
+      "step": 8795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5239204497092824,
+      "learning_rate": 9.138220052352606e-06,
+      "loss": 0.3316,
+      "step": 8796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9886825920899696,
+      "learning_rate": 9.137998900517793e-06,
+      "loss": 0.5407,
+      "step": 8797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3756106250805658,
+      "learning_rate": 9.137777722987128e-06,
+      "loss": 0.2181,
+      "step": 8798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6758841725289968,
+      "learning_rate": 9.137556519761985e-06,
+      "loss": 0.4501,
+      "step": 8799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3474919260434253,
+      "learning_rate": 9.13733529084374e-06,
+      "loss": 0.3587,
+      "step": 8800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2790299872472508,
+      "learning_rate": 9.137114036233765e-06,
+      "loss": 0.3311,
+      "step": 8801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3763530979488208,
+      "learning_rate": 9.136892755933433e-06,
+      "loss": 0.2627,
+      "step": 8802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9374739660544391,
+      "learning_rate": 9.13667144994412e-06,
+      "loss": 0.292,
+      "step": 8803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9173629826468557,
+      "learning_rate": 9.136450118267197e-06,
+      "loss": 0.2805,
+      "step": 8804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2103076250326295,
+      "learning_rate": 9.13622876090404e-06,
+      "loss": 0.2096,
+      "step": 8805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3154745880144214,
+      "learning_rate": 9.136007377856027e-06,
+      "loss": 0.2531,
+      "step": 8806
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6513205893341396,
+      "learning_rate": 9.13578596912453e-06,
+      "loss": 0.4991,
+      "step": 8807
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2764826498364914,
+      "learning_rate": 9.135564534710921e-06,
+      "loss": 0.1948,
+      "step": 8808
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7984464884615405,
+      "learning_rate": 9.135343074616579e-06,
+      "loss": 0.308,
+      "step": 8809
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.20641063734842,
+      "learning_rate": 9.135121588842879e-06,
+      "loss": 0.1736,
+      "step": 8810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7596231671758906,
+      "learning_rate": 9.134900077391196e-06,
+      "loss": 0.4109,
+      "step": 8811
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5928112004502868,
+      "learning_rate": 9.134678540262902e-06,
+      "loss": 0.37,
+      "step": 8812
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.889933680412772,
+      "learning_rate": 9.134456977459379e-06,
+      "loss": 0.3733,
+      "step": 8813
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4532237608993657,
+      "learning_rate": 9.134235388981997e-06,
+      "loss": 0.3858,
+      "step": 8814
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4092234091211693,
+      "learning_rate": 9.134013774832135e-06,
+      "loss": 0.182,
+      "step": 8815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2982137178233373,
+      "learning_rate": 9.13379213501117e-06,
+      "loss": 0.6169,
+      "step": 8816
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5750974860088667,
+      "learning_rate": 9.133570469520476e-06,
+      "loss": 0.24,
+      "step": 8817
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7147643857330386,
+      "learning_rate": 9.13334877836143e-06,
+      "loss": 0.2849,
+      "step": 8818
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.339882307419533,
+      "learning_rate": 9.13312706153541e-06,
+      "loss": 0.6749,
+      "step": 8819
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7655833585436682,
+      "learning_rate": 9.13290531904379e-06,
+      "loss": 0.2171,
+      "step": 8820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7220018878368726,
+      "learning_rate": 9.13268355088795e-06,
+      "loss": 0.2321,
+      "step": 8821
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1382850662966737,
+      "learning_rate": 9.132461757069267e-06,
+      "loss": 0.3454,
+      "step": 8822
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7564294502783218,
+      "learning_rate": 9.132239937589116e-06,
+      "loss": 0.3203,
+      "step": 8823
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.992631516809745,
+      "learning_rate": 9.132018092448878e-06,
+      "loss": 0.5035,
+      "step": 8824
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7782097226425195,
+      "learning_rate": 9.131796221649925e-06,
+      "loss": 0.2708,
+      "step": 8825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1646148783960049,
+      "learning_rate": 9.13157432519364e-06,
+      "loss": 0.1368,
+      "step": 8826
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4626038987186403,
+      "learning_rate": 9.131352403081398e-06,
+      "loss": 0.3354,
+      "step": 8827
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7759668487346945,
+      "learning_rate": 9.131130455314578e-06,
+      "loss": 0.3531,
+      "step": 8828
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8116069640776118,
+      "learning_rate": 9.13090848189456e-06,
+      "loss": 0.3563,
+      "step": 8829
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3012356093460176,
+      "learning_rate": 9.13068648282272e-06,
+      "loss": 0.3492,
+      "step": 8830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7919706047068065,
+      "learning_rate": 9.130464458100436e-06,
+      "loss": 0.3695,
+      "step": 8831
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.844765563277929,
+      "learning_rate": 9.130242407729088e-06,
+      "loss": 0.3568,
+      "step": 8832
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.874291452998888,
+      "learning_rate": 9.130020331710056e-06,
+      "loss": 0.5327,
+      "step": 8833
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0309100746947935,
+      "learning_rate": 9.129798230044717e-06,
+      "loss": 0.3589,
+      "step": 8834
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9068133194111507,
+      "learning_rate": 9.12957610273445e-06,
+      "loss": 0.4456,
+      "step": 8835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.486975242010528,
+      "learning_rate": 9.129353949780636e-06,
+      "loss": 0.4727,
+      "step": 8836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1946892549392873,
+      "learning_rate": 9.129131771184655e-06,
+      "loss": 0.4463,
+      "step": 8837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8148669088251466,
+      "learning_rate": 9.128909566947886e-06,
+      "loss": 0.4945,
+      "step": 8838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.09971647477345,
+      "learning_rate": 9.128687337071707e-06,
+      "loss": 0.7342,
+      "step": 8839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5702573543563725,
+      "learning_rate": 9.1284650815575e-06,
+      "loss": 0.3393,
+      "step": 8840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.239127626487379,
+      "learning_rate": 9.128242800406645e-06,
+      "loss": 0.264,
+      "step": 8841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.223589149954448,
+      "learning_rate": 9.128020493620521e-06,
+      "loss": 0.5032,
+      "step": 8842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2928877888712422,
+      "learning_rate": 9.12779816120051e-06,
+      "loss": 0.3281,
+      "step": 8843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1100875781175645,
+      "learning_rate": 9.12757580314799e-06,
+      "loss": 0.4209,
+      "step": 8844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0106885327614017,
+      "learning_rate": 9.127353419464346e-06,
+      "loss": 0.5877,
+      "step": 8845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.300156979785857,
+      "learning_rate": 9.127131010150956e-06,
+      "loss": 0.2207,
+      "step": 8846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7855166440819314,
+      "learning_rate": 9.126908575209201e-06,
+      "loss": 0.5603,
+      "step": 8847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.781928970365202,
+      "learning_rate": 9.126686114640463e-06,
+      "loss": 0.5013,
+      "step": 8848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6937980567617843,
+      "learning_rate": 9.126463628446124e-06,
+      "loss": 0.4914,
+      "step": 8849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.577796968038733,
+      "learning_rate": 9.126241116627566e-06,
+      "loss": 0.1551,
+      "step": 8850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3356700010354254,
+      "learning_rate": 9.126018579186167e-06,
+      "loss": 0.3561,
+      "step": 8851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7251273486564553,
+      "learning_rate": 9.125796016123313e-06,
+      "loss": 0.4794,
+      "step": 8852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.671673147547752,
+      "learning_rate": 9.125573427440387e-06,
+      "loss": 0.3986,
+      "step": 8853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.458121548896692,
+      "learning_rate": 9.125350813138765e-06,
+      "loss": 0.1396,
+      "step": 8854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.236399377333506,
+      "learning_rate": 9.125128173219835e-06,
+      "loss": 0.2451,
+      "step": 8855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5888570844989074,
+      "learning_rate": 9.124905507684977e-06,
+      "loss": 0.2533,
+      "step": 8856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4114771017067824,
+      "learning_rate": 9.124682816535577e-06,
+      "loss": 0.1597,
+      "step": 8857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3382714335209054,
+      "learning_rate": 9.124460099773013e-06,
+      "loss": 0.2549,
+      "step": 8858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8123820755677904,
+      "learning_rate": 9.12423735739867e-06,
+      "loss": 0.2673,
+      "step": 8859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7787772465681244,
+      "learning_rate": 9.124014589413934e-06,
+      "loss": 0.3492,
+      "step": 8860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4933791532710974,
+      "learning_rate": 9.123791795820184e-06,
+      "loss": 0.2812,
+      "step": 8861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3106737013356824,
+      "learning_rate": 9.123568976618806e-06,
+      "loss": 0.4808,
+      "step": 8862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9373099710734902,
+      "learning_rate": 9.123346131811183e-06,
+      "loss": 0.3886,
+      "step": 8863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8512713811065966,
+      "learning_rate": 9.1231232613987e-06,
+      "loss": 0.3311,
+      "step": 8864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1287702920743576,
+      "learning_rate": 9.12290036538274e-06,
+      "loss": 0.2336,
+      "step": 8865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5627183713905448,
+      "learning_rate": 9.122677443764683e-06,
+      "loss": 0.2187,
+      "step": 8866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.240017905079582,
+      "learning_rate": 9.12245449654592e-06,
+      "loss": 0.241,
+      "step": 8867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.764579007377261,
+      "learning_rate": 9.12223152372783e-06,
+      "loss": 0.3894,
+      "step": 8868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8434306045871454,
+      "learning_rate": 9.122008525311804e-06,
+      "loss": 0.3588,
+      "step": 8869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.761715337069096,
+      "learning_rate": 9.121785501299221e-06,
+      "loss": 0.6612,
+      "step": 8870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3126396391586912,
+      "learning_rate": 9.12156245169147e-06,
+      "loss": 0.3441,
+      "step": 8871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9252998570573658,
+      "learning_rate": 9.12133937648993e-06,
+      "loss": 0.4708,
+      "step": 8872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4921230182396463,
+      "learning_rate": 9.121116275695994e-06,
+      "loss": 0.3108,
+      "step": 8873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.172839636609214,
+      "learning_rate": 9.120893149311041e-06,
+      "loss": 0.2298,
+      "step": 8874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7603307614631238,
+      "learning_rate": 9.120669997336463e-06,
+      "loss": 0.3918,
+      "step": 8875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3030526059896805,
+      "learning_rate": 9.120446819773637e-06,
+      "loss": 0.3068,
+      "step": 8876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.736976553949758,
+      "learning_rate": 9.120223616623957e-06,
+      "loss": 0.3603,
+      "step": 8877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0354674535178976,
+      "learning_rate": 9.120000387888803e-06,
+      "loss": 0.3922,
+      "step": 8878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3891660559450183,
+      "learning_rate": 9.119777133569568e-06,
+      "loss": 0.1973,
+      "step": 8879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6312988420094978,
+      "learning_rate": 9.11955385366763e-06,
+      "loss": 0.3381,
+      "step": 8880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.352887258893627,
+      "learning_rate": 9.119330548184382e-06,
+      "loss": 0.7073,
+      "step": 8881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7029123523493543,
+      "learning_rate": 9.119107217121209e-06,
+      "loss": 0.3083,
+      "step": 8882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1644701803852924,
+      "learning_rate": 9.118883860479496e-06,
+      "loss": 0.1504,
+      "step": 8883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.524091438400425,
+      "learning_rate": 9.118660478260633e-06,
+      "loss": 0.3354,
+      "step": 8884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.565379493560714,
+      "learning_rate": 9.118437070466003e-06,
+      "loss": 0.3767,
+      "step": 8885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.176749397975864,
+      "learning_rate": 9.118213637096998e-06,
+      "loss": 0.2108,
+      "step": 8886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3867030639835831,
+      "learning_rate": 9.117990178155002e-06,
+      "loss": 0.5066,
+      "step": 8887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5405325428531895,
+      "learning_rate": 9.117766693641403e-06,
+      "loss": 0.2899,
+      "step": 8888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.910340785842575,
+      "learning_rate": 9.11754318355759e-06,
+      "loss": 0.6099,
+      "step": 8889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.680043213904428,
+      "learning_rate": 9.117319647904953e-06,
+      "loss": 0.4395,
+      "step": 8890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6779268574016417,
+      "learning_rate": 9.117096086684876e-06,
+      "loss": 0.4768,
+      "step": 8891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9046335402020158,
+      "learning_rate": 9.116872499898748e-06,
+      "loss": 0.6475,
+      "step": 8892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.115344797564943,
+      "learning_rate": 9.11664888754796e-06,
+      "loss": 0.4585,
+      "step": 8893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7884600574531253,
+      "learning_rate": 9.116425249633898e-06,
+      "loss": 0.2555,
+      "step": 8894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9704015555891,
+      "learning_rate": 9.116201586157952e-06,
+      "loss": 0.5592,
+      "step": 8895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4492551896688535,
+      "learning_rate": 9.115977897121508e-06,
+      "loss": 0.158,
+      "step": 8896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.703601236866616,
+      "learning_rate": 9.115754182525962e-06,
+      "loss": 0.3025,
+      "step": 8897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2174917950977004,
+      "learning_rate": 9.115530442372696e-06,
+      "loss": 0.2943,
+      "step": 8898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9978067835534028,
+      "learning_rate": 9.115306676663102e-06,
+      "loss": 0.3273,
+      "step": 8899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8081420536239172,
+      "learning_rate": 9.11508288539857e-06,
+      "loss": 0.2904,
+      "step": 8900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1304659292274435,
+      "learning_rate": 9.11485906858049e-06,
+      "loss": 0.2309,
+      "step": 8901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5934993275619131,
+      "learning_rate": 9.11463522621025e-06,
+      "loss": 0.436,
+      "step": 8902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.488569777392345,
+      "learning_rate": 9.114411358289242e-06,
+      "loss": 0.2985,
+      "step": 8903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4371685807023162,
+      "learning_rate": 9.114187464818855e-06,
+      "loss": 0.3698,
+      "step": 8904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5420005561355103,
+      "learning_rate": 9.113963545800479e-06,
+      "loss": 0.3715,
+      "step": 8905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6334232894501177,
+      "learning_rate": 9.113739601235508e-06,
+      "loss": 0.3364,
+      "step": 8906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.761170972427809,
+      "learning_rate": 9.113515631125328e-06,
+      "loss": 0.3062,
+      "step": 8907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5050331426809649,
+      "learning_rate": 9.113291635471329e-06,
+      "loss": 0.2952,
+      "step": 8908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9330887106033925,
+      "learning_rate": 9.113067614274908e-06,
+      "loss": 0.3224,
+      "step": 8909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.285265429476774,
+      "learning_rate": 9.112843567537451e-06,
+      "loss": 0.3214,
+      "step": 8910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.814156552192232,
+      "learning_rate": 9.11261949526035e-06,
+      "loss": 0.3894,
+      "step": 8911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6452363240213115,
+      "learning_rate": 9.112395397445e-06,
+      "loss": 0.3324,
+      "step": 8912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.976458139642375,
+      "learning_rate": 9.112171274092789e-06,
+      "loss": 0.2284,
+      "step": 8913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1652818276592827,
+      "learning_rate": 9.111947125205107e-06,
+      "loss": 0.1881,
+      "step": 8914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7560871776614717,
+      "learning_rate": 9.111722950783351e-06,
+      "loss": 0.1903,
+      "step": 8915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5425535210690118,
+      "learning_rate": 9.11149875082891e-06,
+      "loss": 0.1858,
+      "step": 8916
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4235515267596672,
+      "learning_rate": 9.111274525343177e-06,
+      "loss": 0.3708,
+      "step": 8917
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.082055140925423,
+      "learning_rate": 9.111050274327546e-06,
+      "loss": 0.2457,
+      "step": 8918
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8255495734861518,
+      "learning_rate": 9.110825997783405e-06,
+      "loss": 0.3846,
+      "step": 8919
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.980037457567709,
+      "learning_rate": 9.110601695712151e-06,
+      "loss": 0.1528,
+      "step": 8920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2675914320444392,
+      "learning_rate": 9.110377368115175e-06,
+      "loss": 0.3143,
+      "step": 8921
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9952787332956108,
+      "learning_rate": 9.11015301499387e-06,
+      "loss": 0.3078,
+      "step": 8922
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2723929269200267,
+      "learning_rate": 9.10992863634963e-06,
+      "loss": 0.1382,
+      "step": 8923
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.418169201850904,
+      "learning_rate": 9.109704232183846e-06,
+      "loss": 0.19,
+      "step": 8924
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7986208167718412,
+      "learning_rate": 9.109479802497916e-06,
+      "loss": 0.3655,
+      "step": 8925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5386229894445023,
+      "learning_rate": 9.109255347293228e-06,
+      "loss": 0.3098,
+      "step": 8926
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5670761426344662,
+      "learning_rate": 9.109030866571181e-06,
+      "loss": 0.4462,
+      "step": 8927
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.692904793024576,
+      "learning_rate": 9.108806360333167e-06,
+      "loss": 0.1817,
+      "step": 8928
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6666690265122237,
+      "learning_rate": 9.108581828580578e-06,
+      "loss": 0.3561,
+      "step": 8929
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.899185692559496,
+      "learning_rate": 9.108357271314812e-06,
+      "loss": 0.372,
+      "step": 8930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.239499043823218,
+      "learning_rate": 9.10813268853726e-06,
+      "loss": 0.2451,
+      "step": 8931
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5429552681249412,
+      "learning_rate": 9.10790808024932e-06,
+      "loss": 0.3889,
+      "step": 8932
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.092815688895373,
+      "learning_rate": 9.107683446452382e-06,
+      "loss": 0.621,
+      "step": 8933
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6672314615719057,
+      "learning_rate": 9.107458787147846e-06,
+      "loss": 0.4554,
+      "step": 8934
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.729214271690586,
+      "learning_rate": 9.107234102337104e-06,
+      "loss": 0.3007,
+      "step": 8935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.084155232112022,
+      "learning_rate": 9.107009392021553e-06,
+      "loss": 0.3712,
+      "step": 8936
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1282090471562767,
+      "learning_rate": 9.106784656202587e-06,
+      "loss": 0.7328,
+      "step": 8937
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1454545727645393,
+      "learning_rate": 9.106559894881603e-06,
+      "loss": 0.5491,
+      "step": 8938
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5114312311421814,
+      "learning_rate": 9.106335108059995e-06,
+      "loss": 0.5223,
+      "step": 8939
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.822702121968978,
+      "learning_rate": 9.10611029573916e-06,
+      "loss": 0.371,
+      "step": 8940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.762160562994995,
+      "learning_rate": 9.105885457920493e-06,
+      "loss": 0.2349,
+      "step": 8941
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.004961542738155,
+      "learning_rate": 9.105660594605391e-06,
+      "loss": 0.3004,
+      "step": 8942
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6221269003995205,
+      "learning_rate": 9.105435705795252e-06,
+      "loss": 0.3584,
+      "step": 8943
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6642485232542208,
+      "learning_rate": 9.105210791491468e-06,
+      "loss": 0.4406,
+      "step": 8944
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3995398380381885,
+      "learning_rate": 9.104985851695437e-06,
+      "loss": 0.1835,
+      "step": 8945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8336650135639545,
+      "learning_rate": 9.104760886408562e-06,
+      "loss": 0.4362,
+      "step": 8946
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3186396435126935,
+      "learning_rate": 9.104535895632231e-06,
+      "loss": 0.1868,
+      "step": 8947
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.023476335160202,
+      "learning_rate": 9.104310879367848e-06,
+      "loss": 0.5591,
+      "step": 8948
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5515890443803553,
+      "learning_rate": 9.104085837616808e-06,
+      "loss": 0.2411,
+      "step": 8949
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.925334027341348,
+      "learning_rate": 9.103860770380506e-06,
+      "loss": 0.3771,
+      "step": 8950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1471746144289372,
+      "learning_rate": 9.103635677660342e-06,
+      "loss": 0.1336,
+      "step": 8951
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8365534134815766,
+      "learning_rate": 9.103410559457714e-06,
+      "loss": 0.317,
+      "step": 8952
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1798338136190152,
+      "learning_rate": 9.103185415774018e-06,
+      "loss": 0.1608,
+      "step": 8953
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2957105148192098,
+      "learning_rate": 9.102960246610656e-06,
+      "loss": 0.3237,
+      "step": 8954
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.174159762438681,
+      "learning_rate": 9.102735051969021e-06,
+      "loss": 0.6117,
+      "step": 8955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1377813593205284,
+      "learning_rate": 9.102509831850515e-06,
+      "loss": 0.2028,
+      "step": 8956
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8298507760914418,
+      "learning_rate": 9.102284586256536e-06,
+      "loss": 0.5477,
+      "step": 8957
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.142785617792707,
+      "learning_rate": 9.102059315188481e-06,
+      "loss": 0.5037,
+      "step": 8958
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.49122720046253,
+      "learning_rate": 9.10183401864775e-06,
+      "loss": 0.4611,
+      "step": 8959
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7724809713781497,
+      "learning_rate": 9.101608696635744e-06,
+      "loss": 0.4836,
+      "step": 8960
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2391287479063675,
+      "learning_rate": 9.10138334915386e-06,
+      "loss": 0.5042,
+      "step": 8961
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0678236414341282,
+      "learning_rate": 9.101157976203497e-06,
+      "loss": 0.5259,
+      "step": 8962
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.229229878631576,
+      "learning_rate": 9.100932577786056e-06,
+      "loss": 0.4161,
+      "step": 8963
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2167201347093355,
+      "learning_rate": 9.100707153902935e-06,
+      "loss": 0.2409,
+      "step": 8964
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5442616683297736,
+      "learning_rate": 9.100481704555534e-06,
+      "loss": 0.4912,
+      "step": 8965
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0028643809012454,
+      "learning_rate": 9.100256229745257e-06,
+      "loss": 0.5662,
+      "step": 8966
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5559068000723681,
+      "learning_rate": 9.100030729473498e-06,
+      "loss": 0.3557,
+      "step": 8967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6536842628328954,
+      "learning_rate": 9.09980520374166e-06,
+      "loss": 0.2433,
+      "step": 8968
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3824927632517963,
+      "learning_rate": 9.099579652551144e-06,
+      "loss": 0.2693,
+      "step": 8969
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5507262794303074,
+      "learning_rate": 9.099354075903351e-06,
+      "loss": 0.3621,
+      "step": 8970
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5032176633573613,
+      "learning_rate": 9.09912847379968e-06,
+      "loss": 0.3283,
+      "step": 8971
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7104769374246886,
+      "learning_rate": 9.098902846241535e-06,
+      "loss": 0.3606,
+      "step": 8972
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3103772485659455,
+      "learning_rate": 9.098677193230315e-06,
+      "loss": 0.5578,
+      "step": 8973
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.675370563965984,
+      "learning_rate": 9.098451514767418e-06,
+      "loss": 0.2919,
+      "step": 8974
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9572992788491284,
+      "learning_rate": 9.098225810854251e-06,
+      "loss": 0.4194,
+      "step": 8975
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2850848601319727,
+      "learning_rate": 9.098000081492211e-06,
+      "loss": 0.235,
+      "step": 8976
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5032111741482874,
+      "learning_rate": 9.097774326682705e-06,
+      "loss": 0.2538,
+      "step": 8977
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.976393632120952,
+      "learning_rate": 9.09754854642713e-06,
+      "loss": 0.609,
+      "step": 8978
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4653227569683185,
+      "learning_rate": 9.09732274072689e-06,
+      "loss": 0.3075,
+      "step": 8979
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9661178306875966,
+      "learning_rate": 9.097096909583387e-06,
+      "loss": 0.3483,
+      "step": 8980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.199829182847874,
+      "learning_rate": 9.096871052998023e-06,
+      "loss": 0.3189,
+      "step": 8981
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6659197718246437,
+      "learning_rate": 9.096645170972201e-06,
+      "loss": 0.342,
+      "step": 8982
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.691015572833199,
+      "learning_rate": 9.096419263507324e-06,
+      "loss": 0.2488,
+      "step": 8983
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.719702664696116,
+      "learning_rate": 9.096193330604795e-06,
+      "loss": 0.3016,
+      "step": 8984
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6703224058722865,
+      "learning_rate": 9.095967372266015e-06,
+      "loss": 0.3027,
+      "step": 8985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0206877239767915,
+      "learning_rate": 9.09574138849239e-06,
+      "loss": 0.4994,
+      "step": 8986
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8259996139143653,
+      "learning_rate": 9.09551537928532e-06,
+      "loss": 0.3249,
+      "step": 8987
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8121571380971444,
+      "learning_rate": 9.09528934464621e-06,
+      "loss": 0.2381,
+      "step": 8988
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6702429512949628,
+      "learning_rate": 9.095063284576466e-06,
+      "loss": 0.3437,
+      "step": 8989
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.646613739289367,
+      "learning_rate": 9.094837199077487e-06,
+      "loss": 0.2235,
+      "step": 8990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5855401415140344,
+      "learning_rate": 9.094611088150683e-06,
+      "loss": 0.3436,
+      "step": 8991
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.169588298548944,
+      "learning_rate": 9.094384951797452e-06,
+      "loss": 0.3769,
+      "step": 8992
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0477611974771666,
+      "learning_rate": 9.0941587900192e-06,
+      "loss": 0.3941,
+      "step": 8993
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3285661497662198,
+      "learning_rate": 9.093932602817335e-06,
+      "loss": 0.2578,
+      "step": 8994
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6828807374229802,
+      "learning_rate": 9.093706390193258e-06,
+      "loss": 0.4374,
+      "step": 8995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5625621354146264,
+      "learning_rate": 9.093480152148373e-06,
+      "loss": 0.5191,
+      "step": 8996
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9172439273871615,
+      "learning_rate": 9.093253888684088e-06,
+      "loss": 0.3211,
+      "step": 8997
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5888460904662696,
+      "learning_rate": 9.093027599801807e-06,
+      "loss": 0.442,
+      "step": 8998
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7996616724724812,
+      "learning_rate": 9.092801285502932e-06,
+      "loss": 0.1344,
+      "step": 8999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9657712780276092,
+      "learning_rate": 9.092574945788875e-06,
+      "loss": 0.1145,
+      "step": 9000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.474542851242246,
+      "learning_rate": 9.092348580661035e-06,
+      "loss": 0.3455,
+      "step": 9001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7000899241134284,
+      "learning_rate": 9.09212219012082e-06,
+      "loss": 0.3738,
+      "step": 9002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2281615454898853,
+      "learning_rate": 9.091895774169636e-06,
+      "loss": 0.5763,
+      "step": 9003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0223997503775335,
+      "learning_rate": 9.091669332808891e-06,
+      "loss": 0.3793,
+      "step": 9004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3025859722499833,
+      "learning_rate": 9.091442866039986e-06,
+      "loss": 0.1647,
+      "step": 9005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5607050527872077,
+      "learning_rate": 9.091216373864333e-06,
+      "loss": 0.2904,
+      "step": 9006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5171146335547014,
+      "learning_rate": 9.090989856283335e-06,
+      "loss": 0.3559,
+      "step": 9007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.515429486612591,
+      "learning_rate": 9.090763313298398e-06,
+      "loss": 0.4182,
+      "step": 9008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.623917049150723,
+      "learning_rate": 9.090536744910932e-06,
+      "loss": 0.2401,
+      "step": 9009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6554308206012514,
+      "learning_rate": 9.090310151122344e-06,
+      "loss": 0.3712,
+      "step": 9010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7734111813344224,
+      "learning_rate": 9.090083531934036e-06,
+      "loss": 0.4072,
+      "step": 9011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1702954599907596,
+      "learning_rate": 9.08985688734742e-06,
+      "loss": 0.7262,
+      "step": 9012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3867100325834256,
+      "learning_rate": 9.089630217363901e-06,
+      "loss": 0.3126,
+      "step": 9013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2964592376368476,
+      "learning_rate": 9.089403521984889e-06,
+      "loss": 0.171,
+      "step": 9014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.160184760017645,
+      "learning_rate": 9.08917680121179e-06,
+      "loss": 0.5703,
+      "step": 9015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8048734424629784,
+      "learning_rate": 9.08895005504601e-06,
+      "loss": 0.2129,
+      "step": 9016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8069113616161174,
+      "learning_rate": 9.088723283488962e-06,
+      "loss": 0.3798,
+      "step": 9017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.124833763155753,
+      "learning_rate": 9.08849648654205e-06,
+      "loss": 0.7923,
+      "step": 9018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.498675601129136,
+      "learning_rate": 9.088269664206684e-06,
+      "loss": 0.3798,
+      "step": 9019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.567424619123073,
+      "learning_rate": 9.08804281648427e-06,
+      "loss": 0.1921,
+      "step": 9020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.243670972273455,
+      "learning_rate": 9.087815943376222e-06,
+      "loss": 0.2138,
+      "step": 9021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3993877813050288,
+      "learning_rate": 9.087589044883945e-06,
+      "loss": 0.2202,
+      "step": 9022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6831424001919735,
+      "learning_rate": 9.087362121008846e-06,
+      "loss": 0.2711,
+      "step": 9023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9541731005856215,
+      "learning_rate": 9.08713517175234e-06,
+      "loss": 0.3623,
+      "step": 9024
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6383704907474645,
+      "learning_rate": 9.086908197115832e-06,
+      "loss": 0.3242,
+      "step": 9025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.843713848923406,
+      "learning_rate": 9.08668119710073e-06,
+      "loss": 0.3258,
+      "step": 9026
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0293407539410466,
+      "learning_rate": 9.08645417170845e-06,
+      "loss": 0.1376,
+      "step": 9027
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6697858286535625,
+      "learning_rate": 9.086227120940396e-06,
+      "loss": 0.2705,
+      "step": 9028
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7904799376808247,
+      "learning_rate": 9.086000044797981e-06,
+      "loss": 0.274,
+      "step": 9029
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.399047550673686,
+      "learning_rate": 9.085772943282611e-06,
+      "loss": 0.257,
+      "step": 9030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4539348687363884,
+      "learning_rate": 9.0855458163957e-06,
+      "loss": 0.185,
+      "step": 9031
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5094987502615709,
+      "learning_rate": 9.08531866413866e-06,
+      "loss": 0.2545,
+      "step": 9032
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0840855731636496,
+      "learning_rate": 9.085091486512897e-06,
+      "loss": 0.1737,
+      "step": 9033
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.736907990888019,
+      "learning_rate": 9.084864283519823e-06,
+      "loss": 0.2465,
+      "step": 9034
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6299607045170303,
+      "learning_rate": 9.084637055160851e-06,
+      "loss": 0.4431,
+      "step": 9035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5138448210328708,
+      "learning_rate": 9.084409801437391e-06,
+      "loss": 0.4047,
+      "step": 9036
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3655893262645096,
+      "learning_rate": 9.084182522350854e-06,
+      "loss": 0.2948,
+      "step": 9037
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.137008535759908,
+      "learning_rate": 9.083955217902651e-06,
+      "loss": 0.3569,
+      "step": 9038
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5897508636258704,
+      "learning_rate": 9.083727888094194e-06,
+      "loss": 0.2528,
+      "step": 9039
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6162247533294145,
+      "learning_rate": 9.083500532926893e-06,
+      "loss": 0.3813,
+      "step": 9040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4634588128753006,
+      "learning_rate": 9.083273152402162e-06,
+      "loss": 0.1695,
+      "step": 9041
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4089283127467636,
+      "learning_rate": 9.083045746521413e-06,
+      "loss": 0.3862,
+      "step": 9042
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8388760870881764,
+      "learning_rate": 9.082818315286054e-06,
+      "loss": 0.3569,
+      "step": 9043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8987749686743494,
+      "learning_rate": 9.082590858697503e-06,
+      "loss": 0.3918,
+      "step": 9044
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.508631548020321,
+      "learning_rate": 9.08236337675717e-06,
+      "loss": 0.1788,
+      "step": 9045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.871471108150893,
+      "learning_rate": 9.082135869466469e-06,
+      "loss": 0.5599,
+      "step": 9046
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5543643073951634,
+      "learning_rate": 9.08190833682681e-06,
+      "loss": 0.273,
+      "step": 9047
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7894842866667082,
+      "learning_rate": 9.081680778839607e-06,
+      "loss": 0.411,
+      "step": 9048
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0272804039473644,
+      "learning_rate": 9.081453195506274e-06,
+      "loss": 0.1703,
+      "step": 9049
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7205629735499324,
+      "learning_rate": 9.081225586828221e-06,
+      "loss": 0.358,
+      "step": 9050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.141027517134322,
+      "learning_rate": 9.080997952806868e-06,
+      "loss": 0.5297,
+      "step": 9051
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1487191605171767,
+      "learning_rate": 9.080770293443622e-06,
+      "loss": 0.2667,
+      "step": 9052
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.478950776600164,
+      "learning_rate": 9.0805426087399e-06,
+      "loss": 0.2312,
+      "step": 9053
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4789000632453397,
+      "learning_rate": 9.080314898697114e-06,
+      "loss": 0.363,
+      "step": 9054
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5659062355679607,
+      "learning_rate": 9.08008716331668e-06,
+      "loss": 0.4531,
+      "step": 9055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3297131271447824,
+      "learning_rate": 9.07985940260001e-06,
+      "loss": 0.2779,
+      "step": 9056
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8293533823953567,
+      "learning_rate": 9.079631616548521e-06,
+      "loss": 0.3811,
+      "step": 9057
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4735862327681646,
+      "learning_rate": 9.079403805163627e-06,
+      "loss": 0.2723,
+      "step": 9058
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.628505870271063,
+      "learning_rate": 9.07917596844674e-06,
+      "loss": 0.2706,
+      "step": 9059
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8284436583949075,
+      "learning_rate": 9.078948106399278e-06,
+      "loss": 0.3983,
+      "step": 9060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1660846565054053,
+      "learning_rate": 9.078720219022654e-06,
+      "loss": 0.5054,
+      "step": 9061
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6047777187685996,
+      "learning_rate": 9.078492306318282e-06,
+      "loss": 0.4736,
+      "step": 9062
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6044325257328576,
+      "learning_rate": 9.078264368287582e-06,
+      "loss": 0.673,
+      "step": 9063
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2872999018431284,
+      "learning_rate": 9.078036404931963e-06,
+      "loss": 0.1637,
+      "step": 9064
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8156152335936722,
+      "learning_rate": 9.077808416252848e-06,
+      "loss": 0.2858,
+      "step": 9065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.346465115123084,
+      "learning_rate": 9.077580402251645e-06,
+      "loss": 0.1585,
+      "step": 9066
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4191896748694264,
+      "learning_rate": 9.077352362929775e-06,
+      "loss": 0.2184,
+      "step": 9067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9306702948987258,
+      "learning_rate": 9.077124298288652e-06,
+      "loss": 0.3666,
+      "step": 9068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1487313017661283,
+      "learning_rate": 9.076896208329693e-06,
+      "loss": 0.3464,
+      "step": 9069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.515391267863967,
+      "learning_rate": 9.076668093054314e-06,
+      "loss": 0.1349,
+      "step": 9070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8312565806664456,
+      "learning_rate": 9.076439952463931e-06,
+      "loss": 0.2482,
+      "step": 9071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1135639810430247,
+      "learning_rate": 9.076211786559963e-06,
+      "loss": 0.1706,
+      "step": 9072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5150302231169193,
+      "learning_rate": 9.075983595343825e-06,
+      "loss": 0.2642,
+      "step": 9073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4957028568055377,
+      "learning_rate": 9.075755378816935e-06,
+      "loss": 0.4455,
+      "step": 9074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4173762341951575,
+      "learning_rate": 9.075527136980708e-06,
+      "loss": 0.2346,
+      "step": 9075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1638794793511977,
+      "learning_rate": 9.075298869836564e-06,
+      "loss": 0.2597,
+      "step": 9076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.110345901669421,
+      "learning_rate": 9.07507057738592e-06,
+      "loss": 0.1628,
+      "step": 9077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7903261303095659,
+      "learning_rate": 9.07484225963019e-06,
+      "loss": 0.3539,
+      "step": 9078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.562029448227045,
+      "learning_rate": 9.074613916570798e-06,
+      "loss": 0.3361,
+      "step": 9079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7439595349526844,
+      "learning_rate": 9.074385548209157e-06,
+      "loss": 0.576,
+      "step": 9080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8265520934174766,
+      "learning_rate": 9.074157154546685e-06,
+      "loss": 0.6208,
+      "step": 9081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6433955368803796,
+      "learning_rate": 9.073928735584805e-06,
+      "loss": 0.3297,
+      "step": 9082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.678818863999342,
+      "learning_rate": 9.073700291324932e-06,
+      "loss": 0.4727,
+      "step": 9083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6025362892289605,
+      "learning_rate": 9.073471821768483e-06,
+      "loss": 0.4199,
+      "step": 9084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7669154468753847,
+      "learning_rate": 9.073243326916881e-06,
+      "loss": 0.4188,
+      "step": 9085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.600530207968891,
+      "learning_rate": 9.073014806771541e-06,
+      "loss": 0.2507,
+      "step": 9086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3767470780548212,
+      "learning_rate": 9.072786261333885e-06,
+      "loss": 0.2298,
+      "step": 9087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2804738229089019,
+      "learning_rate": 9.072557690605331e-06,
+      "loss": 0.1995,
+      "step": 9088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6155653341185021,
+      "learning_rate": 9.072329094587297e-06,
+      "loss": 0.3589,
+      "step": 9089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5058707126729924,
+      "learning_rate": 9.072100473281205e-06,
+      "loss": 0.3642,
+      "step": 9090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2425512914749277,
+      "learning_rate": 9.071871826688472e-06,
+      "loss": 0.318,
+      "step": 9091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6764726207412999,
+      "learning_rate": 9.071643154810518e-06,
+      "loss": 0.2104,
+      "step": 9092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9788436881265168,
+      "learning_rate": 9.071414457648768e-06,
+      "loss": 0.2843,
+      "step": 9093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.739322179238333,
+      "learning_rate": 9.071185735204636e-06,
+      "loss": 0.5063,
+      "step": 9094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4409011586474572,
+      "learning_rate": 9.070956987479546e-06,
+      "loss": 0.3149,
+      "step": 9095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6928720752062951,
+      "learning_rate": 9.070728214474916e-06,
+      "loss": 0.3996,
+      "step": 9096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4759463609142314,
+      "learning_rate": 9.070499416192169e-06,
+      "loss": 0.3274,
+      "step": 9097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4186815054677062,
+      "learning_rate": 9.070270592632724e-06,
+      "loss": 0.2062,
+      "step": 9098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9557595470742317,
+      "learning_rate": 9.070041743798003e-06,
+      "loss": 0.3765,
+      "step": 9099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.400984966508672,
+      "learning_rate": 9.069812869689425e-06,
+      "loss": 0.3138,
+      "step": 9100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5104654227532166,
+      "learning_rate": 9.069583970308415e-06,
+      "loss": 0.2847,
+      "step": 9101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6869689980161464,
+      "learning_rate": 9.069355045656392e-06,
+      "loss": 0.5002,
+      "step": 9102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0847063976324252,
+      "learning_rate": 9.069126095734776e-06,
+      "loss": 0.2474,
+      "step": 9103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.606678833844517,
+      "learning_rate": 9.068897120544993e-06,
+      "loss": 0.3521,
+      "step": 9104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8579247893911357,
+      "learning_rate": 9.06866812008846e-06,
+      "loss": 0.4501,
+      "step": 9105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6767359728271227,
+      "learning_rate": 9.068439094366602e-06,
+      "loss": 0.3587,
+      "step": 9106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.887269640595856,
+      "learning_rate": 9.068210043380841e-06,
+      "loss": 0.4121,
+      "step": 9107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5550461575849983,
+      "learning_rate": 9.067980967132599e-06,
+      "loss": 0.2879,
+      "step": 9108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7978789427421689,
+      "learning_rate": 9.067751865623299e-06,
+      "loss": 0.3038,
+      "step": 9109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.083434800220215,
+      "learning_rate": 9.067522738854362e-06,
+      "loss": 0.5886,
+      "step": 9110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7536198312070193,
+      "learning_rate": 9.067293586827212e-06,
+      "loss": 0.3917,
+      "step": 9111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8123995120046288,
+      "learning_rate": 9.067064409543274e-06,
+      "loss": 0.3222,
+      "step": 9112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6028789273778916,
+      "learning_rate": 9.066835207003966e-06,
+      "loss": 0.4693,
+      "step": 9113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4199966243079183,
+      "learning_rate": 9.066605979210717e-06,
+      "loss": 0.2554,
+      "step": 9114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3162879302505868,
+      "learning_rate": 9.066376726164947e-06,
+      "loss": 0.4142,
+      "step": 9115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5897409138705545,
+      "learning_rate": 9.06614744786808e-06,
+      "loss": 0.3593,
+      "step": 9116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4499718799001922,
+      "learning_rate": 9.06591814432154e-06,
+      "loss": 0.183,
+      "step": 9117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.762511204115785,
+      "learning_rate": 9.06568881552675e-06,
+      "loss": 0.3534,
+      "step": 9118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.100838628703864,
+      "learning_rate": 9.065459461485135e-06,
+      "loss": 0.4999,
+      "step": 9119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2207786353598493,
+      "learning_rate": 9.065230082198122e-06,
+      "loss": 0.1845,
+      "step": 9120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5619359834274495,
+      "learning_rate": 9.065000677667131e-06,
+      "loss": 0.3499,
+      "step": 9121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7016920439743934,
+      "learning_rate": 9.064771247893588e-06,
+      "loss": 0.2763,
+      "step": 9122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0469044872518674,
+      "learning_rate": 9.064541792878919e-06,
+      "loss": 0.3725,
+      "step": 9123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4103028915960427,
+      "learning_rate": 9.064312312624547e-06,
+      "loss": 0.2148,
+      "step": 9124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.027809387465021,
+      "learning_rate": 9.064082807131898e-06,
+      "loss": 0.2102,
+      "step": 9125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5587660758758068,
+      "learning_rate": 9.063853276402397e-06,
+      "loss": 0.2253,
+      "step": 9126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9083997767623997,
+      "learning_rate": 9.063623720437468e-06,
+      "loss": 0.4158,
+      "step": 9127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5160773978181386,
+      "learning_rate": 9.063394139238541e-06,
+      "loss": 0.4977,
+      "step": 9128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9296071734678986,
+      "learning_rate": 9.063164532807036e-06,
+      "loss": 0.2899,
+      "step": 9129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8105108160812804,
+      "learning_rate": 9.062934901144382e-06,
+      "loss": 0.2298,
+      "step": 9130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3884672880701423,
+      "learning_rate": 9.062705244252004e-06,
+      "loss": 0.2311,
+      "step": 9131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0029769177682817,
+      "learning_rate": 9.062475562131328e-06,
+      "loss": 0.6566,
+      "step": 9132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3260933764319973,
+      "learning_rate": 9.06224585478378e-06,
+      "loss": 0.3563,
+      "step": 9133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.29042195825469,
+      "learning_rate": 9.062016122210789e-06,
+      "loss": 0.1943,
+      "step": 9134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6225443451101091,
+      "learning_rate": 9.061786364413778e-06,
+      "loss": 0.1999,
+      "step": 9135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5802368725601204,
+      "learning_rate": 9.061556581394175e-06,
+      "loss": 0.4011,
+      "step": 9136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.940895443191812,
+      "learning_rate": 9.061326773153407e-06,
+      "loss": 0.157,
+      "step": 9137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3797165270351537,
+      "learning_rate": 9.061096939692901e-06,
+      "loss": 0.2264,
+      "step": 9138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.435219399086121,
+      "learning_rate": 9.060867081014086e-06,
+      "loss": 0.2718,
+      "step": 9139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7942335720557931,
+      "learning_rate": 9.060637197118388e-06,
+      "loss": 0.4413,
+      "step": 9140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7318686942843038,
+      "learning_rate": 9.060407288007234e-06,
+      "loss": 0.2333,
+      "step": 9141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7206270502011036,
+      "learning_rate": 9.06017735368205e-06,
+      "loss": 0.3666,
+      "step": 9142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4323392896248826,
+      "learning_rate": 9.059947394144268e-06,
+      "loss": 0.3595,
+      "step": 9143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.101612356367396,
+      "learning_rate": 9.059717409395312e-06,
+      "loss": 0.2746,
+      "step": 9144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.355495644997904,
+      "learning_rate": 9.059487399436612e-06,
+      "loss": 0.4585,
+      "step": 9145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.025109691871999,
+      "learning_rate": 9.059257364269598e-06,
+      "loss": 0.2514,
+      "step": 9146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.722455806705071,
+      "learning_rate": 9.059027303895695e-06,
+      "loss": 0.2522,
+      "step": 9147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6352506212102973,
+      "learning_rate": 9.058797218316334e-06,
+      "loss": 0.4717,
+      "step": 9148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4762188294517138,
+      "learning_rate": 9.058567107532943e-06,
+      "loss": 0.3779,
+      "step": 9149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9244075638368925,
+      "learning_rate": 9.058336971546951e-06,
+      "loss": 0.3126,
+      "step": 9150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4142259837850544,
+      "learning_rate": 9.058106810359789e-06,
+      "loss": 0.1772,
+      "step": 9151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5019676445902823,
+      "learning_rate": 9.057876623972881e-06,
+      "loss": 0.2144,
+      "step": 9152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8193677161282895,
+      "learning_rate": 9.05764641238766e-06,
+      "loss": 0.5369,
+      "step": 9153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5051236984081608,
+      "learning_rate": 9.057416175605557e-06,
+      "loss": 0.2349,
+      "step": 9154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.219507996498126,
+      "learning_rate": 9.057185913628e-06,
+      "loss": 0.2264,
+      "step": 9155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9349299698453154,
+      "learning_rate": 9.056955626456417e-06,
+      "loss": 0.1787,
+      "step": 9156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.498242024837043,
+      "learning_rate": 9.05672531409224e-06,
+      "loss": 0.262,
+      "step": 9157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8327723571888637,
+      "learning_rate": 9.056494976536901e-06,
+      "loss": 0.5023,
+      "step": 9158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3012979300372056,
+      "learning_rate": 9.056264613791827e-06,
+      "loss": 0.3053,
+      "step": 9159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.953928402649663,
+      "learning_rate": 9.05603422585845e-06,
+      "loss": 0.3467,
+      "step": 9160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4043492611171886,
+      "learning_rate": 9.055803812738202e-06,
+      "loss": 0.1559,
+      "step": 9161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.200604568431214,
+      "learning_rate": 9.05557337443251e-06,
+      "loss": 0.167,
+      "step": 9162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6098338833994996,
+      "learning_rate": 9.055342910942807e-06,
+      "loss": 0.3395,
+      "step": 9163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5650839896866504,
+      "learning_rate": 9.055112422270527e-06,
+      "loss": 0.2503,
+      "step": 9164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6743228808767598,
+      "learning_rate": 9.054881908417096e-06,
+      "loss": 0.1941,
+      "step": 9165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9619893961029546,
+      "learning_rate": 9.05465136938395e-06,
+      "loss": 0.2753,
+      "step": 9166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5722454377767978,
+      "learning_rate": 9.054420805172517e-06,
+      "loss": 0.3617,
+      "step": 9167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0327839257872706,
+      "learning_rate": 9.054190215784232e-06,
+      "loss": 0.1645,
+      "step": 9168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.395331217231705,
+      "learning_rate": 9.053959601220525e-06,
+      "loss": 0.2286,
+      "step": 9169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7452336055303614,
+      "learning_rate": 9.053728961482827e-06,
+      "loss": 0.303,
+      "step": 9170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6571898088217916,
+      "learning_rate": 9.053498296572573e-06,
+      "loss": 0.3628,
+      "step": 9171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7641798325029185,
+      "learning_rate": 9.053267606491193e-06,
+      "loss": 0.3562,
+      "step": 9172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9119424562004974,
+      "learning_rate": 9.053036891240122e-06,
+      "loss": 0.6639,
+      "step": 9173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3438393097407557,
+      "learning_rate": 9.052806150820788e-06,
+      "loss": 0.2943,
+      "step": 9174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4386266647662462,
+      "learning_rate": 9.05257538523463e-06,
+      "loss": 0.4124,
+      "step": 9175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0509607908585,
+      "learning_rate": 9.052344594483076e-06,
+      "loss": 0.5168,
+      "step": 9176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3456490786790485,
+      "learning_rate": 9.052113778567562e-06,
+      "loss": 0.27,
+      "step": 9177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8012496810650724,
+      "learning_rate": 9.05188293748952e-06,
+      "loss": 0.3838,
+      "step": 9178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9186126976586377,
+      "learning_rate": 9.051652071250386e-06,
+      "loss": 0.1701,
+      "step": 9179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8378514418443541,
+      "learning_rate": 9.051421179851588e-06,
+      "loss": 0.4653,
+      "step": 9180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.871026223065176,
+      "learning_rate": 9.051190263294566e-06,
+      "loss": 0.4467,
+      "step": 9181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.304834432282465,
+      "learning_rate": 9.05095932158075e-06,
+      "loss": 0.6736,
+      "step": 9182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.000907649151635,
+      "learning_rate": 9.050728354711575e-06,
+      "loss": 0.262,
+      "step": 9183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.900164568198791,
+      "learning_rate": 9.050497362688476e-06,
+      "loss": 0.2788,
+      "step": 9184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.656512509541438,
+      "learning_rate": 9.050266345512887e-06,
+      "loss": 0.2415,
+      "step": 9185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8024266435984118,
+      "learning_rate": 9.050035303186242e-06,
+      "loss": 0.5108,
+      "step": 9186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0592614044413842,
+      "learning_rate": 9.049804235709976e-06,
+      "loss": 0.1439,
+      "step": 9187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.765916715846276,
+      "learning_rate": 9.049573143085525e-06,
+      "loss": 0.3781,
+      "step": 9188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.077299950063676,
+      "learning_rate": 9.049342025314323e-06,
+      "loss": 0.1764,
+      "step": 9189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6804440502300568,
+      "learning_rate": 9.049110882397805e-06,
+      "loss": 0.4457,
+      "step": 9190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.381855132168475,
+      "learning_rate": 9.048879714337406e-06,
+      "loss": 0.2641,
+      "step": 9191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0084960976079134,
+      "learning_rate": 9.048648521134562e-06,
+      "loss": 0.5679,
+      "step": 9192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4071393353273436,
+      "learning_rate": 9.048417302790709e-06,
+      "loss": 0.2861,
+      "step": 9193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6262896857744826,
+      "learning_rate": 9.048186059307283e-06,
+      "loss": 0.3427,
+      "step": 9194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6951012743490739,
+      "learning_rate": 9.047954790685718e-06,
+      "loss": 0.343,
+      "step": 9195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9286846535738718,
+      "learning_rate": 9.047723496927453e-06,
+      "loss": 0.4296,
+      "step": 9196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.937377869117673,
+      "learning_rate": 9.047492178033922e-06,
+      "loss": 0.2657,
+      "step": 9197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.252265299774838,
+      "learning_rate": 9.047260834006564e-06,
+      "loss": 0.4317,
+      "step": 9198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.447662858473979,
+      "learning_rate": 9.047029464846812e-06,
+      "loss": 0.4423,
+      "step": 9199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1780920213161339,
+      "learning_rate": 9.046798070556107e-06,
+      "loss": 0.1538,
+      "step": 9200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2388832354359431,
+      "learning_rate": 9.046566651135881e-06,
+      "loss": 0.2014,
+      "step": 9201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.100729775569526,
+      "learning_rate": 9.046335206587574e-06,
+      "loss": 0.2267,
+      "step": 9202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2214876202331901,
+      "learning_rate": 9.046103736912624e-06,
+      "loss": 0.151,
+      "step": 9203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7359283617845254,
+      "learning_rate": 9.045872242112466e-06,
+      "loss": 0.4455,
+      "step": 9204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7556751651054263,
+      "learning_rate": 9.045640722188537e-06,
+      "loss": 0.4877,
+      "step": 9205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8711085906266456,
+      "learning_rate": 9.04540917714228e-06,
+      "loss": 0.3325,
+      "step": 9206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.789785717872991,
+      "learning_rate": 9.045177606975128e-06,
+      "loss": 0.3801,
+      "step": 9207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9661584478988354,
+      "learning_rate": 9.044946011688519e-06,
+      "loss": 0.5387,
+      "step": 9208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4292519080403872,
+      "learning_rate": 9.044714391283893e-06,
+      "loss": 0.2101,
+      "step": 9209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2421873338546425,
+      "learning_rate": 9.044482745762686e-06,
+      "loss": 0.4252,
+      "step": 9210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.698354172263671,
+      "learning_rate": 9.04425107512634e-06,
+      "loss": 0.2014,
+      "step": 9211
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5782589935392597,
+      "learning_rate": 9.044019379376291e-06,
+      "loss": 0.4899,
+      "step": 9212
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.516792376244117,
+      "learning_rate": 9.04378765851398e-06,
+      "loss": 0.3303,
+      "step": 9213
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4482732474969953,
+      "learning_rate": 9.043555912540842e-06,
+      "loss": 0.2964,
+      "step": 9214
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7124121159605636,
+      "learning_rate": 9.04332414145832e-06,
+      "loss": 0.5014,
+      "step": 9215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6495809747273245,
+      "learning_rate": 9.04309234526785e-06,
+      "loss": 0.3513,
+      "step": 9216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0384276797683665,
+      "learning_rate": 9.042860523970876e-06,
+      "loss": 0.5358,
+      "step": 9217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.038693198377312,
+      "learning_rate": 9.042628677568833e-06,
+      "loss": 0.3756,
+      "step": 9218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4879669473331312,
+      "learning_rate": 9.04239680606316e-06,
+      "loss": 0.1636,
+      "step": 9219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.875291072584466,
+      "learning_rate": 9.042164909455302e-06,
+      "loss": 0.4402,
+      "step": 9220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9004637967770952,
+      "learning_rate": 9.041932987746695e-06,
+      "loss": 0.5563,
+      "step": 9221
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7544710443609883,
+      "learning_rate": 9.04170104093878e-06,
+      "loss": 0.4703,
+      "step": 9222
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5603245372678445,
+      "learning_rate": 9.041469069032999e-06,
+      "loss": 0.3083,
+      "step": 9223
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0405750218882663,
+      "learning_rate": 9.041237072030791e-06,
+      "loss": 0.4492,
+      "step": 9224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1335716852634663,
+      "learning_rate": 9.041005049933596e-06,
+      "loss": 0.333,
+      "step": 9225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3148679036343236,
+      "learning_rate": 9.040773002742855e-06,
+      "loss": 0.1857,
+      "step": 9226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4027605128156928,
+      "learning_rate": 9.040540930460011e-06,
+      "loss": 0.3336,
+      "step": 9227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.662235046817598,
+      "learning_rate": 9.040308833086505e-06,
+      "loss": 0.3179,
+      "step": 9228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3553663553154673,
+      "learning_rate": 9.040076710623772e-06,
+      "loss": 0.2416,
+      "step": 9229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3214240051065442,
+      "learning_rate": 9.039844563073263e-06,
+      "loss": 0.29,
+      "step": 9230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6030494555953538,
+      "learning_rate": 9.039612390436413e-06,
+      "loss": 0.2242,
+      "step": 9231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.206035937422204,
+      "learning_rate": 9.039380192714667e-06,
+      "loss": 0.1929,
+      "step": 9232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8007658587478612,
+      "learning_rate": 9.039147969909464e-06,
+      "loss": 0.4056,
+      "step": 9233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4463519572736903,
+      "learning_rate": 9.038915722022248e-06,
+      "loss": 0.2056,
+      "step": 9234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0216830481618173,
+      "learning_rate": 9.03868344905446e-06,
+      "loss": 0.2039,
+      "step": 9235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9832132084882999,
+      "learning_rate": 9.038451151007544e-06,
+      "loss": 0.4781,
+      "step": 9236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.22313587238955,
+      "learning_rate": 9.03821882788294e-06,
+      "loss": 0.198,
+      "step": 9237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0018813031967073,
+      "learning_rate": 9.037986479682096e-06,
+      "loss": 0.7762,
+      "step": 9238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6306566146623847,
+      "learning_rate": 9.037754106406448e-06,
+      "loss": 0.4316,
+      "step": 9239
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3746231277874883,
+      "learning_rate": 9.037521708057443e-06,
+      "loss": 0.2138,
+      "step": 9240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6454567468343222,
+      "learning_rate": 9.037289284636524e-06,
+      "loss": 0.3684,
+      "step": 9241
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0535503120072676,
+      "learning_rate": 9.037056836145131e-06,
+      "loss": 0.1676,
+      "step": 9242
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5630956229782467,
+      "learning_rate": 9.036824362584713e-06,
+      "loss": 0.3267,
+      "step": 9243
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3249575205507245,
+      "learning_rate": 9.036591863956709e-06,
+      "loss": 0.2247,
+      "step": 9244
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.836601867591376,
+      "learning_rate": 9.036359340262564e-06,
+      "loss": 0.5207,
+      "step": 9245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6087497589211923,
+      "learning_rate": 9.036126791503721e-06,
+      "loss": 0.4295,
+      "step": 9246
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9258482594467354,
+      "learning_rate": 9.03589421768163e-06,
+      "loss": 0.3406,
+      "step": 9247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8552125372420467,
+      "learning_rate": 9.035661618797725e-06,
+      "loss": 0.5285,
+      "step": 9248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.85273049129149,
+      "learning_rate": 9.03542899485346e-06,
+      "loss": 0.3793,
+      "step": 9249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.163376233845439,
+      "learning_rate": 9.035196345850273e-06,
+      "loss": 0.2149,
+      "step": 9250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.45241150977537,
+      "learning_rate": 9.03496367178961e-06,
+      "loss": 0.3906,
+      "step": 9251
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8192370680165524,
+      "learning_rate": 9.03473097267292e-06,
+      "loss": 0.2798,
+      "step": 9252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0332937522720154,
+      "learning_rate": 9.034498248501643e-06,
+      "loss": 0.4906,
+      "step": 9253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1436066847787796,
+      "learning_rate": 9.034265499277227e-06,
+      "loss": 0.2698,
+      "step": 9254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2747930437839576,
+      "learning_rate": 9.034032725001115e-06,
+      "loss": 0.5123,
+      "step": 9255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7136006205040077,
+      "learning_rate": 9.033799925674754e-06,
+      "loss": 0.2349,
+      "step": 9256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5907289983442754,
+      "learning_rate": 9.033567101299592e-06,
+      "loss": 0.1965,
+      "step": 9257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9210016565173005,
+      "learning_rate": 9.03333425187707e-06,
+      "loss": 0.3336,
+      "step": 9258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.764851554745273,
+      "learning_rate": 9.033101377408634e-06,
+      "loss": 0.1795,
+      "step": 9259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6752736449899166,
+      "learning_rate": 9.032868477895737e-06,
+      "loss": 0.1995,
+      "step": 9260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8697099924623455,
+      "learning_rate": 9.032635553339817e-06,
+      "loss": 0.4572,
+      "step": 9261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5547937143626545,
+      "learning_rate": 9.032402603742326e-06,
+      "loss": 0.4099,
+      "step": 9262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6885929740458514,
+      "learning_rate": 9.032169629104707e-06,
+      "loss": 0.3714,
+      "step": 9263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3829301148305044,
+      "learning_rate": 9.031936629428409e-06,
+      "loss": 0.2895,
+      "step": 9264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3336955549537726,
+      "learning_rate": 9.031703604714878e-06,
+      "loss": 0.2451,
+      "step": 9265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5367428125612421,
+      "learning_rate": 9.031470554965561e-06,
+      "loss": 0.3019,
+      "step": 9266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.717275394415206,
+      "learning_rate": 9.031237480181904e-06,
+      "loss": 0.3065,
+      "step": 9267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3123168036924489,
+      "learning_rate": 9.031004380365357e-06,
+      "loss": 0.2148,
+      "step": 9268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0143422769225605,
+      "learning_rate": 9.030771255517367e-06,
+      "loss": 0.551,
+      "step": 9269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.079641402637456,
+      "learning_rate": 9.03053810563938e-06,
+      "loss": 0.2218,
+      "step": 9270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1226144074066737,
+      "learning_rate": 9.030304930732843e-06,
+      "loss": 0.1436,
+      "step": 9271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7043241732551269,
+      "learning_rate": 9.030071730799206e-06,
+      "loss": 0.4037,
+      "step": 9272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8308184255127016,
+      "learning_rate": 9.029838505839918e-06,
+      "loss": 0.3927,
+      "step": 9273
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.381904832180653,
+      "learning_rate": 9.029605255856423e-06,
+      "loss": 0.1877,
+      "step": 9274
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.04597541463974,
+      "learning_rate": 9.029371980850174e-06,
+      "loss": 0.3527,
+      "step": 9275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6817883970438823,
+      "learning_rate": 9.029138680822618e-06,
+      "loss": 0.2743,
+      "step": 9276
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4567454418260322,
+      "learning_rate": 9.028905355775203e-06,
+      "loss": 0.1669,
+      "step": 9277
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5021147423986236,
+      "learning_rate": 9.02867200570938e-06,
+      "loss": 0.2264,
+      "step": 9278
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.736659392776681,
+      "learning_rate": 9.028438630626596e-06,
+      "loss": 0.485,
+      "step": 9279
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.060853717035142,
+      "learning_rate": 9.028205230528299e-06,
+      "loss": 0.2288,
+      "step": 9280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5842398873678056,
+      "learning_rate": 9.02797180541594e-06,
+      "loss": 0.516,
+      "step": 9281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7584362671700173,
+      "learning_rate": 9.027738355290969e-06,
+      "loss": 0.2293,
+      "step": 9282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7832245569518466,
+      "learning_rate": 9.027504880154836e-06,
+      "loss": 0.1408,
+      "step": 9283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8866892352273599,
+      "learning_rate": 9.02727138000899e-06,
+      "loss": 0.3365,
+      "step": 9284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9498615657439706,
+      "learning_rate": 9.02703785485488e-06,
+      "loss": 0.3348,
+      "step": 9285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5438253480591437,
+      "learning_rate": 9.026804304693957e-06,
+      "loss": 0.2385,
+      "step": 9286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7409100339390613,
+      "learning_rate": 9.026570729527672e-06,
+      "loss": 0.3081,
+      "step": 9287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3471087539174533,
+      "learning_rate": 9.026337129357474e-06,
+      "loss": 0.2261,
+      "step": 9288
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.856185854821388,
+      "learning_rate": 9.026103504184815e-06,
+      "loss": 0.5311,
+      "step": 9289
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6597971052985434,
+      "learning_rate": 9.025869854011145e-06,
+      "loss": 0.2888,
+      "step": 9290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8434574776081236,
+      "learning_rate": 9.025636178837916e-06,
+      "loss": 0.3224,
+      "step": 9291
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0156206530147074,
+      "learning_rate": 9.025402478666578e-06,
+      "loss": 0.3097,
+      "step": 9292
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8230899365139248,
+      "learning_rate": 9.02516875349858e-06,
+      "loss": 0.2246,
+      "step": 9293
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5773291493156214,
+      "learning_rate": 9.024935003335378e-06,
+      "loss": 0.3348,
+      "step": 9294
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9732874323190537,
+      "learning_rate": 9.02470122817842e-06,
+      "loss": 0.4159,
+      "step": 9295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5118627528916326,
+      "learning_rate": 9.02446742802916e-06,
+      "loss": 0.3521,
+      "step": 9296
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9358380480722703,
+      "learning_rate": 9.024233602889049e-06,
+      "loss": 0.5987,
+      "step": 9297
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7586088887837867,
+      "learning_rate": 9.023999752759538e-06,
+      "loss": 0.2394,
+      "step": 9298
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8029831728038932,
+      "learning_rate": 9.02376587764208e-06,
+      "loss": 0.5227,
+      "step": 9299
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9338495701804492,
+      "learning_rate": 9.023531977538126e-06,
+      "loss": 0.1985,
+      "step": 9300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0626226229773312,
+      "learning_rate": 9.023298052449132e-06,
+      "loss": 0.3478,
+      "step": 9301
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.729317552353723,
+      "learning_rate": 9.023064102376546e-06,
+      "loss": 0.3033,
+      "step": 9302
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5975845922428453,
+      "learning_rate": 9.022830127321825e-06,
+      "loss": 0.4561,
+      "step": 9303
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.890383287221578,
+      "learning_rate": 9.022596127286419e-06,
+      "loss": 0.2858,
+      "step": 9304
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1143646847537845,
+      "learning_rate": 9.022362102271781e-06,
+      "loss": 0.8695,
+      "step": 9305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.742077503238689,
+      "learning_rate": 9.022128052279366e-06,
+      "loss": 0.3713,
+      "step": 9306
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5814527890439825,
+      "learning_rate": 9.021893977310628e-06,
+      "loss": 0.2338,
+      "step": 9307
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2826869944877122,
+      "learning_rate": 9.021659877367016e-06,
+      "loss": 0.1869,
+      "step": 9308
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1862736381052015,
+      "learning_rate": 9.02142575244999e-06,
+      "loss": 0.289,
+      "step": 9309
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6941688081395052,
+      "learning_rate": 9.021191602560998e-06,
+      "loss": 0.3184,
+      "step": 9310
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.163050845444926,
+      "learning_rate": 9.0209574277015e-06,
+      "loss": 0.3805,
+      "step": 9311
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.056642776879383,
+      "learning_rate": 9.020723227872945e-06,
+      "loss": 0.601,
+      "step": 9312
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3928903996003488,
+      "learning_rate": 9.020489003076789e-06,
+      "loss": 0.197,
+      "step": 9313
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3771510989669684,
+      "learning_rate": 9.020254753314488e-06,
+      "loss": 0.3466,
+      "step": 9314
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8518711186991768,
+      "learning_rate": 9.020020478587494e-06,
+      "loss": 0.4157,
+      "step": 9315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6152831558001979,
+      "learning_rate": 9.019786178897263e-06,
+      "loss": 0.3994,
+      "step": 9316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5284584787333333,
+      "learning_rate": 9.019551854245252e-06,
+      "loss": 0.2202,
+      "step": 9317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3397557213005862,
+      "learning_rate": 9.019317504632912e-06,
+      "loss": 0.1868,
+      "step": 9318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.183660554102198,
+      "learning_rate": 9.0190831300617e-06,
+      "loss": 0.1431,
+      "step": 9319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.768517138347407,
+      "learning_rate": 9.018848730533073e-06,
+      "loss": 0.2703,
+      "step": 9320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8559922293015294,
+      "learning_rate": 9.018614306048486e-06,
+      "loss": 0.5444,
+      "step": 9321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.774865781048033,
+      "learning_rate": 9.018379856609392e-06,
+      "loss": 0.484,
+      "step": 9322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6567714518074512,
+      "learning_rate": 9.01814538221725e-06,
+      "loss": 0.3713,
+      "step": 9323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.867135172872852,
+      "learning_rate": 9.017910882873515e-06,
+      "loss": 0.272,
+      "step": 9324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4174778419546505,
+      "learning_rate": 9.017676358579644e-06,
+      "loss": 0.1745,
+      "step": 9325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.021471213330162,
+      "learning_rate": 9.017441809337091e-06,
+      "loss": 0.4209,
+      "step": 9326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6518727999913219,
+      "learning_rate": 9.017207235147313e-06,
+      "loss": 0.3506,
+      "step": 9327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6157590943254225,
+      "learning_rate": 9.01697263601177e-06,
+      "loss": 0.2967,
+      "step": 9328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3139241615215,
+      "learning_rate": 9.016738011931916e-06,
+      "loss": 0.2838,
+      "step": 9329
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5469003010616913,
+      "learning_rate": 9.016503362909209e-06,
+      "loss": 0.2311,
+      "step": 9330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3491789370164473,
+      "learning_rate": 9.016268688945103e-06,
+      "loss": 0.1728,
+      "step": 9331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3963911172098715,
+      "learning_rate": 9.016033990041057e-06,
+      "loss": 0.2167,
+      "step": 9332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8170475391730798,
+      "learning_rate": 9.015799266198533e-06,
+      "loss": 0.5148,
+      "step": 9333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.533010855239296,
+      "learning_rate": 9.015564517418982e-06,
+      "loss": 0.1642,
+      "step": 9334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7857410883595024,
+      "learning_rate": 9.015329743703865e-06,
+      "loss": 0.5659,
+      "step": 9335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4959335540183933,
+      "learning_rate": 9.015094945054639e-06,
+      "loss": 0.218,
+      "step": 9336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2570191066859544,
+      "learning_rate": 9.014860121472762e-06,
+      "loss": 0.2068,
+      "step": 9337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5912269903201781,
+      "learning_rate": 9.014625272959695e-06,
+      "loss": 0.186,
+      "step": 9338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9243119614111075,
+      "learning_rate": 9.01439039951689e-06,
+      "loss": 0.4935,
+      "step": 9339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4665518757979712,
+      "learning_rate": 9.014155501145811e-06,
+      "loss": 0.3508,
+      "step": 9340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7061814390457932,
+      "learning_rate": 9.013920577847916e-06,
+      "loss": 0.4781,
+      "step": 9341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6790073392743419,
+      "learning_rate": 9.013685629624662e-06,
+      "loss": 0.2358,
+      "step": 9342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4986412430914768,
+      "learning_rate": 9.01345065647751e-06,
+      "loss": 0.3878,
+      "step": 9343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0785531386352645,
+      "learning_rate": 9.013215658407917e-06,
+      "loss": 0.3136,
+      "step": 9344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.21848406702456,
+      "learning_rate": 9.012980635417343e-06,
+      "loss": 0.2044,
+      "step": 9345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5167846421802849,
+      "learning_rate": 9.012745587507249e-06,
+      "loss": 0.3821,
+      "step": 9346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4525268641374236,
+      "learning_rate": 9.012510514679089e-06,
+      "loss": 0.3404,
+      "step": 9347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5494286478179113,
+      "learning_rate": 9.012275416934331e-06,
+      "loss": 0.3513,
+      "step": 9348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8112793709168378,
+      "learning_rate": 9.012040294274429e-06,
+      "loss": 0.2666,
+      "step": 9349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.58920410419917,
+      "learning_rate": 9.011805146700846e-06,
+      "loss": 0.353,
+      "step": 9350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3268470674612243,
+      "learning_rate": 9.01156997421504e-06,
+      "loss": 0.166,
+      "step": 9351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3295952327169367,
+      "learning_rate": 9.011334776818472e-06,
+      "loss": 0.2389,
+      "step": 9352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.821858285922139,
+      "learning_rate": 9.011099554512604e-06,
+      "loss": 0.3568,
+      "step": 9353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.420972238832886,
+      "learning_rate": 9.010864307298897e-06,
+      "loss": 0.2931,
+      "step": 9354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1185570746639963,
+      "learning_rate": 9.010629035178809e-06,
+      "loss": 0.2229,
+      "step": 9355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4189967948831257,
+      "learning_rate": 9.010393738153803e-06,
+      "loss": 0.159,
+      "step": 9356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2911334799154663,
+      "learning_rate": 9.010158416225337e-06,
+      "loss": 0.2339,
+      "step": 9357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2967220440252865,
+      "learning_rate": 9.009923069394876e-06,
+      "loss": 0.2552,
+      "step": 9358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7273363783751103,
+      "learning_rate": 9.009687697663883e-06,
+      "loss": 0.196,
+      "step": 9359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4548867912714336,
+      "learning_rate": 9.009452301033815e-06,
+      "loss": 0.218,
+      "step": 9360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7569712639216275,
+      "learning_rate": 9.009216879506134e-06,
+      "loss": 0.4888,
+      "step": 9361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.471981268284139,
+      "learning_rate": 9.008981433082304e-06,
+      "loss": 0.2137,
+      "step": 9362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7418965622751652,
+      "learning_rate": 9.008745961763789e-06,
+      "loss": 0.5314,
+      "step": 9363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5945155184547755,
+      "learning_rate": 9.008510465552046e-06,
+      "loss": 0.401,
+      "step": 9364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6119628511526527,
+      "learning_rate": 9.008274944448543e-06,
+      "loss": 0.2988,
+      "step": 9365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9687905023586632,
+      "learning_rate": 9.008039398454738e-06,
+      "loss": 0.3729,
+      "step": 9366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.471734264250292,
+      "learning_rate": 9.007803827572095e-06,
+      "loss": 0.3069,
+      "step": 9367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4769803646736643,
+      "learning_rate": 9.007568231802078e-06,
+      "loss": 0.2959,
+      "step": 9368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1909754800612755,
+      "learning_rate": 9.007332611146149e-06,
+      "loss": 0.5781,
+      "step": 9369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9198327824827368,
+      "learning_rate": 9.00709696560577e-06,
+      "loss": 0.4113,
+      "step": 9370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.454670735107909,
+      "learning_rate": 9.006861295182407e-06,
+      "loss": 0.2295,
+      "step": 9371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4562270166769093,
+      "learning_rate": 9.006625599877523e-06,
+      "loss": 0.2214,
+      "step": 9372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8843025528090616,
+      "learning_rate": 9.006389879692578e-06,
+      "loss": 0.2199,
+      "step": 9373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6591883272391759,
+      "learning_rate": 9.00615413462904e-06,
+      "loss": 0.3227,
+      "step": 9374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3274102194916861,
+      "learning_rate": 9.005918364688372e-06,
+      "loss": 0.2698,
+      "step": 9375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5350500754411192,
+      "learning_rate": 9.005682569872035e-06,
+      "loss": 0.3458,
+      "step": 9376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6822196792006412,
+      "learning_rate": 9.005446750181499e-06,
+      "loss": 0.4744,
+      "step": 9377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6441386925451913,
+      "learning_rate": 9.005210905618222e-06,
+      "loss": 0.18,
+      "step": 9378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4213001063647626,
+      "learning_rate": 9.004975036183675e-06,
+      "loss": 0.3804,
+      "step": 9379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0241705770586966,
+      "learning_rate": 9.004739141879315e-06,
+      "loss": 0.416,
+      "step": 9380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6978648305559048,
+      "learning_rate": 9.004503222706613e-06,
+      "loss": 0.4562,
+      "step": 9381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9939454942483263,
+      "learning_rate": 9.004267278667032e-06,
+      "loss": 0.4131,
+      "step": 9382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8900667412629146,
+      "learning_rate": 9.004031309762038e-06,
+      "loss": 0.2366,
+      "step": 9383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2832117573928206,
+      "learning_rate": 9.003795315993095e-06,
+      "loss": 0.2073,
+      "step": 9384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3765016613786385,
+      "learning_rate": 9.003559297361668e-06,
+      "loss": 0.3074,
+      "step": 9385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8190343159384657,
+      "learning_rate": 9.003323253869225e-06,
+      "loss": 0.5065,
+      "step": 9386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3352764345633208,
+      "learning_rate": 9.003087185517229e-06,
+      "loss": 0.2027,
+      "step": 9387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4845527492307844,
+      "learning_rate": 9.002851092307148e-06,
+      "loss": 0.2878,
+      "step": 9388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6424750217461055,
+      "learning_rate": 9.002614974240448e-06,
+      "loss": 0.1673,
+      "step": 9389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7025134799860544,
+      "learning_rate": 9.002378831318592e-06,
+      "loss": 0.5754,
+      "step": 9390
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.98122383711114,
+      "learning_rate": 9.00214266354305e-06,
+      "loss": 0.387,
+      "step": 9391
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3171737164209474,
+      "learning_rate": 9.00190647091529e-06,
+      "loss": 0.2055,
+      "step": 9392
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8886882994291692,
+      "learning_rate": 9.001670253436774e-06,
+      "loss": 0.3611,
+      "step": 9393
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8014131728135319,
+      "learning_rate": 9.00143401110897e-06,
+      "loss": 0.2653,
+      "step": 9394
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.886251545748068,
+      "learning_rate": 9.001197743933347e-06,
+      "loss": 0.5705,
+      "step": 9395
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5554398221210008,
+      "learning_rate": 9.000961451911372e-06,
+      "loss": 0.2956,
+      "step": 9396
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.610410010281715,
+      "learning_rate": 9.000725135044511e-06,
+      "loss": 0.2161,
+      "step": 9397
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.840770347785414,
+      "learning_rate": 9.000488793334232e-06,
+      "loss": 0.1429,
+      "step": 9398
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6833651934062106,
+      "learning_rate": 9.000252426782002e-06,
+      "loss": 0.4058,
+      "step": 9399
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.631623990469174,
+      "learning_rate": 9.00001603538929e-06,
+      "loss": 0.4699,
+      "step": 9400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.019783506004789,
+      "learning_rate": 8.999779619157563e-06,
+      "loss": 0.2079,
+      "step": 9401
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7937590160209742,
+      "learning_rate": 8.999543178088291e-06,
+      "loss": 0.4757,
+      "step": 9402
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.084583827250022,
+      "learning_rate": 8.999306712182937e-06,
+      "loss": 0.1498,
+      "step": 9403
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3104805537788002,
+      "learning_rate": 8.999070221442976e-06,
+      "loss": 0.4211,
+      "step": 9404
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2250026157717384,
+      "learning_rate": 8.998833705869871e-06,
+      "loss": 0.5481,
+      "step": 9405
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7516516529231383,
+      "learning_rate": 8.998597165465095e-06,
+      "loss": 0.3461,
+      "step": 9406
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.905282830508965,
+      "learning_rate": 8.998360600230115e-06,
+      "loss": 0.5256,
+      "step": 9407
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.019171383889674,
+      "learning_rate": 8.9981240101664e-06,
+      "loss": 0.3104,
+      "step": 9408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.926750121219711,
+      "learning_rate": 8.997887395275418e-06,
+      "loss": 0.4201,
+      "step": 9409
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9693360062046525,
+      "learning_rate": 8.99765075555864e-06,
+      "loss": 0.5556,
+      "step": 9410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6368569700377653,
+      "learning_rate": 8.997414091017537e-06,
+      "loss": 0.3615,
+      "step": 9411
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.875230745184024,
+      "learning_rate": 8.997177401653574e-06,
+      "loss": 0.2193,
+      "step": 9412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.024929770382119,
+      "learning_rate": 8.996940687468225e-06,
+      "loss": 0.6221,
+      "step": 9413
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0957485621341938,
+      "learning_rate": 8.996703948462959e-06,
+      "loss": 0.2426,
+      "step": 9414
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9650865948881677,
+      "learning_rate": 8.996467184639244e-06,
+      "loss": 0.2052,
+      "step": 9415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6639688134125081,
+      "learning_rate": 8.99623039599855e-06,
+      "loss": 0.2929,
+      "step": 9416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6497895494906627,
+      "learning_rate": 8.995993582542353e-06,
+      "loss": 0.3888,
+      "step": 9417
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3158748640881308,
+      "learning_rate": 8.995756744272117e-06,
+      "loss": 0.3948,
+      "step": 9418
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.834596149998122,
+      "learning_rate": 8.995519881189316e-06,
+      "loss": 0.3402,
+      "step": 9419
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.088269764325449,
+      "learning_rate": 8.99528299329542e-06,
+      "loss": 0.2065,
+      "step": 9420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3736961449836036,
+      "learning_rate": 8.9950460805919e-06,
+      "loss": 0.2172,
+      "step": 9421
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.48214903872665,
+      "learning_rate": 8.994809143080228e-06,
+      "loss": 0.1896,
+      "step": 9422
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.552168626703177,
+      "learning_rate": 8.994572180761873e-06,
+      "loss": 0.263,
+      "step": 9423
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5882731442636338,
+      "learning_rate": 8.99433519363831e-06,
+      "loss": 0.2733,
+      "step": 9424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7168415549825318,
+      "learning_rate": 8.994098181711009e-06,
+      "loss": 0.2601,
+      "step": 9425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5677778370131252,
+      "learning_rate": 8.99386114498144e-06,
+      "loss": 0.3441,
+      "step": 9426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9897636646828902,
+      "learning_rate": 8.993624083451076e-06,
+      "loss": 0.4116,
+      "step": 9427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6469444761062269,
+      "learning_rate": 8.993386997121391e-06,
+      "loss": 0.2934,
+      "step": 9428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2345423645322608,
+      "learning_rate": 8.993149885993855e-06,
+      "loss": 0.2987,
+      "step": 9429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4071173800932841,
+      "learning_rate": 8.992912750069943e-06,
+      "loss": 0.3155,
+      "step": 9430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1413550114733697,
+      "learning_rate": 8.992675589351123e-06,
+      "loss": 0.1397,
+      "step": 9431
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.715604083484378,
+      "learning_rate": 8.992438403838873e-06,
+      "loss": 0.2166,
+      "step": 9432
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7641609069417186,
+      "learning_rate": 8.992201193534661e-06,
+      "loss": 0.1662,
+      "step": 9433
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9073348866692164,
+      "learning_rate": 8.991963958439964e-06,
+      "loss": 0.3162,
+      "step": 9434
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.719181327450669,
+      "learning_rate": 8.991726698556252e-06,
+      "loss": 0.3749,
+      "step": 9435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.435741406190006,
+      "learning_rate": 8.991489413885001e-06,
+      "loss": 0.1849,
+      "step": 9436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.991163158732188,
+      "learning_rate": 8.991252104427682e-06,
+      "loss": 0.4052,
+      "step": 9437
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7840039104530538,
+      "learning_rate": 8.99101477018577e-06,
+      "loss": 0.4837,
+      "step": 9438
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7319565979317364,
+      "learning_rate": 8.99077741116074e-06,
+      "loss": 0.3119,
+      "step": 9439
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.351202161526151,
+      "learning_rate": 8.990540027354064e-06,
+      "loss": 0.2192,
+      "step": 9440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.764716387202047,
+      "learning_rate": 8.990302618767217e-06,
+      "loss": 0.2627,
+      "step": 9441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6727551144634467,
+      "learning_rate": 8.990065185401673e-06,
+      "loss": 0.4549,
+      "step": 9442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8634666404623899,
+      "learning_rate": 8.989827727258906e-06,
+      "loss": 0.2466,
+      "step": 9443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5789067225429578,
+      "learning_rate": 8.989590244340391e-06,
+      "loss": 0.3445,
+      "step": 9444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9137955226379995,
+      "learning_rate": 8.989352736647603e-06,
+      "loss": 0.4853,
+      "step": 9445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4646523414170127,
+      "learning_rate": 8.989115204182014e-06,
+      "loss": 0.4815,
+      "step": 9446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4089425509352596,
+      "learning_rate": 8.988877646945105e-06,
+      "loss": 0.2625,
+      "step": 9447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6200339288455217,
+      "learning_rate": 8.988640064938345e-06,
+      "loss": 0.4661,
+      "step": 9448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8848354583610982,
+      "learning_rate": 8.988402458163214e-06,
+      "loss": 0.532,
+      "step": 9449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9738792367519742,
+      "learning_rate": 8.988164826621184e-06,
+      "loss": 0.568,
+      "step": 9450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.596499268957246,
+      "learning_rate": 8.987927170313733e-06,
+      "loss": 0.2692,
+      "step": 9451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2192907681778786,
+      "learning_rate": 8.987689489242334e-06,
+      "loss": 0.4016,
+      "step": 9452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.790725647093314,
+      "learning_rate": 8.987451783408468e-06,
+      "loss": 0.375,
+      "step": 9453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4758994364630966,
+      "learning_rate": 8.987214052813605e-06,
+      "loss": 0.3518,
+      "step": 9454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.518110341503475,
+      "learning_rate": 8.986976297459224e-06,
+      "loss": 0.4546,
+      "step": 9455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7272617576839138,
+      "learning_rate": 8.986738517346803e-06,
+      "loss": 0.5455,
+      "step": 9456
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.179151097297101,
+      "learning_rate": 8.986500712477818e-06,
+      "loss": 0.2119,
+      "step": 9457
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.110485602267477,
+      "learning_rate": 8.986262882853742e-06,
+      "loss": 0.1986,
+      "step": 9458
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7925161027185041,
+      "learning_rate": 8.986025028476057e-06,
+      "loss": 0.5466,
+      "step": 9459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5796512597716164,
+      "learning_rate": 8.985787149346236e-06,
+      "loss": 0.2507,
+      "step": 9460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6529317220039563,
+      "learning_rate": 8.985549245465758e-06,
+      "loss": 0.412,
+      "step": 9461
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5017342674621357,
+      "learning_rate": 8.985311316836101e-06,
+      "loss": 0.3758,
+      "step": 9462
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5030169446177852,
+      "learning_rate": 8.98507336345874e-06,
+      "loss": 0.2089,
+      "step": 9463
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4973093179228754,
+      "learning_rate": 8.984835385335156e-06,
+      "loss": 0.2433,
+      "step": 9464
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8603235728243348,
+      "learning_rate": 8.984597382466825e-06,
+      "loss": 0.2831,
+      "step": 9465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5314124517487446,
+      "learning_rate": 8.984359354855223e-06,
+      "loss": 0.3383,
+      "step": 9466
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4192668198753329,
+      "learning_rate": 8.984121302501832e-06,
+      "loss": 0.1697,
+      "step": 9467
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.759219009571512,
+      "learning_rate": 8.983883225408126e-06,
+      "loss": 0.5376,
+      "step": 9468
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.60344301517031,
+      "learning_rate": 8.983645123575586e-06,
+      "loss": 0.3522,
+      "step": 9469
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7634582728139772,
+      "learning_rate": 8.983406997005693e-06,
+      "loss": 0.5325,
+      "step": 9470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.839092372531754,
+      "learning_rate": 8.98316884569992e-06,
+      "loss": 0.5878,
+      "step": 9471
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.378368933995882,
+      "learning_rate": 8.982930669659749e-06,
+      "loss": 0.4073,
+      "step": 9472
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8887135363996796,
+      "learning_rate": 8.98269246888666e-06,
+      "loss": 0.1983,
+      "step": 9473
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2536037579258474,
+      "learning_rate": 8.98245424338213e-06,
+      "loss": 0.3605,
+      "step": 9474
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.921081421771057,
+      "learning_rate": 8.982215993147641e-06,
+      "loss": 0.3461,
+      "step": 9475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7393723139055066,
+      "learning_rate": 8.981977718184669e-06,
+      "loss": 0.3899,
+      "step": 9476
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.43089658130091,
+      "learning_rate": 8.981739418494696e-06,
+      "loss": 0.2214,
+      "step": 9477
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.026725090316326,
+      "learning_rate": 8.981501094079199e-06,
+      "loss": 0.4072,
+      "step": 9478
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5536745020384346,
+      "learning_rate": 8.981262744939662e-06,
+      "loss": 0.2693,
+      "step": 9479
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4743598171032906,
+      "learning_rate": 8.981024371077563e-06,
+      "loss": 0.1221,
+      "step": 9480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.845935645418154,
+      "learning_rate": 8.980785972494383e-06,
+      "loss": 0.501,
+      "step": 9481
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8857154034332837,
+      "learning_rate": 8.9805475491916e-06,
+      "loss": 0.3532,
+      "step": 9482
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5278079115419703,
+      "learning_rate": 8.980309101170698e-06,
+      "loss": 0.2685,
+      "step": 9483
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9844513799159804,
+      "learning_rate": 8.980070628433154e-06,
+      "loss": 0.1485,
+      "step": 9484
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4804792026362148,
+      "learning_rate": 8.979832130980453e-06,
+      "loss": 0.3249,
+      "step": 9485
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5348580145947273,
+      "learning_rate": 8.979593608814072e-06,
+      "loss": 0.3069,
+      "step": 9486
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3984345990156208,
+      "learning_rate": 8.979355061935496e-06,
+      "loss": 0.2531,
+      "step": 9487
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5437200720971023,
+      "learning_rate": 8.979116490346203e-06,
+      "loss": 0.513,
+      "step": 9488
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7199212375131119,
+      "learning_rate": 8.978877894047675e-06,
+      "loss": 0.3181,
+      "step": 9489
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6497801289221405,
+      "learning_rate": 8.978639273041397e-06,
+      "loss": 0.2541,
+      "step": 9490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9957240147754498,
+      "learning_rate": 8.978400627328845e-06,
+      "loss": 0.1586,
+      "step": 9491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.896694494839168,
+      "learning_rate": 8.978161956911506e-06,
+      "loss": 0.3185,
+      "step": 9492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0221227365863,
+      "learning_rate": 8.977923261790859e-06,
+      "loss": 0.377,
+      "step": 9493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5659161654807812,
+      "learning_rate": 8.977684541968388e-06,
+      "loss": 0.3953,
+      "step": 9494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.964022626778723,
+      "learning_rate": 8.977445797445575e-06,
+      "loss": 0.1648,
+      "step": 9495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1758139643759622,
+      "learning_rate": 8.977207028223902e-06,
+      "loss": 0.2017,
+      "step": 9496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7620628426897567,
+      "learning_rate": 8.976968234304853e-06,
+      "loss": 0.2827,
+      "step": 9497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5170889229183884,
+      "learning_rate": 8.976729415689909e-06,
+      "loss": 0.3151,
+      "step": 9498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9348006442971154,
+      "learning_rate": 8.976490572380555e-06,
+      "loss": 0.1515,
+      "step": 9499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6095876321785394,
+      "learning_rate": 8.976251704378269e-06,
+      "loss": 0.1775,
+      "step": 9500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5723688127289979,
+      "learning_rate": 8.976012811684542e-06,
+      "loss": 0.4631,
+      "step": 9501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.09394255151274,
+      "learning_rate": 8.975773894300853e-06,
+      "loss": 0.2095,
+      "step": 9502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8771852276342034,
+      "learning_rate": 8.975534952228687e-06,
+      "loss": 0.4346,
+      "step": 9503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8567293081326066,
+      "learning_rate": 8.975295985469525e-06,
+      "loss": 0.4127,
+      "step": 9504
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6895407461035161,
+      "learning_rate": 8.975056994024856e-06,
+      "loss": 0.286,
+      "step": 9505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3798724131587132,
+      "learning_rate": 8.97481797789616e-06,
+      "loss": 0.3214,
+      "step": 9506
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9054987279043905,
+      "learning_rate": 8.974578937084922e-06,
+      "loss": 0.2188,
+      "step": 9507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6492227360918021,
+      "learning_rate": 8.974339871592625e-06,
+      "loss": 0.4194,
+      "step": 9508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1363824891022054,
+      "learning_rate": 8.974100781420758e-06,
+      "loss": 0.1443,
+      "step": 9509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9725417283307556,
+      "learning_rate": 8.973861666570803e-06,
+      "loss": 0.4899,
+      "step": 9510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7016487832695009,
+      "learning_rate": 8.973622527044244e-06,
+      "loss": 0.2426,
+      "step": 9511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1143268459405875,
+      "learning_rate": 8.973383362842565e-06,
+      "loss": 0.7121,
+      "step": 9512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4588874116974204,
+      "learning_rate": 8.973144173967255e-06,
+      "loss": 0.1979,
+      "step": 9513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7080495010269943,
+      "learning_rate": 8.972904960419797e-06,
+      "loss": 0.3545,
+      "step": 9514
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.447242324785686,
+      "learning_rate": 8.972665722201677e-06,
+      "loss": 0.3525,
+      "step": 9515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5724080726115182,
+      "learning_rate": 8.972426459314378e-06,
+      "loss": 0.3598,
+      "step": 9516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6058565307235677,
+      "learning_rate": 8.972187171759391e-06,
+      "loss": 0.1777,
+      "step": 9517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.635810354150417,
+      "learning_rate": 8.971947859538198e-06,
+      "loss": 0.339,
+      "step": 9518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2107397133287632,
+      "learning_rate": 8.971708522652284e-06,
+      "loss": 0.2106,
+      "step": 9519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.582865106738347,
+      "learning_rate": 8.97146916110314e-06,
+      "loss": 0.3417,
+      "step": 9520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4358067802727412,
+      "learning_rate": 8.971229774892248e-06,
+      "loss": 0.211,
+      "step": 9521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8189547882275638,
+      "learning_rate": 8.970990364021097e-06,
+      "loss": 0.3774,
+      "step": 9522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8983553173973653,
+      "learning_rate": 8.970750928491173e-06,
+      "loss": 0.314,
+      "step": 9523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6556977034950584,
+      "learning_rate": 8.970511468303962e-06,
+      "loss": 0.4675,
+      "step": 9524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9411486871212376,
+      "learning_rate": 8.970271983460951e-06,
+      "loss": 0.4407,
+      "step": 9525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8011478539120316,
+      "learning_rate": 8.970032473963628e-06,
+      "loss": 0.4554,
+      "step": 9526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9487022392122713,
+      "learning_rate": 8.969792939813482e-06,
+      "loss": 0.2289,
+      "step": 9527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0403287489243,
+      "learning_rate": 8.969553381011996e-06,
+      "loss": 0.258,
+      "step": 9528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5873188429818865,
+      "learning_rate": 8.969313797560661e-06,
+      "loss": 0.2857,
+      "step": 9529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.475149510471952,
+      "learning_rate": 8.969074189460964e-06,
+      "loss": 0.2427,
+      "step": 9530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3487466003389303,
+      "learning_rate": 8.968834556714394e-06,
+      "loss": 0.7809,
+      "step": 9531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3912460480361726,
+      "learning_rate": 8.968594899322435e-06,
+      "loss": 0.2589,
+      "step": 9532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.918314349081668,
+      "learning_rate": 8.96835521728658e-06,
+      "loss": 0.2914,
+      "step": 9533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1631024706619852,
+      "learning_rate": 8.968115510608316e-06,
+      "loss": 0.3384,
+      "step": 9534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.330564647988374,
+      "learning_rate": 8.967875779289128e-06,
+      "loss": 0.2394,
+      "step": 9535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.655502641051008,
+      "learning_rate": 8.96763602333051e-06,
+      "loss": 0.4347,
+      "step": 9536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8125246638231336,
+      "learning_rate": 8.967396242733948e-06,
+      "loss": 0.388,
+      "step": 9537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7730243386439275,
+      "learning_rate": 8.96715643750093e-06,
+      "loss": 0.394,
+      "step": 9538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6534707484834419,
+      "learning_rate": 8.966916607632946e-06,
+      "loss": 0.5188,
+      "step": 9539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.079486178850479,
+      "learning_rate": 8.966676753131489e-06,
+      "loss": 0.4429,
+      "step": 9540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7702883517578705,
+      "learning_rate": 8.966436873998044e-06,
+      "loss": 0.5085,
+      "step": 9541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5633522475104302,
+      "learning_rate": 8.966196970234101e-06,
+      "loss": 0.2965,
+      "step": 9542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8085243055222724,
+      "learning_rate": 8.965957041841148e-06,
+      "loss": 0.3934,
+      "step": 9543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5697537538366912,
+      "learning_rate": 8.96571708882068e-06,
+      "loss": 0.2769,
+      "step": 9544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7520627600941452,
+      "learning_rate": 8.965477111174185e-06,
+      "loss": 0.2499,
+      "step": 9545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6554393741401046,
+      "learning_rate": 8.965237108903153e-06,
+      "loss": 0.2661,
+      "step": 9546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5691935647358117,
+      "learning_rate": 8.964997082009073e-06,
+      "loss": 0.4483,
+      "step": 9547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.539712139448908,
+      "learning_rate": 8.964757030493435e-06,
+      "loss": 0.3554,
+      "step": 9548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5799426856540417,
+      "learning_rate": 8.964516954357733e-06,
+      "loss": 0.2206,
+      "step": 9549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3062847376954567,
+      "learning_rate": 8.964276853603456e-06,
+      "loss": 0.2956,
+      "step": 9550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5595936448006635,
+      "learning_rate": 8.964036728232093e-06,
+      "loss": 0.2247,
+      "step": 9551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2457290010196045,
+      "learning_rate": 8.963796578245139e-06,
+      "loss": 0.1675,
+      "step": 9552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7736202998623631,
+      "learning_rate": 8.963556403644081e-06,
+      "loss": 0.334,
+      "step": 9553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.148390162553337,
+      "learning_rate": 8.963316204430414e-06,
+      "loss": 0.507,
+      "step": 9554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7625202921576808,
+      "learning_rate": 8.963075980605629e-06,
+      "loss": 0.2116,
+      "step": 9555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1303028127262023,
+      "learning_rate": 8.962835732171216e-06,
+      "loss": 0.2256,
+      "step": 9556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9150921990756644,
+      "learning_rate": 8.962595459128669e-06,
+      "loss": 0.5286,
+      "step": 9557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0454560923751126,
+      "learning_rate": 8.962355161479478e-06,
+      "loss": 0.636,
+      "step": 9558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5265458967034142,
+      "learning_rate": 8.962114839225135e-06,
+      "loss": 0.2949,
+      "step": 9559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.510105395854583,
+      "learning_rate": 8.961874492367135e-06,
+      "loss": 0.2312,
+      "step": 9560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.611722564970116,
+      "learning_rate": 8.961634120906969e-06,
+      "loss": 0.3066,
+      "step": 9561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.728785212555358,
+      "learning_rate": 8.961393724846128e-06,
+      "loss": 0.2771,
+      "step": 9562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9170847834932807,
+      "learning_rate": 8.961153304186107e-06,
+      "loss": 0.2963,
+      "step": 9563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6627914270980946,
+      "learning_rate": 8.9609128589284e-06,
+      "loss": 0.4683,
+      "step": 9564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9869333018604476,
+      "learning_rate": 8.960672389074497e-06,
+      "loss": 0.6314,
+      "step": 9565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6048732729894417,
+      "learning_rate": 8.960431894625892e-06,
+      "loss": 0.3522,
+      "step": 9566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8637656775722784,
+      "learning_rate": 8.96019137558408e-06,
+      "loss": 0.5058,
+      "step": 9567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3228716412401316,
+      "learning_rate": 8.959950831950554e-06,
+      "loss": 0.2054,
+      "step": 9568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.492524730562945,
+      "learning_rate": 8.959710263726806e-06,
+      "loss": 0.3388,
+      "step": 9569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7766737125583851,
+      "learning_rate": 8.959469670914331e-06,
+      "loss": 0.5812,
+      "step": 9570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6056957165653847,
+      "learning_rate": 8.959229053514625e-06,
+      "loss": 0.3452,
+      "step": 9571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4830299607894737,
+      "learning_rate": 8.95898841152918e-06,
+      "loss": 0.3294,
+      "step": 9572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4391245732293134,
+      "learning_rate": 8.958747744959488e-06,
+      "loss": 0.2857,
+      "step": 9573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4962657152352885,
+      "learning_rate": 8.958507053807048e-06,
+      "loss": 0.3566,
+      "step": 9574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.023161217521777,
+      "learning_rate": 8.958266338073354e-06,
+      "loss": 0.4662,
+      "step": 9575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.899206147177593,
+      "learning_rate": 8.958025597759899e-06,
+      "loss": 0.5213,
+      "step": 9576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9184455317819864,
+      "learning_rate": 8.957784832868177e-06,
+      "loss": 0.6292,
+      "step": 9577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.864346193883284,
+      "learning_rate": 8.957544043399687e-06,
+      "loss": 0.4186,
+      "step": 9578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8091507974546126,
+      "learning_rate": 8.957303229355919e-06,
+      "loss": 0.4009,
+      "step": 9579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2600156343763567,
+      "learning_rate": 8.957062390738373e-06,
+      "loss": 0.1806,
+      "step": 9580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.829891313363528,
+      "learning_rate": 8.956821527548543e-06,
+      "loss": 0.4047,
+      "step": 9581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.950149465871105,
+      "learning_rate": 8.956580639787922e-06,
+      "loss": 0.2574,
+      "step": 9582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6902245122111783,
+      "learning_rate": 8.95633972745801e-06,
+      "loss": 0.3025,
+      "step": 9583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.812310019763889,
+      "learning_rate": 8.956098790560302e-06,
+      "loss": 0.4637,
+      "step": 9584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6236649815559387,
+      "learning_rate": 8.955857829096294e-06,
+      "loss": 0.1716,
+      "step": 9585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4153867653726053,
+      "learning_rate": 8.95561684306748e-06,
+      "loss": 0.3116,
+      "step": 9586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6638522686934851,
+      "learning_rate": 8.955375832475357e-06,
+      "loss": 0.4013,
+      "step": 9587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4821822144854262,
+      "learning_rate": 8.955134797321425e-06,
+      "loss": 0.2973,
+      "step": 9588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7380712468025508,
+      "learning_rate": 8.954893737607177e-06,
+      "loss": 0.2432,
+      "step": 9589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5301725077985124,
+      "learning_rate": 8.954652653334113e-06,
+      "loss": 0.2809,
+      "step": 9590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.116237427211666,
+      "learning_rate": 8.95441154450373e-06,
+      "loss": 0.2879,
+      "step": 9591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5046682545212502,
+      "learning_rate": 8.95417041111752e-06,
+      "loss": 0.163,
+      "step": 9592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6918508548796305,
+      "learning_rate": 8.953929253176988e-06,
+      "loss": 0.4148,
+      "step": 9593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6919711317196355,
+      "learning_rate": 8.953688070683623e-06,
+      "loss": 0.4315,
+      "step": 9594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4637999554159238,
+      "learning_rate": 8.953446863638932e-06,
+      "loss": 0.2578,
+      "step": 9595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8375626919068597,
+      "learning_rate": 8.953205632044406e-06,
+      "loss": 0.3526,
+      "step": 9596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2649823181591633,
+      "learning_rate": 8.952964375901545e-06,
+      "loss": 0.3066,
+      "step": 9597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3069105682020765,
+      "learning_rate": 8.952723095211847e-06,
+      "loss": 0.377,
+      "step": 9598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.003243867940839,
+      "learning_rate": 8.952481789976812e-06,
+      "loss": 0.1415,
+      "step": 9599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2583446680067942,
+      "learning_rate": 8.952240460197937e-06,
+      "loss": 0.3371,
+      "step": 9600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.02327413525451,
+      "learning_rate": 8.951999105876719e-06,
+      "loss": 0.522,
+      "step": 9601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0676097209873527,
+      "learning_rate": 8.951757727014659e-06,
+      "loss": 0.1161,
+      "step": 9602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.817771222375635,
+      "learning_rate": 8.951516323613257e-06,
+      "loss": 0.3801,
+      "step": 9603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.732271000809006,
+      "learning_rate": 8.951274895674006e-06,
+      "loss": 0.4504,
+      "step": 9604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5237583898386875,
+      "learning_rate": 8.951033443198414e-06,
+      "loss": 0.2109,
+      "step": 9605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.121740460379199,
+      "learning_rate": 8.950791966187975e-06,
+      "loss": 0.2423,
+      "step": 9606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7913920465597526,
+      "learning_rate": 8.950550464644187e-06,
+      "loss": 0.4503,
+      "step": 9607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.350078056979582,
+      "learning_rate": 8.950308938568555e-06,
+      "loss": 0.2449,
+      "step": 9608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1068196112255766,
+      "learning_rate": 8.950067387962575e-06,
+      "loss": 0.584,
+      "step": 9609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2183937337437625,
+      "learning_rate": 8.949825812827746e-06,
+      "loss": 0.2416,
+      "step": 9610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9527744283763486,
+      "learning_rate": 8.949584213165571e-06,
+      "loss": 0.6691,
+      "step": 9611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3023626062988092,
+      "learning_rate": 8.94934258897755e-06,
+      "loss": 0.2835,
+      "step": 9612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3589382785524178,
+      "learning_rate": 8.949100940265183e-06,
+      "loss": 0.1914,
+      "step": 9613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5969424716856695,
+      "learning_rate": 8.948859267029969e-06,
+      "loss": 0.355,
+      "step": 9614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.997192857610444,
+      "learning_rate": 8.94861756927341e-06,
+      "loss": 0.1483,
+      "step": 9615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6797340941065675,
+      "learning_rate": 8.948375846997006e-06,
+      "loss": 0.3127,
+      "step": 9616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1011484963215143,
+      "learning_rate": 8.94813410020226e-06,
+      "loss": 0.4166,
+      "step": 9617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6920722152523486,
+      "learning_rate": 8.947892328890673e-06,
+      "loss": 0.3695,
+      "step": 9618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9840781898497022,
+      "learning_rate": 8.947650533063743e-06,
+      "loss": 0.332,
+      "step": 9619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2359047491723332,
+      "learning_rate": 8.947408712722975e-06,
+      "loss": 0.1952,
+      "step": 9620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2536928538979237,
+      "learning_rate": 8.947166867869871e-06,
+      "loss": 0.1942,
+      "step": 9621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4114251068352224,
+      "learning_rate": 8.94692499850593e-06,
+      "loss": 0.3219,
+      "step": 9622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3220574061479056,
+      "learning_rate": 8.946683104632654e-06,
+      "loss": 0.2328,
+      "step": 9623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.515883695540165,
+      "learning_rate": 8.946441186251547e-06,
+      "loss": 0.3445,
+      "step": 9624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8842648527745094,
+      "learning_rate": 8.94619924336411e-06,
+      "loss": 0.4107,
+      "step": 9625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5468617616915412,
+      "learning_rate": 8.945957275971848e-06,
+      "loss": 0.2732,
+      "step": 9626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3846446281697316,
+      "learning_rate": 8.945715284076259e-06,
+      "loss": 0.2135,
+      "step": 9627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3844425163424428,
+      "learning_rate": 8.945473267678851e-06,
+      "loss": 0.1909,
+      "step": 9628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7812613690604677,
+      "learning_rate": 8.945231226781122e-06,
+      "loss": 0.5074,
+      "step": 9629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.028419642801041,
+      "learning_rate": 8.944989161384577e-06,
+      "loss": 0.4075,
+      "step": 9630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7013771839831293,
+      "learning_rate": 8.94474707149072e-06,
+      "loss": 0.4056,
+      "step": 9631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3907828455719657,
+      "learning_rate": 8.944504957101053e-06,
+      "loss": 0.3296,
+      "step": 9632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6052166260495397,
+      "learning_rate": 8.94426281821708e-06,
+      "loss": 0.3455,
+      "step": 9633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.481405864937222,
+      "learning_rate": 8.944020654840305e-06,
+      "loss": 0.4887,
+      "step": 9634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.17411036342887,
+      "learning_rate": 8.943778466972232e-06,
+      "loss": 0.2198,
+      "step": 9635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.67378276198176,
+      "learning_rate": 8.943536254614364e-06,
+      "loss": 0.3468,
+      "step": 9636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8806580131703658,
+      "learning_rate": 8.943294017768205e-06,
+      "loss": 0.2024,
+      "step": 9637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.450799307302166,
+      "learning_rate": 8.943051756435259e-06,
+      "loss": 0.1796,
+      "step": 9638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8863850726965303,
+      "learning_rate": 8.942809470617031e-06,
+      "loss": 0.3215,
+      "step": 9639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.487269636071201,
+      "learning_rate": 8.942567160315027e-06,
+      "loss": 0.2831,
+      "step": 9640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7899187425299028,
+      "learning_rate": 8.942324825530748e-06,
+      "loss": 0.3829,
+      "step": 9641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.912037734110852,
+      "learning_rate": 8.942082466265702e-06,
+      "loss": 0.3278,
+      "step": 9642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.293385066978612,
+      "learning_rate": 8.941840082521393e-06,
+      "loss": 0.2159,
+      "step": 9643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7806058250076329,
+      "learning_rate": 8.941597674299327e-06,
+      "loss": 0.1954,
+      "step": 9644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4014607388309555,
+      "learning_rate": 8.941355241601009e-06,
+      "loss": 0.2175,
+      "step": 9645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8573980921557438,
+      "learning_rate": 8.941112784427942e-06,
+      "loss": 0.3809,
+      "step": 9646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1730207230723397,
+      "learning_rate": 8.940870302781635e-06,
+      "loss": 0.1682,
+      "step": 9647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7281018898851774,
+      "learning_rate": 8.94062779666359e-06,
+      "loss": 0.4519,
+      "step": 9648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.015301375913749,
+      "learning_rate": 8.940385266075318e-06,
+      "loss": 0.1458,
+      "step": 9649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.485373259215991,
+      "learning_rate": 8.94014271101832e-06,
+      "loss": 0.1457,
+      "step": 9650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6171728262488978,
+      "learning_rate": 8.939900131494105e-06,
+      "loss": 0.4783,
+      "step": 9651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.592189686523823,
+      "learning_rate": 8.939657527504178e-06,
+      "loss": 0.3074,
+      "step": 9652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.378696250729239,
+      "learning_rate": 8.939414899050048e-06,
+      "loss": 0.6834,
+      "step": 9653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4796829352464527,
+      "learning_rate": 8.939172246133218e-06,
+      "loss": 0.2171,
+      "step": 9654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.559392052998688,
+      "learning_rate": 8.938929568755198e-06,
+      "loss": 0.4033,
+      "step": 9655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.417901994957707,
+      "learning_rate": 8.938686866917494e-06,
+      "loss": 0.6249,
+      "step": 9656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9259996212997785,
+      "learning_rate": 8.938444140621612e-06,
+      "loss": 0.4416,
+      "step": 9657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5679094018239017,
+      "learning_rate": 8.938201389869058e-06,
+      "loss": 0.1437,
+      "step": 9658
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.177011764792483,
+      "learning_rate": 8.937958614661345e-06,
+      "loss": 0.2234,
+      "step": 9659
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9917216108264953,
+      "learning_rate": 8.937715814999976e-06,
+      "loss": 0.1586,
+      "step": 9660
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8694219526996454,
+      "learning_rate": 8.937472990886459e-06,
+      "loss": 0.3414,
+      "step": 9661
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.401041940833234,
+      "learning_rate": 8.937230142322301e-06,
+      "loss": 0.3245,
+      "step": 9662
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3654220468799156,
+      "learning_rate": 8.936987269309015e-06,
+      "loss": 0.3663,
+      "step": 9663
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6143368022629407,
+      "learning_rate": 8.936744371848103e-06,
+      "loss": 0.2943,
+      "step": 9664
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.981385002139623,
+      "learning_rate": 8.936501449941079e-06,
+      "loss": 0.2174,
+      "step": 9665
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9824859683946103,
+      "learning_rate": 8.936258503589445e-06,
+      "loss": 0.6768,
+      "step": 9666
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.656283184834685,
+      "learning_rate": 8.936015532794716e-06,
+      "loss": 0.4501,
+      "step": 9667
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0736453957529353,
+      "learning_rate": 8.935772537558398e-06,
+      "loss": 0.6488,
+      "step": 9668
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8366939041618493,
+      "learning_rate": 8.935529517881999e-06,
+      "loss": 0.4887,
+      "step": 9669
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7703553858484034,
+      "learning_rate": 8.93528647376703e-06,
+      "loss": 0.1787,
+      "step": 9670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.752977856157532,
+      "learning_rate": 8.935043405214998e-06,
+      "loss": 0.2955,
+      "step": 9671
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2842249118978746,
+      "learning_rate": 8.934800312227415e-06,
+      "loss": 0.1877,
+      "step": 9672
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2878974779575543,
+      "learning_rate": 8.934557194805787e-06,
+      "loss": 0.1827,
+      "step": 9673
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.170397776322186,
+      "learning_rate": 8.934314052951628e-06,
+      "loss": 0.1636,
+      "step": 9674
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.907359340724945,
+      "learning_rate": 8.934070886666443e-06,
+      "loss": 0.4532,
+      "step": 9675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9288292030097531,
+      "learning_rate": 8.93382769595175e-06,
+      "loss": 0.291,
+      "step": 9676
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9489162302048393,
+      "learning_rate": 8.933584480809049e-06,
+      "loss": 0.668,
+      "step": 9677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6641718331538375,
+      "learning_rate": 8.933341241239857e-06,
+      "loss": 0.5029,
+      "step": 9678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7405541037429901,
+      "learning_rate": 8.933097977245682e-06,
+      "loss": 0.3228,
+      "step": 9679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6342220268541268,
+      "learning_rate": 8.932854688828036e-06,
+      "loss": 0.3171,
+      "step": 9680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.281391647696629,
+      "learning_rate": 8.93261137598843e-06,
+      "loss": 0.3017,
+      "step": 9681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6281511708735328,
+      "learning_rate": 8.932368038728371e-06,
+      "loss": 0.1835,
+      "step": 9682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9044169187611406,
+      "learning_rate": 8.932124677049375e-06,
+      "loss": 0.427,
+      "step": 9683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1710566277820162,
+      "learning_rate": 8.93188129095295e-06,
+      "loss": 0.2016,
+      "step": 9684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0863046903352926,
+      "learning_rate": 8.931637880440612e-06,
+      "loss": 0.2497,
+      "step": 9685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4833476677569875,
+      "learning_rate": 8.931394445513866e-06,
+      "loss": 0.3475,
+      "step": 9686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7161353836455358,
+      "learning_rate": 8.931150986174226e-06,
+      "loss": 0.2099,
+      "step": 9687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.899300359544111,
+      "learning_rate": 8.930907502423205e-06,
+      "loss": 0.2559,
+      "step": 9688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9440211822550646,
+      "learning_rate": 8.930663994262316e-06,
+      "loss": 0.2322,
+      "step": 9689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0673681302781337,
+      "learning_rate": 8.930420461693068e-06,
+      "loss": 0.3714,
+      "step": 9690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.245858062320182,
+      "learning_rate": 8.930176904716976e-06,
+      "loss": 0.3214,
+      "step": 9691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.781603008559821,
+      "learning_rate": 8.929933323335552e-06,
+      "loss": 0.452,
+      "step": 9692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6820451557808929,
+      "learning_rate": 8.929689717550307e-06,
+      "loss": 0.422,
+      "step": 9693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3967467535062101,
+      "learning_rate": 8.929446087362755e-06,
+      "loss": 0.3575,
+      "step": 9694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9156965870883935,
+      "learning_rate": 8.929202432774408e-06,
+      "loss": 0.306,
+      "step": 9695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3141208089735172,
+      "learning_rate": 8.928958753786779e-06,
+      "loss": 0.1703,
+      "step": 9696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5384025465849347,
+      "learning_rate": 8.928715050401382e-06,
+      "loss": 0.3067,
+      "step": 9697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6952093769826777,
+      "learning_rate": 8.92847132261973e-06,
+      "loss": 0.3884,
+      "step": 9698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3634277424452645,
+      "learning_rate": 8.928227570443338e-06,
+      "loss": 0.6059,
+      "step": 9699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4157618538597536,
+      "learning_rate": 8.927983793873715e-06,
+      "loss": 0.3912,
+      "step": 9700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0938023133199166,
+      "learning_rate": 8.927739992912381e-06,
+      "loss": 0.1632,
+      "step": 9701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0968094578148717,
+      "learning_rate": 8.927496167560845e-06,
+      "loss": 0.574,
+      "step": 9702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1974068558127073,
+      "learning_rate": 8.927252317820625e-06,
+      "loss": 0.3923,
+      "step": 9703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9288645468275276,
+      "learning_rate": 8.927008443693232e-06,
+      "loss": 0.3795,
+      "step": 9704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7023838644216371,
+      "learning_rate": 8.926764545180182e-06,
+      "loss": 0.5286,
+      "step": 9705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.758282587984302,
+      "learning_rate": 8.926520622282988e-06,
+      "loss": 0.401,
+      "step": 9706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7924219099562806,
+      "learning_rate": 8.926276675003167e-06,
+      "loss": 0.2358,
+      "step": 9707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.652321907782875,
+      "learning_rate": 8.926032703342232e-06,
+      "loss": 0.4603,
+      "step": 9708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2387567251456932,
+      "learning_rate": 8.925788707301699e-06,
+      "loss": 0.434,
+      "step": 9709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6940393501604176,
+      "learning_rate": 8.925544686883083e-06,
+      "loss": 0.4529,
+      "step": 9710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8242637559293167,
+      "learning_rate": 8.9253006420879e-06,
+      "loss": 0.2057,
+      "step": 9711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2257333884912764,
+      "learning_rate": 8.925056572917663e-06,
+      "loss": 0.3651,
+      "step": 9712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8715898656444754,
+      "learning_rate": 8.92481247937389e-06,
+      "loss": 0.3372,
+      "step": 9713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3660756549293436,
+      "learning_rate": 8.924568361458095e-06,
+      "loss": 0.2028,
+      "step": 9714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.074123107371791,
+      "learning_rate": 8.924324219171795e-06,
+      "loss": 0.5093,
+      "step": 9715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.25290042132407,
+      "learning_rate": 8.924080052516505e-06,
+      "loss": 0.2166,
+      "step": 9716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.261169940176879,
+      "learning_rate": 8.923835861493746e-06,
+      "loss": 0.3358,
+      "step": 9717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5946928722702995,
+      "learning_rate": 8.923591646105027e-06,
+      "loss": 0.3295,
+      "step": 9718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4265948234481034,
+      "learning_rate": 8.92334740635187e-06,
+      "loss": 0.2318,
+      "step": 9719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.716811484924575,
+      "learning_rate": 8.923103142235788e-06,
+      "loss": 0.1862,
+      "step": 9720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1589095420322615,
+      "learning_rate": 8.9228588537583e-06,
+      "loss": 0.14,
+      "step": 9721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0715904969725893,
+      "learning_rate": 8.922614540920922e-06,
+      "loss": 0.2425,
+      "step": 9722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2217757707923689,
+      "learning_rate": 8.922370203725172e-06,
+      "loss": 0.2415,
+      "step": 9723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.70463402416853,
+      "learning_rate": 8.922125842172568e-06,
+      "loss": 0.4251,
+      "step": 9724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8510433628143914,
+      "learning_rate": 8.921881456264626e-06,
+      "loss": 0.3001,
+      "step": 9725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4489589648539056,
+      "learning_rate": 8.921637046002861e-06,
+      "loss": 0.2489,
+      "step": 9726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8805663292758044,
+      "learning_rate": 8.921392611388797e-06,
+      "loss": 0.2548,
+      "step": 9727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.182442731947109,
+      "learning_rate": 8.921148152423946e-06,
+      "loss": 0.2685,
+      "step": 9728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0519205507270524,
+      "learning_rate": 8.920903669109829e-06,
+      "loss": 0.4486,
+      "step": 9729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8408463008270473,
+      "learning_rate": 8.920659161447965e-06,
+      "loss": 0.3321,
+      "step": 9730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4850358721585963,
+      "learning_rate": 8.920414629439868e-06,
+      "loss": 0.3113,
+      "step": 9731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6307334211674323,
+      "learning_rate": 8.920170073087062e-06,
+      "loss": 0.2792,
+      "step": 9732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0790692672263864,
+      "learning_rate": 8.919925492391063e-06,
+      "loss": 0.3743,
+      "step": 9733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7007028346981663,
+      "learning_rate": 8.919680887353388e-06,
+      "loss": 0.5104,
+      "step": 9734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7148887263419574,
+      "learning_rate": 8.91943625797556e-06,
+      "loss": 0.2333,
+      "step": 9735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7909652573919421,
+      "learning_rate": 8.919191604259094e-06,
+      "loss": 0.3034,
+      "step": 9736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.276635061349986,
+      "learning_rate": 8.918946926205511e-06,
+      "loss": 0.4086,
+      "step": 9737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8991481702769941,
+      "learning_rate": 8.91870222381633e-06,
+      "loss": 0.3737,
+      "step": 9738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4721483928334138,
+      "learning_rate": 8.918457497093074e-06,
+      "loss": 0.3301,
+      "step": 9739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.119677532756935,
+      "learning_rate": 8.918212746037258e-06,
+      "loss": 0.2748,
+      "step": 9740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5985006511962538,
+      "learning_rate": 8.917967970650402e-06,
+      "loss": 0.3018,
+      "step": 9741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1572385220349397,
+      "learning_rate": 8.91772317093403e-06,
+      "loss": 0.1842,
+      "step": 9742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7087357966974077,
+      "learning_rate": 8.917478346889659e-06,
+      "loss": 0.3194,
+      "step": 9743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.768479154459893,
+      "learning_rate": 8.917233498518809e-06,
+      "loss": 0.5694,
+      "step": 9744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6378404146991585,
+      "learning_rate": 8.916988625823002e-06,
+      "loss": 0.3609,
+      "step": 9745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.840569253896617,
+      "learning_rate": 8.916743728803756e-06,
+      "loss": 0.5913,
+      "step": 9746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8738622829686942,
+      "learning_rate": 8.916498807462597e-06,
+      "loss": 0.4838,
+      "step": 9747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4278985423564483,
+      "learning_rate": 8.91625386180104e-06,
+      "loss": 0.1836,
+      "step": 9748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.011506656465983,
+      "learning_rate": 8.91600889182061e-06,
+      "loss": 0.5292,
+      "step": 9749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0117387196558902,
+      "learning_rate": 8.915763897522827e-06,
+      "loss": 0.3435,
+      "step": 9750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0832835684236746,
+      "learning_rate": 8.91551887890921e-06,
+      "loss": 0.4007,
+      "step": 9751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8464252451965264,
+      "learning_rate": 8.915273835981284e-06,
+      "loss": 0.3944,
+      "step": 9752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3940753378285513,
+      "learning_rate": 8.91502876874057e-06,
+      "loss": 0.2198,
+      "step": 9753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8158473354715627,
+      "learning_rate": 8.914783677188589e-06,
+      "loss": 0.5666,
+      "step": 9754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4643127343587383,
+      "learning_rate": 8.914538561326864e-06,
+      "loss": 0.3344,
+      "step": 9755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3842702946155439,
+      "learning_rate": 8.914293421156914e-06,
+      "loss": 0.1516,
+      "step": 9756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.925861975968432,
+      "learning_rate": 8.914048256680264e-06,
+      "loss": 0.4023,
+      "step": 9757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4653144550635862,
+      "learning_rate": 8.913803067898436e-06,
+      "loss": 0.2548,
+      "step": 9758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5942385125853762,
+      "learning_rate": 8.913557854812953e-06,
+      "loss": 0.2445,
+      "step": 9759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8613058651276573,
+      "learning_rate": 8.913312617425339e-06,
+      "loss": 0.3087,
+      "step": 9760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9795436466643965,
+      "learning_rate": 8.91306735573711e-06,
+      "loss": 0.2397,
+      "step": 9761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8593770095269377,
+      "learning_rate": 8.912822069749798e-06,
+      "loss": 0.3437,
+      "step": 9762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5569861781156669,
+      "learning_rate": 8.912576759464922e-06,
+      "loss": 0.2088,
+      "step": 9763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5546860575070232,
+      "learning_rate": 8.912331424884002e-06,
+      "loss": 0.3172,
+      "step": 9764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5381792627510604,
+      "learning_rate": 8.912086066008567e-06,
+      "loss": 0.3172,
+      "step": 9765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.54004462125618,
+      "learning_rate": 8.91184068284014e-06,
+      "loss": 0.388,
+      "step": 9766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8686591593585955,
+      "learning_rate": 8.911595275380243e-06,
+      "loss": 0.0965,
+      "step": 9767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5936189064638269,
+      "learning_rate": 8.911349843630399e-06,
+      "loss": 0.2232,
+      "step": 9768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8846957020809603,
+      "learning_rate": 8.911104387592134e-06,
+      "loss": 0.4199,
+      "step": 9769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.031745314729155,
+      "learning_rate": 8.910858907266971e-06,
+      "loss": 0.3111,
+      "step": 9770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4444281225008746,
+      "learning_rate": 8.910613402656434e-06,
+      "loss": 0.5643,
+      "step": 9771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0821163913473726,
+      "learning_rate": 8.91036787376205e-06,
+      "loss": 0.1708,
+      "step": 9772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.457992226580579,
+      "learning_rate": 8.91012232058534e-06,
+      "loss": 0.2616,
+      "step": 9773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.177994846795469,
+      "learning_rate": 8.909876743127833e-06,
+      "loss": 0.4568,
+      "step": 9774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2918619022145967,
+      "learning_rate": 8.909631141391051e-06,
+      "loss": 0.7138,
+      "step": 9775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0489637822714024,
+      "learning_rate": 8.909385515376518e-06,
+      "loss": 0.1903,
+      "step": 9776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5538790052661673,
+      "learning_rate": 8.909139865085763e-06,
+      "loss": 0.2799,
+      "step": 9777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4577868499216533,
+      "learning_rate": 8.90889419052031e-06,
+      "loss": 0.3174,
+      "step": 9778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.445841161594947,
+      "learning_rate": 8.908648491681683e-06,
+      "loss": 0.189,
+      "step": 9779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.354943827915517,
+      "learning_rate": 8.908402768571411e-06,
+      "loss": 0.2121,
+      "step": 9780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.263602144301424,
+      "learning_rate": 8.908157021191017e-06,
+      "loss": 0.336,
+      "step": 9781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7057961604864469,
+      "learning_rate": 8.907911249542027e-06,
+      "loss": 0.2612,
+      "step": 9782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2961317079358756,
+      "learning_rate": 8.90766545362597e-06,
+      "loss": 0.5249,
+      "step": 9783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.475667129780262,
+      "learning_rate": 8.907419633444368e-06,
+      "loss": 0.4226,
+      "step": 9784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7883608495779444,
+      "learning_rate": 8.90717378899875e-06,
+      "loss": 0.4437,
+      "step": 9785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.102518006257869,
+      "learning_rate": 8.906927920290641e-06,
+      "loss": 0.3913,
+      "step": 9786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4028803056589227,
+      "learning_rate": 8.906682027321574e-06,
+      "loss": 0.2313,
+      "step": 9787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.03804773533418,
+      "learning_rate": 8.906436110093067e-06,
+      "loss": 0.3222,
+      "step": 9788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7909588321180143,
+      "learning_rate": 8.906190168606654e-06,
+      "loss": 0.471,
+      "step": 9789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.655682617295932,
+      "learning_rate": 8.905944202863856e-06,
+      "loss": 0.2849,
+      "step": 9790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2146479202163594,
+      "learning_rate": 8.905698212866208e-06,
+      "loss": 0.691,
+      "step": 9791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7451977276268247,
+      "learning_rate": 8.90545219861523e-06,
+      "loss": 0.492,
+      "step": 9792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.261655928508489,
+      "learning_rate": 8.905206160112453e-06,
+      "loss": 0.2735,
+      "step": 9793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6549286182177763,
+      "learning_rate": 8.904960097359408e-06,
+      "loss": 0.336,
+      "step": 9794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9714283539726438,
+      "learning_rate": 8.904714010357617e-06,
+      "loss": 0.3579,
+      "step": 9795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.293480898769675,
+      "learning_rate": 8.904467899108611e-06,
+      "loss": 0.4892,
+      "step": 9796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7660910679077042,
+      "learning_rate": 8.90422176361392e-06,
+      "loss": 0.2409,
+      "step": 9797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3296744844671506,
+      "learning_rate": 8.90397560387507e-06,
+      "loss": 0.1675,
+      "step": 9798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7992048480001581,
+      "learning_rate": 8.90372941989359e-06,
+      "loss": 0.2037,
+      "step": 9799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.772805340727403,
+      "learning_rate": 8.903483211671008e-06,
+      "loss": 0.2971,
+      "step": 9800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2587832930900993,
+      "learning_rate": 8.903236979208856e-06,
+      "loss": 0.1755,
+      "step": 9801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2929728845028883,
+      "learning_rate": 8.90299072250866e-06,
+      "loss": 0.2773,
+      "step": 9802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5423175137606804,
+      "learning_rate": 8.90274444157195e-06,
+      "loss": 0.2822,
+      "step": 9803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7155061541669014,
+      "learning_rate": 8.902498136400258e-06,
+      "loss": 0.6021,
+      "step": 9804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9190107902246607,
+      "learning_rate": 8.902251806995106e-06,
+      "loss": 0.3289,
+      "step": 9805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.565476039074126,
+      "learning_rate": 8.902005453358033e-06,
+      "loss": 0.245,
+      "step": 9806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2567825306035683,
+      "learning_rate": 8.90175907549056e-06,
+      "loss": 0.1342,
+      "step": 9807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0813927449714003,
+      "learning_rate": 8.901512673394226e-06,
+      "loss": 0.7286,
+      "step": 9808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7415904146671675,
+      "learning_rate": 8.901266247070555e-06,
+      "loss": 0.3145,
+      "step": 9809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9965888657876218,
+      "learning_rate": 8.901019796521078e-06,
+      "loss": 0.269,
+      "step": 9810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9494831005331579,
+      "learning_rate": 8.900773321747324e-06,
+      "loss": 0.5603,
+      "step": 9811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.933290924094924,
+      "learning_rate": 8.900526822750828e-06,
+      "loss": 0.5621,
+      "step": 9812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8093190427883428,
+      "learning_rate": 8.900280299533119e-06,
+      "loss": 0.2115,
+      "step": 9813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9244194154727455,
+      "learning_rate": 8.900033752095726e-06,
+      "loss": 0.191,
+      "step": 9814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7823120389324358,
+      "learning_rate": 8.899787180440181e-06,
+      "loss": 0.5515,
+      "step": 9815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5858548358927633,
+      "learning_rate": 8.899540584568016e-06,
+      "loss": 0.363,
+      "step": 9816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6277279697380984,
+      "learning_rate": 8.899293964480761e-06,
+      "loss": 0.4598,
+      "step": 9817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3730438905252114,
+      "learning_rate": 8.899047320179947e-06,
+      "loss": 0.3286,
+      "step": 9818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1252882881114763,
+      "learning_rate": 8.898800651667109e-06,
+      "loss": 0.5008,
+      "step": 9819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.867843823820463,
+      "learning_rate": 8.898553958943775e-06,
+      "loss": 0.453,
+      "step": 9820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.981075517450323,
+      "learning_rate": 8.898307242011479e-06,
+      "loss": 0.278,
+      "step": 9821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7582431392866689,
+      "learning_rate": 8.89806050087175e-06,
+      "loss": 0.4425,
+      "step": 9822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2259124989504322,
+      "learning_rate": 8.897813735526124e-06,
+      "loss": 0.1832,
+      "step": 9823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7510714763185675,
+      "learning_rate": 8.897566945976133e-06,
+      "loss": 0.268,
+      "step": 9824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.506634459558343,
+      "learning_rate": 8.897320132223307e-06,
+      "loss": 0.4329,
+      "step": 9825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4382614977203967,
+      "learning_rate": 8.89707329426918e-06,
+      "loss": 0.3136,
+      "step": 9826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8348786486354143,
+      "learning_rate": 8.896826432115285e-06,
+      "loss": 0.3313,
+      "step": 9827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.353250909766232,
+      "learning_rate": 8.896579545763153e-06,
+      "loss": 0.4717,
+      "step": 9828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1972299386908827,
+      "learning_rate": 8.896332635214321e-06,
+      "loss": 0.1579,
+      "step": 9829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.513180372364595,
+      "learning_rate": 8.89608570047032e-06,
+      "loss": 0.27,
+      "step": 9830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7849045707534048,
+      "learning_rate": 8.895838741532682e-06,
+      "loss": 0.3475,
+      "step": 9831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6673237851255698,
+      "learning_rate": 8.895591758402944e-06,
+      "loss": 0.3634,
+      "step": 9832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7513218553601633,
+      "learning_rate": 8.895344751082635e-06,
+      "loss": 0.3474,
+      "step": 9833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5477162732226921,
+      "learning_rate": 8.895097719573294e-06,
+      "loss": 0.3897,
+      "step": 9834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8011030273747914,
+      "learning_rate": 8.894850663876452e-06,
+      "loss": 0.3983,
+      "step": 9835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8769635330213308,
+      "learning_rate": 8.894603583993645e-06,
+      "loss": 0.3308,
+      "step": 9836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5561371327648168,
+      "learning_rate": 8.894356479926403e-06,
+      "loss": 0.223,
+      "step": 9837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8222455997736429,
+      "learning_rate": 8.894109351676265e-06,
+      "loss": 0.245,
+      "step": 9838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5204538275727622,
+      "learning_rate": 8.893862199244762e-06,
+      "loss": 0.3019,
+      "step": 9839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4787685979107341,
+      "learning_rate": 8.893615022633434e-06,
+      "loss": 0.3831,
+      "step": 9840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6373505886359894,
+      "learning_rate": 8.89336782184381e-06,
+      "loss": 0.4654,
+      "step": 9841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2895243019814455,
+      "learning_rate": 8.893120596877429e-06,
+      "loss": 0.2016,
+      "step": 9842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4127101213904432,
+      "learning_rate": 8.892873347735826e-06,
+      "loss": 0.3252,
+      "step": 9843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6678279844111688,
+      "learning_rate": 8.892626074420535e-06,
+      "loss": 0.2274,
+      "step": 9844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6884610190263534,
+      "learning_rate": 8.89237877693309e-06,
+      "loss": 0.5718,
+      "step": 9845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8232785074695737,
+      "learning_rate": 8.892131455275031e-06,
+      "loss": 0.2468,
+      "step": 9846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.765670480395577,
+      "learning_rate": 8.891884109447888e-06,
+      "loss": 0.3954,
+      "step": 9847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6237742954900376,
+      "learning_rate": 8.891636739453202e-06,
+      "loss": 0.3745,
+      "step": 9848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3689896395551095,
+      "learning_rate": 8.891389345292507e-06,
+      "loss": 0.2376,
+      "step": 9849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.781340395992782,
+      "learning_rate": 8.891141926967342e-06,
+      "loss": 0.343,
+      "step": 9850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4210662428095284,
+      "learning_rate": 8.890894484479237e-06,
+      "loss": 0.2375,
+      "step": 9851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.387890219507177,
+      "learning_rate": 8.890647017829735e-06,
+      "loss": 0.2016,
+      "step": 9852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4591261196094596,
+      "learning_rate": 8.890399527020369e-06,
+      "loss": 0.3715,
+      "step": 9853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5954898997597258,
+      "learning_rate": 8.890152012052678e-06,
+      "loss": 0.2013,
+      "step": 9854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3122634930055947,
+      "learning_rate": 8.889904472928197e-06,
+      "loss": 0.1829,
+      "step": 9855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9721404581590245,
+      "learning_rate": 8.889656909648466e-06,
+      "loss": 0.5911,
+      "step": 9856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7659041217855327,
+      "learning_rate": 8.889409322215017e-06,
+      "loss": 0.4548,
+      "step": 9857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.603337300057603,
+      "learning_rate": 8.889161710629395e-06,
+      "loss": 0.3629,
+      "step": 9858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3513501854034784,
+      "learning_rate": 8.88891407489313e-06,
+      "loss": 0.3035,
+      "step": 9859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3132693776422184,
+      "learning_rate": 8.888666415007766e-06,
+      "loss": 0.8672,
+      "step": 9860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9890175563352446,
+      "learning_rate": 8.888418730974838e-06,
+      "loss": 0.4379,
+      "step": 9861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5377523078564692,
+      "learning_rate": 8.888171022795882e-06,
+      "loss": 0.3445,
+      "step": 9862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.082379181438003,
+      "learning_rate": 8.88792329047244e-06,
+      "loss": 0.2866,
+      "step": 9863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8583104708645128,
+      "learning_rate": 8.88767553400605e-06,
+      "loss": 0.2727,
+      "step": 9864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7752345481311425,
+      "learning_rate": 8.887427753398249e-06,
+      "loss": 0.2881,
+      "step": 9865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1675665020802883,
+      "learning_rate": 8.887179948650573e-06,
+      "loss": 0.21,
+      "step": 9866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4533040195346931,
+      "learning_rate": 8.886932119764566e-06,
+      "loss": 0.2368,
+      "step": 9867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.06285034440856,
+      "learning_rate": 8.886684266741765e-06,
+      "loss": 0.6253,
+      "step": 9868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8167995396268304,
+      "learning_rate": 8.886436389583707e-06,
+      "loss": 0.1675,
+      "step": 9869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4689252677800635,
+      "learning_rate": 8.886188488291936e-06,
+      "loss": 0.3294,
+      "step": 9870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9045346194804007,
+      "learning_rate": 8.885940562867986e-06,
+      "loss": 0.3681,
+      "step": 9871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3888137197683568,
+      "learning_rate": 8.8856926133134e-06,
+      "loss": 0.458,
+      "step": 9872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.562629789683947,
+      "learning_rate": 8.885444639629717e-06,
+      "loss": 0.3516,
+      "step": 9873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.099128630847093,
+      "learning_rate": 8.885196641818475e-06,
+      "loss": 0.4874,
+      "step": 9874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7485407360747405,
+      "learning_rate": 8.884948619881218e-06,
+      "loss": 0.3072,
+      "step": 9875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8768851777003064,
+      "learning_rate": 8.884700573819482e-06,
+      "loss": 0.3768,
+      "step": 9876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9686218401656859,
+      "learning_rate": 8.884452503634808e-06,
+      "loss": 0.4928,
+      "step": 9877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7663382760508355,
+      "learning_rate": 8.88420440932874e-06,
+      "loss": 0.5221,
+      "step": 9878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3403706293159332,
+      "learning_rate": 8.883956290902813e-06,
+      "loss": 0.2646,
+      "step": 9879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6718209783792668,
+      "learning_rate": 8.883708148358574e-06,
+      "loss": 0.435,
+      "step": 9880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7501065570883163,
+      "learning_rate": 8.883459981697559e-06,
+      "loss": 0.391,
+      "step": 9881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9822114729450246,
+      "learning_rate": 8.88321179092131e-06,
+      "loss": 0.1826,
+      "step": 9882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8560208151840654,
+      "learning_rate": 8.882963576031371e-06,
+      "loss": 0.386,
+      "step": 9883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2546202982462582,
+      "learning_rate": 8.88271533702928e-06,
+      "loss": 0.1567,
+      "step": 9884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7574781841678133,
+      "learning_rate": 8.882467073916579e-06,
+      "loss": 0.3055,
+      "step": 9885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.021121397002599,
+      "learning_rate": 8.882218786694811e-06,
+      "loss": 0.5809,
+      "step": 9886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.400349593791766,
+      "learning_rate": 8.881970475365519e-06,
+      "loss": 0.1712,
+      "step": 9887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9213269118970353,
+      "learning_rate": 8.881722139930243e-06,
+      "loss": 0.2274,
+      "step": 9888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0573970494059024,
+      "learning_rate": 8.881473780390523e-06,
+      "loss": 0.3916,
+      "step": 9889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.100616791807582,
+      "learning_rate": 8.881225396747904e-06,
+      "loss": 0.5473,
+      "step": 9890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.256984301402621,
+      "learning_rate": 8.880976989003929e-06,
+      "loss": 0.7532,
+      "step": 9891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9145640338963414,
+      "learning_rate": 8.880728557160139e-06,
+      "loss": 0.329,
+      "step": 9892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0786073813198986,
+      "learning_rate": 8.880480101218077e-06,
+      "loss": 0.4038,
+      "step": 9893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1603147273243908,
+      "learning_rate": 8.880231621179288e-06,
+      "loss": 0.3004,
+      "step": 9894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4660684679216207,
+      "learning_rate": 8.87998311704531e-06,
+      "loss": 0.2208,
+      "step": 9895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9187432956889072,
+      "learning_rate": 8.87973458881769e-06,
+      "loss": 0.5129,
+      "step": 9896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4718765020869018,
+      "learning_rate": 8.87948603649797e-06,
+      "loss": 0.1988,
+      "step": 9897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.255266446883812,
+      "learning_rate": 8.879237460087695e-06,
+      "loss": 0.2894,
+      "step": 9898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1721490911730124,
+      "learning_rate": 8.878988859588406e-06,
+      "loss": 0.3535,
+      "step": 9899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5145938993826151,
+      "learning_rate": 8.87874023500165e-06,
+      "loss": 0.3606,
+      "step": 9900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.018277160730167,
+      "learning_rate": 8.878491586328968e-06,
+      "loss": 0.6186,
+      "step": 9901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5437155099541058,
+      "learning_rate": 8.878242913571906e-06,
+      "loss": 0.1971,
+      "step": 9902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4642744458492682,
+      "learning_rate": 8.877994216732006e-06,
+      "loss": 0.1892,
+      "step": 9903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8274432932735352,
+      "learning_rate": 8.877745495810815e-06,
+      "loss": 0.4893,
+      "step": 9904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5187478362762674,
+      "learning_rate": 8.877496750809876e-06,
+      "loss": 0.1843,
+      "step": 9905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0578884530732107,
+      "learning_rate": 8.877247981730733e-06,
+      "loss": 0.5161,
+      "step": 9906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4309775103498543,
+      "learning_rate": 8.87699918857493e-06,
+      "loss": 0.3327,
+      "step": 9907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3990936052139933,
+      "learning_rate": 8.876750371344016e-06,
+      "loss": 0.3465,
+      "step": 9908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7904307368223307,
+      "learning_rate": 8.876501530039533e-06,
+      "loss": 0.4786,
+      "step": 9909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4825665374454196,
+      "learning_rate": 8.876252664663027e-06,
+      "loss": 0.3479,
+      "step": 9910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2510921386169065,
+      "learning_rate": 8.876003775216045e-06,
+      "loss": 0.166,
+      "step": 9911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.283953717076389,
+      "learning_rate": 8.875754861700127e-06,
+      "loss": 0.6662,
+      "step": 9912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0500086609727215,
+      "learning_rate": 8.875505924116825e-06,
+      "loss": 0.5543,
+      "step": 9913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5277635529125628,
+      "learning_rate": 8.875256962467682e-06,
+      "loss": 0.4049,
+      "step": 9914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0613907012780985,
+      "learning_rate": 8.875007976754244e-06,
+      "loss": 0.5192,
+      "step": 9915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8163096612455116,
+      "learning_rate": 8.874758966978056e-06,
+      "loss": 0.2438,
+      "step": 9916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5711808460778611,
+      "learning_rate": 8.874509933140667e-06,
+      "loss": 0.4593,
+      "step": 9917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7514081573033522,
+      "learning_rate": 8.874260875243621e-06,
+      "loss": 0.3915,
+      "step": 9918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0907270792913843,
+      "learning_rate": 8.874011793288468e-06,
+      "loss": 0.5618,
+      "step": 9919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3132398122539457,
+      "learning_rate": 8.87376268727675e-06,
+      "loss": 0.3033,
+      "step": 9920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3616616950062035,
+      "learning_rate": 8.873513557210019e-06,
+      "loss": 0.1962,
+      "step": 9921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4060071311609763,
+      "learning_rate": 8.873264403089817e-06,
+      "loss": 0.2615,
+      "step": 9922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2627121761225575,
+      "learning_rate": 8.873015224917696e-06,
+      "loss": 0.2422,
+      "step": 9923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7734588084012057,
+      "learning_rate": 8.872766022695199e-06,
+      "loss": 0.6609,
+      "step": 9924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2750829496824068,
+      "learning_rate": 8.872516796423874e-06,
+      "loss": 0.6108,
+      "step": 9925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4696778457273005,
+      "learning_rate": 8.872267546105271e-06,
+      "loss": 0.2328,
+      "step": 9926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5505814627801342,
+      "learning_rate": 8.87201827174094e-06,
+      "loss": 0.4627,
+      "step": 9927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8317112673728895,
+      "learning_rate": 8.87176897333242e-06,
+      "loss": 0.6035,
+      "step": 9928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3114311884564704,
+      "learning_rate": 8.871519650881269e-06,
+      "loss": 0.352,
+      "step": 9929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.404734573806038,
+      "learning_rate": 8.871270304389029e-06,
+      "loss": 0.278,
+      "step": 9930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4450377409330888,
+      "learning_rate": 8.871020933857251e-06,
+      "loss": 0.2835,
+      "step": 9931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8115991083056915,
+      "learning_rate": 8.870771539287482e-06,
+      "loss": 0.2038,
+      "step": 9932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5710070827050238,
+      "learning_rate": 8.870522120681273e-06,
+      "loss": 0.1966,
+      "step": 9933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5225887257188706,
+      "learning_rate": 8.870272678040171e-06,
+      "loss": 0.2677,
+      "step": 9934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0203221098869695,
+      "learning_rate": 8.870023211365723e-06,
+      "loss": 0.5808,
+      "step": 9935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7834574798980793,
+      "learning_rate": 8.869773720659483e-06,
+      "loss": 0.3731,
+      "step": 9936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9542830734177323,
+      "learning_rate": 8.869524205922997e-06,
+      "loss": 0.4877,
+      "step": 9937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5328297105653188,
+      "learning_rate": 8.869274667157815e-06,
+      "loss": 0.3687,
+      "step": 9938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7183998444811077,
+      "learning_rate": 8.869025104365487e-06,
+      "loss": 0.5602,
+      "step": 9939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7247968768345878,
+      "learning_rate": 8.868775517547563e-06,
+      "loss": 0.5108,
+      "step": 9940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.416265186811704,
+      "learning_rate": 8.86852590670559e-06,
+      "loss": 0.2547,
+      "step": 9941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.654835106193444,
+      "learning_rate": 8.868276271841123e-06,
+      "loss": 0.2529,
+      "step": 9942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6508518471766669,
+      "learning_rate": 8.868026612955707e-06,
+      "loss": 0.2375,
+      "step": 9943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8769704121280828,
+      "learning_rate": 8.867776930050898e-06,
+      "loss": 0.3191,
+      "step": 9944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2651828225972772,
+      "learning_rate": 8.867527223128241e-06,
+      "loss": 0.2585,
+      "step": 9945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5309052809178592,
+      "learning_rate": 8.86727749218929e-06,
+      "loss": 0.3409,
+      "step": 9946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5864114663842273,
+      "learning_rate": 8.867027737235592e-06,
+      "loss": 0.4269,
+      "step": 9947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.697213241242866,
+      "learning_rate": 8.866777958268704e-06,
+      "loss": 0.3857,
+      "step": 9948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8452283299022167,
+      "learning_rate": 8.866528155290171e-06,
+      "loss": 0.382,
+      "step": 9949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.239803838224577,
+      "learning_rate": 8.866278328301547e-06,
+      "loss": 0.6234,
+      "step": 9950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8837941408459689,
+      "learning_rate": 8.866028477304385e-06,
+      "loss": 0.39,
+      "step": 9951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.201319208096959,
+      "learning_rate": 8.865778602300232e-06,
+      "loss": 0.3159,
+      "step": 9952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2599575125760523,
+      "learning_rate": 8.865528703290645e-06,
+      "loss": 0.2226,
+      "step": 9953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5364886609548798,
+      "learning_rate": 8.86527878027717e-06,
+      "loss": 0.3285,
+      "step": 9954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0779178128896847,
+      "learning_rate": 8.865028833261366e-06,
+      "loss": 0.606,
+      "step": 9955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7180734256429047,
+      "learning_rate": 8.864778862244778e-06,
+      "loss": 0.362,
+      "step": 9956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4478382761464128,
+      "learning_rate": 8.864528867228963e-06,
+      "loss": 0.1972,
+      "step": 9957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.164717915848445,
+      "learning_rate": 8.864278848215472e-06,
+      "loss": 0.5665,
+      "step": 9958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0814252404448492,
+      "learning_rate": 8.864028805205857e-06,
+      "loss": 0.692,
+      "step": 9959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0777820645565983,
+      "learning_rate": 8.863778738201671e-06,
+      "loss": 0.2176,
+      "step": 9960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6742234714599908,
+      "learning_rate": 8.863528647204465e-06,
+      "loss": 0.5217,
+      "step": 9961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2063346100919863,
+      "learning_rate": 8.863278532215797e-06,
+      "loss": 0.5492,
+      "step": 9962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3401653415442265,
+      "learning_rate": 8.863028393237215e-06,
+      "loss": 0.4623,
+      "step": 9963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1058981915583783,
+      "learning_rate": 8.862778230270276e-06,
+      "loss": 0.1386,
+      "step": 9964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1614992118670922,
+      "learning_rate": 8.86252804331653e-06,
+      "loss": 0.1664,
+      "step": 9965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.674779169983722,
+      "learning_rate": 8.862277832377534e-06,
+      "loss": 0.2207,
+      "step": 9966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9243754891088,
+      "learning_rate": 8.862027597454838e-06,
+      "loss": 0.3764,
+      "step": 9967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5600116322691984,
+      "learning_rate": 8.861777338550002e-06,
+      "loss": 0.4375,
+      "step": 9968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3774013493178945,
+      "learning_rate": 8.861527055664573e-06,
+      "loss": 0.2254,
+      "step": 9969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5356267718731353,
+      "learning_rate": 8.861276748800109e-06,
+      "loss": 0.2484,
+      "step": 9970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7905464764299195,
+      "learning_rate": 8.861026417958164e-06,
+      "loss": 0.4433,
+      "step": 9971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8373125223126765,
+      "learning_rate": 8.86077606314029e-06,
+      "loss": 0.4681,
+      "step": 9972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.369767320408112,
+      "learning_rate": 8.860525684348046e-06,
+      "loss": 0.1664,
+      "step": 9973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.716990943191591,
+      "learning_rate": 8.860275281582983e-06,
+      "loss": 0.3335,
+      "step": 9974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1560146182412205,
+      "learning_rate": 8.860024854846659e-06,
+      "loss": 0.1694,
+      "step": 9975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2446880363943988,
+      "learning_rate": 8.859774404140627e-06,
+      "loss": 0.2476,
+      "step": 9976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6223036697523014,
+      "learning_rate": 8.859523929466443e-06,
+      "loss": 0.3418,
+      "step": 9977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5237740071653156,
+      "learning_rate": 8.859273430825661e-06,
+      "loss": 0.3766,
+      "step": 9978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.839535478816951,
+      "learning_rate": 8.859022908219839e-06,
+      "loss": 0.3692,
+      "step": 9979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6833112174255744,
+      "learning_rate": 8.85877236165053e-06,
+      "loss": 0.3963,
+      "step": 9980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1229621954544595,
+      "learning_rate": 8.858521791119292e-06,
+      "loss": 0.6496,
+      "step": 9981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4123045890595995,
+      "learning_rate": 8.858271196627679e-06,
+      "loss": 0.1687,
+      "step": 9982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5698641022974378,
+      "learning_rate": 8.858020578177248e-06,
+      "loss": 0.2149,
+      "step": 9983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2355158820955074,
+      "learning_rate": 8.857769935769557e-06,
+      "loss": 0.2364,
+      "step": 9984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.370703410694596,
+      "learning_rate": 8.857519269406162e-06,
+      "loss": 0.3499,
+      "step": 9985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0749504354467043,
+      "learning_rate": 8.857268579088618e-06,
+      "loss": 0.1652,
+      "step": 9986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.839810340723797,
+      "learning_rate": 8.85701786481848e-06,
+      "loss": 0.6826,
+      "step": 9987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5899942570521146,
+      "learning_rate": 8.85676712659731e-06,
+      "loss": 0.3391,
+      "step": 9988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8703173936096267,
+      "learning_rate": 8.85651636442666e-06,
+      "loss": 0.7189,
+      "step": 9989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7782892206648722,
+      "learning_rate": 8.85626557830809e-06,
+      "loss": 0.3625,
+      "step": 9990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.093144692276784,
+      "learning_rate": 8.856014768243157e-06,
+      "loss": 0.5392,
+      "step": 9991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.06770801824444,
+      "learning_rate": 8.855763934233417e-06,
+      "loss": 0.4711,
+      "step": 9992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.707192858503054,
+      "learning_rate": 8.85551307628043e-06,
+      "loss": 0.2888,
+      "step": 9993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.15292430185916,
+      "learning_rate": 8.855262194385753e-06,
+      "loss": 0.2771,
+      "step": 9994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2284431375487235,
+      "learning_rate": 8.855011288550941e-06,
+      "loss": 0.1494,
+      "step": 9995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.32277386813276,
+      "learning_rate": 8.854760358777555e-06,
+      "loss": 0.1828,
+      "step": 9996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4169641961265005,
+      "learning_rate": 8.854509405067153e-06,
+      "loss": 0.2586,
+      "step": 9997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.337612189953174,
+      "learning_rate": 8.854258427421293e-06,
+      "loss": 0.3629,
+      "step": 9998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.95810342873917,
+      "learning_rate": 8.854007425841536e-06,
+      "loss": 0.507,
+      "step": 9999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3198377811075177,
+      "learning_rate": 8.853756400329434e-06,
+      "loss": 0.1732,
+      "step": 10000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4727596631154813,
+      "learning_rate": 8.853505350886554e-06,
+      "loss": 0.2562,
+      "step": 10001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3643722348924316,
+      "learning_rate": 8.853254277514448e-06,
+      "loss": 0.4114,
+      "step": 10002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2176701762723343,
+      "learning_rate": 8.853003180214677e-06,
+      "loss": 0.1365,
+      "step": 10003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8952303392947638,
+      "learning_rate": 8.852752058988802e-06,
+      "loss": 0.4394,
+      "step": 10004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.728572941492524,
+      "learning_rate": 8.852500913838384e-06,
+      "loss": 0.2575,
+      "step": 10005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.427251929976922,
+      "learning_rate": 8.852249744764978e-06,
+      "loss": 0.204,
+      "step": 10006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.85370788340849,
+      "learning_rate": 8.851998551770147e-06,
+      "loss": 0.2124,
+      "step": 10007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0615666401653687,
+      "learning_rate": 8.851747334855448e-06,
+      "loss": 0.2599,
+      "step": 10008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.357543015048807,
+      "learning_rate": 8.851496094022445e-06,
+      "loss": 0.2794,
+      "step": 10009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0005188581845412,
+      "learning_rate": 8.851244829272694e-06,
+      "loss": 0.3775,
+      "step": 10010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0537657247464907,
+      "learning_rate": 8.850993540607757e-06,
+      "loss": 0.5122,
+      "step": 10011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7777122870028408,
+      "learning_rate": 8.850742228029197e-06,
+      "loss": 0.3452,
+      "step": 10012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5838426956991538,
+      "learning_rate": 8.85049089153857e-06,
+      "loss": 0.2564,
+      "step": 10013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.751217954711374,
+      "learning_rate": 8.85023953113744e-06,
+      "loss": 0.5246,
+      "step": 10014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7694261753388472,
+      "learning_rate": 8.849988146827365e-06,
+      "loss": 0.1881,
+      "step": 10015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6977656673696557,
+      "learning_rate": 8.849736738609907e-06,
+      "loss": 0.1833,
+      "step": 10016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.293355938541936,
+      "learning_rate": 8.84948530648663e-06,
+      "loss": 0.2753,
+      "step": 10017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0989826408367667,
+      "learning_rate": 8.849233850459094e-06,
+      "loss": 0.193,
+      "step": 10018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.125392583037423,
+      "learning_rate": 8.848982370528857e-06,
+      "loss": 0.4562,
+      "step": 10019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6252180056986973,
+      "learning_rate": 8.848730866697485e-06,
+      "loss": 0.2341,
+      "step": 10020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4786499288728874,
+      "learning_rate": 8.848479338966539e-06,
+      "loss": 0.262,
+      "step": 10021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.689227151917757,
+      "learning_rate": 8.84822778733758e-06,
+      "loss": 0.3976,
+      "step": 10022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3775885377293617,
+      "learning_rate": 8.847976211812167e-06,
+      "loss": 0.1495,
+      "step": 10023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.025500604481585,
+      "learning_rate": 8.847724612391869e-06,
+      "loss": 0.3009,
+      "step": 10024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7669647949404477,
+      "learning_rate": 8.847472989078243e-06,
+      "loss": 0.2186,
+      "step": 10025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7372270087491009,
+      "learning_rate": 8.847221341872853e-06,
+      "loss": 0.3863,
+      "step": 10026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9213556289581113,
+      "learning_rate": 8.846969670777263e-06,
+      "loss": 0.2583,
+      "step": 10027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9439037902285397,
+      "learning_rate": 8.846717975793034e-06,
+      "loss": 0.2464,
+      "step": 10028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9209722592976046,
+      "learning_rate": 8.84646625692173e-06,
+      "loss": 0.4433,
+      "step": 10029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.640161185324847,
+      "learning_rate": 8.846214514164913e-06,
+      "loss": 0.2321,
+      "step": 10030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8766464316269258,
+      "learning_rate": 8.845962747524148e-06,
+      "loss": 0.5494,
+      "step": 10031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.274720212914567,
+      "learning_rate": 8.845710957000997e-06,
+      "loss": 0.2466,
+      "step": 10032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0279769541582198,
+      "learning_rate": 8.845459142597024e-06,
+      "loss": 0.1593,
+      "step": 10033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0444127580544884,
+      "learning_rate": 8.845207304313792e-06,
+      "loss": 0.2201,
+      "step": 10034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5844518933456702,
+      "learning_rate": 8.844955442152865e-06,
+      "loss": 0.314,
+      "step": 10035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5067679784831027,
+      "learning_rate": 8.844703556115807e-06,
+      "loss": 0.3875,
+      "step": 10036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0659674843877185,
+      "learning_rate": 8.844451646204186e-06,
+      "loss": 0.3107,
+      "step": 10037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.587815558616967,
+      "learning_rate": 8.84419971241956e-06,
+      "loss": 0.3848,
+      "step": 10038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.112211968745403,
+      "learning_rate": 8.843947754763497e-06,
+      "loss": 0.4069,
+      "step": 10039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.789891509006263,
+      "learning_rate": 8.843695773237561e-06,
+      "loss": 0.5356,
+      "step": 10040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8959187736723002,
+      "learning_rate": 8.843443767843319e-06,
+      "loss": 0.1789,
+      "step": 10041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7782627768847896,
+      "learning_rate": 8.84319173858233e-06,
+      "loss": 0.3496,
+      "step": 10042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8903308276995243,
+      "learning_rate": 8.842939685456164e-06,
+      "loss": 0.569,
+      "step": 10043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6820965679365574,
+      "learning_rate": 8.842687608466386e-06,
+      "loss": 0.2188,
+      "step": 10044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0184106078676765,
+      "learning_rate": 8.84243550761456e-06,
+      "loss": 0.2094,
+      "step": 10045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.804343913345331,
+      "learning_rate": 8.84218338290225e-06,
+      "loss": 0.6333,
+      "step": 10046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4160274728739761,
+      "learning_rate": 8.841931234331023e-06,
+      "loss": 0.2634,
+      "step": 10047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.006934794197802,
+      "learning_rate": 8.841679061902446e-06,
+      "loss": 0.4888,
+      "step": 10048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5884992294167068,
+      "learning_rate": 8.841426865618084e-06,
+      "loss": 0.3055,
+      "step": 10049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8176817199777253,
+      "learning_rate": 8.841174645479502e-06,
+      "loss": 0.4693,
+      "step": 10050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1746871394760419,
+      "learning_rate": 8.840922401488268e-06,
+      "loss": 0.3012,
+      "step": 10051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0105731368735285,
+      "learning_rate": 8.840670133645947e-06,
+      "loss": 0.6748,
+      "step": 10052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7416059353003386,
+      "learning_rate": 8.840417841954108e-06,
+      "loss": 0.3732,
+      "step": 10053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4333238152658774,
+      "learning_rate": 8.840165526414312e-06,
+      "loss": 0.2361,
+      "step": 10054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6015925706971192,
+      "learning_rate": 8.839913187028133e-06,
+      "loss": 0.6098,
+      "step": 10055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8521808806871174,
+      "learning_rate": 8.839660823797134e-06,
+      "loss": 0.6153,
+      "step": 10056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6104607530615354,
+      "learning_rate": 8.839408436722883e-06,
+      "loss": 0.3541,
+      "step": 10057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3398947638553214,
+      "learning_rate": 8.839156025806945e-06,
+      "loss": 0.6369,
+      "step": 10058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3304683460183462,
+      "learning_rate": 8.83890359105089e-06,
+      "loss": 0.3422,
+      "step": 10059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.568472196670615,
+      "learning_rate": 8.838651132456285e-06,
+      "loss": 0.3631,
+      "step": 10060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.025105904235246,
+      "learning_rate": 8.838398650024698e-06,
+      "loss": 0.4472,
+      "step": 10061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3647832793763834,
+      "learning_rate": 8.838146143757695e-06,
+      "loss": 0.3048,
+      "step": 10062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.114900890278786,
+      "learning_rate": 8.837893613656847e-06,
+      "loss": 0.4468,
+      "step": 10063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.038119052811142,
+      "learning_rate": 8.837641059723721e-06,
+      "loss": 0.3575,
+      "step": 10064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5228888660836213,
+      "learning_rate": 8.837388481959883e-06,
+      "loss": 0.1799,
+      "step": 10065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2287765371419928,
+      "learning_rate": 8.837135880366905e-06,
+      "loss": 0.1564,
+      "step": 10066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8428848447211355,
+      "learning_rate": 8.836883254946353e-06,
+      "loss": 0.2774,
+      "step": 10067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4936926874633898,
+      "learning_rate": 8.836630605699796e-06,
+      "loss": 0.2908,
+      "step": 10068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8615233968355107,
+      "learning_rate": 8.836377932628804e-06,
+      "loss": 0.1829,
+      "step": 10069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8892664540532027,
+      "learning_rate": 8.836125235734944e-06,
+      "loss": 0.4423,
+      "step": 10070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3432490846675265,
+      "learning_rate": 8.835872515019789e-06,
+      "loss": 0.243,
+      "step": 10071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7512673659351943,
+      "learning_rate": 8.835619770484905e-06,
+      "loss": 0.2873,
+      "step": 10072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5047190290536632,
+      "learning_rate": 8.835367002131863e-06,
+      "loss": 0.3292,
+      "step": 10073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7614440576940007,
+      "learning_rate": 8.835114209962232e-06,
+      "loss": 0.3725,
+      "step": 10074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1060862686738397,
+      "learning_rate": 8.834861393977581e-06,
+      "loss": 0.2124,
+      "step": 10075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8039271437309323,
+      "learning_rate": 8.834608554179482e-06,
+      "loss": 0.2717,
+      "step": 10076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6754205010886176,
+      "learning_rate": 8.834355690569502e-06,
+      "loss": 0.2814,
+      "step": 10077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2004902384209286,
+      "learning_rate": 8.834102803149217e-06,
+      "loss": 0.6568,
+      "step": 10078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8148934288959933,
+      "learning_rate": 8.833849891920188e-06,
+      "loss": 0.4389,
+      "step": 10079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5407498351661362,
+      "learning_rate": 8.833596956883995e-06,
+      "loss": 0.3135,
+      "step": 10080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6227938205731864,
+      "learning_rate": 8.833343998042203e-06,
+      "loss": 0.1952,
+      "step": 10081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1019020302628464,
+      "learning_rate": 8.833091015396385e-06,
+      "loss": 0.7663,
+      "step": 10082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8775272225756927,
+      "learning_rate": 8.83283800894811e-06,
+      "loss": 0.4749,
+      "step": 10083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8179215704185603,
+      "learning_rate": 8.83258497869895e-06,
+      "loss": 0.3855,
+      "step": 10084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9484192578915618,
+      "learning_rate": 8.832331924650481e-06,
+      "loss": 0.5463,
+      "step": 10085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8294095616615869,
+      "learning_rate": 8.832078846804266e-06,
+      "loss": 0.5009,
+      "step": 10086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4192934678982183,
+      "learning_rate": 8.831825745161883e-06,
+      "loss": 0.2113,
+      "step": 10087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5012935087869503,
+      "learning_rate": 8.8315726197249e-06,
+      "loss": 0.3984,
+      "step": 10088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5641402266221645,
+      "learning_rate": 8.83131947049489e-06,
+      "loss": 0.3516,
+      "step": 10089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.766221646123835,
+      "learning_rate": 8.831066297473425e-06,
+      "loss": 0.5055,
+      "step": 10090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5430270630645855,
+      "learning_rate": 8.830813100662079e-06,
+      "loss": 0.355,
+      "step": 10091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.301666011929908,
+      "learning_rate": 8.830559880062422e-06,
+      "loss": 0.2559,
+      "step": 10092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.758345412959358,
+      "learning_rate": 8.830306635676025e-06,
+      "loss": 0.2926,
+      "step": 10093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6483347440394711,
+      "learning_rate": 8.830053367504465e-06,
+      "loss": 0.1962,
+      "step": 10094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8785818141065334,
+      "learning_rate": 8.829800075549311e-06,
+      "loss": 0.6119,
+      "step": 10095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8266707447829051,
+      "learning_rate": 8.829546759812138e-06,
+      "loss": 0.5182,
+      "step": 10096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6189381710985224,
+      "learning_rate": 8.829293420294517e-06,
+      "loss": 0.4056,
+      "step": 10097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5877892061809864,
+      "learning_rate": 8.829040056998025e-06,
+      "loss": 0.17,
+      "step": 10098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0411710835074715,
+      "learning_rate": 8.828786669924232e-06,
+      "loss": 0.2034,
+      "step": 10099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7389842637004325,
+      "learning_rate": 8.828533259074709e-06,
+      "loss": 0.2825,
+      "step": 10100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9432070495102056,
+      "learning_rate": 8.828279824451036e-06,
+      "loss": 0.465,
+      "step": 10101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5300678085805213,
+      "learning_rate": 8.828026366054782e-06,
+      "loss": 0.1494,
+      "step": 10102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6561832414524318,
+      "learning_rate": 8.827772883887524e-06,
+      "loss": 0.4595,
+      "step": 10103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.762231303930716,
+      "learning_rate": 8.827519377950834e-06,
+      "loss": 0.4172,
+      "step": 10104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5603357871739933,
+      "learning_rate": 8.827265848246287e-06,
+      "loss": 0.2401,
+      "step": 10105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4394249570411959,
+      "learning_rate": 8.827012294775455e-06,
+      "loss": 0.3934,
+      "step": 10106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7113385361923128,
+      "learning_rate": 8.826758717539916e-06,
+      "loss": 0.3224,
+      "step": 10107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9332192087111408,
+      "learning_rate": 8.826505116541243e-06,
+      "loss": 0.3124,
+      "step": 10108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8153596241745122,
+      "learning_rate": 8.826251491781012e-06,
+      "loss": 0.3246,
+      "step": 10109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1436077825532545,
+      "learning_rate": 8.825997843260796e-06,
+      "loss": 0.1456,
+      "step": 10110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4088736113818205,
+      "learning_rate": 8.825744170982172e-06,
+      "loss": 0.2936,
+      "step": 10111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2854087083045673,
+      "learning_rate": 8.825490474946711e-06,
+      "loss": 0.2968,
+      "step": 10112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.830401918149365,
+      "learning_rate": 8.825236755155995e-06,
+      "loss": 0.646,
+      "step": 10113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1510386239893147,
+      "learning_rate": 8.824983011611596e-06,
+      "loss": 0.4418,
+      "step": 10114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.514298022801686,
+      "learning_rate": 8.82472924431509e-06,
+      "loss": 0.1435,
+      "step": 10115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7082252778405769,
+      "learning_rate": 8.824475453268052e-06,
+      "loss": 0.3005,
+      "step": 10116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.674200100378189,
+      "learning_rate": 8.824221638472057e-06,
+      "loss": 0.4748,
+      "step": 10117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9153342673579856,
+      "learning_rate": 8.823967799928683e-06,
+      "loss": 0.4422,
+      "step": 10118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5469711928851047,
+      "learning_rate": 8.823713937639508e-06,
+      "loss": 0.3607,
+      "step": 10119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7732274334383336,
+      "learning_rate": 8.823460051606106e-06,
+      "loss": 0.369,
+      "step": 10120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2281655478396458,
+      "learning_rate": 8.823206141830055e-06,
+      "loss": 0.2061,
+      "step": 10121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.15303868119352,
+      "learning_rate": 8.822952208312929e-06,
+      "loss": 0.382,
+      "step": 10122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.215027372092131,
+      "learning_rate": 8.822698251056307e-06,
+      "loss": 0.2406,
+      "step": 10123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4945461146826364,
+      "learning_rate": 8.822444270061765e-06,
+      "loss": 0.34,
+      "step": 10124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.161067794463007,
+      "learning_rate": 8.822190265330883e-06,
+      "loss": 0.4902,
+      "step": 10125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8311284246220627,
+      "learning_rate": 8.821936236865235e-06,
+      "loss": 0.6067,
+      "step": 10126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5503510327616878,
+      "learning_rate": 8.821682184666398e-06,
+      "loss": 0.2708,
+      "step": 10127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5395901109660675,
+      "learning_rate": 8.821428108735952e-06,
+      "loss": 0.3099,
+      "step": 10128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6194165550319675,
+      "learning_rate": 8.821174009075475e-06,
+      "loss": 0.2645,
+      "step": 10129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7751568366493724,
+      "learning_rate": 8.820919885686543e-06,
+      "loss": 0.2977,
+      "step": 10130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9259099607805918,
+      "learning_rate": 8.820665738570733e-06,
+      "loss": 0.3893,
+      "step": 10131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2800921116036574,
+      "learning_rate": 8.820411567729628e-06,
+      "loss": 0.3095,
+      "step": 10132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4099498832893906,
+      "learning_rate": 8.820157373164801e-06,
+      "loss": 0.1639,
+      "step": 10133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3384125380689986,
+      "learning_rate": 8.819903154877833e-06,
+      "loss": 0.3628,
+      "step": 10134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4919016124819224,
+      "learning_rate": 8.819648912870305e-06,
+      "loss": 0.2978,
+      "step": 10135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5225532080465953,
+      "learning_rate": 8.819394647143791e-06,
+      "loss": 0.2037,
+      "step": 10136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8378103402872146,
+      "learning_rate": 8.819140357699872e-06,
+      "loss": 0.5674,
+      "step": 10137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2468970961871582,
+      "learning_rate": 8.818886044540128e-06,
+      "loss": 0.1602,
+      "step": 10138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.590400699003226,
+      "learning_rate": 8.818631707666136e-06,
+      "loss": 0.3195,
+      "step": 10139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8370581960230874,
+      "learning_rate": 8.818377347079477e-06,
+      "loss": 0.4186,
+      "step": 10140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7090715909228889,
+      "learning_rate": 8.81812296278173e-06,
+      "loss": 0.2596,
+      "step": 10141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.911654849868837,
+      "learning_rate": 8.817868554774476e-06,
+      "loss": 0.4917,
+      "step": 10142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4012018165973195,
+      "learning_rate": 8.817614123059295e-06,
+      "loss": 0.1791,
+      "step": 10143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.170716407203918,
+      "learning_rate": 8.817359667637763e-06,
+      "loss": 0.3254,
+      "step": 10144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0787608424351278,
+      "learning_rate": 8.817105188511463e-06,
+      "loss": 0.1197,
+      "step": 10145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.337839534925657,
+      "learning_rate": 8.816850685681977e-06,
+      "loss": 0.3176,
+      "step": 10146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7617800209234236,
+      "learning_rate": 8.816596159150883e-06,
+      "loss": 0.2039,
+      "step": 10147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9613549445468126,
+      "learning_rate": 8.81634160891976e-06,
+      "loss": 0.5044,
+      "step": 10148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4428284433841887,
+      "learning_rate": 8.816087034990192e-06,
+      "loss": 0.3585,
+      "step": 10149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8707380334954926,
+      "learning_rate": 8.815832437363758e-06,
+      "loss": 0.3834,
+      "step": 10150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0165422223022027,
+      "learning_rate": 8.81557781604204e-06,
+      "loss": 0.6678,
+      "step": 10151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.435802151565943,
+      "learning_rate": 8.81532317102662e-06,
+      "loss": 0.2394,
+      "step": 10152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6356397007392525,
+      "learning_rate": 8.815068502319075e-06,
+      "loss": 0.2067,
+      "step": 10153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0632719415976855,
+      "learning_rate": 8.81481380992099e-06,
+      "loss": 0.663,
+      "step": 10154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7585989686794454,
+      "learning_rate": 8.814559093833947e-06,
+      "loss": 0.2047,
+      "step": 10155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.673494891154045,
+      "learning_rate": 8.814304354059527e-06,
+      "loss": 0.3627,
+      "step": 10156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9477055166587764,
+      "learning_rate": 8.81404959059931e-06,
+      "loss": 0.5626,
+      "step": 10157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8884046423695515,
+      "learning_rate": 8.813794803454878e-06,
+      "loss": 0.2242,
+      "step": 10158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1008451117704958,
+      "learning_rate": 8.813539992627817e-06,
+      "loss": 0.2078,
+      "step": 10159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1354818693594435,
+      "learning_rate": 8.813285158119706e-06,
+      "loss": 0.1577,
+      "step": 10160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.534749093552878,
+      "learning_rate": 8.813030299932128e-06,
+      "loss": 0.182,
+      "step": 10161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6359411192674507,
+      "learning_rate": 8.812775418066668e-06,
+      "loss": 0.2786,
+      "step": 10162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.675758752916398,
+      "learning_rate": 8.812520512524905e-06,
+      "loss": 0.312,
+      "step": 10163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8021246849306387,
+      "learning_rate": 8.812265583308424e-06,
+      "loss": 0.3268,
+      "step": 10164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4233253596044901,
+      "learning_rate": 8.812010630418808e-06,
+      "loss": 0.2249,
+      "step": 10165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8860236382111746,
+      "learning_rate": 8.811755653857639e-06,
+      "loss": 0.3395,
+      "step": 10166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9527286356234772,
+      "learning_rate": 8.811500653626501e-06,
+      "loss": 0.5608,
+      "step": 10167
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.078687971626778,
+      "learning_rate": 8.81124562972698e-06,
+      "loss": 0.682,
+      "step": 10168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.899385207871487,
+      "learning_rate": 8.810990582160654e-06,
+      "loss": 0.5395,
+      "step": 10169
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.543113083266573,
+      "learning_rate": 8.810735510929111e-06,
+      "loss": 0.3803,
+      "step": 10170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.312242135929647,
+      "learning_rate": 8.810480416033935e-06,
+      "loss": 0.5438,
+      "step": 10171
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0290919476702736,
+      "learning_rate": 8.810225297476708e-06,
+      "loss": 0.199,
+      "step": 10172
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.447141895378305,
+      "learning_rate": 8.809970155259018e-06,
+      "loss": 0.2611,
+      "step": 10173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7224763162110004,
+      "learning_rate": 8.809714989382443e-06,
+      "loss": 0.2628,
+      "step": 10174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.737845995175831,
+      "learning_rate": 8.809459799848574e-06,
+      "loss": 0.097,
+      "step": 10175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.887100402267435,
+      "learning_rate": 8.80920458665899e-06,
+      "loss": 0.3136,
+      "step": 10176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1745384839050705,
+      "learning_rate": 8.808949349815282e-06,
+      "loss": 0.2114,
+      "step": 10177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5368424781284125,
+      "learning_rate": 8.808694089319029e-06,
+      "loss": 0.4047,
+      "step": 10178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4125644602140213,
+      "learning_rate": 8.80843880517182e-06,
+      "loss": 0.1282,
+      "step": 10179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9284460441147868,
+      "learning_rate": 8.808183497375236e-06,
+      "loss": 0.4041,
+      "step": 10180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5894259857613287,
+      "learning_rate": 8.807928165930868e-06,
+      "loss": 0.4087,
+      "step": 10181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6414098451802872,
+      "learning_rate": 8.807672810840299e-06,
+      "loss": 0.2812,
+      "step": 10182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8453044925099285,
+      "learning_rate": 8.807417432105113e-06,
+      "loss": 0.5261,
+      "step": 10183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7267220324177102,
+      "learning_rate": 8.8071620297269e-06,
+      "loss": 0.4622,
+      "step": 10184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.580023287185797,
+      "learning_rate": 8.806906603707241e-06,
+      "loss": 0.227,
+      "step": 10185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2743965103869987,
+      "learning_rate": 8.806651154047725e-06,
+      "loss": 0.163,
+      "step": 10186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.925742793134456,
+      "learning_rate": 8.806395680749938e-06,
+      "loss": 0.3729,
+      "step": 10187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4842862152852472,
+      "learning_rate": 8.806140183815465e-06,
+      "loss": 0.3292,
+      "step": 10188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0518128625092698,
+      "learning_rate": 8.805884663245894e-06,
+      "loss": 0.3848,
+      "step": 10189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6431385630936204,
+      "learning_rate": 8.805629119042812e-06,
+      "loss": 0.3411,
+      "step": 10190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5363332781555004,
+      "learning_rate": 8.805373551207805e-06,
+      "loss": 0.5725,
+      "step": 10191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4032808695156465,
+      "learning_rate": 8.805117959742461e-06,
+      "loss": 0.1713,
+      "step": 10192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7066058909941013,
+      "learning_rate": 8.804862344648366e-06,
+      "loss": 0.452,
+      "step": 10193
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5971616023562742,
+      "learning_rate": 8.804606705927108e-06,
+      "loss": 0.2152,
+      "step": 10194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.037104614852069,
+      "learning_rate": 8.804351043580273e-06,
+      "loss": 0.3587,
+      "step": 10195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9741934612784504,
+      "learning_rate": 8.804095357609452e-06,
+      "loss": 0.4488,
+      "step": 10196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.522019581390494,
+      "learning_rate": 8.80383964801623e-06,
+      "loss": 0.5061,
+      "step": 10197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6092165711578248,
+      "learning_rate": 8.803583914802193e-06,
+      "loss": 0.2359,
+      "step": 10198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1078548958758514,
+      "learning_rate": 8.803328157968933e-06,
+      "loss": 0.5364,
+      "step": 10199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8808254504288606,
+      "learning_rate": 8.803072377518038e-06,
+      "loss": 0.3331,
+      "step": 10200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3868119987512846,
+      "learning_rate": 8.802816573451094e-06,
+      "loss": 0.1473,
+      "step": 10201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.314630143657889,
+      "learning_rate": 8.802560745769689e-06,
+      "loss": 0.1727,
+      "step": 10202
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1384176719723127,
+      "learning_rate": 8.802304894475415e-06,
+      "loss": 0.2081,
+      "step": 10203
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6022796327883617,
+      "learning_rate": 8.802049019569858e-06,
+      "loss": 0.1659,
+      "step": 10204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9640533609412825,
+      "learning_rate": 8.801793121054607e-06,
+      "loss": 0.2934,
+      "step": 10205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.008795145236066,
+      "learning_rate": 8.801537198931254e-06,
+      "loss": 0.5741,
+      "step": 10206
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.545472491846712,
+      "learning_rate": 8.801281253201383e-06,
+      "loss": 0.2672,
+      "step": 10207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4674588823032884,
+      "learning_rate": 8.801025283866588e-06,
+      "loss": 0.8325,
+      "step": 10208
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7830236664266828,
+      "learning_rate": 8.800769290928457e-06,
+      "loss": 0.2176,
+      "step": 10209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7850504787504156,
+      "learning_rate": 8.80051327438858e-06,
+      "loss": 0.6442,
+      "step": 10210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6704223329066241,
+      "learning_rate": 8.800257234248546e-06,
+      "loss": 0.4048,
+      "step": 10211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.054330143837759,
+      "learning_rate": 8.800001170509945e-06,
+      "loss": 0.2366,
+      "step": 10212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.724857971228847,
+      "learning_rate": 8.799745083174368e-06,
+      "loss": 0.3098,
+      "step": 10213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7279568733533626,
+      "learning_rate": 8.799488972243402e-06,
+      "loss": 0.3548,
+      "step": 10214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5574955429362392,
+      "learning_rate": 8.799232837718642e-06,
+      "loss": 0.3377,
+      "step": 10215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4785473617312161,
+      "learning_rate": 8.798976679601677e-06,
+      "loss": 0.2949,
+      "step": 10216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5370341116523842,
+      "learning_rate": 8.798720497894095e-06,
+      "loss": 0.2473,
+      "step": 10217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7318941364036429,
+      "learning_rate": 8.798464292597491e-06,
+      "loss": 0.1662,
+      "step": 10218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6647141796872187,
+      "learning_rate": 8.798208063713452e-06,
+      "loss": 0.183,
+      "step": 10219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6876081789628041,
+      "learning_rate": 8.797951811243573e-06,
+      "loss": 0.3289,
+      "step": 10220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3858015089396163,
+      "learning_rate": 8.797695535189442e-06,
+      "loss": 0.2256,
+      "step": 10221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.148145766825181,
+      "learning_rate": 8.797439235552653e-06,
+      "loss": 0.4838,
+      "step": 10222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5989196231103475,
+      "learning_rate": 8.797182912334795e-06,
+      "loss": 0.3192,
+      "step": 10223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3519899163617937,
+      "learning_rate": 8.796926565537461e-06,
+      "loss": 0.6026,
+      "step": 10224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6692057630102823,
+      "learning_rate": 8.796670195162244e-06,
+      "loss": 0.342,
+      "step": 10225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8425462559492851,
+      "learning_rate": 8.796413801210733e-06,
+      "loss": 0.5393,
+      "step": 10226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.382010950928142,
+      "learning_rate": 8.796157383684522e-06,
+      "loss": 0.2025,
+      "step": 10227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.259586206123823,
+      "learning_rate": 8.795900942585204e-06,
+      "loss": 0.2193,
+      "step": 10228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2040113870937543,
+      "learning_rate": 8.795644477914372e-06,
+      "loss": 0.5068,
+      "step": 10229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6738465930182518,
+      "learning_rate": 8.795387989673616e-06,
+      "loss": 0.3526,
+      "step": 10230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.643789060476663,
+      "learning_rate": 8.79513147786453e-06,
+      "loss": 0.4828,
+      "step": 10231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3501254446194748,
+      "learning_rate": 8.794874942488705e-06,
+      "loss": 0.203,
+      "step": 10232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3602932919566464,
+      "learning_rate": 8.794618383547738e-06,
+      "loss": 0.2224,
+      "step": 10233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4473862436870726,
+      "learning_rate": 8.794361801043219e-06,
+      "loss": 0.2134,
+      "step": 10234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7409427412738556,
+      "learning_rate": 8.794105194976744e-06,
+      "loss": 0.5403,
+      "step": 10235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8129248039050125,
+      "learning_rate": 8.793848565349903e-06,
+      "loss": 0.338,
+      "step": 10236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.434907166031834,
+      "learning_rate": 8.793591912164292e-06,
+      "loss": 0.2192,
+      "step": 10237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7235502336728667,
+      "learning_rate": 8.793335235421503e-06,
+      "loss": 0.2664,
+      "step": 10238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6761988717445182,
+      "learning_rate": 8.793078535123133e-06,
+      "loss": 0.2028,
+      "step": 10239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.321889198729586,
+      "learning_rate": 8.792821811270771e-06,
+      "loss": 0.2417,
+      "step": 10240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1259806137501953,
+      "learning_rate": 8.792565063866017e-06,
+      "loss": 0.2713,
+      "step": 10241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6164696016798528,
+      "learning_rate": 8.79230829291046e-06,
+      "loss": 0.3428,
+      "step": 10242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.004794037530866,
+      "learning_rate": 8.7920514984057e-06,
+      "loss": 0.2376,
+      "step": 10243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.806619130599543,
+      "learning_rate": 8.791794680353328e-06,
+      "loss": 0.3818,
+      "step": 10244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6977627007665608,
+      "learning_rate": 8.791537838754938e-06,
+      "loss": 0.4453,
+      "step": 10245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8980753517880202,
+      "learning_rate": 8.791280973612126e-06,
+      "loss": 0.3022,
+      "step": 10246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9946383790032867,
+      "learning_rate": 8.791024084926488e-06,
+      "loss": 0.2047,
+      "step": 10247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7725953778860541,
+      "learning_rate": 8.790767172699618e-06,
+      "loss": 0.3702,
+      "step": 10248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0173411013142855,
+      "learning_rate": 8.790510236933114e-06,
+      "loss": 0.5182,
+      "step": 10249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.23882244883722,
+      "learning_rate": 8.790253277628568e-06,
+      "loss": 0.6157,
+      "step": 10250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7993711093159597,
+      "learning_rate": 8.789996294787577e-06,
+      "loss": 0.3825,
+      "step": 10251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6281327885809376,
+      "learning_rate": 8.789739288411739e-06,
+      "loss": 0.3105,
+      "step": 10252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4567121611440723,
+      "learning_rate": 8.789482258502645e-06,
+      "loss": 0.4409,
+      "step": 10253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5794259931199435,
+      "learning_rate": 8.789225205061895e-06,
+      "loss": 0.2596,
+      "step": 10254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4496314625559403,
+      "learning_rate": 8.788968128091084e-06,
+      "loss": 0.2144,
+      "step": 10255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6588983729494262,
+      "learning_rate": 8.788711027591808e-06,
+      "loss": 0.3152,
+      "step": 10256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6212861647796564,
+      "learning_rate": 8.788453903565666e-06,
+      "loss": 0.2948,
+      "step": 10257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4824719225812284,
+      "learning_rate": 8.788196756014252e-06,
+      "loss": 0.3391,
+      "step": 10258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8525213583591429,
+      "learning_rate": 8.787939584939162e-06,
+      "loss": 0.3292,
+      "step": 10259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4557308077305353,
+      "learning_rate": 8.787682390341994e-06,
+      "loss": 0.3111,
+      "step": 10260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.565796007879189,
+      "learning_rate": 8.787425172224348e-06,
+      "loss": 0.3752,
+      "step": 10261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2365284677564958,
+      "learning_rate": 8.787167930587817e-06,
+      "loss": 0.3422,
+      "step": 10262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5143591520304382,
+      "learning_rate": 8.786910665434002e-06,
+      "loss": 0.3018,
+      "step": 10263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6405021939409763,
+      "learning_rate": 8.786653376764498e-06,
+      "loss": 0.3254,
+      "step": 10264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5914594866088483,
+      "learning_rate": 8.786396064580904e-06,
+      "loss": 0.3242,
+      "step": 10265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6211574105321602,
+      "learning_rate": 8.786138728884817e-06,
+      "loss": 0.4352,
+      "step": 10266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.942272276151991,
+      "learning_rate": 8.785881369677834e-06,
+      "loss": 0.3092,
+      "step": 10267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.322423305564288,
+      "learning_rate": 8.785623986961556e-06,
+      "loss": 0.2424,
+      "step": 10268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7693959114916789,
+      "learning_rate": 8.785366580737579e-06,
+      "loss": 0.2257,
+      "step": 10269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8745365702598185,
+      "learning_rate": 8.785109151007502e-06,
+      "loss": 0.365,
+      "step": 10270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5281148217843246,
+      "learning_rate": 8.784851697772925e-06,
+      "loss": 0.3301,
+      "step": 10271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9346622404761107,
+      "learning_rate": 8.784594221035446e-06,
+      "loss": 0.5513,
+      "step": 10272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.398468914957794,
+      "learning_rate": 8.784336720796661e-06,
+      "loss": 0.328,
+      "step": 10273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5154796658472942,
+      "learning_rate": 8.784079197058172e-06,
+      "loss": 0.2392,
+      "step": 10274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4380537345251363,
+      "learning_rate": 8.783821649821578e-06,
+      "loss": 0.1543,
+      "step": 10275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9764591696431559,
+      "learning_rate": 8.783564079088478e-06,
+      "loss": 0.2252,
+      "step": 10276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.869059661578619,
+      "learning_rate": 8.783306484860469e-06,
+      "loss": 0.5156,
+      "step": 10277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8337144256916824,
+      "learning_rate": 8.783048867139155e-06,
+      "loss": 0.2018,
+      "step": 10278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5544919796783083,
+      "learning_rate": 8.782791225926132e-06,
+      "loss": 0.3746,
+      "step": 10279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9940102821367656,
+      "learning_rate": 8.782533561223004e-06,
+      "loss": 0.6839,
+      "step": 10280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7956190695765946,
+      "learning_rate": 8.782275873031365e-06,
+      "loss": 0.3473,
+      "step": 10281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2727875321159057,
+      "learning_rate": 8.782018161352821e-06,
+      "loss": 0.7116,
+      "step": 10282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8974950162033049,
+      "learning_rate": 8.78176042618897e-06,
+      "loss": 0.4286,
+      "step": 10283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7595947428220562,
+      "learning_rate": 8.781502667541413e-06,
+      "loss": 0.3861,
+      "step": 10284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4294323823867332,
+      "learning_rate": 8.781244885411748e-06,
+      "loss": 0.3875,
+      "step": 10285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.378537147550693,
+      "learning_rate": 8.780987079801578e-06,
+      "loss": 0.2693,
+      "step": 10286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7932936180853554,
+      "learning_rate": 8.780729250712506e-06,
+      "loss": 0.4744,
+      "step": 10287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3048403501917414,
+      "learning_rate": 8.780471398146128e-06,
+      "loss": 0.2261,
+      "step": 10288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6764192055990825,
+      "learning_rate": 8.780213522104048e-06,
+      "loss": 0.5284,
+      "step": 10289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8327093109975867,
+      "learning_rate": 8.779955622587869e-06,
+      "loss": 0.4394,
+      "step": 10290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8443588205323973,
+      "learning_rate": 8.77969769959919e-06,
+      "loss": 0.1786,
+      "step": 10291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4833799835674508,
+      "learning_rate": 8.779439753139612e-06,
+      "loss": 0.1911,
+      "step": 10292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9780072541018279,
+      "learning_rate": 8.77918178321074e-06,
+      "loss": 0.1756,
+      "step": 10293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3490598639152527,
+      "learning_rate": 8.778923789814174e-06,
+      "loss": 0.3484,
+      "step": 10294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9162913013868687,
+      "learning_rate": 8.778665772951516e-06,
+      "loss": 0.1627,
+      "step": 10295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2223681332841687,
+      "learning_rate": 8.778407732624368e-06,
+      "loss": 0.2716,
+      "step": 10296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4546295140117622,
+      "learning_rate": 8.778149668834333e-06,
+      "loss": 0.3743,
+      "step": 10297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5801343707434556,
+      "learning_rate": 8.777891581583011e-06,
+      "loss": 0.2727,
+      "step": 10298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2268670124828245,
+      "learning_rate": 8.77763347087201e-06,
+      "loss": 0.2264,
+      "step": 10299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5732425047994303,
+      "learning_rate": 8.777375336702928e-06,
+      "loss": 0.3528,
+      "step": 10300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2974491342228467,
+      "learning_rate": 8.77711717907737e-06,
+      "loss": 0.2001,
+      "step": 10301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.947625434025757,
+      "learning_rate": 8.776858997996938e-06,
+      "loss": 0.2493,
+      "step": 10302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1777377025753353,
+      "learning_rate": 8.776600793463236e-06,
+      "loss": 0.318,
+      "step": 10303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.740976757534079,
+      "learning_rate": 8.77634256547787e-06,
+      "loss": 0.3237,
+      "step": 10304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1380196792486117,
+      "learning_rate": 8.776084314042439e-06,
+      "loss": 0.1569,
+      "step": 10305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8775616313403687,
+      "learning_rate": 8.775826039158548e-06,
+      "loss": 0.3439,
+      "step": 10306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.579973302224046,
+      "learning_rate": 8.775567740827802e-06,
+      "loss": 0.1544,
+      "step": 10307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.554216490855406,
+      "learning_rate": 8.775309419051804e-06,
+      "loss": 0.3939,
+      "step": 10308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6288821358119343,
+      "learning_rate": 8.775051073832158e-06,
+      "loss": 0.35,
+      "step": 10309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5512081910980544,
+      "learning_rate": 8.774792705170469e-06,
+      "loss": 0.189,
+      "step": 10310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9247978129039143,
+      "learning_rate": 8.774534313068341e-06,
+      "loss": 0.5264,
+      "step": 10311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8649463930556223,
+      "learning_rate": 8.77427589752738e-06,
+      "loss": 0.3029,
+      "step": 10312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.203039119801241,
+      "learning_rate": 8.774017458549186e-06,
+      "loss": 0.2211,
+      "step": 10313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4294639110309406,
+      "learning_rate": 8.773758996135372e-06,
+      "loss": 0.5116,
+      "step": 10314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4112218518528947,
+      "learning_rate": 8.773500510287533e-06,
+      "loss": 0.2386,
+      "step": 10315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6234306267380816,
+      "learning_rate": 8.773242001007282e-06,
+      "loss": 0.2451,
+      "step": 10316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6871434095464586,
+      "learning_rate": 8.772983468296221e-06,
+      "loss": 0.5145,
+      "step": 10317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9044836725414915,
+      "learning_rate": 8.772724912155956e-06,
+      "loss": 0.3433,
+      "step": 10318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2310471560371745,
+      "learning_rate": 8.772466332588091e-06,
+      "loss": 0.2237,
+      "step": 10319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6252743553595261,
+      "learning_rate": 8.772207729594236e-06,
+      "loss": 0.3522,
+      "step": 10320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.70786615899922,
+      "learning_rate": 8.771949103175991e-06,
+      "loss": 0.3878,
+      "step": 10321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.887654652875476,
+      "learning_rate": 8.77169045333497e-06,
+      "loss": 0.3117,
+      "step": 10322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.810057342421843,
+      "learning_rate": 8.771431780072769e-06,
+      "loss": 0.4376,
+      "step": 10323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9549919056054086,
+      "learning_rate": 8.771173083391001e-06,
+      "loss": 0.3538,
+      "step": 10324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4855182110300287,
+      "learning_rate": 8.770914363291272e-06,
+      "loss": 0.3436,
+      "step": 10325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.985976072637617,
+      "learning_rate": 8.770655619775187e-06,
+      "loss": 0.2459,
+      "step": 10326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9633711117672974,
+      "learning_rate": 8.770396852844355e-06,
+      "loss": 0.5491,
+      "step": 10327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.332943992344586,
+      "learning_rate": 8.770138062500377e-06,
+      "loss": 0.2817,
+      "step": 10328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5771603374209417,
+      "learning_rate": 8.769879248744867e-06,
+      "loss": 0.1724,
+      "step": 10329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.396183896583201,
+      "learning_rate": 8.769620411579428e-06,
+      "loss": 0.3086,
+      "step": 10330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4020236029757032,
+      "learning_rate": 8.769361551005669e-06,
+      "loss": 0.2849,
+      "step": 10331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6261208879723212,
+      "learning_rate": 8.769102667025198e-06,
+      "loss": 0.3285,
+      "step": 10332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.26621870550138,
+      "learning_rate": 8.76884375963962e-06,
+      "loss": 0.1997,
+      "step": 10333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7265714066910047,
+      "learning_rate": 8.768584828850545e-06,
+      "loss": 0.4237,
+      "step": 10334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1409980769030734,
+      "learning_rate": 8.768325874659581e-06,
+      "loss": 0.537,
+      "step": 10335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8854298345271063,
+      "learning_rate": 8.768066897068334e-06,
+      "loss": 0.3094,
+      "step": 10336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.516652791078555,
+      "learning_rate": 8.767807896078415e-06,
+      "loss": 0.2913,
+      "step": 10337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9325391687159488,
+      "learning_rate": 8.767548871691432e-06,
+      "loss": 0.2363,
+      "step": 10338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7583475930323476,
+      "learning_rate": 8.76728982390899e-06,
+      "loss": 0.2599,
+      "step": 10339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8161724904088408,
+      "learning_rate": 8.767030752732701e-06,
+      "loss": 0.3727,
+      "step": 10340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.312487025482278,
+      "learning_rate": 8.766771658164171e-06,
+      "loss": 0.3726,
+      "step": 10341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.141039390550382,
+      "learning_rate": 8.766512540205012e-06,
+      "loss": 0.878,
+      "step": 10342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.948232909828219,
+      "learning_rate": 8.766253398856831e-06,
+      "loss": 0.3183,
+      "step": 10343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0320797509283146,
+      "learning_rate": 8.765994234121237e-06,
+      "loss": 0.5542,
+      "step": 10344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.441495365396367,
+      "learning_rate": 8.765735045999843e-06,
+      "loss": 0.1508,
+      "step": 10345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5475165970390663,
+      "learning_rate": 8.765475834494253e-06,
+      "loss": 0.2018,
+      "step": 10346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.435250175073065,
+      "learning_rate": 8.765216599606081e-06,
+      "loss": 0.2021,
+      "step": 10347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8057610675013656,
+      "learning_rate": 8.764957341336935e-06,
+      "loss": 0.5175,
+      "step": 10348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6595973902995396,
+      "learning_rate": 8.764698059688423e-06,
+      "loss": 0.3695,
+      "step": 10349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5647255193665566,
+      "learning_rate": 8.764438754662159e-06,
+      "loss": 0.1942,
+      "step": 10350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8400072077812122,
+      "learning_rate": 8.764179426259752e-06,
+      "loss": 0.4216,
+      "step": 10351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2949313556099693,
+      "learning_rate": 8.76392007448281e-06,
+      "loss": 0.2833,
+      "step": 10352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8420535300141294,
+      "learning_rate": 8.763660699332945e-06,
+      "loss": 0.3925,
+      "step": 10353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7929449557718067,
+      "learning_rate": 8.763401300811768e-06,
+      "loss": 0.295,
+      "step": 10354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6854832505719604,
+      "learning_rate": 8.76314187892089e-06,
+      "loss": 0.2644,
+      "step": 10355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6552004187578555,
+      "learning_rate": 8.76288243366192e-06,
+      "loss": 0.3978,
+      "step": 10356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7839534555540972,
+      "learning_rate": 8.762622965036473e-06,
+      "loss": 0.571,
+      "step": 10357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.731360971856939,
+      "learning_rate": 8.762363473046157e-06,
+      "loss": 0.5709,
+      "step": 10358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3230849929287678,
+      "learning_rate": 8.762103957692584e-06,
+      "loss": 0.1771,
+      "step": 10359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4214426578061994,
+      "learning_rate": 8.761844418977366e-06,
+      "loss": 0.2648,
+      "step": 10360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7394986218872297,
+      "learning_rate": 8.761584856902115e-06,
+      "loss": 0.3441,
+      "step": 10361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.899954135083958,
+      "learning_rate": 8.761325271468442e-06,
+      "loss": 0.2344,
+      "step": 10362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4765388951105165,
+      "learning_rate": 8.761065662677958e-06,
+      "loss": 0.2182,
+      "step": 10363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4270938683791554,
+      "learning_rate": 8.760806030532276e-06,
+      "loss": 0.3197,
+      "step": 10364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.528933534295896,
+      "learning_rate": 8.76054637503301e-06,
+      "loss": 0.2012,
+      "step": 10365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3807219716170849,
+      "learning_rate": 8.76028669618177e-06,
+      "loss": 0.2694,
+      "step": 10366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.178181827233488,
+      "learning_rate": 8.76002699398017e-06,
+      "loss": 0.454,
+      "step": 10367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9248861366275938,
+      "learning_rate": 8.759767268429822e-06,
+      "loss": 0.4061,
+      "step": 10368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1667014791293147,
+      "learning_rate": 8.75950751953234e-06,
+      "loss": 0.4853,
+      "step": 10369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.375172812807793,
+      "learning_rate": 8.759247747289333e-06,
+      "loss": 0.2716,
+      "step": 10370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3887810825108657,
+      "learning_rate": 8.75898795170242e-06,
+      "loss": 0.2199,
+      "step": 10371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6081530158496906,
+      "learning_rate": 8.758728132773209e-06,
+      "loss": 0.4918,
+      "step": 10372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.355484855182403,
+      "learning_rate": 8.758468290503317e-06,
+      "loss": 0.3717,
+      "step": 10373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6949957714337516,
+      "learning_rate": 8.758208424894356e-06,
+      "loss": 0.609,
+      "step": 10374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.368019180740783,
+      "learning_rate": 8.75794853594794e-06,
+      "loss": 0.209,
+      "step": 10375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.514528509590317,
+      "learning_rate": 8.757688623665683e-06,
+      "loss": 0.2159,
+      "step": 10376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1256488763039716,
+      "learning_rate": 8.757428688049198e-06,
+      "loss": 0.2499,
+      "step": 10377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7767167001802493,
+      "learning_rate": 8.7571687291001e-06,
+      "loss": 0.457,
+      "step": 10378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2827778115734187,
+      "learning_rate": 8.756908746820004e-06,
+      "loss": 0.1348,
+      "step": 10379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7431251735249484,
+      "learning_rate": 8.756648741210522e-06,
+      "loss": 0.283,
+      "step": 10380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7052207965741026,
+      "learning_rate": 8.756388712273272e-06,
+      "loss": 0.196,
+      "step": 10381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8676392316046853,
+      "learning_rate": 8.756128660009865e-06,
+      "loss": 0.6682,
+      "step": 10382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1831094597699165,
+      "learning_rate": 8.75586858442192e-06,
+      "loss": 0.2358,
+      "step": 10383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.820054123923229,
+      "learning_rate": 8.755608485511049e-06,
+      "loss": 0.3757,
+      "step": 10384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.973654479753511,
+      "learning_rate": 8.755348363278867e-06,
+      "loss": 0.2002,
+      "step": 10385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2072375918454488,
+      "learning_rate": 8.75508821772699e-06,
+      "loss": 0.7904,
+      "step": 10386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8850026589407747,
+      "learning_rate": 8.754828048857034e-06,
+      "loss": 0.4548,
+      "step": 10387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.031719503794953,
+      "learning_rate": 8.754567856670616e-06,
+      "loss": 0.3742,
+      "step": 10388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.779824688423995,
+      "learning_rate": 8.754307641169348e-06,
+      "loss": 0.3142,
+      "step": 10389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7768092515034088,
+      "learning_rate": 8.754047402354848e-06,
+      "loss": 0.2001,
+      "step": 10390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.180290167698007,
+      "learning_rate": 8.75378714022873e-06,
+      "loss": 0.2113,
+      "step": 10391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7059885764455078,
+      "learning_rate": 8.753526854792616e-06,
+      "loss": 0.2319,
+      "step": 10392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6906742511845139,
+      "learning_rate": 8.753266546048114e-06,
+      "loss": 0.4821,
+      "step": 10393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.04874704380532,
+      "learning_rate": 8.753006213996847e-06,
+      "loss": 0.2417,
+      "step": 10394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6120696286366387,
+      "learning_rate": 8.752745858640427e-06,
+      "loss": 0.3651,
+      "step": 10395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8388284019719274,
+      "learning_rate": 8.752485479980476e-06,
+      "loss": 0.3095,
+      "step": 10396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6331251292296167,
+      "learning_rate": 8.752225078018606e-06,
+      "loss": 0.1995,
+      "step": 10397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.031859717192799,
+      "learning_rate": 8.751964652756437e-06,
+      "loss": 0.3003,
+      "step": 10398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0158349822325294,
+      "learning_rate": 8.751704204195585e-06,
+      "loss": 0.1056,
+      "step": 10399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.990957680873712,
+      "learning_rate": 8.751443732337668e-06,
+      "loss": 0.6623,
+      "step": 10400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8004626366682746,
+      "learning_rate": 8.751183237184301e-06,
+      "loss": 0.4592,
+      "step": 10401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.282187692901288,
+      "learning_rate": 8.750922718737105e-06,
+      "loss": 0.4174,
+      "step": 10402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1360594815079605,
+      "learning_rate": 8.750662176997695e-06,
+      "loss": 0.2438,
+      "step": 10403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.342270648804354,
+      "learning_rate": 8.75040161196769e-06,
+      "loss": 0.2537,
+      "step": 10404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5390643686196075,
+      "learning_rate": 8.75014102364871e-06,
+      "loss": 0.3232,
+      "step": 10405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8684088571014714,
+      "learning_rate": 8.749880412042368e-06,
+      "loss": 0.5314,
+      "step": 10406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8346949694687242,
+      "learning_rate": 8.749619777150288e-06,
+      "loss": 0.5487,
+      "step": 10407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.240158369333591,
+      "learning_rate": 8.749359118974086e-06,
+      "loss": 0.2297,
+      "step": 10408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1513909344221855,
+      "learning_rate": 8.74909843751538e-06,
+      "loss": 0.3735,
+      "step": 10409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5103152854920583,
+      "learning_rate": 8.74883773277579e-06,
+      "loss": 0.3923,
+      "step": 10410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6499830656916588,
+      "learning_rate": 8.748577004756935e-06,
+      "loss": 0.4478,
+      "step": 10411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.461531739923523,
+      "learning_rate": 8.748316253460435e-06,
+      "loss": 0.4295,
+      "step": 10412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4243679985403246,
+      "learning_rate": 8.748055478887905e-06,
+      "loss": 0.3532,
+      "step": 10413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9982681450207906,
+      "learning_rate": 8.747794681040966e-06,
+      "loss": 0.3806,
+      "step": 10414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7818540063535857,
+      "learning_rate": 8.74753385992124e-06,
+      "loss": 0.3901,
+      "step": 10415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6563681974161208,
+      "learning_rate": 8.747273015530345e-06,
+      "loss": 0.4332,
+      "step": 10416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8509364216440858,
+      "learning_rate": 8.747012147869902e-06,
+      "loss": 0.4031,
+      "step": 10417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8636951746510584,
+      "learning_rate": 8.746751256941529e-06,
+      "loss": 0.2953,
+      "step": 10418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.502299235413961,
+      "learning_rate": 8.746490342746845e-06,
+      "loss": 0.2222,
+      "step": 10419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8538325514132272,
+      "learning_rate": 8.746229405287475e-06,
+      "loss": 0.3887,
+      "step": 10420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.123443596469217,
+      "learning_rate": 8.745968444565034e-06,
+      "loss": 0.2148,
+      "step": 10421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2551604858026841,
+      "learning_rate": 8.745707460581145e-06,
+      "loss": 0.2068,
+      "step": 10422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5086741431460562,
+      "learning_rate": 8.745446453337429e-06,
+      "loss": 0.3486,
+      "step": 10423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5781354951037423,
+      "learning_rate": 8.745185422835508e-06,
+      "loss": 0.2285,
+      "step": 10424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.684158690652485,
+      "learning_rate": 8.744924369077e-06,
+      "loss": 0.4818,
+      "step": 10425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.280490219626753,
+      "learning_rate": 8.744663292063526e-06,
+      "loss": 0.5928,
+      "step": 10426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7373162194695366,
+      "learning_rate": 8.74440219179671e-06,
+      "loss": 0.3866,
+      "step": 10427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7142552493448817,
+      "learning_rate": 8.744141068278173e-06,
+      "loss": 0.2537,
+      "step": 10428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1155281314681766,
+      "learning_rate": 8.743879921509534e-06,
+      "loss": 0.1703,
+      "step": 10429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.419141093875711,
+      "learning_rate": 8.743618751492415e-06,
+      "loss": 0.5207,
+      "step": 10430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.454527005004897,
+      "learning_rate": 8.74335755822844e-06,
+      "loss": 0.3992,
+      "step": 10431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6483816526091701,
+      "learning_rate": 8.74309634171923e-06,
+      "loss": 0.4573,
+      "step": 10432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6824488143234362,
+      "learning_rate": 8.742835101966407e-06,
+      "loss": 0.2965,
+      "step": 10433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8952356189252113,
+      "learning_rate": 8.742573838971594e-06,
+      "loss": 0.5065,
+      "step": 10434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.753211529529166,
+      "learning_rate": 8.74231255273641e-06,
+      "loss": 0.4524,
+      "step": 10435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7390325615663078,
+      "learning_rate": 8.742051243262482e-06,
+      "loss": 0.3979,
+      "step": 10436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7314630089042136,
+      "learning_rate": 8.741789910551431e-06,
+      "loss": 0.4987,
+      "step": 10437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.053366624499832,
+      "learning_rate": 8.741528554604877e-06,
+      "loss": 0.4465,
+      "step": 10438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6560442004910456,
+      "learning_rate": 8.741267175424447e-06,
+      "loss": 0.3186,
+      "step": 10439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.962104789991298,
+      "learning_rate": 8.741005773011761e-06,
+      "loss": 0.3844,
+      "step": 10440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.746942629554235,
+      "learning_rate": 8.740744347368447e-06,
+      "loss": 0.2877,
+      "step": 10441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.808530522141404,
+      "learning_rate": 8.740482898496122e-06,
+      "loss": 0.231,
+      "step": 10442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7602080689704058,
+      "learning_rate": 8.740221426396413e-06,
+      "loss": 0.2426,
+      "step": 10443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6669614059386297,
+      "learning_rate": 8.739959931070946e-06,
+      "loss": 0.448,
+      "step": 10444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.234572081777297,
+      "learning_rate": 8.73969841252134e-06,
+      "loss": 0.6471,
+      "step": 10445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0162909427856284,
+      "learning_rate": 8.73943687074922e-06,
+      "loss": 0.5187,
+      "step": 10446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.040599557913684,
+      "learning_rate": 8.739175305756213e-06,
+      "loss": 0.2292,
+      "step": 10447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4791926442694847,
+      "learning_rate": 8.738913717543942e-06,
+      "loss": 0.3218,
+      "step": 10448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6962164400462698,
+      "learning_rate": 8.73865210611403e-06,
+      "loss": 0.2318,
+      "step": 10449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9387416053051598,
+      "learning_rate": 8.738390471468102e-06,
+      "loss": 0.603,
+      "step": 10450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2971921642102904,
+      "learning_rate": 8.738128813607782e-06,
+      "loss": 0.2242,
+      "step": 10451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5738728950288876,
+      "learning_rate": 8.7378671325347e-06,
+      "loss": 0.1812,
+      "step": 10452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7296748266815447,
+      "learning_rate": 8.737605428250473e-06,
+      "loss": 0.3692,
+      "step": 10453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7412756256907056,
+      "learning_rate": 8.737343700756731e-06,
+      "loss": 0.2341,
+      "step": 10454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.419152666708114,
+      "learning_rate": 8.737081950055098e-06,
+      "loss": 0.2125,
+      "step": 10455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7076383728786537,
+      "learning_rate": 8.736820176147199e-06,
+      "loss": 0.3374,
+      "step": 10456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3999750419823065,
+      "learning_rate": 8.73655837903466e-06,
+      "loss": 0.1901,
+      "step": 10457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.224801072861263,
+      "learning_rate": 8.736296558719109e-06,
+      "loss": 0.3744,
+      "step": 10458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9295013005852053,
+      "learning_rate": 8.736034715202168e-06,
+      "loss": 0.1704,
+      "step": 10459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6641198263681407,
+      "learning_rate": 8.735772848485467e-06,
+      "loss": 0.2699,
+      "step": 10460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5173257203003656,
+      "learning_rate": 8.735510958570627e-06,
+      "loss": 0.3982,
+      "step": 10461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5415103732176718,
+      "learning_rate": 8.735249045459277e-06,
+      "loss": 0.2677,
+      "step": 10462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.089956850930884,
+      "learning_rate": 8.734987109153046e-06,
+      "loss": 0.2424,
+      "step": 10463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.178934562009507,
+      "learning_rate": 8.734725149653555e-06,
+      "loss": 0.4193,
+      "step": 10464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.78227232155779,
+      "learning_rate": 8.734463166962437e-06,
+      "loss": 0.297,
+      "step": 10465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5927526588417746,
+      "learning_rate": 8.734201161081312e-06,
+      "loss": 0.2489,
+      "step": 10466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4610304436465027,
+      "learning_rate": 8.733939132011812e-06,
+      "loss": 0.3938,
+      "step": 10467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4159690402030374,
+      "learning_rate": 8.733677079755565e-06,
+      "loss": 0.3477,
+      "step": 10468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9235634229548562,
+      "learning_rate": 8.733415004314194e-06,
+      "loss": 0.2056,
+      "step": 10469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5599976194880125,
+      "learning_rate": 8.733152905689328e-06,
+      "loss": 0.2107,
+      "step": 10470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5706241023314977,
+      "learning_rate": 8.732890783882596e-06,
+      "loss": 0.2592,
+      "step": 10471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.761600887982693,
+      "learning_rate": 8.732628638895622e-06,
+      "loss": 0.4735,
+      "step": 10472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1114738018866306,
+      "learning_rate": 8.732366470730038e-06,
+      "loss": 0.3449,
+      "step": 10473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.533223981776182,
+      "learning_rate": 8.732104279387472e-06,
+      "loss": 0.3769,
+      "step": 10474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0696300816205158,
+      "learning_rate": 8.731842064869549e-06,
+      "loss": 0.175,
+      "step": 10475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8971031585932088,
+      "learning_rate": 8.731579827177899e-06,
+      "loss": 0.5101,
+      "step": 10476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.812533357740786,
+      "learning_rate": 8.73131756631415e-06,
+      "loss": 0.5237,
+      "step": 10477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4965667474909066,
+      "learning_rate": 8.731055282279932e-06,
+      "loss": 0.3247,
+      "step": 10478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8897039755217662,
+      "learning_rate": 8.730792975076872e-06,
+      "loss": 0.3114,
+      "step": 10479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2933046532716737,
+      "learning_rate": 8.730530644706601e-06,
+      "loss": 0.2237,
+      "step": 10480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7069906441596818,
+      "learning_rate": 8.730268291170744e-06,
+      "loss": 0.193,
+      "step": 10481
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6659213505643578,
+      "learning_rate": 8.730005914470934e-06,
+      "loss": 0.3702,
+      "step": 10482
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.25586216927747,
+      "learning_rate": 8.729743514608797e-06,
+      "loss": 0.2223,
+      "step": 10483
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6388131150131435,
+      "learning_rate": 8.729481091585966e-06,
+      "loss": 0.2014,
+      "step": 10484
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7629426481489794,
+      "learning_rate": 8.729218645404068e-06,
+      "loss": 0.3559,
+      "step": 10485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3668092367872975,
+      "learning_rate": 8.728956176064735e-06,
+      "loss": 0.3211,
+      "step": 10486
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.576235825424231,
+      "learning_rate": 8.728693683569594e-06,
+      "loss": 0.405,
+      "step": 10487
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8341204721736402,
+      "learning_rate": 8.728431167920278e-06,
+      "loss": 0.4667,
+      "step": 10488
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3462790431960445,
+      "learning_rate": 8.728168629118414e-06,
+      "loss": 0.1833,
+      "step": 10489
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1787691346903053,
+      "learning_rate": 8.727906067165636e-06,
+      "loss": 0.2034,
+      "step": 10490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.350924623187772,
+      "learning_rate": 8.72764348206357e-06,
+      "loss": 0.2161,
+      "step": 10491
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.071582753584102,
+      "learning_rate": 8.727380873813851e-06,
+      "loss": 0.393,
+      "step": 10492
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6725741934214486,
+      "learning_rate": 8.727118242418106e-06,
+      "loss": 0.3293,
+      "step": 10493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4949600167015098,
+      "learning_rate": 8.726855587877968e-06,
+      "loss": 0.3476,
+      "step": 10494
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.94365227950408,
+      "learning_rate": 8.726592910195068e-06,
+      "loss": 0.3192,
+      "step": 10495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.554300504470733,
+      "learning_rate": 8.726330209371037e-06,
+      "loss": 0.5082,
+      "step": 10496
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6940885203772784,
+      "learning_rate": 8.726067485407505e-06,
+      "loss": 0.4636,
+      "step": 10497
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.637841087955046,
+      "learning_rate": 8.725804738306105e-06,
+      "loss": 0.2869,
+      "step": 10498
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.859036046075735,
+      "learning_rate": 8.725541968068467e-06,
+      "loss": 0.2779,
+      "step": 10499
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7229659003319546,
+      "learning_rate": 8.725279174696226e-06,
+      "loss": 0.2864,
+      "step": 10500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0066392119946443,
+      "learning_rate": 8.72501635819101e-06,
+      "loss": 0.6394,
+      "step": 10501
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.442445853062499,
+      "learning_rate": 8.724753518554454e-06,
+      "loss": 0.4132,
+      "step": 10502
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5077237503956888,
+      "learning_rate": 8.724490655788187e-06,
+      "loss": 0.2383,
+      "step": 10503
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7788900712206914,
+      "learning_rate": 8.724227769893846e-06,
+      "loss": 0.4894,
+      "step": 10504
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4934542276082619,
+      "learning_rate": 8.723964860873057e-06,
+      "loss": 0.1817,
+      "step": 10505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.700747949705552,
+      "learning_rate": 8.723701928727458e-06,
+      "loss": 0.3606,
+      "step": 10506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6891109627221084,
+      "learning_rate": 8.72343897345868e-06,
+      "loss": 0.5423,
+      "step": 10507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5915236396546926,
+      "learning_rate": 8.723175995068355e-06,
+      "loss": 0.3652,
+      "step": 10508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.644089136475263,
+      "learning_rate": 8.722912993558118e-06,
+      "loss": 0.4876,
+      "step": 10509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3870680436272065,
+      "learning_rate": 8.7226499689296e-06,
+      "loss": 0.2031,
+      "step": 10510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4437030072171038,
+      "learning_rate": 8.722386921184436e-06,
+      "loss": 0.1972,
+      "step": 10511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.185522186047362,
+      "learning_rate": 8.722123850324258e-06,
+      "loss": 0.1059,
+      "step": 10512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6645426729004222,
+      "learning_rate": 8.7218607563507e-06,
+      "loss": 0.2298,
+      "step": 10513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.499142729644276,
+      "learning_rate": 8.721597639265398e-06,
+      "loss": 0.3023,
+      "step": 10514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.180986557384309,
+      "learning_rate": 8.721334499069982e-06,
+      "loss": 0.506,
+      "step": 10515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.627269994508339,
+      "learning_rate": 8.721071335766089e-06,
+      "loss": 0.2129,
+      "step": 10516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5149225519428726,
+      "learning_rate": 8.720808149355353e-06,
+      "loss": 0.2633,
+      "step": 10517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9671222906235213,
+      "learning_rate": 8.720544939839406e-06,
+      "loss": 0.2626,
+      "step": 10518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7751033454010985,
+      "learning_rate": 8.720281707219885e-06,
+      "loss": 0.4769,
+      "step": 10519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.389915043616057,
+      "learning_rate": 8.720018451498423e-06,
+      "loss": 0.2898,
+      "step": 10520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8553360259119596,
+      "learning_rate": 8.719755172676656e-06,
+      "loss": 0.4366,
+      "step": 10521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2573840978933286,
+      "learning_rate": 8.719491870756218e-06,
+      "loss": 0.7562,
+      "step": 10522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7897896949359757,
+      "learning_rate": 8.719228545738743e-06,
+      "loss": 0.213,
+      "step": 10523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1618690874096624,
+      "learning_rate": 8.718965197625868e-06,
+      "loss": 0.6829,
+      "step": 10524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.739804802481356,
+      "learning_rate": 8.71870182641923e-06,
+      "loss": 0.2675,
+      "step": 10525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7016894213487357,
+      "learning_rate": 8.71843843212046e-06,
+      "loss": 0.2126,
+      "step": 10526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5952736919887383,
+      "learning_rate": 8.718175014731194e-06,
+      "loss": 0.4776,
+      "step": 10527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2681054471733149,
+      "learning_rate": 8.717911574253073e-06,
+      "loss": 0.1766,
+      "step": 10528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.681996816202058,
+      "learning_rate": 8.717648110687729e-06,
+      "loss": 0.2622,
+      "step": 10529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4351832286061972,
+      "learning_rate": 8.717384624036796e-06,
+      "loss": 0.3453,
+      "step": 10530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6671306438636533,
+      "learning_rate": 8.717121114301915e-06,
+      "loss": 0.1985,
+      "step": 10531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9937130416386355,
+      "learning_rate": 8.716857581484719e-06,
+      "loss": 0.2694,
+      "step": 10532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6203773386240288,
+      "learning_rate": 8.716594025586847e-06,
+      "loss": 0.3716,
+      "step": 10533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9413971277095536,
+      "learning_rate": 8.716330446609931e-06,
+      "loss": 0.2145,
+      "step": 10534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.436851837358124,
+      "learning_rate": 8.716066844555613e-06,
+      "loss": 0.1977,
+      "step": 10535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7489609187979689,
+      "learning_rate": 8.715803219425527e-06,
+      "loss": 0.2377,
+      "step": 10536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0061015648685556,
+      "learning_rate": 8.715539571221311e-06,
+      "loss": 0.2496,
+      "step": 10537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.79180956749322,
+      "learning_rate": 8.715275899944603e-06,
+      "loss": 0.3946,
+      "step": 10538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4718877724986275,
+      "learning_rate": 8.715012205597038e-06,
+      "loss": 0.1746,
+      "step": 10539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3765960000277753,
+      "learning_rate": 8.714748488180256e-06,
+      "loss": 0.2568,
+      "step": 10540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3817681784579752,
+      "learning_rate": 8.714484747695892e-06,
+      "loss": 0.206,
+      "step": 10541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.081132980723216,
+      "learning_rate": 8.714220984145586e-06,
+      "loss": 0.1736,
+      "step": 10542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2594909412860695,
+      "learning_rate": 8.713957197530974e-06,
+      "loss": 0.2232,
+      "step": 10543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4124989611908798,
+      "learning_rate": 8.713693387853695e-06,
+      "loss": 0.2772,
+      "step": 10544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.356413340282479,
+      "learning_rate": 8.713429555115388e-06,
+      "loss": 0.2123,
+      "step": 10545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.780591345194849,
+      "learning_rate": 8.71316569931769e-06,
+      "loss": 0.4887,
+      "step": 10546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9883590712990538,
+      "learning_rate": 8.71290182046224e-06,
+      "loss": 0.371,
+      "step": 10547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6954093457172752,
+      "learning_rate": 8.712637918550678e-06,
+      "loss": 0.2866,
+      "step": 10548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2988065531380515,
+      "learning_rate": 8.71237399358464e-06,
+      "loss": 0.1891,
+      "step": 10549
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.739143931169496,
+      "learning_rate": 8.712110045565768e-06,
+      "loss": 0.3386,
+      "step": 10550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5376954009750032,
+      "learning_rate": 8.711846074495698e-06,
+      "loss": 0.2519,
+      "step": 10551
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4094618093902749,
+      "learning_rate": 8.711582080376072e-06,
+      "loss": 0.2542,
+      "step": 10552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7328607869794364,
+      "learning_rate": 8.711318063208526e-06,
+      "loss": 0.2078,
+      "step": 10553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9700805290796365,
+      "learning_rate": 8.711054022994703e-06,
+      "loss": 0.1696,
+      "step": 10554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2025826104396877,
+      "learning_rate": 8.710789959736239e-06,
+      "loss": 0.2389,
+      "step": 10555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1421836824261993,
+      "learning_rate": 8.710525873434777e-06,
+      "loss": 0.1771,
+      "step": 10556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.582980415158926,
+      "learning_rate": 8.710261764091956e-06,
+      "loss": 0.2953,
+      "step": 10557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6643841549008827,
+      "learning_rate": 8.709997631709415e-06,
+      "loss": 0.474,
+      "step": 10558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1346410676416943,
+      "learning_rate": 8.709733476288796e-06,
+      "loss": 0.1184,
+      "step": 10559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6227476392713227,
+      "learning_rate": 8.709469297831736e-06,
+      "loss": 0.2493,
+      "step": 10560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9061963511562636,
+      "learning_rate": 8.709205096339879e-06,
+      "loss": 0.3432,
+      "step": 10561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.890916766360507,
+      "learning_rate": 8.708940871814865e-06,
+      "loss": 0.5771,
+      "step": 10562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2626644965687195,
+      "learning_rate": 8.708676624258333e-06,
+      "loss": 0.1235,
+      "step": 10563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7624132559444772,
+      "learning_rate": 8.708412353671925e-06,
+      "loss": 0.2688,
+      "step": 10564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5388064413503368,
+      "learning_rate": 8.708148060057282e-06,
+      "loss": 0.372,
+      "step": 10565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7872286147063998,
+      "learning_rate": 8.707883743416047e-06,
+      "loss": 0.189,
+      "step": 10566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5594228102668783,
+      "learning_rate": 8.707619403749858e-06,
+      "loss": 0.2617,
+      "step": 10567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5271977991483177,
+      "learning_rate": 8.707355041060358e-06,
+      "loss": 0.4855,
+      "step": 10568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8189983208230704,
+      "learning_rate": 8.707090655349189e-06,
+      "loss": 0.1901,
+      "step": 10569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5871796746614037,
+      "learning_rate": 8.706826246617992e-06,
+      "loss": 0.3725,
+      "step": 10570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5345875084491192,
+      "learning_rate": 8.70656181486841e-06,
+      "loss": 0.1769,
+      "step": 10571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6623657512956818,
+      "learning_rate": 8.706297360102083e-06,
+      "loss": 0.4021,
+      "step": 10572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4364684807766017,
+      "learning_rate": 8.706032882320657e-06,
+      "loss": 0.3523,
+      "step": 10573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.862688262116532,
+      "learning_rate": 8.70576838152577e-06,
+      "loss": 0.3203,
+      "step": 10574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9826598959411341,
+      "learning_rate": 8.705503857719067e-06,
+      "loss": 0.219,
+      "step": 10575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3050565740089886,
+      "learning_rate": 8.70523931090219e-06,
+      "loss": 0.3052,
+      "step": 10576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.463181951583367,
+      "learning_rate": 8.704974741076782e-06,
+      "loss": 0.4256,
+      "step": 10577
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4317301556953983,
+      "learning_rate": 8.704710148244487e-06,
+      "loss": 0.3244,
+      "step": 10578
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6538072962793848,
+      "learning_rate": 8.704445532406944e-06,
+      "loss": 0.26,
+      "step": 10579
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5232573476057272,
+      "learning_rate": 8.7041808935658e-06,
+      "loss": 0.1936,
+      "step": 10580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5755569908364762,
+      "learning_rate": 8.703916231722697e-06,
+      "loss": 0.3353,
+      "step": 10581
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5531391839697983,
+      "learning_rate": 8.703651546879279e-06,
+      "loss": 0.1772,
+      "step": 10582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.565261036907827,
+      "learning_rate": 8.703386839037189e-06,
+      "loss": 0.2801,
+      "step": 10583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8928763653578902,
+      "learning_rate": 8.703122108198073e-06,
+      "loss": 0.2915,
+      "step": 10584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.127827222159419,
+      "learning_rate": 8.70285735436357e-06,
+      "loss": 0.2995,
+      "step": 10585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.632602626687536,
+      "learning_rate": 8.702592577535329e-06,
+      "loss": 0.2847,
+      "step": 10586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3148106319814785,
+      "learning_rate": 8.702327777714991e-06,
+      "loss": 0.193,
+      "step": 10587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1589852276815962,
+      "learning_rate": 8.702062954904203e-06,
+      "loss": 0.2053,
+      "step": 10588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.472590462710206,
+      "learning_rate": 8.70179810910461e-06,
+      "loss": 0.1971,
+      "step": 10589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9415844792112231,
+      "learning_rate": 8.70153324031785e-06,
+      "loss": 0.1605,
+      "step": 10590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3324069614744458,
+      "learning_rate": 8.701268348545576e-06,
+      "loss": 0.2094,
+      "step": 10591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8816688633677305,
+      "learning_rate": 8.701003433789428e-06,
+      "loss": 0.4936,
+      "step": 10592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7974457000348534,
+      "learning_rate": 8.700738496051053e-06,
+      "loss": 0.5518,
+      "step": 10593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8153354893254927,
+      "learning_rate": 8.700473535332094e-06,
+      "loss": 0.4237,
+      "step": 10594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7044656665146687,
+      "learning_rate": 8.7002085516342e-06,
+      "loss": 0.4204,
+      "step": 10595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6164346568919605,
+      "learning_rate": 8.699943544959013e-06,
+      "loss": 0.3624,
+      "step": 10596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1403535550931516,
+      "learning_rate": 8.69967851530818e-06,
+      "loss": 0.2161,
+      "step": 10597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.787972947222167,
+      "learning_rate": 8.699413462683347e-06,
+      "loss": 0.454,
+      "step": 10598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5537315405649066,
+      "learning_rate": 8.69914838708616e-06,
+      "loss": 0.302,
+      "step": 10599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9489075750601668,
+      "learning_rate": 8.698883288518266e-06,
+      "loss": 0.3605,
+      "step": 10600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.449139975902863,
+      "learning_rate": 8.698618166981308e-06,
+      "loss": 0.4277,
+      "step": 10601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.37819913666888,
+      "learning_rate": 8.698353022476935e-06,
+      "loss": 0.2532,
+      "step": 10602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3427685616144605,
+      "learning_rate": 8.698087855006794e-06,
+      "loss": 0.2249,
+      "step": 10603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.055059522592809,
+      "learning_rate": 8.697822664572528e-06,
+      "loss": 0.3668,
+      "step": 10604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5198570284690374,
+      "learning_rate": 8.697557451175788e-06,
+      "loss": 0.2002,
+      "step": 10605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.258847423943657,
+      "learning_rate": 8.697292214818218e-06,
+      "loss": 0.1984,
+      "step": 10606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2484798620554036,
+      "learning_rate": 8.697026955501467e-06,
+      "loss": 0.2405,
+      "step": 10607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.580081382362452,
+      "learning_rate": 8.69676167322718e-06,
+      "loss": 0.5164,
+      "step": 10608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7834028631449317,
+      "learning_rate": 8.696496367997007e-06,
+      "loss": 0.2126,
+      "step": 10609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6375515464684853,
+      "learning_rate": 8.696231039812593e-06,
+      "loss": 0.3448,
+      "step": 10610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1433726865627247,
+      "learning_rate": 8.695965688675588e-06,
+      "loss": 0.6388,
+      "step": 10611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.716231780415643,
+      "learning_rate": 8.695700314587636e-06,
+      "loss": 0.2412,
+      "step": 10612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.700400529075013,
+      "learning_rate": 8.69543491755039e-06,
+      "loss": 0.285,
+      "step": 10613
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8225021454127364,
+      "learning_rate": 8.695169497565493e-06,
+      "loss": 0.4154,
+      "step": 10614
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3837243080467767,
+      "learning_rate": 8.694904054634597e-06,
+      "loss": 0.2925,
+      "step": 10615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5925502667529658,
+      "learning_rate": 8.69463858875935e-06,
+      "loss": 0.3211,
+      "step": 10616
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0542056893078358,
+      "learning_rate": 8.694373099941398e-06,
+      "loss": 0.5479,
+      "step": 10617
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3523492645512651,
+      "learning_rate": 8.69410758818239e-06,
+      "loss": 0.1651,
+      "step": 10618
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6148114140728338,
+      "learning_rate": 8.693842053483977e-06,
+      "loss": 0.2411,
+      "step": 10619
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.015699791814395,
+      "learning_rate": 8.693576495847807e-06,
+      "loss": 0.5078,
+      "step": 10620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4068735170886149,
+      "learning_rate": 8.693310915275528e-06,
+      "loss": 0.3892,
+      "step": 10621
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2598302181451098,
+      "learning_rate": 8.69304531176879e-06,
+      "loss": 0.1855,
+      "step": 10622
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5237366455985633,
+      "learning_rate": 8.692779685329242e-06,
+      "loss": 0.1855,
+      "step": 10623
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.602562855945463,
+      "learning_rate": 8.692514035958534e-06,
+      "loss": 0.3029,
+      "step": 10624
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1991454390940475,
+      "learning_rate": 8.692248363658316e-06,
+      "loss": 0.1692,
+      "step": 10625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.554109922546936,
+      "learning_rate": 8.691982668430236e-06,
+      "loss": 0.287,
+      "step": 10626
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6197447552792563,
+      "learning_rate": 8.691716950275946e-06,
+      "loss": 0.4573,
+      "step": 10627
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5221418495357733,
+      "learning_rate": 8.691451209197093e-06,
+      "loss": 0.3605,
+      "step": 10628
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2886900855122219,
+      "learning_rate": 8.69118544519533e-06,
+      "loss": 0.2868,
+      "step": 10629
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.464342611421144,
+      "learning_rate": 8.690919658272308e-06,
+      "loss": 0.3232,
+      "step": 10630
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.343585514380959,
+      "learning_rate": 8.690653848429674e-06,
+      "loss": 0.248,
+      "step": 10631
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6106290514565293,
+      "learning_rate": 8.690388015669082e-06,
+      "loss": 0.205,
+      "step": 10632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5433316709348646,
+      "learning_rate": 8.690122159992181e-06,
+      "loss": 0.3598,
+      "step": 10633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5343201170524068,
+      "learning_rate": 8.689856281400621e-06,
+      "loss": 0.1585,
+      "step": 10634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2756498111403396,
+      "learning_rate": 8.689590379896054e-06,
+      "loss": 0.1133,
+      "step": 10635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.798639368349145,
+      "learning_rate": 8.689324455480133e-06,
+      "loss": 0.3649,
+      "step": 10636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.340112151130394,
+      "learning_rate": 8.689058508154506e-06,
+      "loss": 0.3681,
+      "step": 10637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.036446307082573,
+      "learning_rate": 8.688792537920827e-06,
+      "loss": 0.3836,
+      "step": 10638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2729851469260056,
+      "learning_rate": 8.688526544780747e-06,
+      "loss": 0.3188,
+      "step": 10639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9305405381878935,
+      "learning_rate": 8.68826052873592e-06,
+      "loss": 0.2709,
+      "step": 10640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3589864646937222,
+      "learning_rate": 8.687994489787992e-06,
+      "loss": 0.3002,
+      "step": 10641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5247502124349657,
+      "learning_rate": 8.687728427938619e-06,
+      "loss": 0.3691,
+      "step": 10642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9505142386664094,
+      "learning_rate": 8.687462343189453e-06,
+      "loss": 0.3283,
+      "step": 10643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.034677335654866,
+      "learning_rate": 8.687196235542146e-06,
+      "loss": 0.3608,
+      "step": 10644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.725777220904792,
+      "learning_rate": 8.68693010499835e-06,
+      "loss": 0.2045,
+      "step": 10645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0322503165667016,
+      "learning_rate": 8.686663951559718e-06,
+      "loss": 0.0834,
+      "step": 10646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8367534488140917,
+      "learning_rate": 8.686397775227903e-06,
+      "loss": 0.2896,
+      "step": 10647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.569534562175288,
+      "learning_rate": 8.686131576004557e-06,
+      "loss": 0.3278,
+      "step": 10648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1556076358863367,
+      "learning_rate": 8.685865353891335e-06,
+      "loss": 0.1919,
+      "step": 10649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2596463494427415,
+      "learning_rate": 8.685599108889888e-06,
+      "loss": 0.1103,
+      "step": 10650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2909345750708088,
+      "learning_rate": 8.685332841001868e-06,
+      "loss": 0.4518,
+      "step": 10651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4942498534923543,
+      "learning_rate": 8.685066550228933e-06,
+      "loss": 0.3857,
+      "step": 10652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5862518114184083,
+      "learning_rate": 8.684800236572735e-06,
+      "loss": 0.2186,
+      "step": 10653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7296671399242345,
+      "learning_rate": 8.684533900034926e-06,
+      "loss": 0.5087,
+      "step": 10654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7400282192887904,
+      "learning_rate": 8.68426754061716e-06,
+      "loss": 0.3183,
+      "step": 10655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0457129765394084,
+      "learning_rate": 8.684001158321091e-06,
+      "loss": 0.2294,
+      "step": 10656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.873443963793827,
+      "learning_rate": 8.683734753148377e-06,
+      "loss": 0.252,
+      "step": 10657
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3974508422444272,
+      "learning_rate": 8.683468325100666e-06,
+      "loss": 0.3457,
+      "step": 10658
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9798199194599226,
+      "learning_rate": 8.683201874179616e-06,
+      "loss": 0.247,
+      "step": 10659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5211197498197855,
+      "learning_rate": 8.682935400386882e-06,
+      "loss": 0.1901,
+      "step": 10660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.984830730776613,
+      "learning_rate": 8.68266890372412e-06,
+      "loss": 0.4192,
+      "step": 10661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3603588738693415,
+      "learning_rate": 8.68240238419298e-06,
+      "loss": 0.1296,
+      "step": 10662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3936645969709986,
+      "learning_rate": 8.682135841795121e-06,
+      "loss": 0.1993,
+      "step": 10663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1562633916362877,
+      "learning_rate": 8.681869276532197e-06,
+      "loss": 0.1445,
+      "step": 10664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4334766771498553,
+      "learning_rate": 8.681602688405864e-06,
+      "loss": 0.1829,
+      "step": 10665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.695832513347009,
+      "learning_rate": 8.681336077417778e-06,
+      "loss": 0.3357,
+      "step": 10666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0009178536526115,
+      "learning_rate": 8.68106944356959e-06,
+      "loss": 0.3285,
+      "step": 10667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4013683558336232,
+      "learning_rate": 8.680802786862961e-06,
+      "loss": 0.1963,
+      "step": 10668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.64149114973841,
+      "learning_rate": 8.680536107299545e-06,
+      "loss": 0.7607,
+      "step": 10669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8458551697447592,
+      "learning_rate": 8.680269404880997e-06,
+      "loss": 0.2769,
+      "step": 10670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.258565067459324,
+      "learning_rate": 8.680002679608976e-06,
+      "loss": 0.1918,
+      "step": 10671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.866499842025235,
+      "learning_rate": 8.679735931485134e-06,
+      "loss": 0.2475,
+      "step": 10672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6381095403315087,
+      "learning_rate": 8.679469160511131e-06,
+      "loss": 0.4426,
+      "step": 10673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6301922101654522,
+      "learning_rate": 8.679202366688622e-06,
+      "loss": 0.3945,
+      "step": 10674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2353558713205657,
+      "learning_rate": 8.678935550019265e-06,
+      "loss": 0.2031,
+      "step": 10675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4074469876173075,
+      "learning_rate": 8.678668710504717e-06,
+      "loss": 0.2692,
+      "step": 10676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7871868596190046,
+      "learning_rate": 8.678401848146632e-06,
+      "loss": 0.2875,
+      "step": 10677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4252414563497156,
+      "learning_rate": 8.678134962946669e-06,
+      "loss": 0.5548,
+      "step": 10678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4155202380525396,
+      "learning_rate": 8.677868054906486e-06,
+      "loss": 0.2093,
+      "step": 10679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5644111252168686,
+      "learning_rate": 8.677601124027739e-06,
+      "loss": 0.4457,
+      "step": 10680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4592279641946735,
+      "learning_rate": 8.677334170312088e-06,
+      "loss": 0.2743,
+      "step": 10681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7804502147263812,
+      "learning_rate": 8.677067193761189e-06,
+      "loss": 0.4168,
+      "step": 10682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5492421050338387,
+      "learning_rate": 8.6768001943767e-06,
+      "loss": 0.2373,
+      "step": 10683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6867628783495825,
+      "learning_rate": 8.676533172160278e-06,
+      "loss": 0.2114,
+      "step": 10684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5618513167441528,
+      "learning_rate": 8.676266127113581e-06,
+      "loss": 0.3441,
+      "step": 10685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6297947452401822,
+      "learning_rate": 8.675999059238271e-06,
+      "loss": 0.2154,
+      "step": 10686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7739540053751803,
+      "learning_rate": 8.675731968536004e-06,
+      "loss": 0.5695,
+      "step": 10687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6791595782642292,
+      "learning_rate": 8.675464855008435e-06,
+      "loss": 0.3175,
+      "step": 10688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8949762903129315,
+      "learning_rate": 8.67519771865723e-06,
+      "loss": 0.3743,
+      "step": 10689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1693152636587567,
+      "learning_rate": 8.67493055948404e-06,
+      "loss": 0.1982,
+      "step": 10690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5655901678478326,
+      "learning_rate": 8.67466337749053e-06,
+      "loss": 0.3921,
+      "step": 10691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6172903553416325,
+      "learning_rate": 8.67439617267836e-06,
+      "loss": 0.3485,
+      "step": 10692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7628435574483592,
+      "learning_rate": 8.674128945049181e-06,
+      "loss": 0.3392,
+      "step": 10693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7477817995064786,
+      "learning_rate": 8.67386169460466e-06,
+      "loss": 0.2493,
+      "step": 10694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4398805521992848,
+      "learning_rate": 8.673594421346455e-06,
+      "loss": 0.2274,
+      "step": 10695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8930363607199243,
+      "learning_rate": 8.673327125276223e-06,
+      "loss": 0.5688,
+      "step": 10696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.354953377081258,
+      "learning_rate": 8.673059806395628e-06,
+      "loss": 0.4199,
+      "step": 10697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7172105505613684,
+      "learning_rate": 8.672792464706326e-06,
+      "loss": 0.429,
+      "step": 10698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6354859290874881,
+      "learning_rate": 8.672525100209982e-06,
+      "loss": 0.478,
+      "step": 10699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5899989991927146,
+      "learning_rate": 8.67225771290825e-06,
+      "loss": 0.355,
+      "step": 10700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5571399591435828,
+      "learning_rate": 8.671990302802793e-06,
+      "loss": 0.3332,
+      "step": 10701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.440108914623286,
+      "learning_rate": 8.671722869895273e-06,
+      "loss": 0.3292,
+      "step": 10702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8960500429992055,
+      "learning_rate": 8.671455414187351e-06,
+      "loss": 0.4628,
+      "step": 10703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1936240142730603,
+      "learning_rate": 8.671187935680687e-06,
+      "loss": 0.2097,
+      "step": 10704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4826167635665164,
+      "learning_rate": 8.670920434376939e-06,
+      "loss": 0.3667,
+      "step": 10705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6390275694195147,
+      "learning_rate": 8.670652910277772e-06,
+      "loss": 0.2579,
+      "step": 10706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9143967472874928,
+      "learning_rate": 8.670385363384846e-06,
+      "loss": 0.5443,
+      "step": 10707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8899479315727938,
+      "learning_rate": 8.670117793699823e-06,
+      "loss": 0.1864,
+      "step": 10708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3344197323825717,
+      "learning_rate": 8.669850201224364e-06,
+      "loss": 0.3808,
+      "step": 10709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8153487459866444,
+      "learning_rate": 8.669582585960129e-06,
+      "loss": 0.5393,
+      "step": 10710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.039889639225396,
+      "learning_rate": 8.669314947908783e-06,
+      "loss": 0.5242,
+      "step": 10711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.146957590008209,
+      "learning_rate": 8.669047287071985e-06,
+      "loss": 0.3102,
+      "step": 10712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6271632266079399,
+      "learning_rate": 8.668779603451398e-06,
+      "loss": 0.2397,
+      "step": 10713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.084462462253088,
+      "learning_rate": 8.668511897048686e-06,
+      "loss": 0.6327,
+      "step": 10714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1051995249489202,
+      "learning_rate": 8.66824416786551e-06,
+      "loss": 0.1968,
+      "step": 10715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.343601906178625,
+      "learning_rate": 8.667976415903531e-06,
+      "loss": 0.3382,
+      "step": 10716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9976657679229985,
+      "learning_rate": 8.667708641164415e-06,
+      "loss": 0.4432,
+      "step": 10717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5268747586506979,
+      "learning_rate": 8.667440843649822e-06,
+      "loss": 0.1069,
+      "step": 10718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5648197501708951,
+      "learning_rate": 8.667173023361418e-06,
+      "loss": 0.4155,
+      "step": 10719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4089794082702123,
+      "learning_rate": 8.666905180300862e-06,
+      "loss": 0.336,
+      "step": 10720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7756916395268494,
+      "learning_rate": 8.666637314469818e-06,
+      "loss": 0.2456,
+      "step": 10721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.130900878222345,
+      "learning_rate": 8.666369425869954e-06,
+      "loss": 0.423,
+      "step": 10722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7286908012461226,
+      "learning_rate": 8.66610151450293e-06,
+      "loss": 0.3548,
+      "step": 10723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3708670529268847,
+      "learning_rate": 8.665833580370407e-06,
+      "loss": 0.157,
+      "step": 10724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6499684442768288,
+      "learning_rate": 8.665565623474054e-06,
+      "loss": 0.4247,
+      "step": 10725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.123083258196895,
+      "learning_rate": 8.665297643815532e-06,
+      "loss": 0.3951,
+      "step": 10726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6114362888445555,
+      "learning_rate": 8.665029641396506e-06,
+      "loss": 0.1892,
+      "step": 10727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6320139994971874,
+      "learning_rate": 8.664761616218638e-06,
+      "loss": 0.2271,
+      "step": 10728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6605704452826802,
+      "learning_rate": 8.664493568283597e-06,
+      "loss": 0.3554,
+      "step": 10729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.939921849034946,
+      "learning_rate": 8.664225497593045e-06,
+      "loss": 0.1635,
+      "step": 10730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1837357004736007,
+      "learning_rate": 8.663957404148645e-06,
+      "loss": 0.1516,
+      "step": 10731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.678066811360793,
+      "learning_rate": 8.663689287952062e-06,
+      "loss": 0.3585,
+      "step": 10732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2988624116487921,
+      "learning_rate": 8.663421149004964e-06,
+      "loss": 0.2423,
+      "step": 10733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2723127748737655,
+      "learning_rate": 8.663152987309014e-06,
+      "loss": 0.1673,
+      "step": 10734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.021266118305605,
+      "learning_rate": 8.662884802865877e-06,
+      "loss": 0.4533,
+      "step": 10735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5869396266345341,
+      "learning_rate": 8.662616595677217e-06,
+      "loss": 0.3437,
+      "step": 10736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8990410465442964,
+      "learning_rate": 8.662348365744705e-06,
+      "loss": 0.2954,
+      "step": 10737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.153813056188789,
+      "learning_rate": 8.66208011307e-06,
+      "loss": 0.5248,
+      "step": 10738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8942731229194716,
+      "learning_rate": 8.661811837654772e-06,
+      "loss": 0.2741,
+      "step": 10739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.147068530803487,
+      "learning_rate": 8.661543539500686e-06,
+      "loss": 0.6003,
+      "step": 10740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5734201871644902,
+      "learning_rate": 8.661275218609406e-06,
+      "loss": 0.1766,
+      "step": 10741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.560724465550379,
+      "learning_rate": 8.6610068749826e-06,
+      "loss": 0.3845,
+      "step": 10742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.783897166650425,
+      "learning_rate": 8.660738508621937e-06,
+      "loss": 0.2829,
+      "step": 10743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8082741193334937,
+      "learning_rate": 8.660470119529079e-06,
+      "loss": 0.5456,
+      "step": 10744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4548753094715936,
+      "learning_rate": 8.660201707705694e-06,
+      "loss": 0.4833,
+      "step": 10745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6876962291572446,
+      "learning_rate": 8.659933273153448e-06,
+      "loss": 0.3204,
+      "step": 10746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5173630334922936,
+      "learning_rate": 8.659664815874011e-06,
+      "loss": 0.4374,
+      "step": 10747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7316904943278317,
+      "learning_rate": 8.659396335869046e-06,
+      "loss": 0.2908,
+      "step": 10748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.873930355779917,
+      "learning_rate": 8.659127833140223e-06,
+      "loss": 0.4536,
+      "step": 10749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1505921426960926,
+      "learning_rate": 8.65885930768921e-06,
+      "loss": 0.1902,
+      "step": 10750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5608326557153585,
+      "learning_rate": 8.65859075951767e-06,
+      "loss": 0.2216,
+      "step": 10751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5113543384178647,
+      "learning_rate": 8.658322188627278e-06,
+      "loss": 0.3873,
+      "step": 10752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1148301910345746,
+      "learning_rate": 8.658053595019692e-06,
+      "loss": 0.1687,
+      "step": 10753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7229562009529276,
+      "learning_rate": 8.65778497869659e-06,
+      "loss": 0.3285,
+      "step": 10754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4518477930979794,
+      "learning_rate": 8.657516339659633e-06,
+      "loss": 0.3124,
+      "step": 10755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5051902021374244,
+      "learning_rate": 8.657247677910493e-06,
+      "loss": 0.442,
+      "step": 10756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6286892323271935,
+      "learning_rate": 8.656978993450836e-06,
+      "loss": 0.3678,
+      "step": 10757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4851522042281073,
+      "learning_rate": 8.65671028628233e-06,
+      "loss": 0.2596,
+      "step": 10758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4736995934686212,
+      "learning_rate": 8.656441556406645e-06,
+      "loss": 0.3526,
+      "step": 10759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.242389386643478,
+      "learning_rate": 8.656172803825451e-06,
+      "loss": 0.1932,
+      "step": 10760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.73619728968565,
+      "learning_rate": 8.655904028540415e-06,
+      "loss": 0.2364,
+      "step": 10761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.640358358241061,
+      "learning_rate": 8.655635230553206e-06,
+      "loss": 0.244,
+      "step": 10762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5883201883780902,
+      "learning_rate": 8.655366409865495e-06,
+      "loss": 0.1929,
+      "step": 10763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6617768781006812,
+      "learning_rate": 8.65509756647895e-06,
+      "loss": 0.4061,
+      "step": 10764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7702357129903767,
+      "learning_rate": 8.654828700395237e-06,
+      "loss": 0.3199,
+      "step": 10765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5304196305544422,
+      "learning_rate": 8.654559811616033e-06,
+      "loss": 0.2613,
+      "step": 10766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.084347050798574,
+      "learning_rate": 8.654290900143001e-06,
+      "loss": 0.1592,
+      "step": 10767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3650817922262255,
+      "learning_rate": 8.654021965977814e-06,
+      "loss": 0.3785,
+      "step": 10768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9986028969675955,
+      "learning_rate": 8.653753009122143e-06,
+      "loss": 0.4185,
+      "step": 10769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0858391710401596,
+      "learning_rate": 8.653484029577656e-06,
+      "loss": 0.1994,
+      "step": 10770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7722383532781099,
+      "learning_rate": 8.653215027346025e-06,
+      "loss": 0.4921,
+      "step": 10771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.395929148811072,
+      "learning_rate": 8.652946002428918e-06,
+      "loss": 0.1952,
+      "step": 10772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8887859377102358,
+      "learning_rate": 8.652676954828007e-06,
+      "loss": 0.5121,
+      "step": 10773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0508095512253846,
+      "learning_rate": 8.652407884544965e-06,
+      "loss": 0.3854,
+      "step": 10774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.705462933406793,
+      "learning_rate": 8.652138791581458e-06,
+      "loss": 0.1705,
+      "step": 10775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6565907775532158,
+      "learning_rate": 8.65186967593916e-06,
+      "loss": 0.363,
+      "step": 10776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.217743618494665,
+      "learning_rate": 8.651600537619744e-06,
+      "loss": 0.2366,
+      "step": 10777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.375292679133104,
+      "learning_rate": 8.651331376624877e-06,
+      "loss": 0.329,
+      "step": 10778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.085652706978462,
+      "learning_rate": 8.651062192956232e-06,
+      "loss": 0.4821,
+      "step": 10779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9098621046368214,
+      "learning_rate": 8.650792986615481e-06,
+      "loss": 0.2244,
+      "step": 10780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6094871315271324,
+      "learning_rate": 8.650523757604297e-06,
+      "loss": 0.3247,
+      "step": 10781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.825977000890229,
+      "learning_rate": 8.65025450592435e-06,
+      "loss": 0.3545,
+      "step": 10782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8780464655789506,
+      "learning_rate": 8.649985231577311e-06,
+      "loss": 0.4001,
+      "step": 10783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9088514587327625,
+      "learning_rate": 8.649715934564857e-06,
+      "loss": 0.6939,
+      "step": 10784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9575341467479448,
+      "learning_rate": 8.649446614888655e-06,
+      "loss": 0.4369,
+      "step": 10785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7356270893412282,
+      "learning_rate": 8.649177272550378e-06,
+      "loss": 0.4911,
+      "step": 10786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8767624429706573,
+      "learning_rate": 8.648907907551702e-06,
+      "loss": 0.2029,
+      "step": 10787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1080864000751476,
+      "learning_rate": 8.648638519894296e-06,
+      "loss": 0.2002,
+      "step": 10788
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5563210903681068,
+      "learning_rate": 8.648369109579834e-06,
+      "loss": 0.2418,
+      "step": 10789
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6884533432113586,
+      "learning_rate": 8.648099676609991e-06,
+      "loss": 0.5227,
+      "step": 10790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2644195069589421,
+      "learning_rate": 8.647830220986438e-06,
+      "loss": 0.3071,
+      "step": 10791
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3597378958899602,
+      "learning_rate": 8.647560742710847e-06,
+      "loss": 0.3714,
+      "step": 10792
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8275723396313737,
+      "learning_rate": 8.647291241784894e-06,
+      "loss": 0.5823,
+      "step": 10793
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8702170801002447,
+      "learning_rate": 8.647021718210252e-06,
+      "loss": 0.3065,
+      "step": 10794
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1516103299752287,
+      "learning_rate": 8.646752171988595e-06,
+      "loss": 0.1817,
+      "step": 10795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6024030999461316,
+      "learning_rate": 8.646482603121596e-06,
+      "loss": 0.5879,
+      "step": 10796
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5760971959230419,
+      "learning_rate": 8.646213011610927e-06,
+      "loss": 0.3634,
+      "step": 10797
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.597502176368377,
+      "learning_rate": 8.645943397458265e-06,
+      "loss": 0.4277,
+      "step": 10798
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.73359213593018,
+      "learning_rate": 8.645673760665282e-06,
+      "loss": 0.3872,
+      "step": 10799
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.772915103618801,
+      "learning_rate": 8.645404101233657e-06,
+      "loss": 0.2527,
+      "step": 10800
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2423068962357011,
+      "learning_rate": 8.645134419165058e-06,
+      "loss": 0.1986,
+      "step": 10801
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3902421997699008,
+      "learning_rate": 8.644864714461165e-06,
+      "loss": 0.2063,
+      "step": 10802
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7985330602676435,
+      "learning_rate": 8.644594987123649e-06,
+      "loss": 0.5521,
+      "step": 10803
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5451899330914791,
+      "learning_rate": 8.644325237154188e-06,
+      "loss": 0.1871,
+      "step": 10804
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5529869205671747,
+      "learning_rate": 8.644055464554455e-06,
+      "loss": 0.6384,
+      "step": 10805
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.87584358153078,
+      "learning_rate": 8.643785669326125e-06,
+      "loss": 0.3529,
+      "step": 10806
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.125067330537694,
+      "learning_rate": 8.643515851470877e-06,
+      "loss": 0.1699,
+      "step": 10807
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.955605152903541,
+      "learning_rate": 8.643246010990381e-06,
+      "loss": 0.3317,
+      "step": 10808
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1283887614254828,
+      "learning_rate": 8.642976147886316e-06,
+      "loss": 0.1789,
+      "step": 10809
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.841454503426955,
+      "learning_rate": 8.642706262160356e-06,
+      "loss": 0.4228,
+      "step": 10810
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7819664668675375,
+      "learning_rate": 8.64243635381418e-06,
+      "loss": 0.3302,
+      "step": 10811
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.44947550764763,
+      "learning_rate": 8.642166422849461e-06,
+      "loss": 0.6345,
+      "step": 10812
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5342909968624994,
+      "learning_rate": 8.641896469267878e-06,
+      "loss": 0.3449,
+      "step": 10813
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2744904056139985,
+      "learning_rate": 8.641626493071104e-06,
+      "loss": 0.5977,
+      "step": 10814
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2586371542492638,
+      "learning_rate": 8.641356494260817e-06,
+      "loss": 0.1729,
+      "step": 10815
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1822988599888888,
+      "learning_rate": 8.641086472838695e-06,
+      "loss": 0.2381,
+      "step": 10816
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9882415840798171,
+      "learning_rate": 8.640816428806414e-06,
+      "loss": 0.2173,
+      "step": 10817
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4152403553619206,
+      "learning_rate": 8.640546362165647e-06,
+      "loss": 0.405,
+      "step": 10818
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1829271851641945,
+      "learning_rate": 8.640276272918078e-06,
+      "loss": 0.1866,
+      "step": 10819
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.22447518095598,
+      "learning_rate": 8.640006161065378e-06,
+      "loss": 0.2112,
+      "step": 10820
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0002240375862104,
+      "learning_rate": 8.639736026609228e-06,
+      "loss": 0.4156,
+      "step": 10821
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4866752219230168,
+      "learning_rate": 8.639465869551305e-06,
+      "loss": 0.1324,
+      "step": 10822
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6954532028735543,
+      "learning_rate": 8.639195689893285e-06,
+      "loss": 0.2495,
+      "step": 10823
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9973846486626865,
+      "learning_rate": 8.638925487636847e-06,
+      "loss": 0.5596,
+      "step": 10824
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9553855631405883,
+      "learning_rate": 8.63865526278367e-06,
+      "loss": 0.6687,
+      "step": 10825
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5717046364744918,
+      "learning_rate": 8.638385015335429e-06,
+      "loss": 0.2989,
+      "step": 10826
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.012136345639815,
+      "learning_rate": 8.638114745293803e-06,
+      "loss": 0.4012,
+      "step": 10827
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.829787635343462,
+      "learning_rate": 8.637844452660474e-06,
+      "loss": 0.2412,
+      "step": 10828
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7761105797189165,
+      "learning_rate": 8.637574137437117e-06,
+      "loss": 0.3535,
+      "step": 10829
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7132190467790098,
+      "learning_rate": 8.637303799625409e-06,
+      "loss": 0.2205,
+      "step": 10830
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4281962126541696,
+      "learning_rate": 8.637033439227033e-06,
+      "loss": 0.3279,
+      "step": 10831
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4504765851093664,
+      "learning_rate": 8.636763056243664e-06,
+      "loss": 0.1668,
+      "step": 10832
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6011226794467048,
+      "learning_rate": 8.636492650676984e-06,
+      "loss": 0.2405,
+      "step": 10833
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4448462932894726,
+      "learning_rate": 8.636222222528671e-06,
+      "loss": 0.2668,
+      "step": 10834
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8171615014877367,
+      "learning_rate": 8.635951771800403e-06,
+      "loss": 0.3469,
+      "step": 10835
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4782675329607593,
+      "learning_rate": 8.63568129849386e-06,
+      "loss": 0.2475,
+      "step": 10836
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1063670505067837,
+      "learning_rate": 8.635410802610724e-06,
+      "loss": 0.4235,
+      "step": 10837
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3117720509617394,
+      "learning_rate": 8.635140284152671e-06,
+      "loss": 0.2135,
+      "step": 10838
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3714279721123852,
+      "learning_rate": 8.634869743121386e-06,
+      "loss": 0.1936,
+      "step": 10839
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7587913034478821,
+      "learning_rate": 8.634599179518544e-06,
+      "loss": 0.1662,
+      "step": 10840
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.096956223266378,
+      "learning_rate": 8.634328593345827e-06,
+      "loss": 0.2667,
+      "step": 10841
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7812089136262284,
+      "learning_rate": 8.634057984604913e-06,
+      "loss": 0.2052,
+      "step": 10842
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7895410805474412,
+      "learning_rate": 8.633787353297488e-06,
+      "loss": 0.5566,
+      "step": 10843
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6523705800450625,
+      "learning_rate": 8.633516699425227e-06,
+      "loss": 0.2816,
+      "step": 10844
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6142529570038608,
+      "learning_rate": 8.633246022989812e-06,
+      "loss": 0.2614,
+      "step": 10845
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5953886779845008,
+      "learning_rate": 8.632975323992925e-06,
+      "loss": 0.3691,
+      "step": 10846
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.581611591768875,
+      "learning_rate": 8.632704602436248e-06,
+      "loss": 0.2302,
+      "step": 10847
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.875149057344473,
+      "learning_rate": 8.632433858321458e-06,
+      "loss": 0.5658,
+      "step": 10848
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.868032976288418,
+      "learning_rate": 8.632163091650241e-06,
+      "loss": 0.2849,
+      "step": 10849
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.807346465922581,
+      "learning_rate": 8.631892302424277e-06,
+      "loss": 0.3706,
+      "step": 10850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1107854272278508,
+      "learning_rate": 8.631621490645245e-06,
+      "loss": 0.1643,
+      "step": 10851
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5379531463944565,
+      "learning_rate": 8.63135065631483e-06,
+      "loss": 0.2979,
+      "step": 10852
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.820910560201758,
+      "learning_rate": 8.63107979943471e-06,
+      "loss": 0.6431,
+      "step": 10853
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0240299911267767,
+      "learning_rate": 8.630808920006571e-06,
+      "loss": 0.4045,
+      "step": 10854
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9040294146390648,
+      "learning_rate": 8.630538018032093e-06,
+      "loss": 0.3256,
+      "step": 10855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6432319126972725,
+      "learning_rate": 8.630267093512958e-06,
+      "loss": 0.227,
+      "step": 10856
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8507515100977692,
+      "learning_rate": 8.629996146450849e-06,
+      "loss": 0.3072,
+      "step": 10857
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.807667505356354,
+      "learning_rate": 8.62972517684745e-06,
+      "loss": 0.5378,
+      "step": 10858
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8835222402357807,
+      "learning_rate": 8.62945418470444e-06,
+      "loss": 0.1921,
+      "step": 10859
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6692242018265944,
+      "learning_rate": 8.629183170023504e-06,
+      "loss": 0.4409,
+      "step": 10860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4430729939385962,
+      "learning_rate": 8.628912132806327e-06,
+      "loss": 0.3351,
+      "step": 10861
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8598068801608323,
+      "learning_rate": 8.628641073054587e-06,
+      "loss": 0.1733,
+      "step": 10862
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.874811226560863,
+      "learning_rate": 8.628369990769972e-06,
+      "loss": 0.4998,
+      "step": 10863
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7987768888683908,
+      "learning_rate": 8.628098885954163e-06,
+      "loss": 0.251,
+      "step": 10864
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.590680920160971,
+      "learning_rate": 8.627827758608843e-06,
+      "loss": 0.3057,
+      "step": 10865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4661332697920102,
+      "learning_rate": 8.627556608735696e-06,
+      "loss": 0.2802,
+      "step": 10866
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0595062484653366,
+      "learning_rate": 8.627285436336409e-06,
+      "loss": 0.6628,
+      "step": 10867
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1752441738445163,
+      "learning_rate": 8.627014241412661e-06,
+      "loss": 0.1538,
+      "step": 10868
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5724846799851666,
+      "learning_rate": 8.626743023966139e-06,
+      "loss": 0.1727,
+      "step": 10869
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2834382088617207,
+      "learning_rate": 8.626471783998527e-06,
+      "loss": 0.2146,
+      "step": 10870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7931393976388093,
+      "learning_rate": 8.626200521511508e-06,
+      "loss": 0.3132,
+      "step": 10871
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4022209211978236,
+      "learning_rate": 8.625929236506767e-06,
+      "loss": 0.1319,
+      "step": 10872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2210025389634775,
+      "learning_rate": 8.625657928985989e-06,
+      "loss": 0.1639,
+      "step": 10873
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7946943423203683,
+      "learning_rate": 8.625386598950858e-06,
+      "loss": 0.3946,
+      "step": 10874
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9096348953991993,
+      "learning_rate": 8.625115246403061e-06,
+      "loss": 0.3881,
+      "step": 10875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6462897143790904,
+      "learning_rate": 8.624843871344279e-06,
+      "loss": 0.3877,
+      "step": 10876
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.466977378822275,
+      "learning_rate": 8.624572473776202e-06,
+      "loss": 0.1831,
+      "step": 10877
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7896901949380468,
+      "learning_rate": 8.624301053700512e-06,
+      "loss": 0.3381,
+      "step": 10878
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4930414457603205,
+      "learning_rate": 8.624029611118896e-06,
+      "loss": 0.3451,
+      "step": 10879
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5048414319962924,
+      "learning_rate": 8.62375814603304e-06,
+      "loss": 0.3713,
+      "step": 10880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.816538519811774,
+      "learning_rate": 8.623486658444626e-06,
+      "loss": 0.2425,
+      "step": 10881
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5678025868957643,
+      "learning_rate": 8.623215148355344e-06,
+      "loss": 0.2636,
+      "step": 10882
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.638344403488137,
+      "learning_rate": 8.622943615766879e-06,
+      "loss": 0.4761,
+      "step": 10883
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7181400950890293,
+      "learning_rate": 8.622672060680915e-06,
+      "loss": 0.3509,
+      "step": 10884
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1050169996234356,
+      "learning_rate": 8.622400483099142e-06,
+      "loss": 0.3976,
+      "step": 10885
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4857522773826317,
+      "learning_rate": 8.622128883023243e-06,
+      "loss": 0.1965,
+      "step": 10886
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.525155560970395,
+      "learning_rate": 8.621857260454907e-06,
+      "loss": 0.3486,
+      "step": 10887
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5264628618925589,
+      "learning_rate": 8.621585615395819e-06,
+      "loss": 0.3134,
+      "step": 10888
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5216224103412168,
+      "learning_rate": 8.621313947847666e-06,
+      "loss": 0.4111,
+      "step": 10889
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.632177468976948,
+      "learning_rate": 8.621042257812137e-06,
+      "loss": 0.356,
+      "step": 10890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8104937071334,
+      "learning_rate": 8.620770545290918e-06,
+      "loss": 0.5364,
+      "step": 10891
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.914252536448422,
+      "learning_rate": 8.620498810285693e-06,
+      "loss": 0.3069,
+      "step": 10892
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.04542273041518,
+      "learning_rate": 8.620227052798155e-06,
+      "loss": 0.2741,
+      "step": 10893
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.856436181083588,
+      "learning_rate": 8.619955272829987e-06,
+      "loss": 0.2284,
+      "step": 10894
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7424756825308714,
+      "learning_rate": 8.61968347038288e-06,
+      "loss": 0.285,
+      "step": 10895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8748300952224966,
+      "learning_rate": 8.619411645458519e-06,
+      "loss": 0.4922,
+      "step": 10896
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3698764244178,
+      "learning_rate": 8.619139798058595e-06,
+      "loss": 0.1996,
+      "step": 10897
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2802756401875255,
+      "learning_rate": 8.618867928184795e-06,
+      "loss": 0.3489,
+      "step": 10898
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4990204702341736,
+      "learning_rate": 8.618596035838804e-06,
+      "loss": 0.2925,
+      "step": 10899
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.260444166672482,
+      "learning_rate": 8.618324121022315e-06,
+      "loss": 0.6255,
+      "step": 10900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5870339591876483,
+      "learning_rate": 8.618052183737013e-06,
+      "loss": 0.2436,
+      "step": 10901
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7089412956257726,
+      "learning_rate": 8.61778022398459e-06,
+      "loss": 0.2323,
+      "step": 10902
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.68862239280701,
+      "learning_rate": 8.61750824176673e-06,
+      "loss": 0.3651,
+      "step": 10903
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2886575594557033,
+      "learning_rate": 8.617236237085126e-06,
+      "loss": 0.2375,
+      "step": 10904
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7752909071041747,
+      "learning_rate": 8.616964209941469e-06,
+      "loss": 0.3984,
+      "step": 10905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4286969628110153,
+      "learning_rate": 8.616692160337441e-06,
+      "loss": 0.2114,
+      "step": 10906
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9823022627802052,
+      "learning_rate": 8.616420088274739e-06,
+      "loss": 0.5534,
+      "step": 10907
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.818460234273584,
+      "learning_rate": 8.616147993755048e-06,
+      "loss": 0.4077,
+      "step": 10908
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.412444845162931,
+      "learning_rate": 8.615875876780058e-06,
+      "loss": 0.2079,
+      "step": 10909
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7900328628257862,
+      "learning_rate": 8.615603737351458e-06,
+      "loss": 0.354,
+      "step": 10910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8070149956552264,
+      "learning_rate": 8.615331575470942e-06,
+      "loss": 0.4361,
+      "step": 10911
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7148500433699887,
+      "learning_rate": 8.615059391140197e-06,
+      "loss": 0.3933,
+      "step": 10912
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.066772799801372,
+      "learning_rate": 8.614787184360912e-06,
+      "loss": 0.3842,
+      "step": 10913
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1686523886325406,
+      "learning_rate": 8.61451495513478e-06,
+      "loss": 0.1612,
+      "step": 10914
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.292869138422336,
+      "learning_rate": 8.614242703463491e-06,
+      "loss": 0.443,
+      "step": 10915
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7281264174043385,
+      "learning_rate": 8.613970429348735e-06,
+      "loss": 0.4764,
+      "step": 10916
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4747055072901594,
+      "learning_rate": 8.613698132792203e-06,
+      "loss": 0.193,
+      "step": 10917
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.60492282830079,
+      "learning_rate": 8.613425813795584e-06,
+      "loss": 0.2291,
+      "step": 10918
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0444050027465974,
+      "learning_rate": 8.613153472360572e-06,
+      "loss": 0.1389,
+      "step": 10919
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0552236434065296,
+      "learning_rate": 8.612881108488857e-06,
+      "loss": 0.3672,
+      "step": 10920
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7851850019013513,
+      "learning_rate": 8.61260872218213e-06,
+      "loss": 0.2568,
+      "step": 10921
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0201947165411978,
+      "learning_rate": 8.612336313442083e-06,
+      "loss": 0.1902,
+      "step": 10922
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1538916794819714,
+      "learning_rate": 8.612063882270404e-06,
+      "loss": 0.1962,
+      "step": 10923
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5953384003605664,
+      "learning_rate": 8.611791428668792e-06,
+      "loss": 0.2913,
+      "step": 10924
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4835764594172822,
+      "learning_rate": 8.611518952638933e-06,
+      "loss": 0.218,
+      "step": 10925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7492000702571464,
+      "learning_rate": 8.61124645418252e-06,
+      "loss": 0.2916,
+      "step": 10926
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6554638035778195,
+      "learning_rate": 8.610973933301248e-06,
+      "loss": 0.3174,
+      "step": 10927
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.200480369002087,
+      "learning_rate": 8.610701389996805e-06,
+      "loss": 0.2907,
+      "step": 10928
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.042924758335378,
+      "learning_rate": 8.610428824270886e-06,
+      "loss": 0.4708,
+      "step": 10929
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8804627056682393,
+      "learning_rate": 8.610156236125185e-06,
+      "loss": 0.2604,
+      "step": 10930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4122316510667503,
+      "learning_rate": 8.609883625561391e-06,
+      "loss": 0.3464,
+      "step": 10931
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6108414602835257,
+      "learning_rate": 8.6096109925812e-06,
+      "loss": 0.345,
+      "step": 10932
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4091973924546806,
+      "learning_rate": 8.609338337186304e-06,
+      "loss": 0.4413,
+      "step": 10933
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5619864788204385,
+      "learning_rate": 8.609065659378395e-06,
+      "loss": 0.2422,
+      "step": 10934
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6150424063654052,
+      "learning_rate": 8.608792959159167e-06,
+      "loss": 0.3875,
+      "step": 10935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8316140869131565,
+      "learning_rate": 8.608520236530314e-06,
+      "loss": 0.2428,
+      "step": 10936
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0970711862738203,
+      "learning_rate": 8.608247491493526e-06,
+      "loss": 0.1763,
+      "step": 10937
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7958493913152371,
+      "learning_rate": 8.607974724050502e-06,
+      "loss": 0.4487,
+      "step": 10938
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8177214485850282,
+      "learning_rate": 8.607701934202934e-06,
+      "loss": 0.3938,
+      "step": 10939
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6419554674406194,
+      "learning_rate": 8.607429121952517e-06,
+      "loss": 0.3646,
+      "step": 10940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.314645741762722,
+      "learning_rate": 8.60715628730094e-06,
+      "loss": 0.3132,
+      "step": 10941
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6544399740036262,
+      "learning_rate": 8.606883430249903e-06,
+      "loss": 0.3025,
+      "step": 10942
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.879330660497086,
+      "learning_rate": 8.606610550801098e-06,
+      "loss": 0.4033,
+      "step": 10943
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7110720085154065,
+      "learning_rate": 8.606337648956217e-06,
+      "loss": 0.5075,
+      "step": 10944
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.955032729194418,
+      "learning_rate": 8.60606472471696e-06,
+      "loss": 0.2862,
+      "step": 10945
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.568295963283326,
+      "learning_rate": 8.60579177808502e-06,
+      "loss": 0.3758,
+      "step": 10946
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1637021873626359,
+      "learning_rate": 8.605518809062088e-06,
+      "loss": 0.2332,
+      "step": 10947
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6722848960159935,
+      "learning_rate": 8.605245817649864e-06,
+      "loss": 0.4883,
+      "step": 10948
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6411892647193507,
+      "learning_rate": 8.60497280385004e-06,
+      "loss": 0.3456,
+      "step": 10949
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5566381600431025,
+      "learning_rate": 8.604699767664313e-06,
+      "loss": 0.2363,
+      "step": 10950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6071581409783458,
+      "learning_rate": 8.604426709094378e-06,
+      "loss": 0.3954,
+      "step": 10951
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1472008511834857,
+      "learning_rate": 8.604153628141932e-06,
+      "loss": 0.1115,
+      "step": 10952
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9765872350660376,
+      "learning_rate": 8.603880524808669e-06,
+      "loss": 0.4919,
+      "step": 10953
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2687196323326217,
+      "learning_rate": 8.603607399096284e-06,
+      "loss": 0.2346,
+      "step": 10954
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7494186559737983,
+      "learning_rate": 8.603334251006475e-06,
+      "loss": 0.3541,
+      "step": 10955
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.423593852441219,
+      "learning_rate": 8.603061080540938e-06,
+      "loss": 0.3433,
+      "step": 10956
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.251192392854455,
+      "learning_rate": 8.602787887701369e-06,
+      "loss": 0.4227,
+      "step": 10957
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0362694602549418,
+      "learning_rate": 8.602514672489463e-06,
+      "loss": 0.1593,
+      "step": 10958
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9689531902635835,
+      "learning_rate": 8.60224143490692e-06,
+      "loss": 0.1899,
+      "step": 10959
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6052766694304292,
+      "learning_rate": 8.601968174955434e-06,
+      "loss": 0.166,
+      "step": 10960
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4942140748165158,
+      "learning_rate": 8.601694892636701e-06,
+      "loss": 0.1733,
+      "step": 10961
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4704638890680697,
+      "learning_rate": 8.60142158795242e-06,
+      "loss": 0.2894,
+      "step": 10962
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7016403131408953,
+      "learning_rate": 8.601148260904289e-06,
+      "loss": 0.3839,
+      "step": 10963
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9437516473106935,
+      "learning_rate": 8.600874911494003e-06,
+      "loss": 0.4448,
+      "step": 10964
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.562169707674462,
+      "learning_rate": 8.60060153972326e-06,
+      "loss": 0.4236,
+      "step": 10965
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.637564629400197,
+      "learning_rate": 8.600328145593759e-06,
+      "loss": 0.37,
+      "step": 10966
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.458982010600716,
+      "learning_rate": 8.600054729107196e-06,
+      "loss": 0.4805,
+      "step": 10967
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8049316509447546,
+      "learning_rate": 8.59978129026527e-06,
+      "loss": 0.3331,
+      "step": 10968
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7008428200619619,
+      "learning_rate": 8.599507829069678e-06,
+      "loss": 0.4526,
+      "step": 10969
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.817822509247542,
+      "learning_rate": 8.599234345522119e-06,
+      "loss": 0.4212,
+      "step": 10970
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3313599066519421,
+      "learning_rate": 8.598960839624291e-06,
+      "loss": 0.2372,
+      "step": 10971
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8909905017302064,
+      "learning_rate": 8.59868731137789e-06,
+      "loss": 0.3936,
+      "step": 10972
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4581278695962452,
+      "learning_rate": 8.59841376078462e-06,
+      "loss": 0.2119,
+      "step": 10973
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.839415656336194,
+      "learning_rate": 8.598140187846174e-06,
+      "loss": 0.4664,
+      "step": 10974
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8065874741442565,
+      "learning_rate": 8.597866592564255e-06,
+      "loss": 0.2261,
+      "step": 10975
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.995479029977784,
+      "learning_rate": 8.597592974940559e-06,
+      "loss": 0.3311,
+      "step": 10976
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.329348186003693,
+      "learning_rate": 8.597319334976787e-06,
+      "loss": 0.1674,
+      "step": 10977
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0587400937011016,
+      "learning_rate": 8.597045672674638e-06,
+      "loss": 0.167,
+      "step": 10978
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8463220557651916,
+      "learning_rate": 8.596771988035811e-06,
+      "loss": 0.228,
+      "step": 10979
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4686156754247617,
+      "learning_rate": 8.596498281062004e-06,
+      "loss": 0.331,
+      "step": 10980
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5848492473141016,
+      "learning_rate": 8.596224551754919e-06,
+      "loss": 0.2449,
+      "step": 10981
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3991272293630457,
+      "learning_rate": 8.595950800116255e-06,
+      "loss": 0.1911,
+      "step": 10982
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2923548485698249,
+      "learning_rate": 8.595677026147712e-06,
+      "loss": 0.1607,
+      "step": 10983
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.996946626671066,
+      "learning_rate": 8.59540322985099e-06,
+      "loss": 0.5739,
+      "step": 10984
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.064055095414377,
+      "learning_rate": 8.595129411227787e-06,
+      "loss": 0.1601,
+      "step": 10985
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7287329413233656,
+      "learning_rate": 8.594855570279807e-06,
+      "loss": 0.4366,
+      "step": 10986
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.893956366553261,
+      "learning_rate": 8.594581707008749e-06,
+      "loss": 0.2007,
+      "step": 10987
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1650363461637847,
+      "learning_rate": 8.594307821416312e-06,
+      "loss": 0.1419,
+      "step": 10988
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6683137344837675,
+      "learning_rate": 8.594033913504201e-06,
+      "loss": 0.2293,
+      "step": 10989
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.033497579251892,
+      "learning_rate": 8.593759983274111e-06,
+      "loss": 0.581,
+      "step": 10990
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7685806150342114,
+      "learning_rate": 8.593486030727748e-06,
+      "loss": 0.2858,
+      "step": 10991
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.458597275471107,
+      "learning_rate": 8.593212055866812e-06,
+      "loss": 0.2223,
+      "step": 10992
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9931258665664908,
+      "learning_rate": 8.592938058693003e-06,
+      "loss": 0.5034,
+      "step": 10993
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5050458924925834,
+      "learning_rate": 8.592664039208023e-06,
+      "loss": 0.5163,
+      "step": 10994
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7334750822073706,
+      "learning_rate": 8.592389997413573e-06,
+      "loss": 0.1984,
+      "step": 10995
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5294052368202484,
+      "learning_rate": 8.592115933311355e-06,
+      "loss": 0.1799,
+      "step": 10996
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4010698727482958,
+      "learning_rate": 8.591841846903074e-06,
+      "loss": 0.2913,
+      "step": 10997
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.553340299999786,
+      "learning_rate": 8.591567738190425e-06,
+      "loss": 0.1751,
+      "step": 10998
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8518545186058903,
+      "learning_rate": 8.591293607175117e-06,
+      "loss": 0.385,
+      "step": 10999
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7750291421150486,
+      "learning_rate": 8.59101945385885e-06,
+      "loss": 0.4042,
+      "step": 11000
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4478198005900598,
+      "learning_rate": 8.590745278243325e-06,
+      "loss": 0.3372,
+      "step": 11001
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.291181677730711,
+      "learning_rate": 8.590471080330247e-06,
+      "loss": 0.1567,
+      "step": 11002
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5708596049859886,
+      "learning_rate": 8.590196860121316e-06,
+      "loss": 0.3173,
+      "step": 11003
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6487657798651116,
+      "learning_rate": 8.589922617618238e-06,
+      "loss": 0.3165,
+      "step": 11004
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6666035838904485,
+      "learning_rate": 8.589648352822712e-06,
+      "loss": 0.2253,
+      "step": 11005
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.579490330670132,
+      "learning_rate": 8.589374065736445e-06,
+      "loss": 0.2198,
+      "step": 11006
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9447363417700312,
+      "learning_rate": 8.589099756361136e-06,
+      "loss": 0.4906,
+      "step": 11007
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.294458937665054,
+      "learning_rate": 8.588825424698494e-06,
+      "loss": 0.3719,
+      "step": 11008
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9476426887269604,
+      "learning_rate": 8.588551070750218e-06,
+      "loss": 0.3173,
+      "step": 11009
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.23846954273929,
+      "learning_rate": 8.588276694518014e-06,
+      "loss": 0.1688,
+      "step": 11010
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7427006936423635,
+      "learning_rate": 8.588002296003583e-06,
+      "loss": 0.2485,
+      "step": 11011
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8192805406180586,
+      "learning_rate": 8.587727875208632e-06,
+      "loss": 0.3805,
+      "step": 11012
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.528999104747797,
+      "learning_rate": 8.587453432134865e-06,
+      "loss": 0.2547,
+      "step": 11013
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1106304882883027,
+      "learning_rate": 8.587178966783983e-06,
+      "loss": 0.2918,
+      "step": 11014
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.727604745637254,
+      "learning_rate": 8.586904479157694e-06,
+      "loss": 0.4205,
+      "step": 11015
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.615227216084309,
+      "learning_rate": 8.5866299692577e-06,
+      "loss": 0.2795,
+      "step": 11016
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.164890693739273,
+      "learning_rate": 8.586355437085708e-06,
+      "loss": 0.2762,
+      "step": 11017
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9026169453461632,
+      "learning_rate": 8.58608088264342e-06,
+      "loss": 0.415,
+      "step": 11018
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.202727840012735,
+      "learning_rate": 8.585806305932542e-06,
+      "loss": 0.7043,
+      "step": 11019
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7400850473849498,
+      "learning_rate": 8.585531706954781e-06,
+      "loss": 0.3299,
+      "step": 11020
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.563712360201995,
+      "learning_rate": 8.58525708571184e-06,
+      "loss": 0.4431,
+      "step": 11021
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.019549094883127,
+      "learning_rate": 8.584982442205423e-06,
+      "loss": 0.3932,
+      "step": 11022
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8480772421919478,
+      "learning_rate": 8.58470777643724e-06,
+      "loss": 0.407,
+      "step": 11023
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4504907531861042,
+      "learning_rate": 8.584433088408992e-06,
+      "loss": 0.3117,
+      "step": 11024
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3763157147033533,
+      "learning_rate": 8.584158378122387e-06,
+      "loss": 0.2286,
+      "step": 11025
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3620579923450935,
+      "learning_rate": 8.583883645579132e-06,
+      "loss": 0.2132,
+      "step": 11026
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.914055065218441,
+      "learning_rate": 8.58360889078093e-06,
+      "loss": 0.2195,
+      "step": 11027
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2682303739968614,
+      "learning_rate": 8.583334113729489e-06,
+      "loss": 0.4161,
+      "step": 11028
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6692479630945327,
+      "learning_rate": 8.583059314426515e-06,
+      "loss": 0.3184,
+      "step": 11029
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6535387001977842,
+      "learning_rate": 8.582784492873715e-06,
+      "loss": 0.2923,
+      "step": 11030
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7683383718784949,
+      "learning_rate": 8.582509649072793e-06,
+      "loss": 0.5085,
+      "step": 11031
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7924020095899538,
+      "learning_rate": 8.582234783025461e-06,
+      "loss": 0.1818,
+      "step": 11032
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3622349040982284,
+      "learning_rate": 8.58195989473342e-06,
+      "loss": 0.2426,
+      "step": 11033
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3138051192875828,
+      "learning_rate": 8.58168498419838e-06,
+      "loss": 0.2147,
+      "step": 11034
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.427709921067752,
+      "learning_rate": 8.581410051422048e-06,
+      "loss": 0.1788,
+      "step": 11035
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3800804768727721,
+      "learning_rate": 8.58113509640613e-06,
+      "loss": 0.3002,
+      "step": 11036
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5132219064151513,
+      "learning_rate": 8.580860119152335e-06,
+      "loss": 0.197,
+      "step": 11037
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3989201464048546,
+      "learning_rate": 8.58058511966237e-06,
+      "loss": 0.3545,
+      "step": 11038
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6369332466598936,
+      "learning_rate": 8.580310097937943e-06,
+      "loss": 0.5135,
+      "step": 11039
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7362938202772167,
+      "learning_rate": 8.580035053980759e-06,
+      "loss": 0.3877,
+      "step": 11040
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.763199380673492,
+      "learning_rate": 8.57975998779253e-06,
+      "loss": 0.4537,
+      "step": 11041
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7158394510138644,
+      "learning_rate": 8.579484899374963e-06,
+      "loss": 0.2049,
+      "step": 11042
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5267443435019197,
+      "learning_rate": 8.579209788729766e-06,
+      "loss": 0.18,
+      "step": 11043
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9992389446511422,
+      "learning_rate": 8.578934655858644e-06,
+      "loss": 0.2262,
+      "step": 11044
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.759275886901371,
+      "learning_rate": 8.578659500763311e-06,
+      "loss": 0.5989,
+      "step": 11045
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8925029653106447,
+      "learning_rate": 8.578384323445472e-06,
+      "loss": 0.447,
+      "step": 11046
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3189140664659782,
+      "learning_rate": 8.578109123906837e-06,
+      "loss": 0.2161,
+      "step": 11047
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.417216308872928,
+      "learning_rate": 8.577833902149115e-06,
+      "loss": 0.3582,
+      "step": 11048
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1724425546837205,
+      "learning_rate": 8.577558658174015e-06,
+      "loss": 0.2878,
+      "step": 11049
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.38302699542954,
+      "learning_rate": 8.577283391983245e-06,
+      "loss": 0.3122,
+      "step": 11050
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.480838040112901,
+      "learning_rate": 8.577008103578516e-06,
+      "loss": 0.2545,
+      "step": 11051
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.981271268928999,
+      "learning_rate": 8.576732792961536e-06,
+      "loss": 0.6746,
+      "step": 11052
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3698598140063132,
+      "learning_rate": 8.576457460134015e-06,
+      "loss": 0.3435,
+      "step": 11053
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6232283475852938,
+      "learning_rate": 8.576182105097666e-06,
+      "loss": 0.3206,
+      "step": 11054
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6178209566098316,
+      "learning_rate": 8.575906727854194e-06,
+      "loss": 0.3974,
+      "step": 11055
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4314299021123125,
+      "learning_rate": 8.575631328405312e-06,
+      "loss": 0.2503,
+      "step": 11056
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.932879934622485,
+      "learning_rate": 8.575355906752729e-06,
+      "loss": 0.5184,
+      "step": 11057
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6833898770479483,
+      "learning_rate": 8.575080462898153e-06,
+      "loss": 0.5157,
+      "step": 11058
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1186429735878973,
+      "learning_rate": 8.5748049968433e-06,
+      "loss": 0.2467,
+      "step": 11059
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6535984623248994,
+      "learning_rate": 8.574529508589876e-06,
+      "loss": 0.1941,
+      "step": 11060
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5618855274632795,
+      "learning_rate": 8.574253998139595e-06,
+      "loss": 0.6133,
+      "step": 11061
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6384270227072883,
+      "learning_rate": 8.573978465494164e-06,
+      "loss": 0.236,
+      "step": 11062
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6925477367413826,
+      "learning_rate": 8.573702910655296e-06,
+      "loss": 0.2529,
+      "step": 11063
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3479750739856053,
+      "learning_rate": 8.573427333624703e-06,
+      "loss": 0.317,
+      "step": 11064
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4445572000754647,
+      "learning_rate": 8.573151734404095e-06,
+      "loss": 0.2422,
+      "step": 11065
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1775406445637975,
+      "learning_rate": 8.572876112995183e-06,
+      "loss": 0.2254,
+      "step": 11066
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.252744956904111,
+      "learning_rate": 8.57260046939968e-06,
+      "loss": 0.184,
+      "step": 11067
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.144288939629744,
+      "learning_rate": 8.572324803619298e-06,
+      "loss": 0.3237,
+      "step": 11068
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.985561391659927,
+      "learning_rate": 8.572049115655746e-06,
+      "loss": 0.6111,
+      "step": 11069
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1769662078955765,
+      "learning_rate": 8.571773405510741e-06,
+      "loss": 0.2417,
+      "step": 11070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.530721376121696,
+      "learning_rate": 8.571497673185988e-06,
+      "loss": 0.3122,
+      "step": 11071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.940346915226006,
+      "learning_rate": 8.571221918683206e-06,
+      "loss": 0.1539,
+      "step": 11072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5867164934856948,
+      "learning_rate": 8.570946142004103e-06,
+      "loss": 0.3605,
+      "step": 11073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.086206411680028,
+      "learning_rate": 8.570670343150393e-06,
+      "loss": 0.7781,
+      "step": 11074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8679205959659388,
+      "learning_rate": 8.570394522123787e-06,
+      "loss": 0.4008,
+      "step": 11075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8130156465905691,
+      "learning_rate": 8.570118678926003e-06,
+      "loss": 0.4287,
+      "step": 11076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8867640361524103,
+      "learning_rate": 8.56984281355875e-06,
+      "loss": 0.3439,
+      "step": 11077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6548060024430855,
+      "learning_rate": 8.569566926023739e-06,
+      "loss": 0.4107,
+      "step": 11078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9567218619716884,
+      "learning_rate": 8.569291016322686e-06,
+      "loss": 0.6375,
+      "step": 11079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0543174999759053,
+      "learning_rate": 8.569015084457303e-06,
+      "loss": 0.3107,
+      "step": 11080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9452340002958517,
+      "learning_rate": 8.568739130429304e-06,
+      "loss": 0.1859,
+      "step": 11081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.598912169225499,
+      "learning_rate": 8.568463154240405e-06,
+      "loss": 0.415,
+      "step": 11082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5821571076333596,
+      "learning_rate": 8.568187155892318e-06,
+      "loss": 0.1187,
+      "step": 11083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3097208620676815,
+      "learning_rate": 8.567911135386754e-06,
+      "loss": 0.1675,
+      "step": 11084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4732386755266118,
+      "learning_rate": 8.56763509272543e-06,
+      "loss": 0.2581,
+      "step": 11085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2068608836583627,
+      "learning_rate": 8.56735902791006e-06,
+      "loss": 0.173,
+      "step": 11086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0634649143711665,
+      "learning_rate": 8.567082940942359e-06,
+      "loss": 0.6161,
+      "step": 11087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4292924340349311,
+      "learning_rate": 8.56680683182404e-06,
+      "loss": 0.4075,
+      "step": 11088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3021462991907768,
+      "learning_rate": 8.566530700556816e-06,
+      "loss": 0.2005,
+      "step": 11089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7603007951858936,
+      "learning_rate": 8.566254547142407e-06,
+      "loss": 0.2502,
+      "step": 11090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4837280535686754,
+      "learning_rate": 8.565978371582523e-06,
+      "loss": 0.224,
+      "step": 11091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6901702196902157,
+      "learning_rate": 8.565702173878878e-06,
+      "loss": 0.3605,
+      "step": 11092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5643409636411014,
+      "learning_rate": 8.565425954033193e-06,
+      "loss": 0.4018,
+      "step": 11093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2690736459427174,
+      "learning_rate": 8.565149712047177e-06,
+      "loss": 0.1999,
+      "step": 11094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4320569597465989,
+      "learning_rate": 8.56487344792255e-06,
+      "loss": 0.2129,
+      "step": 11095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5874761859354163,
+      "learning_rate": 8.564597161661027e-06,
+      "loss": 0.4215,
+      "step": 11096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1769035502367589,
+      "learning_rate": 8.564320853264318e-06,
+      "loss": 0.1742,
+      "step": 11097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.637747807534325,
+      "learning_rate": 8.564044522734147e-06,
+      "loss": 0.3456,
+      "step": 11098
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.54843000079128,
+      "learning_rate": 8.563768170072225e-06,
+      "loss": 0.4578,
+      "step": 11099
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7517635621397465,
+      "learning_rate": 8.56349179528027e-06,
+      "loss": 0.3919,
+      "step": 11100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4552786055327074,
+      "learning_rate": 8.563215398359996e-06,
+      "loss": 0.2162,
+      "step": 11101
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3058459455533868,
+      "learning_rate": 8.562938979313122e-06,
+      "loss": 0.3822,
+      "step": 11102
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3289330100225096,
+      "learning_rate": 8.562662538141363e-06,
+      "loss": 0.2942,
+      "step": 11103
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.027495516349924,
+      "learning_rate": 8.562386074846435e-06,
+      "loss": 0.5383,
+      "step": 11104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0602533538176804,
+      "learning_rate": 8.562109589430054e-06,
+      "loss": 0.4561,
+      "step": 11105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.06443162255471,
+      "learning_rate": 8.561833081893942e-06,
+      "loss": 0.4871,
+      "step": 11106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4900546034226112,
+      "learning_rate": 8.56155655223981e-06,
+      "loss": 0.1503,
+      "step": 11107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6440818947425908,
+      "learning_rate": 8.561280000469379e-06,
+      "loss": 0.3141,
+      "step": 11108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.53814150513075,
+      "learning_rate": 8.561003426584364e-06,
+      "loss": 0.2664,
+      "step": 11109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.31566615769519,
+      "learning_rate": 8.560726830586485e-06,
+      "loss": 0.2588,
+      "step": 11110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2147835750334333,
+      "learning_rate": 8.560450212477458e-06,
+      "loss": 0.2014,
+      "step": 11111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5024803833466382,
+      "learning_rate": 8.560173572259e-06,
+      "loss": 0.2941,
+      "step": 11112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.340799430134437,
+      "learning_rate": 8.55989690993283e-06,
+      "loss": 0.2199,
+      "step": 11113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7516776286120037,
+      "learning_rate": 8.559620225500665e-06,
+      "loss": 0.5388,
+      "step": 11114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.639883916234818,
+      "learning_rate": 8.559343518964224e-06,
+      "loss": 0.3077,
+      "step": 11115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0377244322568266,
+      "learning_rate": 8.559066790325226e-06,
+      "loss": 0.6517,
+      "step": 11116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5858669711624072,
+      "learning_rate": 8.558790039585388e-06,
+      "loss": 0.2131,
+      "step": 11117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.627796201948398,
+      "learning_rate": 8.558513266746428e-06,
+      "loss": 0.3405,
+      "step": 11118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8499242641437252,
+      "learning_rate": 8.558236471810068e-06,
+      "loss": 0.5593,
+      "step": 11119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.161182814966405,
+      "learning_rate": 8.557959654778023e-06,
+      "loss": 0.5494,
+      "step": 11120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7406745764899207,
+      "learning_rate": 8.557682815652013e-06,
+      "loss": 0.4996,
+      "step": 11121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5013873767252124,
+      "learning_rate": 8.55740595443376e-06,
+      "loss": 0.2479,
+      "step": 11122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9927250903459117,
+      "learning_rate": 8.55712907112498e-06,
+      "loss": 0.531,
+      "step": 11123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1749277962023617,
+      "learning_rate": 8.556852165727391e-06,
+      "loss": 0.2414,
+      "step": 11124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.044521846465371,
+      "learning_rate": 8.556575238242716e-06,
+      "loss": 0.2968,
+      "step": 11125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.888910110896312,
+      "learning_rate": 8.556298288672673e-06,
+      "loss": 0.4732,
+      "step": 11126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.603380009332952,
+      "learning_rate": 8.556021317018984e-06,
+      "loss": 0.3907,
+      "step": 11127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4928143118483401,
+      "learning_rate": 8.555744323283364e-06,
+      "loss": 0.2937,
+      "step": 11128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.784013033612259,
+      "learning_rate": 8.555467307467538e-06,
+      "loss": 0.2489,
+      "step": 11129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2156501053397193,
+      "learning_rate": 8.555190269573223e-06,
+      "loss": 0.1962,
+      "step": 11130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6261865841055752,
+      "learning_rate": 8.554913209602142e-06,
+      "loss": 0.3422,
+      "step": 11131
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9860189003998234,
+      "learning_rate": 8.554636127556012e-06,
+      "loss": 0.336,
+      "step": 11132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.720387996087617,
+      "learning_rate": 8.55435902343656e-06,
+      "loss": 0.3738,
+      "step": 11133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4857216022339628,
+      "learning_rate": 8.554081897245499e-06,
+      "loss": 0.2506,
+      "step": 11134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4994663540963096,
+      "learning_rate": 8.553804748984554e-06,
+      "loss": 0.3383,
+      "step": 11135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1251929455002387,
+      "learning_rate": 8.553527578655444e-06,
+      "loss": 0.6597,
+      "step": 11136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6122453830008574,
+      "learning_rate": 8.553250386259895e-06,
+      "loss": 0.402,
+      "step": 11137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4805365478081027,
+      "learning_rate": 8.55297317179962e-06,
+      "loss": 0.2628,
+      "step": 11138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1699639218883195,
+      "learning_rate": 8.55269593527635e-06,
+      "loss": 0.1972,
+      "step": 11139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5519447934214863,
+      "learning_rate": 8.552418676691798e-06,
+      "loss": 0.4158,
+      "step": 11140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8738081173746133,
+      "learning_rate": 8.552141396047693e-06,
+      "loss": 0.3216,
+      "step": 11141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1871799348232164,
+      "learning_rate": 8.55186409334575e-06,
+      "loss": 0.1534,
+      "step": 11142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6407095705669277,
+      "learning_rate": 8.551586768587696e-06,
+      "loss": 0.3305,
+      "step": 11143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2569950509890044,
+      "learning_rate": 8.551309421775252e-06,
+      "loss": 0.2064,
+      "step": 11144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.918096433223576,
+      "learning_rate": 8.551032052910138e-06,
+      "loss": 0.4327,
+      "step": 11145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.105270859472287,
+      "learning_rate": 8.550754661994079e-06,
+      "loss": 0.3242,
+      "step": 11146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3342657002034042,
+      "learning_rate": 8.550477249028797e-06,
+      "loss": 0.3603,
+      "step": 11147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6259831150598778,
+      "learning_rate": 8.550199814016012e-06,
+      "loss": 0.3415,
+      "step": 11148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.951595245190456,
+      "learning_rate": 8.549922356957452e-06,
+      "loss": 0.6113,
+      "step": 11149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.703170607332403,
+      "learning_rate": 8.549644877854835e-06,
+      "loss": 0.1596,
+      "step": 11150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5954200278723556,
+      "learning_rate": 8.549367376709885e-06,
+      "loss": 0.3716,
+      "step": 11151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2604070996743717,
+      "learning_rate": 8.549089853524329e-06,
+      "loss": 0.1848,
+      "step": 11152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.778029877409484,
+      "learning_rate": 8.548812308299886e-06,
+      "loss": 0.5352,
+      "step": 11153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1257808824796123,
+      "learning_rate": 8.548534741038281e-06,
+      "loss": 0.3877,
+      "step": 11154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.491229211461704,
+      "learning_rate": 8.548257151741238e-06,
+      "loss": 0.3428,
+      "step": 11155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.731617989403989,
+      "learning_rate": 8.54797954041048e-06,
+      "loss": 0.3089,
+      "step": 11156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9513694579136347,
+      "learning_rate": 8.547701907047733e-06,
+      "loss": 0.3836,
+      "step": 11157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.141288792795897,
+      "learning_rate": 8.547424251654716e-06,
+      "loss": 0.1123,
+      "step": 11158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.683086301087436,
+      "learning_rate": 8.547146574233158e-06,
+      "loss": 0.3593,
+      "step": 11159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9443652200146342,
+      "learning_rate": 8.546868874784783e-06,
+      "loss": 0.3346,
+      "step": 11160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0908233536666783,
+      "learning_rate": 8.546591153311313e-06,
+      "loss": 0.1498,
+      "step": 11161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.869181380430486,
+      "learning_rate": 8.546313409814474e-06,
+      "loss": 0.1809,
+      "step": 11162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9746230490010128,
+      "learning_rate": 8.54603564429599e-06,
+      "loss": 0.2832,
+      "step": 11163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2559959083782528,
+      "learning_rate": 8.545757856757588e-06,
+      "loss": 0.1918,
+      "step": 11164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2609379067591657,
+      "learning_rate": 8.545480047200989e-06,
+      "loss": 0.2662,
+      "step": 11165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3580718028270315,
+      "learning_rate": 8.545202215627923e-06,
+      "loss": 0.2319,
+      "step": 11166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.047025012568836,
+      "learning_rate": 8.544924362040112e-06,
+      "loss": 0.5064,
+      "step": 11167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0022773777815024,
+      "learning_rate": 8.54464648643928e-06,
+      "loss": 0.3084,
+      "step": 11168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.471453455019622,
+      "learning_rate": 8.544368588827157e-06,
+      "loss": 0.4111,
+      "step": 11169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4691367147526004,
+      "learning_rate": 8.544090669205467e-06,
+      "loss": 0.3244,
+      "step": 11170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8693528607931356,
+      "learning_rate": 8.543812727575934e-06,
+      "loss": 0.3487,
+      "step": 11171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.610874306320819,
+      "learning_rate": 8.543534763940284e-06,
+      "loss": 0.3126,
+      "step": 11172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3459959389811662,
+      "learning_rate": 8.543256778300245e-06,
+      "loss": 0.2912,
+      "step": 11173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1563385811910227,
+      "learning_rate": 8.542978770657542e-06,
+      "loss": 0.1917,
+      "step": 11174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.838402752765947,
+      "learning_rate": 8.542700741013903e-06,
+      "loss": 0.4978,
+      "step": 11175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9532204108775795,
+      "learning_rate": 8.542422689371052e-06,
+      "loss": 0.4878,
+      "step": 11176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5297485259510786,
+      "learning_rate": 8.542144615730717e-06,
+      "loss": 0.3029,
+      "step": 11177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.458648892020801,
+      "learning_rate": 8.541866520094625e-06,
+      "loss": 0.2039,
+      "step": 11178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5374323615842715,
+      "learning_rate": 8.541588402464502e-06,
+      "loss": 0.3472,
+      "step": 11179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4580198390947583,
+      "learning_rate": 8.541310262842077e-06,
+      "loss": 0.1796,
+      "step": 11180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.41358538476622,
+      "learning_rate": 8.541032101229074e-06,
+      "loss": 0.163,
+      "step": 11181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4694148344565265,
+      "learning_rate": 8.540753917627222e-06,
+      "loss": 0.2585,
+      "step": 11182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9851488722026451,
+      "learning_rate": 8.54047571203825e-06,
+      "loss": 0.5444,
+      "step": 11183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.60600770132147,
+      "learning_rate": 8.540197484463882e-06,
+      "loss": 0.3359,
+      "step": 11184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5269376673468609,
+      "learning_rate": 8.53991923490585e-06,
+      "loss": 0.2187,
+      "step": 11185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5377469924459484,
+      "learning_rate": 8.539640963365878e-06,
+      "loss": 0.3352,
+      "step": 11186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9670137060475263,
+      "learning_rate": 8.539362669845695e-06,
+      "loss": 0.5732,
+      "step": 11187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6277581109041277,
+      "learning_rate": 8.539084354347032e-06,
+      "loss": 0.2896,
+      "step": 11188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.362146602195192,
+      "learning_rate": 8.538806016871612e-06,
+      "loss": 0.1858,
+      "step": 11189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5911328129982247,
+      "learning_rate": 8.53852765742117e-06,
+      "loss": 0.3474,
+      "step": 11190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3068569963673013,
+      "learning_rate": 8.538249275997427e-06,
+      "loss": 0.3736,
+      "step": 11191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2591355367085402,
+      "learning_rate": 8.537970872602119e-06,
+      "loss": 0.3001,
+      "step": 11192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.668752544201115,
+      "learning_rate": 8.53769244723697e-06,
+      "loss": 0.4231,
+      "step": 11193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7114403913149119,
+      "learning_rate": 8.53741399990371e-06,
+      "loss": 0.3622,
+      "step": 11194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.013315089647844,
+      "learning_rate": 8.537135530604068e-06,
+      "loss": 0.495,
+      "step": 11195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2714635318832297,
+      "learning_rate": 8.536857039339775e-06,
+      "loss": 0.2063,
+      "step": 11196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7832056068495061,
+      "learning_rate": 8.536578526112559e-06,
+      "loss": 0.4864,
+      "step": 11197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.695943830329464,
+      "learning_rate": 8.536299990924149e-06,
+      "loss": 0.1855,
+      "step": 11198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6886836245443733,
+      "learning_rate": 8.536021433776275e-06,
+      "loss": 0.384,
+      "step": 11199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7499926515833657,
+      "learning_rate": 8.535742854670667e-06,
+      "loss": 0.3834,
+      "step": 11200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.648602409611211,
+      "learning_rate": 8.535464253609054e-06,
+      "loss": 0.3568,
+      "step": 11201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7193540552015194,
+      "learning_rate": 8.535185630593168e-06,
+      "loss": 0.2016,
+      "step": 11202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4789587891158789,
+      "learning_rate": 8.534906985624737e-06,
+      "loss": 0.2434,
+      "step": 11203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9333234810030036,
+      "learning_rate": 8.534628318705494e-06,
+      "loss": 0.3232,
+      "step": 11204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.741960626207837,
+      "learning_rate": 8.534349629837167e-06,
+      "loss": 0.4746,
+      "step": 11205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7783094381704831,
+      "learning_rate": 8.534070919021486e-06,
+      "loss": 0.2513,
+      "step": 11206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2621733977701237,
+      "learning_rate": 8.533792186260186e-06,
+      "loss": 0.5791,
+      "step": 11207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8493753464425482,
+      "learning_rate": 8.533513431554992e-06,
+      "loss": 0.5593,
+      "step": 11208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8025656300016484,
+      "learning_rate": 8.533234654907638e-06,
+      "loss": 0.1506,
+      "step": 11209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9618696676940548,
+      "learning_rate": 8.532955856319857e-06,
+      "loss": 0.1126,
+      "step": 11210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7129082064449443,
+      "learning_rate": 8.532677035793377e-06,
+      "loss": 0.3502,
+      "step": 11211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6233997076223208,
+      "learning_rate": 8.532398193329932e-06,
+      "loss": 0.2528,
+      "step": 11212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5076521976031814,
+      "learning_rate": 8.53211932893125e-06,
+      "loss": 0.2396,
+      "step": 11213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0233799694233796,
+      "learning_rate": 8.531840442599067e-06,
+      "loss": 0.8024,
+      "step": 11214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7878084285111617,
+      "learning_rate": 8.531561534335112e-06,
+      "loss": 0.5388,
+      "step": 11215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5130828843603603,
+      "learning_rate": 8.531282604141116e-06,
+      "loss": 0.3248,
+      "step": 11216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.368844992137079,
+      "learning_rate": 8.531003652018814e-06,
+      "loss": 0.2723,
+      "step": 11217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9416064057511837,
+      "learning_rate": 8.530724677969938e-06,
+      "loss": 0.4277,
+      "step": 11218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6917576996395645,
+      "learning_rate": 8.530445681996218e-06,
+      "loss": 0.3303,
+      "step": 11219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.576765362303185,
+      "learning_rate": 8.530166664099388e-06,
+      "loss": 0.2581,
+      "step": 11220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5290314065175767,
+      "learning_rate": 8.52988762428118e-06,
+      "loss": 0.4091,
+      "step": 11221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7781174736249155,
+      "learning_rate": 8.52960856254333e-06,
+      "loss": 0.2057,
+      "step": 11222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.976213620422247,
+      "learning_rate": 8.529329478887565e-06,
+      "loss": 0.3453,
+      "step": 11223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9711988265654038,
+      "learning_rate": 8.529050373315622e-06,
+      "loss": 0.101,
+      "step": 11224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0171971916873814,
+      "learning_rate": 8.528771245829234e-06,
+      "loss": 0.1059,
+      "step": 11225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9669492600019636,
+      "learning_rate": 8.528492096430133e-06,
+      "loss": 0.1741,
+      "step": 11226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6877258820706997,
+      "learning_rate": 8.528212925120054e-06,
+      "loss": 0.465,
+      "step": 11227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.476407219527862,
+      "learning_rate": 8.527933731900728e-06,
+      "loss": 0.2897,
+      "step": 11228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4769640532954178,
+      "learning_rate": 8.527654516773892e-06,
+      "loss": 0.3858,
+      "step": 11229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7844194440100192,
+      "learning_rate": 8.527375279741276e-06,
+      "loss": 0.3352,
+      "step": 11230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.126419697576828,
+      "learning_rate": 8.527096020804617e-06,
+      "loss": 0.3084,
+      "step": 11231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9043844427858063,
+      "learning_rate": 8.526816739965648e-06,
+      "loss": 0.5117,
+      "step": 11232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3699010024535103,
+      "learning_rate": 8.526537437226105e-06,
+      "loss": 0.2096,
+      "step": 11233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6063254268169587,
+      "learning_rate": 8.52625811258772e-06,
+      "loss": 0.425,
+      "step": 11234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7019970509742264,
+      "learning_rate": 8.52597876605223e-06,
+      "loss": 0.2817,
+      "step": 11235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7228064482382124,
+      "learning_rate": 8.525699397621365e-06,
+      "loss": 0.4155,
+      "step": 11236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2863633941903057,
+      "learning_rate": 8.525420007296865e-06,
+      "loss": 0.1797,
+      "step": 11237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9359789038010697,
+      "learning_rate": 8.525140595080463e-06,
+      "loss": 0.5331,
+      "step": 11238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4714915757171432,
+      "learning_rate": 8.524861160973894e-06,
+      "loss": 0.1867,
+      "step": 11239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.802448560017757,
+      "learning_rate": 8.524581704978892e-06,
+      "loss": 0.4409,
+      "step": 11240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9718908127938126,
+      "learning_rate": 8.524302227097195e-06,
+      "loss": 0.2022,
+      "step": 11241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8500152110428127,
+      "learning_rate": 8.524022727330536e-06,
+      "loss": 0.5309,
+      "step": 11242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0725153404697065,
+      "learning_rate": 8.523743205680652e-06,
+      "loss": 0.1835,
+      "step": 11243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.285444355596688,
+      "learning_rate": 8.523463662149278e-06,
+      "loss": 0.7687,
+      "step": 11244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8164826653926824,
+      "learning_rate": 8.52318409673815e-06,
+      "loss": 0.4615,
+      "step": 11245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2051481917490356,
+      "learning_rate": 8.522904509449005e-06,
+      "loss": 0.1916,
+      "step": 11246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4578502107385372,
+      "learning_rate": 8.522624900283579e-06,
+      "loss": 0.3224,
+      "step": 11247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5078240453446083,
+      "learning_rate": 8.522345269243606e-06,
+      "loss": 0.3902,
+      "step": 11248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3746222497326834,
+      "learning_rate": 8.522065616330824e-06,
+      "loss": 0.3302,
+      "step": 11249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4443590808430224,
+      "learning_rate": 8.521785941546972e-06,
+      "loss": 0.3626,
+      "step": 11250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0250751591376344,
+      "learning_rate": 8.521506244893782e-06,
+      "loss": 0.1773,
+      "step": 11251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9565197281853046,
+      "learning_rate": 8.521226526372996e-06,
+      "loss": 0.493,
+      "step": 11252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2777911884105433,
+      "learning_rate": 8.520946785986347e-06,
+      "loss": 0.1992,
+      "step": 11253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6111539665661914,
+      "learning_rate": 8.520667023735572e-06,
+      "loss": 0.3496,
+      "step": 11254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5949781538250487,
+      "learning_rate": 8.52038723962241e-06,
+      "loss": 0.4288,
+      "step": 11255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0383726264206208,
+      "learning_rate": 8.5201074336486e-06,
+      "loss": 0.1916,
+      "step": 11256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.430564371743345,
+      "learning_rate": 8.519827605815875e-06,
+      "loss": 0.3654,
+      "step": 11257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6393307576334117,
+      "learning_rate": 8.519547756125977e-06,
+      "loss": 0.2697,
+      "step": 11258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7782949312798442,
+      "learning_rate": 8.519267884580642e-06,
+      "loss": 0.3866,
+      "step": 11259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9662489710318265,
+      "learning_rate": 8.518987991181606e-06,
+      "loss": 0.6568,
+      "step": 11260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.157350661791023,
+      "learning_rate": 8.51870807593061e-06,
+      "loss": 0.3564,
+      "step": 11261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.221059074493462,
+      "learning_rate": 8.518428138829392e-06,
+      "loss": 0.4496,
+      "step": 11262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.463706597495862,
+      "learning_rate": 8.51814817987969e-06,
+      "loss": 0.1253,
+      "step": 11263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5117642201132875,
+      "learning_rate": 8.51786819908324e-06,
+      "loss": 0.3754,
+      "step": 11264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5200994159944299,
+      "learning_rate": 8.517588196441784e-06,
+      "loss": 0.3193,
+      "step": 11265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1192918478902858,
+      "learning_rate": 8.517308171957059e-06,
+      "loss": 0.2557,
+      "step": 11266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0516509291975975,
+      "learning_rate": 8.517028125630802e-06,
+      "loss": 0.4275,
+      "step": 11267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.476887432555385,
+      "learning_rate": 8.516748057464757e-06,
+      "loss": 0.2118,
+      "step": 11268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5367515370411073,
+      "learning_rate": 8.51646796746066e-06,
+      "loss": 0.2079,
+      "step": 11269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5124849945301335,
+      "learning_rate": 8.51618785562025e-06,
+      "loss": 0.3233,
+      "step": 11270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.90812930117565,
+      "learning_rate": 8.515907721945266e-06,
+      "loss": 0.4314,
+      "step": 11271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4880645198660107,
+      "learning_rate": 8.51562756643745e-06,
+      "loss": 0.2324,
+      "step": 11272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.032621767770726,
+      "learning_rate": 8.51534738909854e-06,
+      "loss": 0.3769,
+      "step": 11273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6218219959734148,
+      "learning_rate": 8.515067189930276e-06,
+      "loss": 0.4567,
+      "step": 11274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.41874889712984,
+      "learning_rate": 8.514786968934399e-06,
+      "loss": 0.2075,
+      "step": 11275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.244972868168137,
+      "learning_rate": 8.514506726112646e-06,
+      "loss": 0.3963,
+      "step": 11276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.773275153673721,
+      "learning_rate": 8.514226461466762e-06,
+      "loss": 0.557,
+      "step": 11277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3984812817566243,
+      "learning_rate": 8.513946174998482e-06,
+      "loss": 0.1994,
+      "step": 11278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.907947031131333,
+      "learning_rate": 8.51366586670955e-06,
+      "loss": 0.5062,
+      "step": 11279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7276818069302387,
+      "learning_rate": 8.513385536601707e-06,
+      "loss": 0.4415,
+      "step": 11280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.168929064071276,
+      "learning_rate": 8.513105184676693e-06,
+      "loss": 0.1211,
+      "step": 11281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.082178851365436,
+      "learning_rate": 8.512824810936248e-06,
+      "loss": 0.2735,
+      "step": 11282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.148381458913308,
+      "learning_rate": 8.512544415382114e-06,
+      "loss": 0.1439,
+      "step": 11283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3341562375439988,
+      "learning_rate": 8.51226399801603e-06,
+      "loss": 0.2113,
+      "step": 11284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0220417728508413,
+      "learning_rate": 8.51198355883974e-06,
+      "loss": 0.314,
+      "step": 11285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7578797285151582,
+      "learning_rate": 8.511703097854986e-06,
+      "loss": 0.354,
+      "step": 11286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0370448276903828,
+      "learning_rate": 8.511422615063508e-06,
+      "loss": 0.4249,
+      "step": 11287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1754856730688346,
+      "learning_rate": 8.511142110467046e-06,
+      "loss": 0.3728,
+      "step": 11288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2327733211311187,
+      "learning_rate": 8.510861584067346e-06,
+      "loss": 0.2278,
+      "step": 11289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7347563010851865,
+      "learning_rate": 8.510581035866146e-06,
+      "loss": 0.2959,
+      "step": 11290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.62217300751381,
+      "learning_rate": 8.51030046586519e-06,
+      "loss": 0.2107,
+      "step": 11291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6547027301717798,
+      "learning_rate": 8.510019874066219e-06,
+      "loss": 0.3568,
+      "step": 11292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.779086287899114,
+      "learning_rate": 8.509739260470979e-06,
+      "loss": 0.4226,
+      "step": 11293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9144926950116592,
+      "learning_rate": 8.50945862508121e-06,
+      "loss": 0.5407,
+      "step": 11294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.36117646044318,
+      "learning_rate": 8.509177967898654e-06,
+      "loss": 0.1486,
+      "step": 11295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4336873161694625,
+      "learning_rate": 8.508897288925053e-06,
+      "loss": 0.1991,
+      "step": 11296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5761153153076495,
+      "learning_rate": 8.508616588162152e-06,
+      "loss": 0.1979,
+      "step": 11297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5221724759940232,
+      "learning_rate": 8.508335865611697e-06,
+      "loss": 0.3498,
+      "step": 11298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8468142197704316,
+      "learning_rate": 8.508055121275426e-06,
+      "loss": 0.4057,
+      "step": 11299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5681694813657394,
+      "learning_rate": 8.507774355155081e-06,
+      "loss": 0.3189,
+      "step": 11300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9933217026957346,
+      "learning_rate": 8.507493567252412e-06,
+      "loss": 0.3883,
+      "step": 11301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3265934226416298,
+      "learning_rate": 8.507212757569158e-06,
+      "loss": 0.2136,
+      "step": 11302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8837681003351505,
+      "learning_rate": 8.506931926107062e-06,
+      "loss": 0.5751,
+      "step": 11303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3670359568395496,
+      "learning_rate": 8.506651072867873e-06,
+      "loss": 0.3627,
+      "step": 11304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7917266149989466,
+      "learning_rate": 8.50637019785333e-06,
+      "loss": 0.1967,
+      "step": 11305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9064838156589694,
+      "learning_rate": 8.50608930106518e-06,
+      "loss": 0.3646,
+      "step": 11306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2462421959615453,
+      "learning_rate": 8.505808382505167e-06,
+      "loss": 0.2319,
+      "step": 11307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7370073129374262,
+      "learning_rate": 8.505527442175035e-06,
+      "loss": 0.4105,
+      "step": 11308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8042165030140611,
+      "learning_rate": 8.505246480076525e-06,
+      "loss": 0.3115,
+      "step": 11309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4988145552242869,
+      "learning_rate": 8.504965496211387e-06,
+      "loss": 0.2045,
+      "step": 11310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7177493303525548,
+      "learning_rate": 8.504684490581366e-06,
+      "loss": 0.2287,
+      "step": 11311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.778670023636829,
+      "learning_rate": 8.504403463188203e-06,
+      "loss": 0.3641,
+      "step": 11312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5084861142595265,
+      "learning_rate": 8.504122414033644e-06,
+      "loss": 0.2498,
+      "step": 11313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4993294667544925,
+      "learning_rate": 8.503841343119437e-06,
+      "loss": 0.4841,
+      "step": 11314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1565243090420987,
+      "learning_rate": 8.503560250447324e-06,
+      "loss": 0.5594,
+      "step": 11315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4925610389320547,
+      "learning_rate": 8.50327913601905e-06,
+      "loss": 0.2341,
+      "step": 11316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4455153271402854,
+      "learning_rate": 8.502997999836366e-06,
+      "loss": 0.2708,
+      "step": 11317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4887907452696787,
+      "learning_rate": 8.502716841901015e-06,
+      "loss": 0.262,
+      "step": 11318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.221692067669804,
+      "learning_rate": 8.502435662214742e-06,
+      "loss": 0.2332,
+      "step": 11319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6958019982621086,
+      "learning_rate": 8.50215446077929e-06,
+      "loss": 0.3444,
+      "step": 11320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6782834727655518,
+      "learning_rate": 8.501873237596411e-06,
+      "loss": 0.471,
+      "step": 11321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7077642041328915,
+      "learning_rate": 8.50159199266785e-06,
+      "loss": 0.2211,
+      "step": 11322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9540004904257355,
+      "learning_rate": 8.501310725995353e-06,
+      "loss": 0.6448,
+      "step": 11323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5390131889228011,
+      "learning_rate": 8.501029437580663e-06,
+      "loss": 0.3985,
+      "step": 11324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8868860995865537,
+      "learning_rate": 8.50074812742553e-06,
+      "loss": 0.2599,
+      "step": 11325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.871103872060773,
+      "learning_rate": 8.500466795531704e-06,
+      "loss": 0.4627,
+      "step": 11326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3686464563379899,
+      "learning_rate": 8.500185441900925e-06,
+      "loss": 0.3335,
+      "step": 11327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1732870360709444,
+      "learning_rate": 8.499904066534945e-06,
+      "loss": 0.1623,
+      "step": 11328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5699442839796929,
+      "learning_rate": 8.49962266943551e-06,
+      "loss": 0.2666,
+      "step": 11329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7281565665040255,
+      "learning_rate": 8.499341250604368e-06,
+      "loss": 0.5091,
+      "step": 11330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6622078997483833,
+      "learning_rate": 8.499059810043265e-06,
+      "loss": 0.3335,
+      "step": 11331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.988063241964176,
+      "learning_rate": 8.49877834775395e-06,
+      "loss": 0.5512,
+      "step": 11332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.458091845840607,
+      "learning_rate": 8.498496863738167e-06,
+      "loss": 0.3496,
+      "step": 11333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6627725786457856,
+      "learning_rate": 8.498215357997673e-06,
+      "loss": 0.3314,
+      "step": 11334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3758082209017006,
+      "learning_rate": 8.497933830534206e-06,
+      "loss": 0.2307,
+      "step": 11335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7511269142630017,
+      "learning_rate": 8.49765228134952e-06,
+      "loss": 0.4573,
+      "step": 11336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9792786009196943,
+      "learning_rate": 8.497370710445363e-06,
+      "loss": 0.2956,
+      "step": 11337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1966397203561,
+      "learning_rate": 8.497089117823481e-06,
+      "loss": 0.2719,
+      "step": 11338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6953463787218375,
+      "learning_rate": 8.496807503485624e-06,
+      "loss": 0.1927,
+      "step": 11339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1156749660892336,
+      "learning_rate": 8.496525867433543e-06,
+      "loss": 0.1615,
+      "step": 11340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7163809705425797,
+      "learning_rate": 8.496244209668983e-06,
+      "loss": 0.2258,
+      "step": 11341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1361702767747617,
+      "learning_rate": 8.495962530193693e-06,
+      "loss": 0.3459,
+      "step": 11342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4372776833988605,
+      "learning_rate": 8.495680829009425e-06,
+      "loss": 0.2871,
+      "step": 11343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.021524128546028,
+      "learning_rate": 8.495399106117928e-06,
+      "loss": 0.5813,
+      "step": 11344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2358658364711403,
+      "learning_rate": 8.49511736152095e-06,
+      "loss": 0.307,
+      "step": 11345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.549236150058096,
+      "learning_rate": 8.494835595220243e-06,
+      "loss": 0.3907,
+      "step": 11346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5525799572533172,
+      "learning_rate": 8.494553807217552e-06,
+      "loss": 0.2049,
+      "step": 11347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6208672606571073,
+      "learning_rate": 8.494271997514632e-06,
+      "loss": 0.2997,
+      "step": 11348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5110237593397597,
+      "learning_rate": 8.493990166113228e-06,
+      "loss": 0.2363,
+      "step": 11349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6317613859742273,
+      "learning_rate": 8.493708313015095e-06,
+      "loss": 0.2296,
+      "step": 11350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0053905916160115,
+      "learning_rate": 8.493426438221979e-06,
+      "loss": 0.5852,
+      "step": 11351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.144753178854375,
+      "learning_rate": 8.493144541735633e-06,
+      "loss": 0.6936,
+      "step": 11352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.570377599970463,
+      "learning_rate": 8.492862623557808e-06,
+      "loss": 0.3284,
+      "step": 11353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.334458752701734,
+      "learning_rate": 8.49258068369025e-06,
+      "loss": 0.1988,
+      "step": 11354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.160305238797359,
+      "learning_rate": 8.492298722134715e-06,
+      "loss": 0.4546,
+      "step": 11355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.813656869366544,
+      "learning_rate": 8.492016738892954e-06,
+      "loss": 0.4459,
+      "step": 11356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2534938309303334,
+      "learning_rate": 8.491734733966714e-06,
+      "loss": 0.2592,
+      "step": 11357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3543687941625115,
+      "learning_rate": 8.491452707357748e-06,
+      "loss": 0.3087,
+      "step": 11358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0379325670076205,
+      "learning_rate": 8.491170659067809e-06,
+      "loss": 0.3956,
+      "step": 11359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1548531402921016,
+      "learning_rate": 8.490888589098647e-06,
+      "loss": 0.2239,
+      "step": 11360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8788919825022106,
+      "learning_rate": 8.490606497452014e-06,
+      "loss": 0.5339,
+      "step": 11361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.22520915612825,
+      "learning_rate": 8.49032438412966e-06,
+      "loss": 0.5599,
+      "step": 11362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.339422389078585,
+      "learning_rate": 8.490042249133339e-06,
+      "loss": 0.2295,
+      "step": 11363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8877388583556582,
+      "learning_rate": 8.4897600924648e-06,
+      "loss": 0.5021,
+      "step": 11364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4995480057584245,
+      "learning_rate": 8.4894779141258e-06,
+      "loss": 0.1985,
+      "step": 11365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2626202569119416,
+      "learning_rate": 8.489195714118087e-06,
+      "loss": 0.2343,
+      "step": 11366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7934628888207715,
+      "learning_rate": 8.488913492443415e-06,
+      "loss": 0.5341,
+      "step": 11367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.83072248776892,
+      "learning_rate": 8.488631249103538e-06,
+      "loss": 0.5499,
+      "step": 11368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5810028124210598,
+      "learning_rate": 8.488348984100206e-06,
+      "loss": 0.3217,
+      "step": 11369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3622414058084122,
+      "learning_rate": 8.488066697435173e-06,
+      "loss": 0.3485,
+      "step": 11370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8124147099125933,
+      "learning_rate": 8.487784389110192e-06,
+      "loss": 0.2188,
+      "step": 11371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1293967362847017,
+      "learning_rate": 8.487502059127015e-06,
+      "loss": 0.1594,
+      "step": 11372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1515443307307494,
+      "learning_rate": 8.487219707487396e-06,
+      "loss": 0.528,
+      "step": 11373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.415790776783889,
+      "learning_rate": 8.48693733419309e-06,
+      "loss": 0.3016,
+      "step": 11374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3677955909992894,
+      "learning_rate": 8.486654939245847e-06,
+      "loss": 0.2713,
+      "step": 11375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.050668715183457,
+      "learning_rate": 8.486372522647424e-06,
+      "loss": 0.3209,
+      "step": 11376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5971777766086062,
+      "learning_rate": 8.486090084399572e-06,
+      "loss": 0.3782,
+      "step": 11377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6391011677191043,
+      "learning_rate": 8.485807624504046e-06,
+      "loss": 0.5521,
+      "step": 11378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.06607730352922,
+      "learning_rate": 8.4855251429626e-06,
+      "loss": 0.365,
+      "step": 11379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7686897259979977,
+      "learning_rate": 8.48524263977699e-06,
+      "loss": 0.2864,
+      "step": 11380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8922747791110284,
+      "learning_rate": 8.484960114948967e-06,
+      "loss": 0.3163,
+      "step": 11381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3229192529738167,
+      "learning_rate": 8.484677568480287e-06,
+      "loss": 0.2516,
+      "step": 11382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7616265625950467,
+      "learning_rate": 8.484395000372706e-06,
+      "loss": 0.2076,
+      "step": 11383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8518114123851879,
+      "learning_rate": 8.484112410627973e-06,
+      "loss": 0.4132,
+      "step": 11384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9580818086183711,
+      "learning_rate": 8.483829799247849e-06,
+      "loss": 0.2724,
+      "step": 11385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4099481090887334,
+      "learning_rate": 8.483547166234088e-06,
+      "loss": 0.3038,
+      "step": 11386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4528793004187361,
+      "learning_rate": 8.483264511588443e-06,
+      "loss": 0.1656,
+      "step": 11387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8612532385445033,
+      "learning_rate": 8.482981835312669e-06,
+      "loss": 0.2777,
+      "step": 11388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9418124267164092,
+      "learning_rate": 8.482699137408524e-06,
+      "loss": 0.3288,
+      "step": 11389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5632592144871011,
+      "learning_rate": 8.48241641787776e-06,
+      "loss": 0.3619,
+      "step": 11390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5065553716701778,
+      "learning_rate": 8.482133676722137e-06,
+      "loss": 0.3917,
+      "step": 11391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7164547333014217,
+      "learning_rate": 8.481850913943408e-06,
+      "loss": 0.2521,
+      "step": 11392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.887279086692866,
+      "learning_rate": 8.481568129543328e-06,
+      "loss": 0.191,
+      "step": 11393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9760253425135688,
+      "learning_rate": 8.481285323523653e-06,
+      "loss": 0.2242,
+      "step": 11394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1047571887208325,
+      "learning_rate": 8.481002495886141e-06,
+      "loss": 0.1798,
+      "step": 11395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.902953036608549,
+      "learning_rate": 8.480719646632548e-06,
+      "loss": 0.3688,
+      "step": 11396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2790034649252466,
+      "learning_rate": 8.480436775764631e-06,
+      "loss": 0.1714,
+      "step": 11397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1879668997716104,
+      "learning_rate": 8.480153883284145e-06,
+      "loss": 0.1843,
+      "step": 11398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.977148361228663,
+      "learning_rate": 8.479870969192846e-06,
+      "loss": 0.2644,
+      "step": 11399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7644115726019538,
+      "learning_rate": 8.479588033492494e-06,
+      "loss": 0.344,
+      "step": 11400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4408996022423164,
+      "learning_rate": 8.479305076184843e-06,
+      "loss": 0.1866,
+      "step": 11401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2016918886560584,
+      "learning_rate": 8.47902209727165e-06,
+      "loss": 0.2947,
+      "step": 11402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8030265579589693,
+      "learning_rate": 8.478739096754675e-06,
+      "loss": 0.5993,
+      "step": 11403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.373193456278589,
+      "learning_rate": 8.478456074635673e-06,
+      "loss": 0.2428,
+      "step": 11404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7325802615594132,
+      "learning_rate": 8.478173030916402e-06,
+      "loss": 0.3645,
+      "step": 11405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5501371626769374,
+      "learning_rate": 8.47788996559862e-06,
+      "loss": 0.2958,
+      "step": 11406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.571733318279923,
+      "learning_rate": 8.477606878684086e-06,
+      "loss": 0.1475,
+      "step": 11407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.737916218009067,
+      "learning_rate": 8.477323770174554e-06,
+      "loss": 0.3058,
+      "step": 11408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.710514127265839,
+      "learning_rate": 8.477040640071785e-06,
+      "loss": 0.278,
+      "step": 11409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.450490095702056,
+      "learning_rate": 8.476757488377538e-06,
+      "loss": 0.1566,
+      "step": 11410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8672984261861945,
+      "learning_rate": 8.476474315093568e-06,
+      "loss": 0.5844,
+      "step": 11411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5663213350106133,
+      "learning_rate": 8.476191120221637e-06,
+      "loss": 0.3902,
+      "step": 11412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3346884999922353,
+      "learning_rate": 8.475907903763501e-06,
+      "loss": 0.2208,
+      "step": 11413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6575786416642289,
+      "learning_rate": 8.475624665720921e-06,
+      "loss": 0.3774,
+      "step": 11414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8406637696909354,
+      "learning_rate": 8.475341406095653e-06,
+      "loss": 0.2222,
+      "step": 11415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5187756074112337,
+      "learning_rate": 8.475058124889456e-06,
+      "loss": 0.3449,
+      "step": 11416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0621360869573526,
+      "learning_rate": 8.474774822104093e-06,
+      "loss": 0.3799,
+      "step": 11417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0061326664878862,
+      "learning_rate": 8.47449149774132e-06,
+      "loss": 0.1631,
+      "step": 11418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6172276321444083,
+      "learning_rate": 8.474208151802898e-06,
+      "loss": 0.2924,
+      "step": 11419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5848270155196085,
+      "learning_rate": 8.473924784290585e-06,
+      "loss": 0.4111,
+      "step": 11420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4491921586941388,
+      "learning_rate": 8.47364139520614e-06,
+      "loss": 0.1794,
+      "step": 11421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.682740920394281,
+      "learning_rate": 8.473357984551326e-06,
+      "loss": 0.3842,
+      "step": 11422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4572972318239907,
+      "learning_rate": 8.4730745523279e-06,
+      "loss": 0.3416,
+      "step": 11423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8442136294752702,
+      "learning_rate": 8.472791098537623e-06,
+      "loss": 0.3579,
+      "step": 11424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4063932425341705,
+      "learning_rate": 8.472507623182256e-06,
+      "loss": 0.2547,
+      "step": 11425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4821769426570228,
+      "learning_rate": 8.472224126263555e-06,
+      "loss": 0.249,
+      "step": 11426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1738101379253596,
+      "learning_rate": 8.471940607783286e-06,
+      "loss": 0.1834,
+      "step": 11427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6489479689253281,
+      "learning_rate": 8.471657067743208e-06,
+      "loss": 0.3732,
+      "step": 11428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7356802007175467,
+      "learning_rate": 8.471373506145082e-06,
+      "loss": 0.3598,
+      "step": 11429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4591365566320513,
+      "learning_rate": 8.471089922990667e-06,
+      "loss": 0.3249,
+      "step": 11430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5182893870882435,
+      "learning_rate": 8.470806318281725e-06,
+      "loss": 0.2924,
+      "step": 11431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6674145173062092,
+      "learning_rate": 8.470522692020016e-06,
+      "loss": 0.1915,
+      "step": 11432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3032232937001926,
+      "learning_rate": 8.470239044207304e-06,
+      "loss": 0.2956,
+      "step": 11433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7570554522979982,
+      "learning_rate": 8.46995537484535e-06,
+      "loss": 0.3828,
+      "step": 11434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0765633623897604,
+      "learning_rate": 8.46967168393591e-06,
+      "loss": 0.1716,
+      "step": 11435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.675090336142859,
+      "learning_rate": 8.469387971480752e-06,
+      "loss": 0.4224,
+      "step": 11436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6855094094383682,
+      "learning_rate": 8.469104237481636e-06,
+      "loss": 0.2703,
+      "step": 11437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4896182565218745,
+      "learning_rate": 8.468820481940323e-06,
+      "loss": 0.3679,
+      "step": 11438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8254168617327087,
+      "learning_rate": 8.468536704858576e-06,
+      "loss": 0.4082,
+      "step": 11439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7399042268030758,
+      "learning_rate": 8.468252906238155e-06,
+      "loss": 0.3605,
+      "step": 11440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7106053713674831,
+      "learning_rate": 8.467969086080827e-06,
+      "loss": 0.3122,
+      "step": 11441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.155101207494068,
+      "learning_rate": 8.467685244388349e-06,
+      "loss": 0.6339,
+      "step": 11442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8475093541211787,
+      "learning_rate": 8.467401381162487e-06,
+      "loss": 0.3491,
+      "step": 11443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.241789916127912,
+      "learning_rate": 8.467117496405004e-06,
+      "loss": 0.2553,
+      "step": 11444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9246389510257393,
+      "learning_rate": 8.466833590117661e-06,
+      "loss": 0.3582,
+      "step": 11445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6704821020543104,
+      "learning_rate": 8.46654966230222e-06,
+      "loss": 0.2211,
+      "step": 11446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8280420284515932,
+      "learning_rate": 8.466265712960449e-06,
+      "loss": 0.4815,
+      "step": 11447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1920290179346287,
+      "learning_rate": 8.465981742094104e-06,
+      "loss": 0.3982,
+      "step": 11448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4816509172045391,
+      "learning_rate": 8.465697749704954e-06,
+      "loss": 0.1743,
+      "step": 11449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.995596291801198,
+      "learning_rate": 8.465413735794762e-06,
+      "loss": 0.5147,
+      "step": 11450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3191770532744447,
+      "learning_rate": 8.465129700365292e-06,
+      "loss": 0.1882,
+      "step": 11451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.362876558764253,
+      "learning_rate": 8.464845643418302e-06,
+      "loss": 0.2252,
+      "step": 11452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9672611508960371,
+      "learning_rate": 8.464561564955563e-06,
+      "loss": 0.1811,
+      "step": 11453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.275972856678831,
+      "learning_rate": 8.464277464978836e-06,
+      "loss": 0.2469,
+      "step": 11454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.736718664885745,
+      "learning_rate": 8.463993343489886e-06,
+      "loss": 0.4845,
+      "step": 11455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.564336050846428,
+      "learning_rate": 8.463709200490476e-06,
+      "loss": 0.3967,
+      "step": 11456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7392318703303309,
+      "learning_rate": 8.46342503598237e-06,
+      "loss": 0.1962,
+      "step": 11457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.651545074555022,
+      "learning_rate": 8.463140849967337e-06,
+      "loss": 0.2714,
+      "step": 11458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0595552200625376,
+      "learning_rate": 8.462856642447136e-06,
+      "loss": 0.3813,
+      "step": 11459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0908982224333927,
+      "learning_rate": 8.462572413423535e-06,
+      "loss": 0.1099,
+      "step": 11460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.042960028328887,
+      "learning_rate": 8.462288162898297e-06,
+      "loss": 0.3108,
+      "step": 11461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0418463964607287,
+      "learning_rate": 8.462003890873192e-06,
+      "loss": 0.1907,
+      "step": 11462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.777949139961001,
+      "learning_rate": 8.46171959734998e-06,
+      "loss": 0.3229,
+      "step": 11463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6678879754948843,
+      "learning_rate": 8.461435282330427e-06,
+      "loss": 0.4674,
+      "step": 11464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8061217217532117,
+      "learning_rate": 8.4611509458163e-06,
+      "loss": 0.2112,
+      "step": 11465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5047270702296736,
+      "learning_rate": 8.460866587809367e-06,
+      "loss": 0.3883,
+      "step": 11466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8049441481241661,
+      "learning_rate": 8.460582208311388e-06,
+      "loss": 0.4765,
+      "step": 11467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9599869236461247,
+      "learning_rate": 8.460297807324132e-06,
+      "loss": 0.5186,
+      "step": 11468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3672703813226328,
+      "learning_rate": 8.460013384849366e-06,
+      "loss": 0.3187,
+      "step": 11469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.453188394886424,
+      "learning_rate": 8.459728940888856e-06,
+      "loss": 0.1583,
+      "step": 11470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5521272804025072,
+      "learning_rate": 8.459444475444367e-06,
+      "loss": 0.3432,
+      "step": 11471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.016749970006842,
+      "learning_rate": 8.459159988517666e-06,
+      "loss": 0.6002,
+      "step": 11472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3070463226416058,
+      "learning_rate": 8.45887548011052e-06,
+      "loss": 0.3596,
+      "step": 11473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4168117324159464,
+      "learning_rate": 8.458590950224694e-06,
+      "loss": 0.1357,
+      "step": 11474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.004877375666423,
+      "learning_rate": 8.458306398861958e-06,
+      "loss": 0.5062,
+      "step": 11475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8556720200675065,
+      "learning_rate": 8.458021826024077e-06,
+      "loss": 0.2375,
+      "step": 11476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.472648511810614,
+      "learning_rate": 8.457737231712817e-06,
+      "loss": 0.1958,
+      "step": 11477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2360118262491357,
+      "learning_rate": 8.457452615929949e-06,
+      "loss": 0.2575,
+      "step": 11478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5566201005250273,
+      "learning_rate": 8.457167978677235e-06,
+      "loss": 0.2242,
+      "step": 11479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7586128277889834,
+      "learning_rate": 8.456883319956445e-06,
+      "loss": 0.2228,
+      "step": 11480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.062804591561472,
+      "learning_rate": 8.45659863976935e-06,
+      "loss": 0.4309,
+      "step": 11481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9296048779888704,
+      "learning_rate": 8.456313938117713e-06,
+      "loss": 0.4728,
+      "step": 11482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6299122464291866,
+      "learning_rate": 8.456029215003306e-06,
+      "loss": 0.4337,
+      "step": 11483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2982414456325206,
+      "learning_rate": 8.455744470427894e-06,
+      "loss": 0.6263,
+      "step": 11484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2762264518522208,
+      "learning_rate": 8.455459704393244e-06,
+      "loss": 0.1991,
+      "step": 11485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4961276635691196,
+      "learning_rate": 8.455174916901128e-06,
+      "loss": 0.374,
+      "step": 11486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4795283725455535,
+      "learning_rate": 8.454890107953313e-06,
+      "loss": 0.1722,
+      "step": 11487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3001378557107044,
+      "learning_rate": 8.454605277551566e-06,
+      "loss": 0.3809,
+      "step": 11488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6116775041443967,
+      "learning_rate": 8.454320425697659e-06,
+      "loss": 0.2355,
+      "step": 11489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7615249510277557,
+      "learning_rate": 8.454035552393358e-06,
+      "loss": 0.4407,
+      "step": 11490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4647526992474946,
+      "learning_rate": 8.453750657640434e-06,
+      "loss": 0.188,
+      "step": 11491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6296903737756951,
+      "learning_rate": 8.453465741440653e-06,
+      "loss": 0.2075,
+      "step": 11492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7617124654653575,
+      "learning_rate": 8.453180803795789e-06,
+      "loss": 0.3815,
+      "step": 11493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.789166974261402,
+      "learning_rate": 8.452895844707608e-06,
+      "loss": 0.4321,
+      "step": 11494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2973137001906023,
+      "learning_rate": 8.452610864177878e-06,
+      "loss": 0.3553,
+      "step": 11495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.490515721112149,
+      "learning_rate": 8.452325862208373e-06,
+      "loss": 0.1735,
+      "step": 11496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.743551609758336,
+      "learning_rate": 8.452040838800862e-06,
+      "loss": 0.4594,
+      "step": 11497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4411788750755234,
+      "learning_rate": 8.45175579395711e-06,
+      "loss": 0.2205,
+      "step": 11498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4881709701250003,
+      "learning_rate": 8.451470727678893e-06,
+      "loss": 0.3322,
+      "step": 11499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3563687264256385,
+      "learning_rate": 8.45118563996798e-06,
+      "loss": 0.5018,
+      "step": 11500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.053882223506026,
+      "learning_rate": 8.450900530826137e-06,
+      "loss": 0.2226,
+      "step": 11501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7061814608798738,
+      "learning_rate": 8.45061540025514e-06,
+      "loss": 0.5866,
+      "step": 11502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4946475259941459,
+      "learning_rate": 8.450330248256756e-06,
+      "loss": 0.1517,
+      "step": 11503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.741701708407664,
+      "learning_rate": 8.450045074832757e-06,
+      "loss": 0.1895,
+      "step": 11504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.719885546361024,
+      "learning_rate": 8.449759879984914e-06,
+      "loss": 0.3171,
+      "step": 11505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8153826957313102,
+      "learning_rate": 8.449474663714999e-06,
+      "loss": 0.3986,
+      "step": 11506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7273783471547577,
+      "learning_rate": 8.449189426024779e-06,
+      "loss": 0.3355,
+      "step": 11507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9669354568950355,
+      "learning_rate": 8.44890416691603e-06,
+      "loss": 0.2123,
+      "step": 11508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.661633824773058,
+      "learning_rate": 8.448618886390523e-06,
+      "loss": 0.4664,
+      "step": 11509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0593356898279485,
+      "learning_rate": 8.448333584450025e-06,
+      "loss": 0.3059,
+      "step": 11510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3908540585279194,
+      "learning_rate": 8.448048261096313e-06,
+      "loss": 0.3912,
+      "step": 11511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2241742001225533,
+      "learning_rate": 8.447762916331156e-06,
+      "loss": 0.2518,
+      "step": 11512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8165119140409118,
+      "learning_rate": 8.447477550156324e-06,
+      "loss": 0.4724,
+      "step": 11513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0141091746033655,
+      "learning_rate": 8.447192162573594e-06,
+      "loss": 0.4681,
+      "step": 11514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4240124676156274,
+      "learning_rate": 8.446906753584734e-06,
+      "loss": 0.2605,
+      "step": 11515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7839349036961472,
+      "learning_rate": 8.44662132319152e-06,
+      "loss": 0.5462,
+      "step": 11516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0734682539360207,
+      "learning_rate": 8.44633587139572e-06,
+      "loss": 0.3656,
+      "step": 11517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7366935401250252,
+      "learning_rate": 8.44605039819911e-06,
+      "loss": 0.3696,
+      "step": 11518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4200459285733056,
+      "learning_rate": 8.445764903603462e-06,
+      "loss": 0.2843,
+      "step": 11519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3121820566335245,
+      "learning_rate": 8.445479387610547e-06,
+      "loss": 0.2258,
+      "step": 11520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6101744573490087,
+      "learning_rate": 8.445193850222139e-06,
+      "loss": 0.4031,
+      "step": 11521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4350745533473992,
+      "learning_rate": 8.444908291440015e-06,
+      "loss": 0.1625,
+      "step": 11522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.633930631114397,
+      "learning_rate": 8.444622711265942e-06,
+      "loss": 0.4134,
+      "step": 11523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.674324042302004,
+      "learning_rate": 8.444337109701696e-06,
+      "loss": 0.2852,
+      "step": 11524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7921322385853704,
+      "learning_rate": 8.444051486749054e-06,
+      "loss": 0.3798,
+      "step": 11525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6914592716989532,
+      "learning_rate": 8.443765842409783e-06,
+      "loss": 0.3682,
+      "step": 11526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0917513047213454,
+      "learning_rate": 8.443480176685662e-06,
+      "loss": 0.6324,
+      "step": 11527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9236770493108157,
+      "learning_rate": 8.443194489578461e-06,
+      "loss": 0.4883,
+      "step": 11528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6461766725862836,
+      "learning_rate": 8.442908781089957e-06,
+      "loss": 0.2591,
+      "step": 11529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7473440408966523,
+      "learning_rate": 8.442623051221923e-06,
+      "loss": 0.4249,
+      "step": 11530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2693117091280066,
+      "learning_rate": 8.442337299976134e-06,
+      "loss": 0.1899,
+      "step": 11531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.215461459802629,
+      "learning_rate": 8.442051527354364e-06,
+      "loss": 0.5189,
+      "step": 11532
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1563133410876054,
+      "learning_rate": 8.441765733358387e-06,
+      "loss": 0.215,
+      "step": 11533
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2324176340891664,
+      "learning_rate": 8.44147991798998e-06,
+      "loss": 0.2642,
+      "step": 11534
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1369029633443244,
+      "learning_rate": 8.441194081250914e-06,
+      "loss": 0.3168,
+      "step": 11535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4299068126375982,
+      "learning_rate": 8.440908223142965e-06,
+      "loss": 0.308,
+      "step": 11536
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7034379011100165,
+      "learning_rate": 8.440622343667913e-06,
+      "loss": 0.401,
+      "step": 11537
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.634931355396622,
+      "learning_rate": 8.440336442827525e-06,
+      "loss": 0.3829,
+      "step": 11538
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0094167080697622,
+      "learning_rate": 8.440050520623584e-06,
+      "loss": 0.1562,
+      "step": 11539
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9820824393585201,
+      "learning_rate": 8.439764577057861e-06,
+      "loss": 0.4036,
+      "step": 11540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7549708688728232,
+      "learning_rate": 8.439478612132133e-06,
+      "loss": 0.5067,
+      "step": 11541
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.469405738057981,
+      "learning_rate": 8.439192625848176e-06,
+      "loss": 0.3639,
+      "step": 11542
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8655282353301854,
+      "learning_rate": 8.438906618207765e-06,
+      "loss": 0.397,
+      "step": 11543
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5122862916103699,
+      "learning_rate": 8.438620589212676e-06,
+      "loss": 0.2763,
+      "step": 11544
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.908296639406861,
+      "learning_rate": 8.438334538864688e-06,
+      "loss": 0.3497,
+      "step": 11545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5244329672502008,
+      "learning_rate": 8.438048467165573e-06,
+      "loss": 0.3587,
+      "step": 11546
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.138215417633874,
+      "learning_rate": 8.437762374117109e-06,
+      "loss": 0.2009,
+      "step": 11547
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5871427729193535,
+      "learning_rate": 8.437476259721075e-06,
+      "loss": 0.2701,
+      "step": 11548
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2695576356933505,
+      "learning_rate": 8.437190123979244e-06,
+      "loss": 0.2412,
+      "step": 11549
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5299773474493636,
+      "learning_rate": 8.436903966893395e-06,
+      "loss": 0.3539,
+      "step": 11550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4258435067804687,
+      "learning_rate": 8.436617788465304e-06,
+      "loss": 0.4186,
+      "step": 11551
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5628348182529324,
+      "learning_rate": 8.43633158869675e-06,
+      "loss": 0.4691,
+      "step": 11552
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5997924280750588,
+      "learning_rate": 8.436045367589509e-06,
+      "loss": 0.3419,
+      "step": 11553
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3918544375470496,
+      "learning_rate": 8.435759125145358e-06,
+      "loss": 0.5482,
+      "step": 11554
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1880081447911544,
+      "learning_rate": 8.435472861366072e-06,
+      "loss": 0.1566,
+      "step": 11555
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6730927929754817,
+      "learning_rate": 8.435186576253433e-06,
+      "loss": 0.293,
+      "step": 11556
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5834381252880076,
+      "learning_rate": 8.434900269809218e-06,
+      "loss": 0.3842,
+      "step": 11557
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1784069818743017,
+      "learning_rate": 8.434613942035203e-06,
+      "loss": 0.6192,
+      "step": 11558
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0160138546130786,
+      "learning_rate": 8.434327592933167e-06,
+      "loss": 0.2665,
+      "step": 11559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3385606737648934,
+      "learning_rate": 8.434041222504887e-06,
+      "loss": 0.1531,
+      "step": 11560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7085111191648423,
+      "learning_rate": 8.433754830752144e-06,
+      "loss": 0.3892,
+      "step": 11561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5519140439762162,
+      "learning_rate": 8.433468417676714e-06,
+      "loss": 0.3887,
+      "step": 11562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7323836453551558,
+      "learning_rate": 8.433181983280376e-06,
+      "loss": 0.1902,
+      "step": 11563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1906209077038863,
+      "learning_rate": 8.43289552756491e-06,
+      "loss": 0.2015,
+      "step": 11564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1095711087088596,
+      "learning_rate": 8.432609050532093e-06,
+      "loss": 0.4682,
+      "step": 11565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.486542163731428,
+      "learning_rate": 8.432322552183705e-06,
+      "loss": 0.2887,
+      "step": 11566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6759309590653202,
+      "learning_rate": 8.432036032521525e-06,
+      "loss": 0.3262,
+      "step": 11567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6160191263663923,
+      "learning_rate": 8.431749491547333e-06,
+      "loss": 0.199,
+      "step": 11568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1735185637074423,
+      "learning_rate": 8.431462929262904e-06,
+      "loss": 0.2987,
+      "step": 11569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2959733039724712,
+      "learning_rate": 8.431176345670023e-06,
+      "loss": 0.465,
+      "step": 11570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9517808637781866,
+      "learning_rate": 8.430889740770468e-06,
+      "loss": 0.2516,
+      "step": 11571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4478999844168667,
+      "learning_rate": 8.430603114566019e-06,
+      "loss": 0.4023,
+      "step": 11572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.804843919212836,
+      "learning_rate": 8.430316467058453e-06,
+      "loss": 0.1597,
+      "step": 11573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5497536028859964,
+      "learning_rate": 8.430029798249553e-06,
+      "loss": 0.2113,
+      "step": 11574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.634651536731889,
+      "learning_rate": 8.429743108141097e-06,
+      "loss": 0.1995,
+      "step": 11575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6197108655403785,
+      "learning_rate": 8.429456396734868e-06,
+      "loss": 0.2421,
+      "step": 11576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6456992108787207,
+      "learning_rate": 8.429169664032644e-06,
+      "loss": 0.4051,
+      "step": 11577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0265779904940424,
+      "learning_rate": 8.428882910036207e-06,
+      "loss": 0.4755,
+      "step": 11578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8676566169104594,
+      "learning_rate": 8.428596134747336e-06,
+      "loss": 0.2188,
+      "step": 11579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.274932202062946,
+      "learning_rate": 8.428309338167814e-06,
+      "loss": 0.1989,
+      "step": 11580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2307605719524783,
+      "learning_rate": 8.428022520299422e-06,
+      "loss": 0.2432,
+      "step": 11581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5412544771270198,
+      "learning_rate": 8.427735681143936e-06,
+      "loss": 0.3379,
+      "step": 11582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4672054237371845,
+      "learning_rate": 8.427448820703145e-06,
+      "loss": 0.2248,
+      "step": 11583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6467710994250757,
+      "learning_rate": 8.427161938978824e-06,
+      "loss": 0.399,
+      "step": 11584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6567392886085943,
+      "learning_rate": 8.426875035972758e-06,
+      "loss": 0.3609,
+      "step": 11585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8554968821751998,
+      "learning_rate": 8.426588111686727e-06,
+      "loss": 0.4104,
+      "step": 11586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2944393608441302,
+      "learning_rate": 8.426301166122514e-06,
+      "loss": 0.4282,
+      "step": 11587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7920519369283223,
+      "learning_rate": 8.4260141992819e-06,
+      "loss": 0.205,
+      "step": 11588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4767231601523705,
+      "learning_rate": 8.425727211166666e-06,
+      "loss": 0.291,
+      "step": 11589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4912624537594643,
+      "learning_rate": 8.425440201778595e-06,
+      "loss": 0.1948,
+      "step": 11590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6432288805139679,
+      "learning_rate": 8.42515317111947e-06,
+      "loss": 0.2411,
+      "step": 11591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9179076128692156,
+      "learning_rate": 8.424866119191071e-06,
+      "loss": 0.2107,
+      "step": 11592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7456603909902262,
+      "learning_rate": 8.424579045995183e-06,
+      "loss": 0.4572,
+      "step": 11593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.670922471480582,
+      "learning_rate": 8.42429195153359e-06,
+      "loss": 0.3424,
+      "step": 11594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8656507376871523,
+      "learning_rate": 8.42400483580807e-06,
+      "loss": 0.4563,
+      "step": 11595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0770468746110649,
+      "learning_rate": 8.42371769882041e-06,
+      "loss": 0.1765,
+      "step": 11596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9001132021609815,
+      "learning_rate": 8.423430540572393e-06,
+      "loss": 0.7121,
+      "step": 11597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2768208136149928,
+      "learning_rate": 8.423143361065798e-06,
+      "loss": 0.2419,
+      "step": 11598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.661911476358295,
+      "learning_rate": 8.422856160302413e-06,
+      "loss": 0.2665,
+      "step": 11599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9272232924147403,
+      "learning_rate": 8.422568938284019e-06,
+      "loss": 0.3396,
+      "step": 11600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.973713355518448,
+      "learning_rate": 8.4222816950124e-06,
+      "loss": 0.5851,
+      "step": 11601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5425766796840408,
+      "learning_rate": 8.421994430489339e-06,
+      "loss": 0.3227,
+      "step": 11602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9803674133979898,
+      "learning_rate": 8.421707144716623e-06,
+      "loss": 0.3671,
+      "step": 11603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3771368621092994,
+      "learning_rate": 8.421419837696031e-06,
+      "loss": 0.1994,
+      "step": 11604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.723513197598227,
+      "learning_rate": 8.42113250942935e-06,
+      "loss": 0.364,
+      "step": 11605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7764009511296346,
+      "learning_rate": 8.420845159918366e-06,
+      "loss": 0.4109,
+      "step": 11606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5597517162441343,
+      "learning_rate": 8.42055778916486e-06,
+      "loss": 0.4249,
+      "step": 11607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6745455794465682,
+      "learning_rate": 8.420270397170618e-06,
+      "loss": 0.4766,
+      "step": 11608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8716903085933712,
+      "learning_rate": 8.419982983937425e-06,
+      "loss": 0.5782,
+      "step": 11609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4619392986578466,
+      "learning_rate": 8.419695549467064e-06,
+      "loss": 0.2199,
+      "step": 11610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5981161159899508,
+      "learning_rate": 8.41940809376132e-06,
+      "loss": 0.4132,
+      "step": 11611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0772714439453197,
+      "learning_rate": 8.419120616821983e-06,
+      "loss": 0.5299,
+      "step": 11612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1327165540136974,
+      "learning_rate": 8.41883311865083e-06,
+      "loss": 0.4719,
+      "step": 11613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2487827021309075,
+      "learning_rate": 8.418545599249654e-06,
+      "loss": 0.1549,
+      "step": 11614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3839469824941875,
+      "learning_rate": 8.418258058620235e-06,
+      "loss": 0.2242,
+      "step": 11615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.633857465681305,
+      "learning_rate": 8.417970496764362e-06,
+      "loss": 0.2882,
+      "step": 11616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4773928455837055,
+      "learning_rate": 8.417682913683818e-06,
+      "loss": 0.1771,
+      "step": 11617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2621613367588385,
+      "learning_rate": 8.41739530938039e-06,
+      "loss": 0.406,
+      "step": 11618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.104755542677598,
+      "learning_rate": 8.417107683855865e-06,
+      "loss": 0.2923,
+      "step": 11619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6240413536326803,
+      "learning_rate": 8.416820037112027e-06,
+      "loss": 0.2361,
+      "step": 11620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4683997543193887,
+      "learning_rate": 8.416532369150662e-06,
+      "loss": 0.2667,
+      "step": 11621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.272782527367911,
+      "learning_rate": 8.41624467997356e-06,
+      "loss": 0.2121,
+      "step": 11622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8247487076431907,
+      "learning_rate": 8.415956969582504e-06,
+      "loss": 0.4166,
+      "step": 11623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4746868290043795,
+      "learning_rate": 8.415669237979282e-06,
+      "loss": 0.319,
+      "step": 11624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.031747635991958,
+      "learning_rate": 8.415381485165682e-06,
+      "loss": 0.3206,
+      "step": 11625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9063292735063384,
+      "learning_rate": 8.415093711143486e-06,
+      "loss": 0.2188,
+      "step": 11626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.492377813296163,
+      "learning_rate": 8.414805915914486e-06,
+      "loss": 0.2934,
+      "step": 11627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5184618268338088,
+      "learning_rate": 8.414518099480467e-06,
+      "loss": 0.2523,
+      "step": 11628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.390113077341189,
+      "learning_rate": 8.414230261843217e-06,
+      "loss": 0.331,
+      "step": 11629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5080927448789483,
+      "learning_rate": 8.413942403004523e-06,
+      "loss": 0.1704,
+      "step": 11630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5482013150803378,
+      "learning_rate": 8.413654522966171e-06,
+      "loss": 0.1935,
+      "step": 11631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.378110802073447,
+      "learning_rate": 8.413366621729952e-06,
+      "loss": 0.2426,
+      "step": 11632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5832690071122764,
+      "learning_rate": 8.413078699297652e-06,
+      "loss": 0.1562,
+      "step": 11633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6820814881748924,
+      "learning_rate": 8.412790755671058e-06,
+      "loss": 0.4041,
+      "step": 11634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9604146162819369,
+      "learning_rate": 8.412502790851959e-06,
+      "loss": 0.2099,
+      "step": 11635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6188686036510875,
+      "learning_rate": 8.412214804842145e-06,
+      "loss": 0.3645,
+      "step": 11636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4491214578158798,
+      "learning_rate": 8.4119267976434e-06,
+      "loss": 0.2167,
+      "step": 11637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2579593276577283,
+      "learning_rate": 8.411638769257518e-06,
+      "loss": 0.2132,
+      "step": 11638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4569325055777593,
+      "learning_rate": 8.411350719686282e-06,
+      "loss": 0.3606,
+      "step": 11639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8901152278197717,
+      "learning_rate": 8.411062648931485e-06,
+      "loss": 0.4929,
+      "step": 11640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8573743170296613,
+      "learning_rate": 8.410774556994912e-06,
+      "loss": 0.3481,
+      "step": 11641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.44445323167351,
+      "learning_rate": 8.410486443878355e-06,
+      "loss": 0.4845,
+      "step": 11642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9318183493167886,
+      "learning_rate": 8.4101983095836e-06,
+      "loss": 0.3419,
+      "step": 11643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.005236438868016,
+      "learning_rate": 8.409910154112442e-06,
+      "loss": 0.4032,
+      "step": 11644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.580143711440558,
+      "learning_rate": 8.409621977466666e-06,
+      "loss": 0.3112,
+      "step": 11645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.026015918113709,
+      "learning_rate": 8.40933377964806e-06,
+      "loss": 0.3657,
+      "step": 11646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1995556603507365,
+      "learning_rate": 8.409045560658418e-06,
+      "loss": 0.1732,
+      "step": 11647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8598560544427147,
+      "learning_rate": 8.408757320499526e-06,
+      "loss": 0.3477,
+      "step": 11648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.734263708650767,
+      "learning_rate": 8.408469059173177e-06,
+      "loss": 0.2601,
+      "step": 11649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.874226311163142,
+      "learning_rate": 8.40818077668116e-06,
+      "loss": 0.3234,
+      "step": 11650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.451694743452996,
+      "learning_rate": 8.407892473025262e-06,
+      "loss": 0.3846,
+      "step": 11651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.354606343074604,
+      "learning_rate": 8.407604148207279e-06,
+      "loss": 0.3865,
+      "step": 11652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.859394546213714,
+      "learning_rate": 8.407315802228997e-06,
+      "loss": 0.3397,
+      "step": 11653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.624486711479866,
+      "learning_rate": 8.407027435092207e-06,
+      "loss": 0.3182,
+      "step": 11654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2737221450902207,
+      "learning_rate": 8.406739046798703e-06,
+      "loss": 0.1603,
+      "step": 11655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5536351020571109,
+      "learning_rate": 8.406450637350271e-06,
+      "loss": 0.3869,
+      "step": 11656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6321807693314916,
+      "learning_rate": 8.406162206748706e-06,
+      "loss": 0.398,
+      "step": 11657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0645436977125837,
+      "learning_rate": 8.405873754995797e-06,
+      "loss": 0.5061,
+      "step": 11658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2354523788922416,
+      "learning_rate": 8.405585282093336e-06,
+      "loss": 0.4004,
+      "step": 11659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6647897371456688,
+      "learning_rate": 8.405296788043113e-06,
+      "loss": 0.3626,
+      "step": 11660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3760674462176463,
+      "learning_rate": 8.405008272846922e-06,
+      "loss": 0.2544,
+      "step": 11661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7339651164725383,
+      "learning_rate": 8.404719736506552e-06,
+      "loss": 0.2205,
+      "step": 11662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.634378560592974,
+      "learning_rate": 8.404431179023797e-06,
+      "loss": 0.2562,
+      "step": 11663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4867739115437966,
+      "learning_rate": 8.404142600400446e-06,
+      "loss": 0.3215,
+      "step": 11664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6411050866106005,
+      "learning_rate": 8.403854000638294e-06,
+      "loss": 0.4136,
+      "step": 11665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.02058060856411,
+      "learning_rate": 8.403565379739132e-06,
+      "loss": 0.1549,
+      "step": 11666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5776046801625327,
+      "learning_rate": 8.40327673770475e-06,
+      "loss": 0.4701,
+      "step": 11667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9452604914993772,
+      "learning_rate": 8.402988074536944e-06,
+      "loss": 0.4215,
+      "step": 11668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7934852138831818,
+      "learning_rate": 8.402699390237505e-06,
+      "loss": 0.3084,
+      "step": 11669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.56194666601519,
+      "learning_rate": 8.402410684808225e-06,
+      "loss": 0.3909,
+      "step": 11670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5086983267651133,
+      "learning_rate": 8.402121958250899e-06,
+      "loss": 0.3713,
+      "step": 11671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.464681056842505,
+      "learning_rate": 8.401833210567317e-06,
+      "loss": 0.3594,
+      "step": 11672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3511189672892319,
+      "learning_rate": 8.401544441759272e-06,
+      "loss": 0.2673,
+      "step": 11673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6746607549004866,
+      "learning_rate": 8.401255651828562e-06,
+      "loss": 0.3314,
+      "step": 11674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9643858950631694,
+      "learning_rate": 8.400966840776974e-06,
+      "loss": 0.266,
+      "step": 11675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1338150616421463,
+      "learning_rate": 8.400678008606306e-06,
+      "loss": 0.477,
+      "step": 11676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5591176366757549,
+      "learning_rate": 8.400389155318348e-06,
+      "loss": 0.3774,
+      "step": 11677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.858830177803976,
+      "learning_rate": 8.400100280914899e-06,
+      "loss": 0.6172,
+      "step": 11678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5228208724448735,
+      "learning_rate": 8.399811385397745e-06,
+      "loss": 0.2039,
+      "step": 11679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2156755782541997,
+      "learning_rate": 8.399522468768685e-06,
+      "loss": 0.1966,
+      "step": 11680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7321669338284778,
+      "learning_rate": 8.399233531029516e-06,
+      "loss": 0.5277,
+      "step": 11681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3804183158448224,
+      "learning_rate": 8.398944572182027e-06,
+      "loss": 0.3113,
+      "step": 11682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.620834694121614,
+      "learning_rate": 8.398655592228012e-06,
+      "loss": 0.2222,
+      "step": 11683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5551800810127165,
+      "learning_rate": 8.398366591169269e-06,
+      "loss": 0.4249,
+      "step": 11684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8724764033735584,
+      "learning_rate": 8.398077569007591e-06,
+      "loss": 0.2449,
+      "step": 11685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6197622461090078,
+      "learning_rate": 8.397788525744773e-06,
+      "loss": 0.4065,
+      "step": 11686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6268364016093744,
+      "learning_rate": 8.397499461382612e-06,
+      "loss": 0.4677,
+      "step": 11687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1566094439064492,
+      "learning_rate": 8.397210375922897e-06,
+      "loss": 0.2313,
+      "step": 11688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.663221710988561,
+      "learning_rate": 8.396921269367428e-06,
+      "loss": 0.4392,
+      "step": 11689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7022771474791518,
+      "learning_rate": 8.396632141718e-06,
+      "loss": 0.5347,
+      "step": 11690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.135320983852,
+      "learning_rate": 8.396342992976407e-06,
+      "loss": 0.6383,
+      "step": 11691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.680090780416726,
+      "learning_rate": 8.396053823144445e-06,
+      "loss": 0.3608,
+      "step": 11692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.661905650508607,
+      "learning_rate": 8.395764632223911e-06,
+      "loss": 0.261,
+      "step": 11693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7570444081968428,
+      "learning_rate": 8.395475420216599e-06,
+      "loss": 0.3682,
+      "step": 11694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4261412375112803,
+      "learning_rate": 8.395186187124305e-06,
+      "loss": 0.363,
+      "step": 11695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7375675450718948,
+      "learning_rate": 8.394896932948826e-06,
+      "loss": 0.4384,
+      "step": 11696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.818137883710798,
+      "learning_rate": 8.394607657691958e-06,
+      "loss": 0.3764,
+      "step": 11697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.769221647179851,
+      "learning_rate": 8.394318361355498e-06,
+      "loss": 0.5363,
+      "step": 11698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.67132061162542,
+      "learning_rate": 8.394029043941239e-06,
+      "loss": 0.2264,
+      "step": 11699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5181333981686478,
+      "learning_rate": 8.393739705450982e-06,
+      "loss": 0.2657,
+      "step": 11700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.262364913687709,
+      "learning_rate": 8.39345034588652e-06,
+      "loss": 0.1634,
+      "step": 11701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0636175769115757,
+      "learning_rate": 8.393160965249655e-06,
+      "loss": 0.4632,
+      "step": 11702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4102273920452786,
+      "learning_rate": 8.392871563542178e-06,
+      "loss": 0.211,
+      "step": 11703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.609858049193868,
+      "learning_rate": 8.39258214076589e-06,
+      "loss": 0.1792,
+      "step": 11704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.216377191685888,
+      "learning_rate": 8.392292696922586e-06,
+      "loss": 0.3266,
+      "step": 11705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7161869098183444,
+      "learning_rate": 8.392003232014066e-06,
+      "loss": 0.4616,
+      "step": 11706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9075929882537415,
+      "learning_rate": 8.391713746042124e-06,
+      "loss": 0.6075,
+      "step": 11707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3883629038019247,
+      "learning_rate": 8.391424239008559e-06,
+      "loss": 0.3471,
+      "step": 11708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.769142541865386,
+      "learning_rate": 8.391134710915172e-06,
+      "loss": 0.3513,
+      "step": 11709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.791083250360264,
+      "learning_rate": 8.390845161763756e-06,
+      "loss": 0.5312,
+      "step": 11710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9180364150097353,
+      "learning_rate": 8.39055559155611e-06,
+      "loss": 0.6245,
+      "step": 11711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3834916779892015,
+      "learning_rate": 8.390266000294036e-06,
+      "loss": 0.3085,
+      "step": 11712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8681641861336764,
+      "learning_rate": 8.389976387979329e-06,
+      "loss": 0.4011,
+      "step": 11713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.393700427645982,
+      "learning_rate": 8.389686754613788e-06,
+      "loss": 0.1463,
+      "step": 11714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0297013000379724,
+      "learning_rate": 8.389397100199212e-06,
+      "loss": 0.1978,
+      "step": 11715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4901169659421425,
+      "learning_rate": 8.389107424737397e-06,
+      "loss": 0.1509,
+      "step": 11716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3362613447196632,
+      "learning_rate": 8.388817728230146e-06,
+      "loss": 0.225,
+      "step": 11717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.101879929035679,
+      "learning_rate": 8.388528010679255e-06,
+      "loss": 0.573,
+      "step": 11718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.751222738905377,
+      "learning_rate": 8.388238272086525e-06,
+      "loss": 0.3715,
+      "step": 11719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1299282683043677,
+      "learning_rate": 8.387948512453753e-06,
+      "loss": 0.3827,
+      "step": 11720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.714912934526098,
+      "learning_rate": 8.387658731782742e-06,
+      "loss": 0.3631,
+      "step": 11721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.924796113605264,
+      "learning_rate": 8.387368930075288e-06,
+      "loss": 0.5357,
+      "step": 11722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3343752739300772,
+      "learning_rate": 8.38707910733319e-06,
+      "loss": 0.2133,
+      "step": 11723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0748644037638573,
+      "learning_rate": 8.386789263558251e-06,
+      "loss": 0.3632,
+      "step": 11724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5733517893352573,
+      "learning_rate": 8.386499398752268e-06,
+      "loss": 0.3502,
+      "step": 11725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8412155536542354,
+      "learning_rate": 8.386209512917043e-06,
+      "loss": 0.4456,
+      "step": 11726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2479464112920458,
+      "learning_rate": 8.385919606054375e-06,
+      "loss": 0.3427,
+      "step": 11727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.96822831644358,
+      "learning_rate": 8.385629678166064e-06,
+      "loss": 0.6155,
+      "step": 11728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8243156178668094,
+      "learning_rate": 8.385339729253913e-06,
+      "loss": 0.6358,
+      "step": 11729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6581009877159678,
+      "learning_rate": 8.385049759319718e-06,
+      "loss": 0.3367,
+      "step": 11730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8457359109217064,
+      "learning_rate": 8.384759768365283e-06,
+      "loss": 0.5191,
+      "step": 11731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9397031809174672,
+      "learning_rate": 8.384469756392408e-06,
+      "loss": 0.4126,
+      "step": 11732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.541097104586126,
+      "learning_rate": 8.384179723402894e-06,
+      "loss": 0.2224,
+      "step": 11733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5237362397552874,
+      "learning_rate": 8.383889669398542e-06,
+      "loss": 0.303,
+      "step": 11734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6363851871954325,
+      "learning_rate": 8.383599594381152e-06,
+      "loss": 0.3259,
+      "step": 11735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.805543911011959,
+      "learning_rate": 8.383309498352525e-06,
+      "loss": 0.3723,
+      "step": 11736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9401720449021744,
+      "learning_rate": 8.383019381314466e-06,
+      "loss": 0.5313,
+      "step": 11737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4815497037837453,
+      "learning_rate": 8.382729243268773e-06,
+      "loss": 0.3541,
+      "step": 11738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7967295297534345,
+      "learning_rate": 8.382439084217248e-06,
+      "loss": 0.3837,
+      "step": 11739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4831379679970151,
+      "learning_rate": 8.382148904161696e-06,
+      "loss": 0.3427,
+      "step": 11740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6013077486771052,
+      "learning_rate": 8.381858703103915e-06,
+      "loss": 0.1839,
+      "step": 11741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8493606839265644,
+      "learning_rate": 8.381568481045708e-06,
+      "loss": 0.421,
+      "step": 11742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2925811932972067,
+      "learning_rate": 8.381278237988879e-06,
+      "loss": 0.2425,
+      "step": 11743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.360681355313481,
+      "learning_rate": 8.380987973935228e-06,
+      "loss": 0.2927,
+      "step": 11744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7158983263877226,
+      "learning_rate": 8.380697688886559e-06,
+      "loss": 0.2244,
+      "step": 11745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0643533852421454,
+      "learning_rate": 8.380407382844675e-06,
+      "loss": 0.2564,
+      "step": 11746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.484123089395532,
+      "learning_rate": 8.380117055811378e-06,
+      "loss": 0.2147,
+      "step": 11747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3841229145759864,
+      "learning_rate": 8.379826707788471e-06,
+      "loss": 0.3152,
+      "step": 11748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6060961961398224,
+      "learning_rate": 8.379536338777756e-06,
+      "loss": 0.2173,
+      "step": 11749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5157630110365796,
+      "learning_rate": 8.379245948781036e-06,
+      "loss": 0.3291,
+      "step": 11750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6737589326620868,
+      "learning_rate": 8.378955537800118e-06,
+      "loss": 0.1975,
+      "step": 11751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3853891371887492,
+      "learning_rate": 8.378665105836802e-06,
+      "loss": 0.1744,
+      "step": 11752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2511641907374105,
+      "learning_rate": 8.378374652892889e-06,
+      "loss": 0.2178,
+      "step": 11753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.056623884460772,
+      "learning_rate": 8.378084178970188e-06,
+      "loss": 0.4644,
+      "step": 11754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8578767029085166,
+      "learning_rate": 8.377793684070499e-06,
+      "loss": 0.5988,
+      "step": 11755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2663542830378867,
+      "learning_rate": 8.37750316819563e-06,
+      "loss": 0.2836,
+      "step": 11756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4367869321545361,
+      "learning_rate": 8.377212631347377e-06,
+      "loss": 0.2792,
+      "step": 11757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4994039791831997,
+      "learning_rate": 8.376922073527553e-06,
+      "loss": 0.3232,
+      "step": 11758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4782662149803187,
+      "learning_rate": 8.37663149473796e-06,
+      "loss": 0.3276,
+      "step": 11759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5506452313402201,
+      "learning_rate": 8.3763408949804e-06,
+      "loss": 0.3954,
+      "step": 11760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.654604163223568,
+      "learning_rate": 8.376050274256679e-06,
+      "loss": 0.2564,
+      "step": 11761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1603299954893007,
+      "learning_rate": 8.375759632568601e-06,
+      "loss": 0.3391,
+      "step": 11762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6758202634951753,
+      "learning_rate": 8.37546896991797e-06,
+      "loss": 0.4269,
+      "step": 11763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8896514833410987,
+      "learning_rate": 8.375178286306595e-06,
+      "loss": 0.3071,
+      "step": 11764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8725204125680879,
+      "learning_rate": 8.374887581736275e-06,
+      "loss": 0.6069,
+      "step": 11765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5323902818241586,
+      "learning_rate": 8.374596856208821e-06,
+      "loss": 0.3517,
+      "step": 11766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7520415582526268,
+      "learning_rate": 8.374306109726034e-06,
+      "loss": 0.5524,
+      "step": 11767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.264294055815482,
+      "learning_rate": 8.374015342289723e-06,
+      "loss": 0.313,
+      "step": 11768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8248523004088042,
+      "learning_rate": 8.37372455390169e-06,
+      "loss": 0.5575,
+      "step": 11769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.719546129334758,
+      "learning_rate": 8.373433744563743e-06,
+      "loss": 0.1793,
+      "step": 11770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4070599263396901,
+      "learning_rate": 8.37314291427769e-06,
+      "loss": 0.3745,
+      "step": 11771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7457757442629616,
+      "learning_rate": 8.372852063045331e-06,
+      "loss": 0.2096,
+      "step": 11772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8393871344141137,
+      "learning_rate": 8.372561190868475e-06,
+      "loss": 0.3979,
+      "step": 11773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1859280100858653,
+      "learning_rate": 8.372270297748931e-06,
+      "loss": 0.2858,
+      "step": 11774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8844594677138766,
+      "learning_rate": 8.371979383688501e-06,
+      "loss": 0.4113,
+      "step": 11775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8913922683974278,
+      "learning_rate": 8.371688448688996e-06,
+      "loss": 0.2625,
+      "step": 11776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5875054275802885,
+      "learning_rate": 8.371397492752218e-06,
+      "loss": 0.3431,
+      "step": 11777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.743253702375309,
+      "learning_rate": 8.371106515879977e-06,
+      "loss": 0.4327,
+      "step": 11778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4880499284550408,
+      "learning_rate": 8.370815518074079e-06,
+      "loss": 0.246,
+      "step": 11779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.590039098299394,
+      "learning_rate": 8.37052449933633e-06,
+      "loss": 0.3988,
+      "step": 11780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2151836827370124,
+      "learning_rate": 8.370233459668538e-06,
+      "loss": 0.6792,
+      "step": 11781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6204497862237637,
+      "learning_rate": 8.369942399072511e-06,
+      "loss": 0.3291,
+      "step": 11782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3751783268880085,
+      "learning_rate": 8.369651317550055e-06,
+      "loss": 0.1358,
+      "step": 11783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.377048118230421,
+      "learning_rate": 8.369360215102979e-06,
+      "loss": 0.2738,
+      "step": 11784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7049253729600082,
+      "learning_rate": 8.369069091733088e-06,
+      "loss": 0.5732,
+      "step": 11785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7211499451317245,
+      "learning_rate": 8.368777947442195e-06,
+      "loss": 0.4717,
+      "step": 11786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.216496392609728,
+      "learning_rate": 8.3684867822321e-06,
+      "loss": 0.4109,
+      "step": 11787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2578957571587521,
+      "learning_rate": 8.36819559610462e-06,
+      "loss": 0.3661,
+      "step": 11788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2481850516162853,
+      "learning_rate": 8.367904389061556e-06,
+      "loss": 0.5795,
+      "step": 11789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7255908785729297,
+      "learning_rate": 8.367613161104721e-06,
+      "loss": 0.2,
+      "step": 11790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1836747882974095,
+      "learning_rate": 8.36732191223592e-06,
+      "loss": 0.1955,
+      "step": 11791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8996623644300927,
+      "learning_rate": 8.367030642456965e-06,
+      "loss": 0.4385,
+      "step": 11792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.173864486245412,
+      "learning_rate": 8.366739351769662e-06,
+      "loss": 0.1328,
+      "step": 11793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6049256624404493,
+      "learning_rate": 8.36644804017582e-06,
+      "loss": 0.4585,
+      "step": 11794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8150421024331123,
+      "learning_rate": 8.366156707677248e-06,
+      "loss": 0.5023,
+      "step": 11795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.557681469203731,
+      "learning_rate": 8.365865354275757e-06,
+      "loss": 0.3478,
+      "step": 11796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9436787817241332,
+      "learning_rate": 8.365573979973157e-06,
+      "loss": 0.2782,
+      "step": 11797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1526009739974334,
+      "learning_rate": 8.365282584771252e-06,
+      "loss": 0.2922,
+      "step": 11798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.605843349479824,
+      "learning_rate": 8.364991168671857e-06,
+      "loss": 0.3451,
+      "step": 11799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9770715005344341,
+      "learning_rate": 8.364699731676777e-06,
+      "loss": 0.5504,
+      "step": 11800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1682825352747277,
+      "learning_rate": 8.364408273787828e-06,
+      "loss": 0.1121,
+      "step": 11801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0459648097017247,
+      "learning_rate": 8.364116795006812e-06,
+      "loss": 0.2131,
+      "step": 11802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2462712463435017,
+      "learning_rate": 8.363825295335546e-06,
+      "loss": 0.2495,
+      "step": 11803
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6468761720056002,
+      "learning_rate": 8.363533774775837e-06,
+      "loss": 0.321,
+      "step": 11804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3761932169991569,
+      "learning_rate": 8.363242233329494e-06,
+      "loss": 0.2699,
+      "step": 11805
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.232232853136801,
+      "learning_rate": 8.36295067099833e-06,
+      "loss": 0.1449,
+      "step": 11806
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2420022124527608,
+      "learning_rate": 8.362659087784153e-06,
+      "loss": 0.2086,
+      "step": 11807
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9769860443347211,
+      "learning_rate": 8.362367483688776e-06,
+      "loss": 0.4402,
+      "step": 11808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2268517891859685,
+      "learning_rate": 8.36207585871401e-06,
+      "loss": 0.2402,
+      "step": 11809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0867939640195667,
+      "learning_rate": 8.361784212861663e-06,
+      "loss": 0.13,
+      "step": 11810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.553783470201478,
+      "learning_rate": 8.361492546133547e-06,
+      "loss": 0.2021,
+      "step": 11811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9236493874799325,
+      "learning_rate": 8.361200858531476e-06,
+      "loss": 0.2205,
+      "step": 11812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1511910532728191,
+      "learning_rate": 8.36090915005726e-06,
+      "loss": 0.175,
+      "step": 11813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3075925726422022,
+      "learning_rate": 8.360617420712707e-06,
+      "loss": 0.182,
+      "step": 11814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7250217363812486,
+      "learning_rate": 8.360325670499635e-06,
+      "loss": 0.3806,
+      "step": 11815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5142922785101303,
+      "learning_rate": 8.360033899419848e-06,
+      "loss": 0.3788,
+      "step": 11816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.014333481237962,
+      "learning_rate": 8.359742107475161e-06,
+      "loss": 0.3912,
+      "step": 11817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7730985729330975,
+      "learning_rate": 8.35945029466739e-06,
+      "loss": 0.5328,
+      "step": 11818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9148082097964183,
+      "learning_rate": 8.359158460998342e-06,
+      "loss": 0.4722,
+      "step": 11819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8034098383676387,
+      "learning_rate": 8.358866606469831e-06,
+      "loss": 0.3781,
+      "step": 11820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.789909108341048,
+      "learning_rate": 8.358574731083671e-06,
+      "loss": 0.3268,
+      "step": 11821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4936964434435742,
+      "learning_rate": 8.358282834841672e-06,
+      "loss": 0.1755,
+      "step": 11822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0861848868426367,
+      "learning_rate": 8.357990917745648e-06,
+      "loss": 0.1652,
+      "step": 11823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5197847395153592,
+      "learning_rate": 8.357698979797409e-06,
+      "loss": 0.5347,
+      "step": 11824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5784596263087651,
+      "learning_rate": 8.357407020998772e-06,
+      "loss": 0.3866,
+      "step": 11825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8176713947015104,
+      "learning_rate": 8.357115041351548e-06,
+      "loss": 0.4038,
+      "step": 11826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5892864065796397,
+      "learning_rate": 8.35682304085755e-06,
+      "loss": 0.5276,
+      "step": 11827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0816859648057688,
+      "learning_rate": 8.35653101951859e-06,
+      "loss": 0.1357,
+      "step": 11828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.967310913099113,
+      "learning_rate": 8.356238977336483e-06,
+      "loss": 0.6002,
+      "step": 11829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0920543619033554,
+      "learning_rate": 8.355946914313042e-06,
+      "loss": 0.3965,
+      "step": 11830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9232776241499197,
+      "learning_rate": 8.355654830450081e-06,
+      "loss": 0.2171,
+      "step": 11831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.138373798039182,
+      "learning_rate": 8.355362725749414e-06,
+      "loss": 0.1843,
+      "step": 11832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5000852384289827,
+      "learning_rate": 8.355070600212854e-06,
+      "loss": 0.2163,
+      "step": 11833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6095172833072882,
+      "learning_rate": 8.354778453842216e-06,
+      "loss": 0.1868,
+      "step": 11834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0899802332723647,
+      "learning_rate": 8.354486286639313e-06,
+      "loss": 0.3836,
+      "step": 11835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4469576159098025,
+      "learning_rate": 8.35419409860596e-06,
+      "loss": 0.1828,
+      "step": 11836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5233961454304936,
+      "learning_rate": 8.353901889743975e-06,
+      "loss": 0.3161,
+      "step": 11837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0463812663710141,
+      "learning_rate": 8.353609660055165e-06,
+      "loss": 0.2129,
+      "step": 11838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6813237891968411,
+      "learning_rate": 8.353317409541348e-06,
+      "loss": 0.1929,
+      "step": 11839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.883239614548197,
+      "learning_rate": 8.35302513820434e-06,
+      "loss": 0.4879,
+      "step": 11840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5570684895557765,
+      "learning_rate": 8.352732846045957e-06,
+      "loss": 0.154,
+      "step": 11841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7368525816866303,
+      "learning_rate": 8.352440533068011e-06,
+      "loss": 0.4284,
+      "step": 11842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.52960071614655,
+      "learning_rate": 8.352148199272318e-06,
+      "loss": 0.152,
+      "step": 11843
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4365684299956523,
+      "learning_rate": 8.351855844660695e-06,
+      "loss": 0.2249,
+      "step": 11844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.485062986756705,
+      "learning_rate": 8.351563469234955e-06,
+      "loss": 0.6986,
+      "step": 11845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8714691335065026,
+      "learning_rate": 8.351271072996916e-06,
+      "loss": 0.5452,
+      "step": 11846
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.107781831123738,
+      "learning_rate": 8.350978655948393e-06,
+      "loss": 0.2249,
+      "step": 11847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4151672794863694,
+      "learning_rate": 8.3506862180912e-06,
+      "loss": 0.3208,
+      "step": 11848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7386653122256333,
+      "learning_rate": 8.350393759427153e-06,
+      "loss": 0.2852,
+      "step": 11849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5649781030884138,
+      "learning_rate": 8.350101279958071e-06,
+      "loss": 0.3125,
+      "step": 11850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9900705309390399,
+      "learning_rate": 8.349808779685769e-06,
+      "loss": 0.657,
+      "step": 11851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6033493587703531,
+      "learning_rate": 8.349516258612063e-06,
+      "loss": 0.3905,
+      "step": 11852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8639699115125679,
+      "learning_rate": 8.349223716738769e-06,
+      "loss": 0.3416,
+      "step": 11853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1214329042689757,
+      "learning_rate": 8.348931154067706e-06,
+      "loss": 0.5544,
+      "step": 11854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5395671064465257,
+      "learning_rate": 8.348638570600684e-06,
+      "loss": 0.1688,
+      "step": 11855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8915118746156625,
+      "learning_rate": 8.348345966339527e-06,
+      "loss": 0.3208,
+      "step": 11856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.676185593585573,
+      "learning_rate": 8.348053341286051e-06,
+      "loss": 0.2189,
+      "step": 11857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3887135463208415,
+      "learning_rate": 8.347760695442072e-06,
+      "loss": 0.346,
+      "step": 11858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3606482118679855,
+      "learning_rate": 8.347468028809407e-06,
+      "loss": 0.2211,
+      "step": 11859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.524300032755068,
+      "learning_rate": 8.347175341389872e-06,
+      "loss": 0.1889,
+      "step": 11860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.480831046551685,
+      "learning_rate": 8.346882633185287e-06,
+      "loss": 0.2866,
+      "step": 11861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.69325349954139,
+      "learning_rate": 8.346589904197467e-06,
+      "loss": 0.3392,
+      "step": 11862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6584261731856604,
+      "learning_rate": 8.346297154428233e-06,
+      "loss": 0.3597,
+      "step": 11863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6112234925486038,
+      "learning_rate": 8.3460043838794e-06,
+      "loss": 0.4428,
+      "step": 11864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6792403868043275,
+      "learning_rate": 8.345711592552788e-06,
+      "loss": 0.2242,
+      "step": 11865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3570876224560529,
+      "learning_rate": 8.345418780450214e-06,
+      "loss": 0.2826,
+      "step": 11866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9147273310363984,
+      "learning_rate": 8.345125947573497e-06,
+      "loss": 0.3336,
+      "step": 11867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4579543776535024,
+      "learning_rate": 8.344833093924453e-06,
+      "loss": 0.4201,
+      "step": 11868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4232305913384442,
+      "learning_rate": 8.344540219504906e-06,
+      "loss": 0.3841,
+      "step": 11869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.179956983499504,
+      "learning_rate": 8.34424732431667e-06,
+      "loss": 0.7974,
+      "step": 11870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.354406370278529,
+      "learning_rate": 8.343954408361566e-06,
+      "loss": 0.1789,
+      "step": 11871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8805863864651642,
+      "learning_rate": 8.34366147164141e-06,
+      "loss": 0.2114,
+      "step": 11872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4942092829867806,
+      "learning_rate": 8.343368514158023e-06,
+      "loss": 0.2415,
+      "step": 11873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0487211943661319,
+      "learning_rate": 8.343075535913226e-06,
+      "loss": 0.1888,
+      "step": 11874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5145949422512965,
+      "learning_rate": 8.342782536908835e-06,
+      "loss": 0.3661,
+      "step": 11875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3095198343491958,
+      "learning_rate": 8.342489517146674e-06,
+      "loss": 0.1448,
+      "step": 11876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9177880755630325,
+      "learning_rate": 8.342196476628557e-06,
+      "loss": 0.3949,
+      "step": 11877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6725900849855495,
+      "learning_rate": 8.341903415356308e-06,
+      "loss": 0.3578,
+      "step": 11878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1222232045455993,
+      "learning_rate": 8.341610333331743e-06,
+      "loss": 0.192,
+      "step": 11879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6134932611919925,
+      "learning_rate": 8.341317230556685e-06,
+      "loss": 0.3842,
+      "step": 11880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.059522207683222,
+      "learning_rate": 8.341024107032953e-06,
+      "loss": 0.3626,
+      "step": 11881
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.110840810880984,
+      "learning_rate": 8.34073096276237e-06,
+      "loss": 0.6477,
+      "step": 11882
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2155994689633496,
+      "learning_rate": 8.34043779774675e-06,
+      "loss": 0.222,
+      "step": 11883
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0702742292182035,
+      "learning_rate": 8.340144611987918e-06,
+      "loss": 0.1482,
+      "step": 11884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6225065796396159,
+      "learning_rate": 8.339851405487696e-06,
+      "loss": 0.2556,
+      "step": 11885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5522681729983467,
+      "learning_rate": 8.3395581782479e-06,
+      "loss": 0.3579,
+      "step": 11886
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3978344293567475,
+      "learning_rate": 8.339264930270355e-06,
+      "loss": 0.3443,
+      "step": 11887
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4269179737973459,
+      "learning_rate": 8.338971661556879e-06,
+      "loss": 0.3633,
+      "step": 11888
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5151112167370966,
+      "learning_rate": 8.338678372109297e-06,
+      "loss": 0.3456,
+      "step": 11889
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2585543454070227,
+      "learning_rate": 8.338385061929427e-06,
+      "loss": 0.2062,
+      "step": 11890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0580411817255206,
+      "learning_rate": 8.33809173101909e-06,
+      "loss": 0.2945,
+      "step": 11891
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0773109666180818,
+      "learning_rate": 8.337798379380109e-06,
+      "loss": 0.2033,
+      "step": 11892
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3401861080093869,
+      "learning_rate": 8.337505007014305e-06,
+      "loss": 0.2624,
+      "step": 11893
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.846730946152705,
+      "learning_rate": 8.3372116139235e-06,
+      "loss": 0.441,
+      "step": 11894
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6507417745424275,
+      "learning_rate": 8.336918200109517e-06,
+      "loss": 0.3587,
+      "step": 11895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.412385499372822,
+      "learning_rate": 8.336624765574176e-06,
+      "loss": 0.2886,
+      "step": 11896
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8321966897733033,
+      "learning_rate": 8.3363313103193e-06,
+      "loss": 0.4298,
+      "step": 11897
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8915234708643223,
+      "learning_rate": 8.336037834346714e-06,
+      "loss": 0.325,
+      "step": 11898
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.334158331725664,
+      "learning_rate": 8.335744337658233e-06,
+      "loss": 0.3788,
+      "step": 11899
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8956850746281135,
+      "learning_rate": 8.335450820255688e-06,
+      "loss": 0.4155,
+      "step": 11900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5129068718000178,
+      "learning_rate": 8.335157282140898e-06,
+      "loss": 0.5714,
+      "step": 11901
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4544215929365447,
+      "learning_rate": 8.334863723315685e-06,
+      "loss": 0.1383,
+      "step": 11902
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5650975975135724,
+      "learning_rate": 8.334570143781872e-06,
+      "loss": 0.2053,
+      "step": 11903
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2283023776508477,
+      "learning_rate": 8.334276543541285e-06,
+      "loss": 0.2638,
+      "step": 11904
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4456856838638912,
+      "learning_rate": 8.333982922595743e-06,
+      "loss": 0.1927,
+      "step": 11905
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0153442897055047,
+      "learning_rate": 8.33368928094707e-06,
+      "loss": 0.2279,
+      "step": 11906
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4330214233708714,
+      "learning_rate": 8.333395618597093e-06,
+      "loss": 0.2967,
+      "step": 11907
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.49254082695793,
+      "learning_rate": 8.333101935547632e-06,
+      "loss": 0.4419,
+      "step": 11908
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7362072133380442,
+      "learning_rate": 8.332808231800513e-06,
+      "loss": 0.5167,
+      "step": 11909
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0883312670322502,
+      "learning_rate": 8.332514507357558e-06,
+      "loss": 0.4465,
+      "step": 11910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5292462599990635,
+      "learning_rate": 8.332220762220593e-06,
+      "loss": 0.2046,
+      "step": 11911
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5234778129916993,
+      "learning_rate": 8.33192699639144e-06,
+      "loss": 0.3068,
+      "step": 11912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.917551644026431,
+      "learning_rate": 8.331633209871924e-06,
+      "loss": 0.339,
+      "step": 11913
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.691055724051207,
+      "learning_rate": 8.331339402663869e-06,
+      "loss": 0.268,
+      "step": 11914
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6857973887262998,
+      "learning_rate": 8.331045574769099e-06,
+      "loss": 0.2338,
+      "step": 11915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1806117299410985,
+      "learning_rate": 8.330751726189441e-06,
+      "loss": 0.1719,
+      "step": 11916
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9531888855976254,
+      "learning_rate": 8.330457856926717e-06,
+      "loss": 0.3945,
+      "step": 11917
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6979856851266273,
+      "learning_rate": 8.330163966982754e-06,
+      "loss": 0.2853,
+      "step": 11918
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0858461801292763,
+      "learning_rate": 8.329870056359374e-06,
+      "loss": 0.2044,
+      "step": 11919
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3005431740635331,
+      "learning_rate": 8.329576125058406e-06,
+      "loss": 0.1797,
+      "step": 11920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5032646532176408,
+      "learning_rate": 8.329282173081673e-06,
+      "loss": 0.2007,
+      "step": 11921
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3869448329755236,
+      "learning_rate": 8.328988200431e-06,
+      "loss": 0.156,
+      "step": 11922
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.645712118171768,
+      "learning_rate": 8.328694207108213e-06,
+      "loss": 0.4052,
+      "step": 11923
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6635507504146116,
+      "learning_rate": 8.328400193115138e-06,
+      "loss": 0.418,
+      "step": 11924
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4046117127249111,
+      "learning_rate": 8.328106158453602e-06,
+      "loss": 0.2345,
+      "step": 11925
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3869089721448746,
+      "learning_rate": 8.327812103125427e-06,
+      "loss": 0.2773,
+      "step": 11926
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.498606421900615,
+      "learning_rate": 8.327518027132443e-06,
+      "loss": 0.3002,
+      "step": 11927
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.078954667276948,
+      "learning_rate": 8.327223930476474e-06,
+      "loss": 0.2602,
+      "step": 11928
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.727347302400826,
+      "learning_rate": 8.326929813159347e-06,
+      "loss": 0.2709,
+      "step": 11929
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.232686952838874,
+      "learning_rate": 8.326635675182886e-06,
+      "loss": 0.2289,
+      "step": 11930
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6056403836396484,
+      "learning_rate": 8.326341516548924e-06,
+      "loss": 0.3615,
+      "step": 11931
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3054805646643677,
+      "learning_rate": 8.32604733725928e-06,
+      "loss": 0.2645,
+      "step": 11932
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8281593971602905,
+      "learning_rate": 8.325753137315785e-06,
+      "loss": 0.3888,
+      "step": 11933
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.460079627172267,
+      "learning_rate": 8.325458916720264e-06,
+      "loss": 0.1905,
+      "step": 11934
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6561813047845908,
+      "learning_rate": 8.325164675474546e-06,
+      "loss": 0.3731,
+      "step": 11935
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1024985766347575,
+      "learning_rate": 8.324870413580457e-06,
+      "loss": 0.0871,
+      "step": 11936
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9976267144470516,
+      "learning_rate": 8.324576131039825e-06,
+      "loss": 0.6133,
+      "step": 11937
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5076518986216447,
+      "learning_rate": 8.324281827854477e-06,
+      "loss": 0.494,
+      "step": 11938
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9689762746115063,
+      "learning_rate": 8.323987504026238e-06,
+      "loss": 0.3686,
+      "step": 11939
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3194030755930644,
+      "learning_rate": 8.32369315955694e-06,
+      "loss": 0.1779,
+      "step": 11940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4228125038624544,
+      "learning_rate": 8.32339879444841e-06,
+      "loss": 0.2597,
+      "step": 11941
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8651820470050642,
+      "learning_rate": 8.323104408702473e-06,
+      "loss": 0.4609,
+      "step": 11942
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.746113139136616,
+      "learning_rate": 8.322810002320958e-06,
+      "loss": 0.202,
+      "step": 11943
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.371621405313722,
+      "learning_rate": 8.322515575305697e-06,
+      "loss": 0.3061,
+      "step": 11944
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9449697377516277,
+      "learning_rate": 8.322221127658515e-06,
+      "loss": 0.1309,
+      "step": 11945
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4011301301387227,
+      "learning_rate": 8.32192665938124e-06,
+      "loss": 0.2713,
+      "step": 11946
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6948526149822822,
+      "learning_rate": 8.3216321704757e-06,
+      "loss": 0.3626,
+      "step": 11947
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4061839353089887,
+      "learning_rate": 8.321337660943726e-06,
+      "loss": 0.3302,
+      "step": 11948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.841593264897784,
+      "learning_rate": 8.321043130787147e-06,
+      "loss": 0.3686,
+      "step": 11949
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.601310776511945,
+      "learning_rate": 8.320748580007789e-06,
+      "loss": 0.3633,
+      "step": 11950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7458777198027429,
+      "learning_rate": 8.320454008607483e-06,
+      "loss": 0.3725,
+      "step": 11951
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8513215145590778,
+      "learning_rate": 8.32015941658806e-06,
+      "loss": 0.5215,
+      "step": 11952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4446247217069645,
+      "learning_rate": 8.319864803951347e-06,
+      "loss": 0.2817,
+      "step": 11953
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0222301636995224,
+      "learning_rate": 8.319570170699172e-06,
+      "loss": 0.5176,
+      "step": 11954
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.188899301214588,
+      "learning_rate": 8.31927551683337e-06,
+      "loss": 0.1854,
+      "step": 11955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3476109372689575,
+      "learning_rate": 8.318980842355763e-06,
+      "loss": 0.2187,
+      "step": 11956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.474889637669388,
+      "learning_rate": 8.31868614726819e-06,
+      "loss": 0.284,
+      "step": 11957
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4967931327850696,
+      "learning_rate": 8.318391431572472e-06,
+      "loss": 0.372,
+      "step": 11958
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7629778521964707,
+      "learning_rate": 8.318096695270445e-06,
+      "loss": 0.3319,
+      "step": 11959
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.529867796457117,
+      "learning_rate": 8.317801938363936e-06,
+      "loss": 0.3171,
+      "step": 11960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9273625704214048,
+      "learning_rate": 8.31750716085478e-06,
+      "loss": 0.2974,
+      "step": 11961
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.380009882199183,
+      "learning_rate": 8.3172123627448e-06,
+      "loss": 0.472,
+      "step": 11962
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4579305660278987,
+      "learning_rate": 8.316917544035833e-06,
+      "loss": 0.3769,
+      "step": 11963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.445263918498334,
+      "learning_rate": 8.316622704729709e-06,
+      "loss": 0.1634,
+      "step": 11964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.560121845998786,
+      "learning_rate": 8.316327844828255e-06,
+      "loss": 0.2191,
+      "step": 11965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.587258948775694,
+      "learning_rate": 8.316032964333306e-06,
+      "loss": 0.2944,
+      "step": 11966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2365301902669759,
+      "learning_rate": 8.31573806324669e-06,
+      "loss": 0.2673,
+      "step": 11967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7775808747326922,
+      "learning_rate": 8.315443141570242e-06,
+      "loss": 0.1913,
+      "step": 11968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.845976390490298,
+      "learning_rate": 8.315148199305791e-06,
+      "loss": 0.5329,
+      "step": 11969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.353445389812041,
+      "learning_rate": 8.314853236455167e-06,
+      "loss": 0.2062,
+      "step": 11970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3209181417901752,
+      "learning_rate": 8.314558253020205e-06,
+      "loss": 0.2814,
+      "step": 11971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6015655657110879,
+      "learning_rate": 8.314263249002734e-06,
+      "loss": 0.3176,
+      "step": 11972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6663803381141513,
+      "learning_rate": 8.313968224404589e-06,
+      "loss": 0.4918,
+      "step": 11973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.44462573772437,
+      "learning_rate": 8.313673179227599e-06,
+      "loss": 0.2544,
+      "step": 11974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.004851319167427,
+      "learning_rate": 8.313378113473598e-06,
+      "loss": 0.2505,
+      "step": 11975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.946414778841467,
+      "learning_rate": 8.313083027144414e-06,
+      "loss": 0.3702,
+      "step": 11976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.448550347784931,
+      "learning_rate": 8.312787920241887e-06,
+      "loss": 0.2713,
+      "step": 11977
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6807203798760848,
+      "learning_rate": 8.312492792767844e-06,
+      "loss": 0.4539,
+      "step": 11978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.765035277600339,
+      "learning_rate": 8.31219764472412e-06,
+      "loss": 0.6076,
+      "step": 11979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9243729492800998,
+      "learning_rate": 8.311902476112545e-06,
+      "loss": 0.5395,
+      "step": 11980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0976480843454093,
+      "learning_rate": 8.311607286934955e-06,
+      "loss": 0.1967,
+      "step": 11981
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4548299667423272,
+      "learning_rate": 8.311312077193183e-06,
+      "loss": 0.3043,
+      "step": 11982
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4486837582053758,
+      "learning_rate": 8.31101684688906e-06,
+      "loss": 0.32,
+      "step": 11983
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3421995101642272,
+      "learning_rate": 8.31072159602442e-06,
+      "loss": 0.2155,
+      "step": 11984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2942311586583966,
+      "learning_rate": 8.310426324601098e-06,
+      "loss": 0.1704,
+      "step": 11985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0373055865556693,
+      "learning_rate": 8.310131032620924e-06,
+      "loss": 0.6172,
+      "step": 11986
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.035094487210395,
+      "learning_rate": 8.309835720085735e-06,
+      "loss": 0.3119,
+      "step": 11987
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.280088624483364,
+      "learning_rate": 8.309540386997366e-06,
+      "loss": 0.1614,
+      "step": 11988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9927362394521768,
+      "learning_rate": 8.309245033357646e-06,
+      "loss": 0.237,
+      "step": 11989
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6595976574182467,
+      "learning_rate": 8.308949659168412e-06,
+      "loss": 0.4432,
+      "step": 11990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7068167929291025,
+      "learning_rate": 8.3086542644315e-06,
+      "loss": 0.336,
+      "step": 11991
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6348811374265897,
+      "learning_rate": 8.30835884914874e-06,
+      "loss": 0.4498,
+      "step": 11992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.394690461780635,
+      "learning_rate": 8.308063413321971e-06,
+      "loss": 0.1719,
+      "step": 11993
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.905207935421504,
+      "learning_rate": 8.307767956953026e-06,
+      "loss": 0.4896,
+      "step": 11994
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4441848531424637,
+      "learning_rate": 8.307472480043737e-06,
+      "loss": 0.1886,
+      "step": 11995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2778151171756118,
+      "learning_rate": 8.30717698259594e-06,
+      "loss": 0.1599,
+      "step": 11996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2965866466318536,
+      "learning_rate": 8.306881464611474e-06,
+      "loss": 0.2089,
+      "step": 11997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7865861220480457,
+      "learning_rate": 8.306585926092168e-06,
+      "loss": 0.5225,
+      "step": 11998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3200320899740174,
+      "learning_rate": 8.306290367039862e-06,
+      "loss": 0.2026,
+      "step": 11999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3568957564770785,
+      "learning_rate": 8.30599478745639e-06,
+      "loss": 0.2123,
+      "step": 12000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6535785831583827,
+      "learning_rate": 8.305699187343586e-06,
+      "loss": 0.237,
+      "step": 12001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.092142747908934,
+      "learning_rate": 8.305403566703287e-06,
+      "loss": 0.5113,
+      "step": 12002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5952723002062512,
+      "learning_rate": 8.305107925537327e-06,
+      "loss": 0.3778,
+      "step": 12003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8054735032616989,
+      "learning_rate": 8.304812263847544e-06,
+      "loss": 0.4239,
+      "step": 12004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9945304332691138,
+      "learning_rate": 8.304516581635773e-06,
+      "loss": 0.7238,
+      "step": 12005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9728496796603046,
+      "learning_rate": 8.30422087890385e-06,
+      "loss": 0.4015,
+      "step": 12006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3465767088667384,
+      "learning_rate": 8.303925155653612e-06,
+      "loss": 0.2382,
+      "step": 12007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.308432574808396,
+      "learning_rate": 8.303629411886895e-06,
+      "loss": 0.1783,
+      "step": 12008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.512319024409938,
+      "learning_rate": 8.303333647605534e-06,
+      "loss": 0.2575,
+      "step": 12009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4008502085694676,
+      "learning_rate": 8.303037862811368e-06,
+      "loss": 0.3879,
+      "step": 12010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4849591385753835,
+      "learning_rate": 8.302742057506233e-06,
+      "loss": 0.3369,
+      "step": 12011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5989517937649709,
+      "learning_rate": 8.302446231691966e-06,
+      "loss": 0.2295,
+      "step": 12012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3994996802685533,
+      "learning_rate": 8.302150385370402e-06,
+      "loss": 0.1833,
+      "step": 12013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7604631402767748,
+      "learning_rate": 8.30185451854338e-06,
+      "loss": 0.5932,
+      "step": 12014
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2879342227691761,
+      "learning_rate": 8.301558631212739e-06,
+      "loss": 0.1532,
+      "step": 12015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.58136468667435,
+      "learning_rate": 8.301262723380312e-06,
+      "loss": 0.4636,
+      "step": 12016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6823507885144853,
+      "learning_rate": 8.30096679504794e-06,
+      "loss": 0.241,
+      "step": 12017
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2803164968807634,
+      "learning_rate": 8.300670846217457e-06,
+      "loss": 0.2249,
+      "step": 12018
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0695817773453564,
+      "learning_rate": 8.300374876890706e-06,
+      "loss": 0.4423,
+      "step": 12019
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9437806942714917,
+      "learning_rate": 8.300078887069522e-06,
+      "loss": 0.481,
+      "step": 12020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2672450946226395,
+      "learning_rate": 8.299782876755742e-06,
+      "loss": 0.1837,
+      "step": 12021
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6524886119640576,
+      "learning_rate": 8.299486845951206e-06,
+      "loss": 0.3306,
+      "step": 12022
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.46923922192945,
+      "learning_rate": 8.29919079465775e-06,
+      "loss": 0.3303,
+      "step": 12023
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5519031531015217,
+      "learning_rate": 8.298894722877216e-06,
+      "loss": 0.3114,
+      "step": 12024
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.843270756781443,
+      "learning_rate": 8.298598630611438e-06,
+      "loss": 0.5902,
+      "step": 12025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3849141333163455,
+      "learning_rate": 8.29830251786226e-06,
+      "loss": 0.2151,
+      "step": 12026
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5704801811536255,
+      "learning_rate": 8.298006384631516e-06,
+      "loss": 0.4307,
+      "step": 12027
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.567038971926322,
+      "learning_rate": 8.297710230921047e-06,
+      "loss": 0.2766,
+      "step": 12028
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.403196479361523,
+      "learning_rate": 8.297414056732693e-06,
+      "loss": 0.2847,
+      "step": 12029
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6280625162700888,
+      "learning_rate": 8.29711786206829e-06,
+      "loss": 0.2234,
+      "step": 12030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5491382245149021,
+      "learning_rate": 8.296821646929683e-06,
+      "loss": 0.2491,
+      "step": 12031
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8351908281429474,
+      "learning_rate": 8.296525411318704e-06,
+      "loss": 0.4283,
+      "step": 12032
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8076519707744423,
+      "learning_rate": 8.296229155237198e-06,
+      "loss": 0.3408,
+      "step": 12033
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9287935765585567,
+      "learning_rate": 8.295932878687003e-06,
+      "loss": 0.4669,
+      "step": 12034
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9854158465199279,
+      "learning_rate": 8.295636581669957e-06,
+      "loss": 0.5144,
+      "step": 12035
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9603778034882028,
+      "learning_rate": 8.295340264187902e-06,
+      "loss": 0.1739,
+      "step": 12036
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0414291655472474,
+      "learning_rate": 8.295043926242678e-06,
+      "loss": 0.1863,
+      "step": 12037
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8113651833723983,
+      "learning_rate": 8.294747567836126e-06,
+      "loss": 0.2933,
+      "step": 12038
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6081492636905734,
+      "learning_rate": 8.294451188970085e-06,
+      "loss": 0.2238,
+      "step": 12039
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7380128246035593,
+      "learning_rate": 8.294154789646394e-06,
+      "loss": 0.333,
+      "step": 12040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3620793310013055,
+      "learning_rate": 8.293858369866896e-06,
+      "loss": 0.2003,
+      "step": 12041
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5433585724747683,
+      "learning_rate": 8.293561929633432e-06,
+      "loss": 0.2538,
+      "step": 12042
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4579358680384975,
+      "learning_rate": 8.29326546894784e-06,
+      "loss": 0.2289,
+      "step": 12043
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2668086215782353,
+      "learning_rate": 8.292968987811965e-06,
+      "loss": 0.22,
+      "step": 12044
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.840638068206198,
+      "learning_rate": 8.292672486227642e-06,
+      "loss": 0.2832,
+      "step": 12045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7855277353152639,
+      "learning_rate": 8.292375964196718e-06,
+      "loss": 0.4114,
+      "step": 12046
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2597117715606176,
+      "learning_rate": 8.292079421721032e-06,
+      "loss": 0.2126,
+      "step": 12047
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3452428835596375,
+      "learning_rate": 8.291782858802426e-06,
+      "loss": 0.1604,
+      "step": 12048
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9217480136818397,
+      "learning_rate": 8.291486275442741e-06,
+      "loss": 0.2192,
+      "step": 12049
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2387758643872986,
+      "learning_rate": 8.291189671643818e-06,
+      "loss": 0.331,
+      "step": 12050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2890514269355915,
+      "learning_rate": 8.2908930474075e-06,
+      "loss": 0.4857,
+      "step": 12051
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8481606788887281,
+      "learning_rate": 8.29059640273563e-06,
+      "loss": 0.3713,
+      "step": 12052
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6680757412331846,
+      "learning_rate": 8.290299737630048e-06,
+      "loss": 0.2172,
+      "step": 12053
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5309783538892752,
+      "learning_rate": 8.290003052092598e-06,
+      "loss": 0.1678,
+      "step": 12054
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4023411307137768,
+      "learning_rate": 8.289706346125119e-06,
+      "loss": 0.3466,
+      "step": 12055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9198009573190231,
+      "learning_rate": 8.289409619729457e-06,
+      "loss": 0.4863,
+      "step": 12056
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4743321720313025,
+      "learning_rate": 8.289112872907454e-06,
+      "loss": 0.3187,
+      "step": 12057
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.801228990974977,
+      "learning_rate": 8.288816105660951e-06,
+      "loss": 0.6438,
+      "step": 12058
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4920333313315723,
+      "learning_rate": 8.288519317991792e-06,
+      "loss": 0.2752,
+      "step": 12059
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3624345614966262,
+      "learning_rate": 8.28822250990182e-06,
+      "loss": 0.3625,
+      "step": 12060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5943299762240999,
+      "learning_rate": 8.287925681392879e-06,
+      "loss": 0.2389,
+      "step": 12061
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7883553835800257,
+      "learning_rate": 8.28762883246681e-06,
+      "loss": 0.2162,
+      "step": 12062
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2783304566616587,
+      "learning_rate": 8.28733196312546e-06,
+      "loss": 0.3233,
+      "step": 12063
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0073450041492578,
+      "learning_rate": 8.287035073370666e-06,
+      "loss": 0.5298,
+      "step": 12064
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.480549056538095,
+      "learning_rate": 8.28673816320428e-06,
+      "loss": 0.2779,
+      "step": 12065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6758553637497073,
+      "learning_rate": 8.286441232628139e-06,
+      "loss": 0.2126,
+      "step": 12066
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7846872727880398,
+      "learning_rate": 8.286144281644088e-06,
+      "loss": 0.4518,
+      "step": 12067
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4804135390280937,
+      "learning_rate": 8.285847310253975e-06,
+      "loss": 0.1778,
+      "step": 12068
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4323690390640502,
+      "learning_rate": 8.28555031845964e-06,
+      "loss": 0.2774,
+      "step": 12069
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8873201324010933,
+      "learning_rate": 8.285253306262927e-06,
+      "loss": 0.554,
+      "step": 12070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.997379008640483,
+      "learning_rate": 8.284956273665685e-06,
+      "loss": 0.2695,
+      "step": 12071
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6133924565999715,
+      "learning_rate": 8.284659220669754e-06,
+      "loss": 0.1682,
+      "step": 12072
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5513399880810488,
+      "learning_rate": 8.28436214727698e-06,
+      "loss": 0.3723,
+      "step": 12073
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6285649772297563,
+      "learning_rate": 8.28406505348921e-06,
+      "loss": 0.3576,
+      "step": 12074
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4375860898527122,
+      "learning_rate": 8.283767939308283e-06,
+      "loss": 0.3195,
+      "step": 12075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.671588387387812,
+      "learning_rate": 8.28347080473605e-06,
+      "loss": 0.4252,
+      "step": 12076
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8979348392398203,
+      "learning_rate": 8.283173649774353e-06,
+      "loss": 0.3918,
+      "step": 12077
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5533591518968877,
+      "learning_rate": 8.282876474425039e-06,
+      "loss": 0.3609,
+      "step": 12078
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6913983386693858,
+      "learning_rate": 8.28257927868995e-06,
+      "loss": 0.3648,
+      "step": 12079
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7990386789960322,
+      "learning_rate": 8.282282062570936e-06,
+      "loss": 0.1844,
+      "step": 12080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3214148879302923,
+      "learning_rate": 8.281984826069842e-06,
+      "loss": 0.3301,
+      "step": 12081
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8308596657210914,
+      "learning_rate": 8.281687569188509e-06,
+      "loss": 0.3835,
+      "step": 12082
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7627544615059747,
+      "learning_rate": 8.281390291928788e-06,
+      "loss": 0.4311,
+      "step": 12083
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7710432746267235,
+      "learning_rate": 8.281092994292523e-06,
+      "loss": 0.2508,
+      "step": 12084
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5789769205456567,
+      "learning_rate": 8.28079567628156e-06,
+      "loss": 0.3367,
+      "step": 12085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8213316209335983,
+      "learning_rate": 8.280498337897746e-06,
+      "loss": 0.1623,
+      "step": 12086
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4710932159828032,
+      "learning_rate": 8.280200979142927e-06,
+      "loss": 0.3074,
+      "step": 12087
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6087498237591982,
+      "learning_rate": 8.279903600018951e-06,
+      "loss": 0.2671,
+      "step": 12088
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.946314986190759,
+      "learning_rate": 8.27960620052766e-06,
+      "loss": 0.5143,
+      "step": 12089
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9206053680436965,
+      "learning_rate": 8.279308780670907e-06,
+      "loss": 0.64,
+      "step": 12090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4152533919751649,
+      "learning_rate": 8.279011340450534e-06,
+      "loss": 0.3294,
+      "step": 12091
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1591252665603746,
+      "learning_rate": 8.27871387986839e-06,
+      "loss": 0.1617,
+      "step": 12092
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.735060238575742,
+      "learning_rate": 8.278416398926322e-06,
+      "loss": 0.4688,
+      "step": 12093
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3721818009826994,
+      "learning_rate": 8.278118897626178e-06,
+      "loss": 0.1961,
+      "step": 12094
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6286568622278712,
+      "learning_rate": 8.277821375969805e-06,
+      "loss": 0.3768,
+      "step": 12095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9855926487382176,
+      "learning_rate": 8.277523833959049e-06,
+      "loss": 0.4453,
+      "step": 12096
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7420122567380953,
+      "learning_rate": 8.27722627159576e-06,
+      "loss": 0.3484,
+      "step": 12097
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4277893817705696,
+      "learning_rate": 8.276928688881783e-06,
+      "loss": 0.3359,
+      "step": 12098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7982526307990332,
+      "learning_rate": 8.276631085818969e-06,
+      "loss": 0.4485,
+      "step": 12099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.359260145764426,
+      "learning_rate": 8.276333462409165e-06,
+      "loss": 0.5222,
+      "step": 12100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9609811519612335,
+      "learning_rate": 8.276035818654217e-06,
+      "loss": 0.39,
+      "step": 12101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6174069177429504,
+      "learning_rate": 8.275738154555978e-06,
+      "loss": 0.3328,
+      "step": 12102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.422610095073609,
+      "learning_rate": 8.275440470116291e-06,
+      "loss": 0.3192,
+      "step": 12103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9479845866221026,
+      "learning_rate": 8.275142765337007e-06,
+      "loss": 0.3106,
+      "step": 12104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4765602041156016,
+      "learning_rate": 8.274845040219976e-06,
+      "loss": 0.3441,
+      "step": 12105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.980709877609017,
+      "learning_rate": 8.274547294767044e-06,
+      "loss": 0.1223,
+      "step": 12106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3392389659159116,
+      "learning_rate": 8.274249528980063e-06,
+      "loss": 0.1608,
+      "step": 12107
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5222201533154374,
+      "learning_rate": 8.27395174286088e-06,
+      "loss": 0.2744,
+      "step": 12108
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.687816060455697,
+      "learning_rate": 8.273653936411344e-06,
+      "loss": 0.5335,
+      "step": 12109
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8900421126785414,
+      "learning_rate": 8.273356109633305e-06,
+      "loss": 0.2095,
+      "step": 12110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8100761307761324,
+      "learning_rate": 8.273058262528613e-06,
+      "loss": 0.191,
+      "step": 12111
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2416497317048414,
+      "learning_rate": 8.272760395099115e-06,
+      "loss": 0.1462,
+      "step": 12112
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5149412874295403,
+      "learning_rate": 8.272462507346664e-06,
+      "loss": 0.2693,
+      "step": 12113
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8508473878430192,
+      "learning_rate": 8.27216459927311e-06,
+      "loss": 0.4683,
+      "step": 12114
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8395427692544846,
+      "learning_rate": 8.271866670880298e-06,
+      "loss": 0.5481,
+      "step": 12115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.660711328609404,
+      "learning_rate": 8.271568722170082e-06,
+      "loss": 0.201,
+      "step": 12116
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8464012642159011,
+      "learning_rate": 8.271270753144313e-06,
+      "loss": 0.2882,
+      "step": 12117
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2990326455076266,
+      "learning_rate": 8.27097276380484e-06,
+      "loss": 0.3278,
+      "step": 12118
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3297937988350017,
+      "learning_rate": 8.270674754153512e-06,
+      "loss": 0.3418,
+      "step": 12119
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3430824118510423,
+      "learning_rate": 8.270376724192181e-06,
+      "loss": 0.327,
+      "step": 12120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6695325263700842,
+      "learning_rate": 8.270078673922697e-06,
+      "loss": 0.2004,
+      "step": 12121
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6714222508898724,
+      "learning_rate": 8.269780603346913e-06,
+      "loss": 0.4961,
+      "step": 12122
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6951211236793098,
+      "learning_rate": 8.269482512466676e-06,
+      "loss": 0.1866,
+      "step": 12123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6915546136323751,
+      "learning_rate": 8.26918440128384e-06,
+      "loss": 0.235,
+      "step": 12124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3718337360749917,
+      "learning_rate": 8.268886269800258e-06,
+      "loss": 0.2475,
+      "step": 12125
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7510220159533914,
+      "learning_rate": 8.268588118017774e-06,
+      "loss": 0.2728,
+      "step": 12126
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7763539628873002,
+      "learning_rate": 8.268289945938249e-06,
+      "loss": 0.3961,
+      "step": 12127
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7507009869127121,
+      "learning_rate": 8.267991753563527e-06,
+      "loss": 0.1886,
+      "step": 12128
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9760625380313863,
+      "learning_rate": 8.267693540895465e-06,
+      "loss": 0.3414,
+      "step": 12129
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9036020057754075,
+      "learning_rate": 8.26739530793591e-06,
+      "loss": 0.3574,
+      "step": 12130
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6513378720470269,
+      "learning_rate": 8.267097054686718e-06,
+      "loss": 0.1714,
+      "step": 12131
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9814873508075619,
+      "learning_rate": 8.266798781149738e-06,
+      "loss": 0.3423,
+      "step": 12132
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5212294740867212,
+      "learning_rate": 8.266500487326824e-06,
+      "loss": 0.2907,
+      "step": 12133
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3494410748027372,
+      "learning_rate": 8.266202173219828e-06,
+      "loss": 0.1499,
+      "step": 12134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6328060706710175,
+      "learning_rate": 8.265903838830602e-06,
+      "loss": 0.3755,
+      "step": 12135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8447774493346152,
+      "learning_rate": 8.265605484160998e-06,
+      "loss": 0.3953,
+      "step": 12136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3180658988671854,
+      "learning_rate": 8.265307109212872e-06,
+      "loss": 0.3241,
+      "step": 12137
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6455672210926418,
+      "learning_rate": 8.265008713988073e-06,
+      "loss": 0.4482,
+      "step": 12138
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0121539895042417,
+      "learning_rate": 8.264710298488456e-06,
+      "loss": 0.4414,
+      "step": 12139
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4652346619938452,
+      "learning_rate": 8.264411862715872e-06,
+      "loss": 0.3601,
+      "step": 12140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1670341126631194,
+      "learning_rate": 8.264113406672176e-06,
+      "loss": 0.4795,
+      "step": 12141
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2589125768704448,
+      "learning_rate": 8.263814930359223e-06,
+      "loss": 0.1704,
+      "step": 12142
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8389751060288224,
+      "learning_rate": 8.263516433778864e-06,
+      "loss": 0.5692,
+      "step": 12143
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9853684424167322,
+      "learning_rate": 8.263217916932951e-06,
+      "loss": 0.3226,
+      "step": 12144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.710718541526848,
+      "learning_rate": 8.262919379823342e-06,
+      "loss": 0.4439,
+      "step": 12145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9947205203809062,
+      "learning_rate": 8.262620822451886e-06,
+      "loss": 0.4629,
+      "step": 12146
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7266550352146723,
+      "learning_rate": 8.262322244820442e-06,
+      "loss": 0.1789,
+      "step": 12147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6107778869721092,
+      "learning_rate": 8.262023646930859e-06,
+      "loss": 0.4899,
+      "step": 12148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9971363090620068,
+      "learning_rate": 8.261725028784998e-06,
+      "loss": 0.1782,
+      "step": 12149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8068070701014252,
+      "learning_rate": 8.261426390384705e-06,
+      "loss": 0.534,
+      "step": 12150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7198139495569014,
+      "learning_rate": 8.261127731731841e-06,
+      "loss": 0.449,
+      "step": 12151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6962896247120542,
+      "learning_rate": 8.260829052828256e-06,
+      "loss": 0.364,
+      "step": 12152
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9246499373394963,
+      "learning_rate": 8.260530353675809e-06,
+      "loss": 0.2934,
+      "step": 12153
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.320398971329201,
+      "learning_rate": 8.260231634276354e-06,
+      "loss": 0.2159,
+      "step": 12154
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.756697207662064,
+      "learning_rate": 8.259932894631741e-06,
+      "loss": 0.4019,
+      "step": 12155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4085279240939539,
+      "learning_rate": 8.259634134743832e-06,
+      "loss": 0.3587,
+      "step": 12156
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5296275803166315,
+      "learning_rate": 8.259335354614477e-06,
+      "loss": 0.1887,
+      "step": 12157
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9211477400060415,
+      "learning_rate": 8.259036554245535e-06,
+      "loss": 0.5694,
+      "step": 12158
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2447596017812346,
+      "learning_rate": 8.25873773363886e-06,
+      "loss": 0.1876,
+      "step": 12159
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7401117737607703,
+      "learning_rate": 8.258438892796305e-06,
+      "loss": 0.3955,
+      "step": 12160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9567785271202354,
+      "learning_rate": 8.25814003171973e-06,
+      "loss": 0.6045,
+      "step": 12161
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7351425261815696,
+      "learning_rate": 8.257841150410991e-06,
+      "loss": 0.3018,
+      "step": 12162
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8019004992712837,
+      "learning_rate": 8.257542248871937e-06,
+      "loss": 0.1949,
+      "step": 12163
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.109460939316796,
+      "learning_rate": 8.257243327104435e-06,
+      "loss": 0.3001,
+      "step": 12164
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5287724274092052,
+      "learning_rate": 8.256944385110332e-06,
+      "loss": 0.1932,
+      "step": 12165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.453654494004081,
+      "learning_rate": 8.256645422891487e-06,
+      "loss": 0.1799,
+      "step": 12166
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5677253039973924,
+      "learning_rate": 8.256346440449758e-06,
+      "loss": 0.3948,
+      "step": 12167
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3701980768145345,
+      "learning_rate": 8.256047437787002e-06,
+      "loss": 0.23,
+      "step": 12168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1473036248340265,
+      "learning_rate": 8.255748414905073e-06,
+      "loss": 0.1966,
+      "step": 12169
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1744100680046217,
+      "learning_rate": 8.25544937180583e-06,
+      "loss": 0.1548,
+      "step": 12170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2104195368859307,
+      "learning_rate": 8.255150308491129e-06,
+      "loss": 0.4276,
+      "step": 12171
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.718509353783717,
+      "learning_rate": 8.254851224962827e-06,
+      "loss": 0.4151,
+      "step": 12172
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4381805472731208,
+      "learning_rate": 8.254552121222782e-06,
+      "loss": 0.3836,
+      "step": 12173
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.54973610075419,
+      "learning_rate": 8.254252997272852e-06,
+      "loss": 0.1864,
+      "step": 12174
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1410183897706476,
+      "learning_rate": 8.253953853114892e-06,
+      "loss": 0.1717,
+      "step": 12175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0742852708598916,
+      "learning_rate": 8.253654688750761e-06,
+      "loss": 0.456,
+      "step": 12176
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4955173158835648,
+      "learning_rate": 8.25335550418232e-06,
+      "loss": 0.1933,
+      "step": 12177
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5067602201611716,
+      "learning_rate": 8.253056299411419e-06,
+      "loss": 0.1471,
+      "step": 12178
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2812258380844648,
+      "learning_rate": 8.252757074439923e-06,
+      "loss": 0.243,
+      "step": 12179
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.596899832851574,
+      "learning_rate": 8.252457829269687e-06,
+      "loss": 0.2409,
+      "step": 12180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5674927384757413,
+      "learning_rate": 8.25215856390257e-06,
+      "loss": 0.3115,
+      "step": 12181
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.643227211960619,
+      "learning_rate": 8.251859278340433e-06,
+      "loss": 0.4859,
+      "step": 12182
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8019947216724546,
+      "learning_rate": 8.251559972585128e-06,
+      "loss": 0.2092,
+      "step": 12183
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.49187058683461,
+      "learning_rate": 8.251260646638519e-06,
+      "loss": 0.1888,
+      "step": 12184
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.200604335522291,
+      "learning_rate": 8.250961300502464e-06,
+      "loss": 0.3097,
+      "step": 12185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.578161064529017,
+      "learning_rate": 8.25066193417882e-06,
+      "loss": 0.2533,
+      "step": 12186
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5999082279149508,
+      "learning_rate": 8.250362547669448e-06,
+      "loss": 0.1635,
+      "step": 12187
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7442836126172812,
+      "learning_rate": 8.250063140976203e-06,
+      "loss": 0.5327,
+      "step": 12188
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3054265053680993,
+      "learning_rate": 8.24976371410095e-06,
+      "loss": 0.3021,
+      "step": 12189
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5459963056053223,
+      "learning_rate": 8.249464267045545e-06,
+      "loss": 0.35,
+      "step": 12190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.262227003282626,
+      "learning_rate": 8.24916479981185e-06,
+      "loss": 0.1722,
+      "step": 12191
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7114186394252677,
+      "learning_rate": 8.248865312401722e-06,
+      "loss": 0.3207,
+      "step": 12192
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9744413211898195,
+      "learning_rate": 8.24856580481702e-06,
+      "loss": 0.2852,
+      "step": 12193
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9764256316884385,
+      "learning_rate": 8.248266277059607e-06,
+      "loss": 0.356,
+      "step": 12194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1230735531475733,
+      "learning_rate": 8.24796672913134e-06,
+      "loss": 0.2958,
+      "step": 12195
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8212450487479284,
+      "learning_rate": 8.24766716103408e-06,
+      "loss": 0.266,
+      "step": 12196
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.112765786983724,
+      "learning_rate": 8.24736757276969e-06,
+      "loss": 0.5636,
+      "step": 12197
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8042783046211894,
+      "learning_rate": 8.247067964340028e-06,
+      "loss": 0.2607,
+      "step": 12198
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9796270203177526,
+      "learning_rate": 8.246768335746951e-06,
+      "loss": 0.2114,
+      "step": 12199
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7271463628176469,
+      "learning_rate": 8.246468686992327e-06,
+      "loss": 0.2029,
+      "step": 12200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.924496305977304,
+      "learning_rate": 8.246169018078012e-06,
+      "loss": 0.0991,
+      "step": 12201
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.848027754486967,
+      "learning_rate": 8.245869329005866e-06,
+      "loss": 0.4554,
+      "step": 12202
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5958447259294193,
+      "learning_rate": 8.245569619777753e-06,
+      "loss": 0.2644,
+      "step": 12203
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.242405119911984,
+      "learning_rate": 8.245269890395534e-06,
+      "loss": 0.2648,
+      "step": 12204
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5922679555104846,
+      "learning_rate": 8.244970140861069e-06,
+      "loss": 0.3993,
+      "step": 12205
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8558601709395053,
+      "learning_rate": 8.244670371176218e-06,
+      "loss": 0.442,
+      "step": 12206
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0901833820274538,
+      "learning_rate": 8.244370581342843e-06,
+      "loss": 0.136,
+      "step": 12207
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1935805679209597,
+      "learning_rate": 8.244070771362809e-06,
+      "loss": 0.2338,
+      "step": 12208
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6580725170231079,
+      "learning_rate": 8.243770941237975e-06,
+      "loss": 0.1648,
+      "step": 12209
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8121618785846063,
+      "learning_rate": 8.243471090970202e-06,
+      "loss": 0.6653,
+      "step": 12210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.482459868194266,
+      "learning_rate": 8.243171220561355e-06,
+      "loss": 0.4269,
+      "step": 12211
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.687021827943492,
+      "learning_rate": 8.242871330013294e-06,
+      "loss": 0.3785,
+      "step": 12212
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2819039032514905,
+      "learning_rate": 8.24257141932788e-06,
+      "loss": 0.2722,
+      "step": 12213
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7530191977766787,
+      "learning_rate": 8.242271488506977e-06,
+      "loss": 0.348,
+      "step": 12214
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.065789662548802,
+      "learning_rate": 8.241971537552449e-06,
+      "loss": 0.5948,
+      "step": 12215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4377898385339591,
+      "learning_rate": 8.241671566466155e-06,
+      "loss": 0.2171,
+      "step": 12216
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1845203343514819,
+      "learning_rate": 8.24137157524996e-06,
+      "loss": 0.215,
+      "step": 12217
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1217734115862428,
+      "learning_rate": 8.241071563905729e-06,
+      "loss": 0.2098,
+      "step": 12218
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9245492890573228,
+      "learning_rate": 8.24077153243532e-06,
+      "loss": 0.2637,
+      "step": 12219
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8068072206134447,
+      "learning_rate": 8.2404714808406e-06,
+      "loss": 0.2109,
+      "step": 12220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.340881756752039,
+      "learning_rate": 8.240171409123432e-06,
+      "loss": 0.1985,
+      "step": 12221
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5339636076844236,
+      "learning_rate": 8.239871317285675e-06,
+      "loss": 0.3749,
+      "step": 12222
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7684773344483504,
+      "learning_rate": 8.239571205329198e-06,
+      "loss": 0.496,
+      "step": 12223
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6398471759544087,
+      "learning_rate": 8.239271073255863e-06,
+      "loss": 0.4433,
+      "step": 12224
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6496791344286188,
+      "learning_rate": 8.238970921067531e-06,
+      "loss": 0.5068,
+      "step": 12225
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4169692636393072,
+      "learning_rate": 8.23867074876607e-06,
+      "loss": 0.3169,
+      "step": 12226
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5214222269746258,
+      "learning_rate": 8.23837055635334e-06,
+      "loss": 0.2861,
+      "step": 12227
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.577801385595955,
+      "learning_rate": 8.23807034383121e-06,
+      "loss": 0.3207,
+      "step": 12228
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5186984702481399,
+      "learning_rate": 8.23777011120154e-06,
+      "loss": 0.3145,
+      "step": 12229
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0211620165926663,
+      "learning_rate": 8.237469858466197e-06,
+      "loss": 0.3782,
+      "step": 12230
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.737584237931416,
+      "learning_rate": 8.237169585627041e-06,
+      "loss": 0.3557,
+      "step": 12231
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3624804147572598,
+      "learning_rate": 8.236869292685942e-06,
+      "loss": 0.1948,
+      "step": 12232
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7222743497676403,
+      "learning_rate": 8.236568979644762e-06,
+      "loss": 0.2301,
+      "step": 12233
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.365876059292529,
+      "learning_rate": 8.236268646505366e-06,
+      "loss": 0.2291,
+      "step": 12234
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.564842866125062,
+      "learning_rate": 8.235968293269621e-06,
+      "loss": 0.3572,
+      "step": 12235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8269113522415923,
+      "learning_rate": 8.23566791993939e-06,
+      "loss": 0.6211,
+      "step": 12236
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5308090671109813,
+      "learning_rate": 8.235367526516538e-06,
+      "loss": 0.2337,
+      "step": 12237
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0972683581659304,
+      "learning_rate": 8.23506711300293e-06,
+      "loss": 0.481,
+      "step": 12238
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2545990102405384,
+      "learning_rate": 8.234766679400434e-06,
+      "loss": 0.4253,
+      "step": 12239
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7632944436143796,
+      "learning_rate": 8.234466225710915e-06,
+      "loss": 0.3958,
+      "step": 12240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9005081752008137,
+      "learning_rate": 8.234165751936237e-06,
+      "loss": 0.5007,
+      "step": 12241
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6540916387620868,
+      "learning_rate": 8.233865258078268e-06,
+      "loss": 0.3855,
+      "step": 12242
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7251869716246158,
+      "learning_rate": 8.233564744138871e-06,
+      "loss": 0.2518,
+      "step": 12243
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6487697858505317,
+      "learning_rate": 8.233264210119915e-06,
+      "loss": 0.211,
+      "step": 12244
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4655432137401339,
+      "learning_rate": 8.232963656023265e-06,
+      "loss": 0.2677,
+      "step": 12245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5550046314579904,
+      "learning_rate": 8.232663081850786e-06,
+      "loss": 0.398,
+      "step": 12246
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7931428961859668,
+      "learning_rate": 8.232362487604349e-06,
+      "loss": 0.4319,
+      "step": 12247
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9731895670085378,
+      "learning_rate": 8.232061873285816e-06,
+      "loss": 0.4119,
+      "step": 12248
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3009912132374026,
+      "learning_rate": 8.231761238897054e-06,
+      "loss": 0.1905,
+      "step": 12249
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8241739206299104,
+      "learning_rate": 8.231460584439931e-06,
+      "loss": 0.1743,
+      "step": 12250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7253596400872215,
+      "learning_rate": 8.231159909916317e-06,
+      "loss": 0.499,
+      "step": 12251
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5712336263129716,
+      "learning_rate": 8.230859215328075e-06,
+      "loss": 0.5322,
+      "step": 12252
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7485203362853916,
+      "learning_rate": 8.230558500677073e-06,
+      "loss": 0.3538,
+      "step": 12253
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.224603853715195,
+      "learning_rate": 8.23025776596518e-06,
+      "loss": 0.266,
+      "step": 12254
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5225959899001464,
+      "learning_rate": 8.22995701119426e-06,
+      "loss": 0.3733,
+      "step": 12255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2977090022086437,
+      "learning_rate": 8.229656236366185e-06,
+      "loss": 0.2622,
+      "step": 12256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9478211523971911,
+      "learning_rate": 8.22935544148282e-06,
+      "loss": 0.5369,
+      "step": 12257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4916248084456343,
+      "learning_rate": 8.229054626546034e-06,
+      "loss": 0.4884,
+      "step": 12258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.635186627684955,
+      "learning_rate": 8.228753791557695e-06,
+      "loss": 0.1649,
+      "step": 12259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2924332806454564,
+      "learning_rate": 8.228452936519669e-06,
+      "loss": 0.2586,
+      "step": 12260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1892111803103842,
+      "learning_rate": 8.228152061433826e-06,
+      "loss": 0.1908,
+      "step": 12261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8807747644583739,
+      "learning_rate": 8.227851166302036e-06,
+      "loss": 0.4829,
+      "step": 12262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.278828093006186,
+      "learning_rate": 8.227550251126164e-06,
+      "loss": 0.5337,
+      "step": 12263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6191359958191374,
+      "learning_rate": 8.227249315908082e-06,
+      "loss": 0.5678,
+      "step": 12264
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1903044249024597,
+      "learning_rate": 8.226948360649657e-06,
+      "loss": 0.1934,
+      "step": 12265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3515186674805153,
+      "learning_rate": 8.226647385352756e-06,
+      "loss": 0.1691,
+      "step": 12266
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.771731848820994,
+      "learning_rate": 8.22634639001925e-06,
+      "loss": 0.2251,
+      "step": 12267
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7917142647327553,
+      "learning_rate": 8.226045374651008e-06,
+      "loss": 0.5148,
+      "step": 12268
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7571183635510088,
+      "learning_rate": 8.2257443392499e-06,
+      "loss": 0.2809,
+      "step": 12269
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1873148158890658,
+      "learning_rate": 8.225443283817795e-06,
+      "loss": 0.1468,
+      "step": 12270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7278777319391805,
+      "learning_rate": 8.22514220835656e-06,
+      "loss": 0.2025,
+      "step": 12271
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.10047434178313,
+      "learning_rate": 8.224841112868069e-06,
+      "loss": 0.3485,
+      "step": 12272
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.264131810329845,
+      "learning_rate": 8.224539997354188e-06,
+      "loss": 0.3866,
+      "step": 12273
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8654746974291427,
+      "learning_rate": 8.224238861816786e-06,
+      "loss": 0.3803,
+      "step": 12274
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3460950121649495,
+      "learning_rate": 8.223937706257739e-06,
+      "loss": 0.163,
+      "step": 12275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1787382950854104,
+      "learning_rate": 8.223636530678912e-06,
+      "loss": 0.3596,
+      "step": 12276
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.884536934848691,
+      "learning_rate": 8.223335335082174e-06,
+      "loss": 0.3128,
+      "step": 12277
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7824672839491542,
+      "learning_rate": 8.2230341194694e-06,
+      "loss": 0.2251,
+      "step": 12278
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6918776121848622,
+      "learning_rate": 8.222732883842459e-06,
+      "loss": 0.3189,
+      "step": 12279
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.000880178100831,
+      "learning_rate": 8.222431628203218e-06,
+      "loss": 0.3728,
+      "step": 12280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5007440092757423,
+      "learning_rate": 8.222130352553553e-06,
+      "loss": 0.4127,
+      "step": 12281
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7198381858874998,
+      "learning_rate": 8.221829056895333e-06,
+      "loss": 0.3888,
+      "step": 12282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8439836515665948,
+      "learning_rate": 8.221527741230426e-06,
+      "loss": 0.4301,
+      "step": 12283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5182619886029416,
+      "learning_rate": 8.221226405560706e-06,
+      "loss": 0.1773,
+      "step": 12284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.726909895940228,
+      "learning_rate": 8.220925049888045e-06,
+      "loss": 0.3225,
+      "step": 12285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4652903594434932,
+      "learning_rate": 8.220623674214313e-06,
+      "loss": 0.1931,
+      "step": 12286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9285752801629614,
+      "learning_rate": 8.220322278541379e-06,
+      "loss": 0.267,
+      "step": 12287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6474347816294928,
+      "learning_rate": 8.220020862871119e-06,
+      "loss": 0.3099,
+      "step": 12288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6637088395343036,
+      "learning_rate": 8.219719427205401e-06,
+      "loss": 0.3704,
+      "step": 12289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2240788049624771,
+      "learning_rate": 8.2194179715461e-06,
+      "loss": 0.1441,
+      "step": 12290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.012410639324964,
+      "learning_rate": 8.219116495895086e-06,
+      "loss": 0.5663,
+      "step": 12291
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0384633144888595,
+      "learning_rate": 8.218815000254233e-06,
+      "loss": 0.684,
+      "step": 12292
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.933350577620938,
+      "learning_rate": 8.21851348462541e-06,
+      "loss": 0.2194,
+      "step": 12293
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9492714823162507,
+      "learning_rate": 8.218211949010493e-06,
+      "loss": 0.6839,
+      "step": 12294
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8902891704124294,
+      "learning_rate": 8.217910393411351e-06,
+      "loss": 0.4048,
+      "step": 12295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9330117235917919,
+      "learning_rate": 8.21760881782986e-06,
+      "loss": 0.3264,
+      "step": 12296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7787855603896727,
+      "learning_rate": 8.217307222267887e-06,
+      "loss": 0.2222,
+      "step": 12297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7236568438445972,
+      "learning_rate": 8.217005606727313e-06,
+      "loss": 0.4969,
+      "step": 12298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4410634369750117,
+      "learning_rate": 8.216703971210005e-06,
+      "loss": 0.2197,
+      "step": 12299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7948651351511502,
+      "learning_rate": 8.216402315717837e-06,
+      "loss": 0.2775,
+      "step": 12300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8132998986951396,
+      "learning_rate": 8.216100640252683e-06,
+      "loss": 0.226,
+      "step": 12301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1514924195178455,
+      "learning_rate": 8.215798944816418e-06,
+      "loss": 0.3878,
+      "step": 12302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.644494355302073,
+      "learning_rate": 8.215497229410911e-06,
+      "loss": 0.2331,
+      "step": 12303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8378563734829672,
+      "learning_rate": 8.21519549403804e-06,
+      "loss": 0.4931,
+      "step": 12304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0182010135943047,
+      "learning_rate": 8.214893738699677e-06,
+      "loss": 0.5109,
+      "step": 12305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.35141565649733,
+      "learning_rate": 8.214591963397696e-06,
+      "loss": 0.3585,
+      "step": 12306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9446030973998483,
+      "learning_rate": 8.214290168133971e-06,
+      "loss": 0.2554,
+      "step": 12307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2421733447533934,
+      "learning_rate": 8.213988352910375e-06,
+      "loss": 0.2829,
+      "step": 12308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9420283642894052,
+      "learning_rate": 8.213686517728783e-06,
+      "loss": 0.5092,
+      "step": 12309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8983282599396714,
+      "learning_rate": 8.21338466259107e-06,
+      "loss": 0.3795,
+      "step": 12310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6469678576184916,
+      "learning_rate": 8.21308278749911e-06,
+      "loss": 0.2614,
+      "step": 12311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9088323701617638,
+      "learning_rate": 8.212780892454775e-06,
+      "loss": 0.3265,
+      "step": 12312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5080946623702731,
+      "learning_rate": 8.212478977459944e-06,
+      "loss": 0.3777,
+      "step": 12313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9442500445612505,
+      "learning_rate": 8.21217704251649e-06,
+      "loss": 0.3437,
+      "step": 12314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1387179295706575,
+      "learning_rate": 8.211875087626285e-06,
+      "loss": 0.4837,
+      "step": 12315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.035481362907352,
+      "learning_rate": 8.21157311279121e-06,
+      "loss": 0.2254,
+      "step": 12316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.687717397249761,
+      "learning_rate": 8.211271118013136e-06,
+      "loss": 0.4425,
+      "step": 12317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1113455998405963,
+      "learning_rate": 8.210969103293938e-06,
+      "loss": 0.379,
+      "step": 12318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7828908440028246,
+      "learning_rate": 8.210667068635494e-06,
+      "loss": 0.5395,
+      "step": 12319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4833212801648072,
+      "learning_rate": 8.210365014039679e-06,
+      "loss": 0.3894,
+      "step": 12320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0839738621932242,
+      "learning_rate": 8.210062939508368e-06,
+      "loss": 0.2022,
+      "step": 12321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6189838328906174,
+      "learning_rate": 8.209760845043434e-06,
+      "loss": 0.2303,
+      "step": 12322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4927904423782408,
+      "learning_rate": 8.209458730646758e-06,
+      "loss": 0.3239,
+      "step": 12323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3605517947239165,
+      "learning_rate": 8.209156596320213e-06,
+      "loss": 0.143,
+      "step": 12324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.544644768575514,
+      "learning_rate": 8.208854442065676e-06,
+      "loss": 0.4319,
+      "step": 12325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.355448987910198,
+      "learning_rate": 8.208552267885022e-06,
+      "loss": 0.2939,
+      "step": 12326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.044511201823635,
+      "learning_rate": 8.208250073780131e-06,
+      "loss": 0.6467,
+      "step": 12327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.465804377602023,
+      "learning_rate": 8.207947859752877e-06,
+      "loss": 0.2997,
+      "step": 12328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.569093114538665,
+      "learning_rate": 8.207645625805133e-06,
+      "loss": 0.4106,
+      "step": 12329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7195810302988241,
+      "learning_rate": 8.207343371938784e-06,
+      "loss": 0.2391,
+      "step": 12330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.465257033539254,
+      "learning_rate": 8.207041098155701e-06,
+      "loss": 0.1872,
+      "step": 12331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.130695087312306,
+      "learning_rate": 8.206738804457762e-06,
+      "loss": 0.1842,
+      "step": 12332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5884143697565671,
+      "learning_rate": 8.206436490846845e-06,
+      "loss": 0.2183,
+      "step": 12333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8923597770845648,
+      "learning_rate": 8.206134157324827e-06,
+      "loss": 0.3656,
+      "step": 12334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5490579639012092,
+      "learning_rate": 8.205831803893586e-06,
+      "loss": 0.5246,
+      "step": 12335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1686563349741967,
+      "learning_rate": 8.205529430554998e-06,
+      "loss": 0.1786,
+      "step": 12336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4896908227500563,
+      "learning_rate": 8.205227037310942e-06,
+      "loss": 0.2087,
+      "step": 12337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6053555627356222,
+      "learning_rate": 8.204924624163296e-06,
+      "loss": 0.2399,
+      "step": 12338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.618207961426856,
+      "learning_rate": 8.204622191113936e-06,
+      "loss": 0.2318,
+      "step": 12339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1636874191472746,
+      "learning_rate": 8.204319738164743e-06,
+      "loss": 0.2311,
+      "step": 12340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.681102369697484,
+      "learning_rate": 8.204017265317592e-06,
+      "loss": 0.2019,
+      "step": 12341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5821972830502153,
+      "learning_rate": 8.203714772574363e-06,
+      "loss": 0.3559,
+      "step": 12342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3917466472820383,
+      "learning_rate": 8.203412259936935e-06,
+      "loss": 0.3333,
+      "step": 12343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7674127490834046,
+      "learning_rate": 8.203109727407185e-06,
+      "loss": 0.563,
+      "step": 12344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6284137914509647,
+      "learning_rate": 8.202807174986993e-06,
+      "loss": 0.4138,
+      "step": 12345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9800296010648875,
+      "learning_rate": 8.202504602678236e-06,
+      "loss": 0.4589,
+      "step": 12346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.408846519035788,
+      "learning_rate": 8.202202010482795e-06,
+      "loss": 0.2079,
+      "step": 12347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6637550078860492,
+      "learning_rate": 8.201899398402546e-06,
+      "loss": 0.3017,
+      "step": 12348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5078356474250973,
+      "learning_rate": 8.201596766439372e-06,
+      "loss": 0.249,
+      "step": 12349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4330439624227511,
+      "learning_rate": 8.201294114595151e-06,
+      "loss": 0.2058,
+      "step": 12350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7452370165396731,
+      "learning_rate": 8.20099144287176e-06,
+      "loss": 0.5387,
+      "step": 12351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3028887610054167,
+      "learning_rate": 8.200688751271082e-06,
+      "loss": 0.2306,
+      "step": 12352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2205088758410583,
+      "learning_rate": 8.200386039794992e-06,
+      "loss": 0.1904,
+      "step": 12353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.538266092073649,
+      "learning_rate": 8.200083308445376e-06,
+      "loss": 0.2455,
+      "step": 12354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.873184043945238,
+      "learning_rate": 8.199780557224108e-06,
+      "loss": 0.3703,
+      "step": 12355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3911449017718571,
+      "learning_rate": 8.199477786133073e-06,
+      "loss": 0.3221,
+      "step": 12356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4775666072317688,
+      "learning_rate": 8.199174995174146e-06,
+      "loss": 0.3177,
+      "step": 12357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6154364363776468,
+      "learning_rate": 8.198872184349212e-06,
+      "loss": 0.2768,
+      "step": 12358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.202667953212247,
+      "learning_rate": 8.198569353660149e-06,
+      "loss": 0.6693,
+      "step": 12359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.719434307161288,
+      "learning_rate": 8.198266503108835e-06,
+      "loss": 0.338,
+      "step": 12360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1961321934203555,
+      "learning_rate": 8.197963632697158e-06,
+      "loss": 0.2789,
+      "step": 12361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.627530300117251,
+      "learning_rate": 8.19766074242699e-06,
+      "loss": 0.3136,
+      "step": 12362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2129404561468735,
+      "learning_rate": 8.19735783230022e-06,
+      "loss": 0.2238,
+      "step": 12363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9505012283732035,
+      "learning_rate": 8.197054902318722e-06,
+      "loss": 0.5324,
+      "step": 12364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3608984303258858,
+      "learning_rate": 8.19675195248438e-06,
+      "loss": 0.2974,
+      "step": 12365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8378047132529807,
+      "learning_rate": 8.196448982799077e-06,
+      "loss": 0.5542,
+      "step": 12366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.28143752865665,
+      "learning_rate": 8.196145993264692e-06,
+      "loss": 0.2186,
+      "step": 12367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7213277451646318,
+      "learning_rate": 8.195842983883107e-06,
+      "loss": 0.3766,
+      "step": 12368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6595186422788752,
+      "learning_rate": 8.195539954656204e-06,
+      "loss": 0.3247,
+      "step": 12369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7603618784472808,
+      "learning_rate": 8.195236905585864e-06,
+      "loss": 0.3887,
+      "step": 12370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6500237773858195,
+      "learning_rate": 8.194933836673972e-06,
+      "loss": 0.4898,
+      "step": 12371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6987100607097612,
+      "learning_rate": 8.194630747922405e-06,
+      "loss": 0.3571,
+      "step": 12372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7331315837209558,
+      "learning_rate": 8.194327639333048e-06,
+      "loss": 0.4013,
+      "step": 12373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5404640535390781,
+      "learning_rate": 8.194024510907784e-06,
+      "loss": 0.1912,
+      "step": 12374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.946333611452703,
+      "learning_rate": 8.19372136264849e-06,
+      "loss": 0.2683,
+      "step": 12375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6433449088281193,
+      "learning_rate": 8.193418194557057e-06,
+      "loss": 0.5491,
+      "step": 12376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.520566670890098,
+      "learning_rate": 8.193115006635359e-06,
+      "loss": 0.3837,
+      "step": 12377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5380649528888442,
+      "learning_rate": 8.192811798885286e-06,
+      "loss": 0.4901,
+      "step": 12378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7727931486338104,
+      "learning_rate": 8.192508571308717e-06,
+      "loss": 0.3149,
+      "step": 12379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.421241493695037,
+      "learning_rate": 8.192205323907537e-06,
+      "loss": 0.139,
+      "step": 12380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3939542003902514,
+      "learning_rate": 8.191902056683623e-06,
+      "loss": 0.1902,
+      "step": 12381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7137034356989376,
+      "learning_rate": 8.191598769638867e-06,
+      "loss": 0.2887,
+      "step": 12382
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7653834084835438,
+      "learning_rate": 8.191295462775148e-06,
+      "loss": 0.254,
+      "step": 12383
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4573026741785513,
+      "learning_rate": 8.190992136094347e-06,
+      "loss": 0.2084,
+      "step": 12384
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4561253510693402,
+      "learning_rate": 8.190688789598353e-06,
+      "loss": 0.2757,
+      "step": 12385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5857962089702782,
+      "learning_rate": 8.190385423289046e-06,
+      "loss": 0.4095,
+      "step": 12386
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5374933415167928,
+      "learning_rate": 8.190082037168311e-06,
+      "loss": 0.2347,
+      "step": 12387
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.582365921319462,
+      "learning_rate": 8.18977863123803e-06,
+      "loss": 0.3592,
+      "step": 12388
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5215845142410593,
+      "learning_rate": 8.189475205500091e-06,
+      "loss": 0.2212,
+      "step": 12389
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1255663755106116,
+      "learning_rate": 8.189171759956376e-06,
+      "loss": 0.2149,
+      "step": 12390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2040952541560423,
+      "learning_rate": 8.188868294608769e-06,
+      "loss": 0.1896,
+      "step": 12391
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8954550467000448,
+      "learning_rate": 8.188564809459155e-06,
+      "loss": 0.4674,
+      "step": 12392
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.601827380908477,
+      "learning_rate": 8.188261304509416e-06,
+      "loss": 0.3284,
+      "step": 12393
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7486076350611417,
+      "learning_rate": 8.187957779761442e-06,
+      "loss": 0.5423,
+      "step": 12394
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9606304042220004,
+      "learning_rate": 8.187654235217113e-06,
+      "loss": 0.513,
+      "step": 12395
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9273475328399807,
+      "learning_rate": 8.187350670878316e-06,
+      "loss": 0.2743,
+      "step": 12396
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.369055706367167,
+      "learning_rate": 8.187047086746936e-06,
+      "loss": 0.5101,
+      "step": 12397
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.110817917289217,
+      "learning_rate": 8.186743482824857e-06,
+      "loss": 0.6784,
+      "step": 12398
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5904639508232776,
+      "learning_rate": 8.186439859113967e-06,
+      "loss": 0.2401,
+      "step": 12399
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6040303540190415,
+      "learning_rate": 8.186136215616148e-06,
+      "loss": 0.1675,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6902932343925146,
+      "learning_rate": 8.185832552333288e-06,
+      "loss": 0.4934,
+      "step": 12401
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.024300670990486,
+      "learning_rate": 8.185528869267272e-06,
+      "loss": 0.5106,
+      "step": 12402
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5524956381701196,
+      "learning_rate": 8.185225166419986e-06,
+      "loss": 0.3935,
+      "step": 12403
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3451787926469574,
+      "learning_rate": 8.184921443793315e-06,
+      "loss": 0.2681,
+      "step": 12404
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.848528762847438,
+      "learning_rate": 8.184617701389147e-06,
+      "loss": 0.56,
+      "step": 12405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5710280671656829,
+      "learning_rate": 8.184313939209362e-06,
+      "loss": 0.2543,
+      "step": 12406
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7062442196059913,
+      "learning_rate": 8.184010157255857e-06,
+      "loss": 0.3503,
+      "step": 12407
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8506885912764672,
+      "learning_rate": 8.18370635553051e-06,
+      "loss": 0.5981,
+      "step": 12408
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.237349057023009,
+      "learning_rate": 8.183402534035209e-06,
+      "loss": 0.1463,
+      "step": 12409
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5882510448347593,
+      "learning_rate": 8.183098692771843e-06,
+      "loss": 0.3289,
+      "step": 12410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6498977195943634,
+      "learning_rate": 8.182794831742296e-06,
+      "loss": 0.2371,
+      "step": 12411
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.47844926188807,
+      "learning_rate": 8.182490950948458e-06,
+      "loss": 0.34,
+      "step": 12412
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8614181584842102,
+      "learning_rate": 8.182187050392213e-06,
+      "loss": 0.3552,
+      "step": 12413
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6663539404333476,
+      "learning_rate": 8.18188313007545e-06,
+      "loss": 0.1512,
+      "step": 12414
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1032297157926854,
+      "learning_rate": 8.181579190000055e-06,
+      "loss": 0.2727,
+      "step": 12415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9350421412940717,
+      "learning_rate": 8.181275230167918e-06,
+      "loss": 0.5317,
+      "step": 12416
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0546575261239575,
+      "learning_rate": 8.180971250580922e-06,
+      "loss": 0.5755,
+      "step": 12417
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.151507085198747,
+      "learning_rate": 8.18066725124096e-06,
+      "loss": 0.3162,
+      "step": 12418
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3045465170473753,
+      "learning_rate": 8.180363232149916e-06,
+      "loss": 0.2366,
+      "step": 12419
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4366746170708304,
+      "learning_rate": 8.180059193309679e-06,
+      "loss": 0.356,
+      "step": 12420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7360894561582643,
+      "learning_rate": 8.179755134722136e-06,
+      "loss": 0.2635,
+      "step": 12421
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8171135950517177,
+      "learning_rate": 8.179451056389177e-06,
+      "loss": 0.3909,
+      "step": 12422
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.661611502611592,
+      "learning_rate": 8.179146958312687e-06,
+      "loss": 0.296,
+      "step": 12423
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8198818671550685,
+      "learning_rate": 8.178842840494558e-06,
+      "loss": 0.6322,
+      "step": 12424
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.474876191461077,
+      "learning_rate": 8.178538702936678e-06,
+      "loss": 0.1883,
+      "step": 12425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7856144974279187,
+      "learning_rate": 8.178234545640934e-06,
+      "loss": 0.4902,
+      "step": 12426
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4324471698501011,
+      "learning_rate": 8.177930368609217e-06,
+      "loss": 0.2197,
+      "step": 12427
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6613947994218228,
+      "learning_rate": 8.17762617184341e-06,
+      "loss": 0.2595,
+      "step": 12428
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7175133591205938,
+      "learning_rate": 8.177321955345411e-06,
+      "loss": 0.4618,
+      "step": 12429
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5796738143841227,
+      "learning_rate": 8.177017719117103e-06,
+      "loss": 0.3412,
+      "step": 12430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7551333334055053,
+      "learning_rate": 8.176713463160375e-06,
+      "loss": 0.4469,
+      "step": 12431
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6223089994478086,
+      "learning_rate": 8.176409187477119e-06,
+      "loss": 0.3376,
+      "step": 12432
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6696737852466799,
+      "learning_rate": 8.176104892069224e-06,
+      "loss": 0.3516,
+      "step": 12433
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1716134764996848,
+      "learning_rate": 8.175800576938578e-06,
+      "loss": 0.1759,
+      "step": 12434
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7385476326159157,
+      "learning_rate": 8.175496242087072e-06,
+      "loss": 0.3113,
+      "step": 12435
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5120992304925553,
+      "learning_rate": 8.175191887516595e-06,
+      "loss": 0.1821,
+      "step": 12436
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7019351031084908,
+      "learning_rate": 8.174887513229038e-06,
+      "loss": 0.3658,
+      "step": 12437
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.756369723451016,
+      "learning_rate": 8.17458311922629e-06,
+      "loss": 0.3126,
+      "step": 12438
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6051869342284442,
+      "learning_rate": 8.174278705510244e-06,
+      "loss": 0.3812,
+      "step": 12439
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0313411783881943,
+      "learning_rate": 8.173974272082787e-06,
+      "loss": 0.3869,
+      "step": 12440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.017093550390997,
+      "learning_rate": 8.17366981894581e-06,
+      "loss": 0.1837,
+      "step": 12441
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7329338074266214,
+      "learning_rate": 8.173365346101206e-06,
+      "loss": 0.1887,
+      "step": 12442
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5923150325008897,
+      "learning_rate": 8.173060853550861e-06,
+      "loss": 0.4163,
+      "step": 12443
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7654088107399373,
+      "learning_rate": 8.17275634129667e-06,
+      "loss": 0.4757,
+      "step": 12444
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0695453073038943,
+      "learning_rate": 8.172451809340523e-06,
+      "loss": 0.1684,
+      "step": 12445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8926242944776435,
+      "learning_rate": 8.172147257684308e-06,
+      "loss": 0.4789,
+      "step": 12446
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.597839233067366,
+      "learning_rate": 8.171842686329923e-06,
+      "loss": 0.2386,
+      "step": 12447
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2727782994729975,
+      "learning_rate": 8.171538095279252e-06,
+      "loss": 0.1731,
+      "step": 12448
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.943457787863744,
+      "learning_rate": 8.17123348453419e-06,
+      "loss": 0.3563,
+      "step": 12449
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5491142393517363,
+      "learning_rate": 8.17092885409663e-06,
+      "loss": 0.3819,
+      "step": 12450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0285839410888635,
+      "learning_rate": 8.170624203968461e-06,
+      "loss": 0.368,
+      "step": 12451
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4433289177536965,
+      "learning_rate": 8.170319534151574e-06,
+      "loss": 0.2133,
+      "step": 12452
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5227281422815564,
+      "learning_rate": 8.170014844647865e-06,
+      "loss": 0.2069,
+      "step": 12453
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.474708972766439,
+      "learning_rate": 8.169710135459221e-06,
+      "loss": 0.2538,
+      "step": 12454
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2000009567162555,
+      "learning_rate": 8.16940540658754e-06,
+      "loss": 0.1831,
+      "step": 12455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3448612376234776,
+      "learning_rate": 8.169100658034707e-06,
+      "loss": 0.6681,
+      "step": 12456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3048160855093767,
+      "learning_rate": 8.168795889802622e-06,
+      "loss": 0.1667,
+      "step": 12457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3342662418542697,
+      "learning_rate": 8.168491101893173e-06,
+      "loss": 0.1866,
+      "step": 12458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9329277664143263,
+      "learning_rate": 8.168186294308253e-06,
+      "loss": 0.3275,
+      "step": 12459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7044023811195845,
+      "learning_rate": 8.167881467049754e-06,
+      "loss": 0.2722,
+      "step": 12460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4480633281167166,
+      "learning_rate": 8.167576620119572e-06,
+      "loss": 0.2647,
+      "step": 12461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3417259790985114,
+      "learning_rate": 8.167271753519598e-06,
+      "loss": 0.2452,
+      "step": 12462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.955398076083741,
+      "learning_rate": 8.166966867251724e-06,
+      "loss": 0.452,
+      "step": 12463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7398127292287984,
+      "learning_rate": 8.166661961317847e-06,
+      "loss": 0.3604,
+      "step": 12464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6144309992350507,
+      "learning_rate": 8.166357035719856e-06,
+      "loss": 0.5182,
+      "step": 12465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9254352268985344,
+      "learning_rate": 8.16605209045965e-06,
+      "loss": 0.1956,
+      "step": 12466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6899204021384129,
+      "learning_rate": 8.165747125539115e-06,
+      "loss": 0.1747,
+      "step": 12467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.418789935213569,
+      "learning_rate": 8.16544214096015e-06,
+      "loss": 0.1377,
+      "step": 12468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6211893995631197,
+      "learning_rate": 8.16513713672465e-06,
+      "loss": 0.3312,
+      "step": 12469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8415532886514914,
+      "learning_rate": 8.164832112834505e-06,
+      "loss": 0.3347,
+      "step": 12470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2533070645468207,
+      "learning_rate": 8.16452706929161e-06,
+      "loss": 0.3455,
+      "step": 12471
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1485424026274933,
+      "learning_rate": 8.164222006097862e-06,
+      "loss": 0.2215,
+      "step": 12472
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8086138244239869,
+      "learning_rate": 8.163916923255152e-06,
+      "loss": 0.3477,
+      "step": 12473
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9180803571240985,
+      "learning_rate": 8.163611820765377e-06,
+      "loss": 0.2507,
+      "step": 12474
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5536572418908716,
+      "learning_rate": 8.16330669863043e-06,
+      "loss": 0.2338,
+      "step": 12475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8553687018475378,
+      "learning_rate": 8.163001556852206e-06,
+      "loss": 0.2151,
+      "step": 12476
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2347489798466897,
+      "learning_rate": 8.162696395432601e-06,
+      "loss": 0.2702,
+      "step": 12477
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2488351669566402,
+      "learning_rate": 8.162391214373509e-06,
+      "loss": 0.2979,
+      "step": 12478
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8527731418273603,
+      "learning_rate": 8.162086013676825e-06,
+      "loss": 0.3481,
+      "step": 12479
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3381818830229606,
+      "learning_rate": 8.161780793344446e-06,
+      "loss": 0.1148,
+      "step": 12480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2502146819654907,
+      "learning_rate": 8.161475553378264e-06,
+      "loss": 0.1977,
+      "step": 12481
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0519337785842098,
+      "learning_rate": 8.161170293780176e-06,
+      "loss": 0.1566,
+      "step": 12482
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5678030383588892,
+      "learning_rate": 8.160865014552078e-06,
+      "loss": 0.2989,
+      "step": 12483
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8635502427147062,
+      "learning_rate": 8.160559715695866e-06,
+      "loss": 0.4026,
+      "step": 12484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7054121776080269,
+      "learning_rate": 8.160254397213435e-06,
+      "loss": 0.3346,
+      "step": 12485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5625652299616588,
+      "learning_rate": 8.159949059106683e-06,
+      "loss": 0.3677,
+      "step": 12486
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6584996720997403,
+      "learning_rate": 8.159643701377502e-06,
+      "loss": 0.1978,
+      "step": 12487
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7811902521385825,
+      "learning_rate": 8.159338324027792e-06,
+      "loss": 0.2756,
+      "step": 12488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7958820419019461,
+      "learning_rate": 8.159032927059446e-06,
+      "loss": 0.3,
+      "step": 12489
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9104024888292193,
+      "learning_rate": 8.158727510474364e-06,
+      "loss": 0.3708,
+      "step": 12490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0152597576982196,
+      "learning_rate": 8.158422074274442e-06,
+      "loss": 0.4733,
+      "step": 12491
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4007647403604953,
+      "learning_rate": 8.158116618461573e-06,
+      "loss": 0.2867,
+      "step": 12492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7915552547287705,
+      "learning_rate": 8.157811143037658e-06,
+      "loss": 0.4041,
+      "step": 12493
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5244621547857138,
+      "learning_rate": 8.15750564800459e-06,
+      "loss": 0.2948,
+      "step": 12494
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4447613490053826,
+      "learning_rate": 8.15720013336427e-06,
+      "loss": 0.4283,
+      "step": 12495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1834258379032967,
+      "learning_rate": 8.156894599118592e-06,
+      "loss": 0.251,
+      "step": 12496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3224133545632668,
+      "learning_rate": 8.156589045269458e-06,
+      "loss": 0.2701,
+      "step": 12497
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5445216754082083,
+      "learning_rate": 8.156283471818758e-06,
+      "loss": 0.2812,
+      "step": 12498
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.171832441510841,
+      "learning_rate": 8.155977878768397e-06,
+      "loss": 0.1825,
+      "step": 12499
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6441991384602341,
+      "learning_rate": 8.155672266120266e-06,
+      "loss": 0.2036,
+      "step": 12500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8012199922420795,
+      "learning_rate": 8.15536663387627e-06,
+      "loss": 0.3275,
+      "step": 12501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9463705111858458,
+      "learning_rate": 8.155060982038301e-06,
+      "loss": 0.6568,
+      "step": 12502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.670985627031521,
+      "learning_rate": 8.154755310608258e-06,
+      "loss": 0.3204,
+      "step": 12503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8133034672294377,
+      "learning_rate": 8.154449619588042e-06,
+      "loss": 0.2314,
+      "step": 12504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5389714940269517,
+      "learning_rate": 8.154143908979548e-06,
+      "loss": 0.3518,
+      "step": 12505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8510431353978771,
+      "learning_rate": 8.153838178784676e-06,
+      "loss": 0.4741,
+      "step": 12506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8357789174031878,
+      "learning_rate": 8.153532429005323e-06,
+      "loss": 0.4521,
+      "step": 12507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7568971592497349,
+      "learning_rate": 8.153226659643392e-06,
+      "loss": 0.2658,
+      "step": 12508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5429883207214763,
+      "learning_rate": 8.152920870700777e-06,
+      "loss": 0.3747,
+      "step": 12509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.881834634919995,
+      "learning_rate": 8.152615062179379e-06,
+      "loss": 0.695,
+      "step": 12510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5814584849052706,
+      "learning_rate": 8.152309234081095e-06,
+      "loss": 0.256,
+      "step": 12511
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.54555980566474,
+      "learning_rate": 8.152003386407827e-06,
+      "loss": 0.3371,
+      "step": 12512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1273048288416694,
+      "learning_rate": 8.151697519161473e-06,
+      "loss": 0.1907,
+      "step": 12513
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8516506182147954,
+      "learning_rate": 8.15139163234393e-06,
+      "loss": 0.5042,
+      "step": 12514
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7717269105782935,
+      "learning_rate": 8.151085725957102e-06,
+      "loss": 0.1072,
+      "step": 12515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9881759252175013,
+      "learning_rate": 8.150779800002885e-06,
+      "loss": 0.4259,
+      "step": 12516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7003779721287728,
+      "learning_rate": 8.15047385448318e-06,
+      "loss": 0.507,
+      "step": 12517
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5142642627836012,
+      "learning_rate": 8.150167889399887e-06,
+      "loss": 0.4147,
+      "step": 12518
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0922861462015416,
+      "learning_rate": 8.149861904754908e-06,
+      "loss": 0.1856,
+      "step": 12519
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8807941714637835,
+      "learning_rate": 8.149555900550136e-06,
+      "loss": 0.5837,
+      "step": 12520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4583378269489629,
+      "learning_rate": 8.14924987678748e-06,
+      "loss": 0.2673,
+      "step": 12521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7000532753744864,
+      "learning_rate": 8.148943833468835e-06,
+      "loss": 0.2558,
+      "step": 12522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5795124724835468,
+      "learning_rate": 8.148637770596104e-06,
+      "loss": 0.224,
+      "step": 12523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.401078484865669,
+      "learning_rate": 8.148331688171186e-06,
+      "loss": 0.2206,
+      "step": 12524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9693538755753623,
+      "learning_rate": 8.14802558619598e-06,
+      "loss": 0.3406,
+      "step": 12525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1093325405323384,
+      "learning_rate": 8.147719464672392e-06,
+      "loss": 0.144,
+      "step": 12526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3748777383153756,
+      "learning_rate": 8.147413323602317e-06,
+      "loss": 0.4411,
+      "step": 12527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2967117721028805,
+      "learning_rate": 8.14710716298766e-06,
+      "loss": 0.1881,
+      "step": 12528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3857622252689707,
+      "learning_rate": 8.146800982830321e-06,
+      "loss": 0.2196,
+      "step": 12529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6327019065382251,
+      "learning_rate": 8.146494783132202e-06,
+      "loss": 0.3304,
+      "step": 12530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7576033828113533,
+      "learning_rate": 8.146188563895202e-06,
+      "loss": 0.2963,
+      "step": 12531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7102151722034933,
+      "learning_rate": 8.145882325121225e-06,
+      "loss": 0.1814,
+      "step": 12532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0838608673879113,
+      "learning_rate": 8.145576066812173e-06,
+      "loss": 0.1571,
+      "step": 12533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.723521821768395,
+      "learning_rate": 8.145269788969947e-06,
+      "loss": 0.5194,
+      "step": 12534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8257428823038115,
+      "learning_rate": 8.144963491596446e-06,
+      "loss": 0.2189,
+      "step": 12535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.843087404610818,
+      "learning_rate": 8.144657174693577e-06,
+      "loss": 0.3326,
+      "step": 12536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7134847230541912,
+      "learning_rate": 8.14435083826324e-06,
+      "loss": 0.4054,
+      "step": 12537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.798335431414869,
+      "learning_rate": 8.144044482307335e-06,
+      "loss": 0.4468,
+      "step": 12538
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6387930803634094,
+      "learning_rate": 8.143738106827768e-06,
+      "loss": 0.1792,
+      "step": 12539
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2697041437821865,
+      "learning_rate": 8.14343171182644e-06,
+      "loss": 0.1936,
+      "step": 12540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6807487783409836,
+      "learning_rate": 8.143125297305254e-06,
+      "loss": 0.2971,
+      "step": 12541
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0324851902167627,
+      "learning_rate": 8.14281886326611e-06,
+      "loss": 0.6898,
+      "step": 12542
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2698112883184869,
+      "learning_rate": 8.142512409710917e-06,
+      "loss": 0.3087,
+      "step": 12543
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.324390398627384,
+      "learning_rate": 8.14220593664157e-06,
+      "loss": 0.5558,
+      "step": 12544
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6782788069440622,
+      "learning_rate": 8.14189944405998e-06,
+      "loss": 0.5066,
+      "step": 12545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.162106261948477,
+      "learning_rate": 8.141592931968046e-06,
+      "loss": 0.2176,
+      "step": 12546
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1511882209329258,
+      "learning_rate": 8.141286400367669e-06,
+      "loss": 0.1675,
+      "step": 12547
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.687302206951428,
+      "learning_rate": 8.140979849260757e-06,
+      "loss": 0.4628,
+      "step": 12548
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2985128995369777,
+      "learning_rate": 8.140673278649214e-06,
+      "loss": 0.2156,
+      "step": 12549
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5797845122057832,
+      "learning_rate": 8.140366688534939e-06,
+      "loss": 0.347,
+      "step": 12550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6780198890117728,
+      "learning_rate": 8.140060078919841e-06,
+      "loss": 0.5,
+      "step": 12551
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8759583011117669,
+      "learning_rate": 8.139753449805818e-06,
+      "loss": 0.4488,
+      "step": 12552
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7525796716139341,
+      "learning_rate": 8.13944680119478e-06,
+      "loss": 0.4722,
+      "step": 12553
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1640897786061255,
+      "learning_rate": 8.13914013308863e-06,
+      "loss": 0.1602,
+      "step": 12554
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4583122972833142,
+      "learning_rate": 8.138833445489267e-06,
+      "loss": 0.3818,
+      "step": 12555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6250756250720495,
+      "learning_rate": 8.138526738398603e-06,
+      "loss": 0.1932,
+      "step": 12556
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.204801461226259,
+      "learning_rate": 8.138220011818538e-06,
+      "loss": 0.167,
+      "step": 12557
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7065771533341851,
+      "learning_rate": 8.13791326575098e-06,
+      "loss": 0.451,
+      "step": 12558
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9146687773593387,
+      "learning_rate": 8.13760650019783e-06,
+      "loss": 0.3413,
+      "step": 12559
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2660115471303275,
+      "learning_rate": 8.137299715160993e-06,
+      "loss": 0.4271,
+      "step": 12560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.743446895617125,
+      "learning_rate": 8.136992910642377e-06,
+      "loss": 0.4953,
+      "step": 12561
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9213676325663176,
+      "learning_rate": 8.136686086643886e-06,
+      "loss": 0.5374,
+      "step": 12562
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9568647122966893,
+      "learning_rate": 8.136379243167424e-06,
+      "loss": 0.2115,
+      "step": 12563
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4820690162282824,
+      "learning_rate": 8.136072380214899e-06,
+      "loss": 0.2853,
+      "step": 12564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.216619355072077,
+      "learning_rate": 8.135765497788216e-06,
+      "loss": 0.4416,
+      "step": 12565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3957772885702677,
+      "learning_rate": 8.135458595889276e-06,
+      "loss": 0.2965,
+      "step": 12566
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2012314618845588,
+      "learning_rate": 8.135151674519992e-06,
+      "loss": 0.2066,
+      "step": 12567
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6298157945023195,
+      "learning_rate": 8.134844733682266e-06,
+      "loss": 0.3232,
+      "step": 12568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6926046139167181,
+      "learning_rate": 8.134537773378002e-06,
+      "loss": 0.5372,
+      "step": 12569
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.433320445584513,
+      "learning_rate": 8.13423079360911e-06,
+      "loss": 0.3378,
+      "step": 12570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.620003915314005,
+      "learning_rate": 8.133923794377495e-06,
+      "loss": 0.3585,
+      "step": 12571
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.070181601362863,
+      "learning_rate": 8.133616775685063e-06,
+      "loss": 0.2348,
+      "step": 12572
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2225411562023036,
+      "learning_rate": 8.13330973753372e-06,
+      "loss": 0.1679,
+      "step": 12573
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.735680557003299,
+      "learning_rate": 8.133002679925375e-06,
+      "loss": 0.4456,
+      "step": 12574
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7949821744302028,
+      "learning_rate": 8.132695602861931e-06,
+      "loss": 0.5845,
+      "step": 12575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4241208297581713,
+      "learning_rate": 8.132388506345299e-06,
+      "loss": 0.2353,
+      "step": 12576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5326005176635658,
+      "learning_rate": 8.132081390377383e-06,
+      "loss": 0.3302,
+      "step": 12577
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0932528523966862,
+      "learning_rate": 8.13177425496009e-06,
+      "loss": 0.214,
+      "step": 12578
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2096922741065437,
+      "learning_rate": 8.13146710009533e-06,
+      "loss": 0.6952,
+      "step": 12579
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9019611530052263,
+      "learning_rate": 8.131159925785008e-06,
+      "loss": 0.5096,
+      "step": 12580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3184758247882185,
+      "learning_rate": 8.130852732031032e-06,
+      "loss": 0.5194,
+      "step": 12581
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8761365723966892,
+      "learning_rate": 8.13054551883531e-06,
+      "loss": 0.3907,
+      "step": 12582
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7628916665396812,
+      "learning_rate": 8.130238286199747e-06,
+      "loss": 0.3504,
+      "step": 12583
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9668554242386953,
+      "learning_rate": 8.129931034126256e-06,
+      "loss": 0.2155,
+      "step": 12584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.767614981757055,
+      "learning_rate": 8.129623762616741e-06,
+      "loss": 0.3338,
+      "step": 12585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5155267093680964,
+      "learning_rate": 8.12931647167311e-06,
+      "loss": 0.2369,
+      "step": 12586
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3269195668062193,
+      "learning_rate": 8.129009161297275e-06,
+      "loss": 0.25,
+      "step": 12587
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9944345181082945,
+      "learning_rate": 8.12870183149114e-06,
+      "loss": 0.4814,
+      "step": 12588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3018012930268485,
+      "learning_rate": 8.128394482256615e-06,
+      "loss": 0.2592,
+      "step": 12589
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4946692360661975,
+      "learning_rate": 8.128087113595611e-06,
+      "loss": 0.2925,
+      "step": 12590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9681635486495739,
+      "learning_rate": 8.127779725510033e-06,
+      "loss": 0.4359,
+      "step": 12591
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9430785431496966,
+      "learning_rate": 8.12747231800179e-06,
+      "loss": 0.337,
+      "step": 12592
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0295003728377536,
+      "learning_rate": 8.127164891072791e-06,
+      "loss": 0.4368,
+      "step": 12593
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0137000040333697,
+      "learning_rate": 8.12685744472495e-06,
+      "loss": 0.4512,
+      "step": 12594
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9414063190790236,
+      "learning_rate": 8.12654997896017e-06,
+      "loss": 0.2311,
+      "step": 12595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3128735714559328,
+      "learning_rate": 8.126242493780362e-06,
+      "loss": 0.2159,
+      "step": 12596
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4016929938892926,
+      "learning_rate": 8.125934989187438e-06,
+      "loss": 0.3446,
+      "step": 12597
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.096061743407661,
+      "learning_rate": 8.125627465183306e-06,
+      "loss": 0.436,
+      "step": 12598
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4989488570705967,
+      "learning_rate": 8.125319921769871e-06,
+      "loss": 0.2018,
+      "step": 12599
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0229344571159915,
+      "learning_rate": 8.125012358949052e-06,
+      "loss": 0.3707,
+      "step": 12600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4137937991249192,
+      "learning_rate": 8.124704776722751e-06,
+      "loss": 0.2614,
+      "step": 12601
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.907294026562078,
+      "learning_rate": 8.124397175092879e-06,
+      "loss": 0.1966,
+      "step": 12602
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8315127722890774,
+      "learning_rate": 8.124089554061351e-06,
+      "loss": 0.5039,
+      "step": 12603
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.235565522101739,
+      "learning_rate": 8.123781913630074e-06,
+      "loss": 0.1906,
+      "step": 12604
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6215396405761104,
+      "learning_rate": 8.123474253800958e-06,
+      "loss": 0.3113,
+      "step": 12605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5044777304558792,
+      "learning_rate": 8.123166574575912e-06,
+      "loss": 0.1167,
+      "step": 12606
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5600906165810762,
+      "learning_rate": 8.122858875956852e-06,
+      "loss": 0.2067,
+      "step": 12607
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4348443226011909,
+      "learning_rate": 8.122551157945681e-06,
+      "loss": 0.222,
+      "step": 12608
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5611968996258356,
+      "learning_rate": 8.122243420544317e-06,
+      "loss": 0.2535,
+      "step": 12609
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8669671044299398,
+      "learning_rate": 8.121935663754669e-06,
+      "loss": 0.2365,
+      "step": 12610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.035702733873321,
+      "learning_rate": 8.121627887578647e-06,
+      "loss": 0.1775,
+      "step": 12611
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8219609470934783,
+      "learning_rate": 8.12132009201816e-06,
+      "loss": 0.3709,
+      "step": 12612
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3461583368323429,
+      "learning_rate": 8.121012277075124e-06,
+      "loss": 0.1725,
+      "step": 12613
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.592568015318411,
+      "learning_rate": 8.120704442751448e-06,
+      "loss": 0.4991,
+      "step": 12614
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4616135139852453,
+      "learning_rate": 8.120396589049043e-06,
+      "loss": 0.2243,
+      "step": 12615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3391163989873016,
+      "learning_rate": 8.120088715969822e-06,
+      "loss": 0.7342,
+      "step": 12616
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8406425815254235,
+      "learning_rate": 8.119780823515694e-06,
+      "loss": 0.1875,
+      "step": 12617
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9846958329066804,
+      "learning_rate": 8.119472911688577e-06,
+      "loss": 0.4108,
+      "step": 12618
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.158802070188493,
+      "learning_rate": 8.119164980490378e-06,
+      "loss": 0.403,
+      "step": 12619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.490740220247237,
+      "learning_rate": 8.118857029923008e-06,
+      "loss": 0.2378,
+      "step": 12620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.891139622472091,
+      "learning_rate": 8.118549059988384e-06,
+      "loss": 0.3524,
+      "step": 12621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.085839265660491,
+      "learning_rate": 8.118241070688416e-06,
+      "loss": 0.6408,
+      "step": 12622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.817513628745239,
+      "learning_rate": 8.117933062025017e-06,
+      "loss": 0.3724,
+      "step": 12623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8243214743728449,
+      "learning_rate": 8.117625034000099e-06,
+      "loss": 0.5356,
+      "step": 12624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8311233670388005,
+      "learning_rate": 8.117316986615572e-06,
+      "loss": 0.2183,
+      "step": 12625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2257535401902375,
+      "learning_rate": 8.117008919873356e-06,
+      "loss": 0.1863,
+      "step": 12626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6062894844144977,
+      "learning_rate": 8.11670083377536e-06,
+      "loss": 0.3721,
+      "step": 12627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9038302904819105,
+      "learning_rate": 8.116392728323494e-06,
+      "loss": 0.1668,
+      "step": 12628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6950841389509077,
+      "learning_rate": 8.116084603519677e-06,
+      "loss": 0.3683,
+      "step": 12629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3383300059757381,
+      "learning_rate": 8.115776459365819e-06,
+      "loss": 0.201,
+      "step": 12630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.670686131984725,
+      "learning_rate": 8.115468295863833e-06,
+      "loss": 0.3264,
+      "step": 12631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4664749676105608,
+      "learning_rate": 8.115160113015634e-06,
+      "loss": 0.3494,
+      "step": 12632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5544196885399,
+      "learning_rate": 8.114851910823138e-06,
+      "loss": 0.1896,
+      "step": 12633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.341241057882515,
+      "learning_rate": 8.114543689288254e-06,
+      "loss": 0.3576,
+      "step": 12634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.793457866266577,
+      "learning_rate": 8.1142354484129e-06,
+      "loss": 0.4641,
+      "step": 12635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.541339559799286,
+      "learning_rate": 8.113927188198989e-06,
+      "loss": 0.2404,
+      "step": 12636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0595958864843755,
+      "learning_rate": 8.113618908648432e-06,
+      "loss": 0.2143,
+      "step": 12637
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.406432026102534,
+      "learning_rate": 8.113310609763147e-06,
+      "loss": 0.3757,
+      "step": 12638
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3591497826633632,
+      "learning_rate": 8.113002291545048e-06,
+      "loss": 0.3742,
+      "step": 12639
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.453939998917596,
+      "learning_rate": 8.11269395399605e-06,
+      "loss": 0.1903,
+      "step": 12640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5044994292913805,
+      "learning_rate": 8.112385597118064e-06,
+      "loss": 0.3469,
+      "step": 12641
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.900761890508776,
+      "learning_rate": 8.11207722091301e-06,
+      "loss": 0.1839,
+      "step": 12642
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0759973533393916,
+      "learning_rate": 8.111768825382799e-06,
+      "loss": 0.6233,
+      "step": 12643
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.465905436738667,
+      "learning_rate": 8.111460410529349e-06,
+      "loss": 0.2913,
+      "step": 12644
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6936674574766182,
+      "learning_rate": 8.111151976354572e-06,
+      "loss": 0.4613,
+      "step": 12645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5544696576167756,
+      "learning_rate": 8.110843522860385e-06,
+      "loss": 0.1766,
+      "step": 12646
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6451385536513767,
+      "learning_rate": 8.110535050048706e-06,
+      "loss": 0.2842,
+      "step": 12647
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4174533411768038,
+      "learning_rate": 8.110226557921446e-06,
+      "loss": 0.4342,
+      "step": 12648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8567202272692587,
+      "learning_rate": 8.109918046480522e-06,
+      "loss": 0.352,
+      "step": 12649
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9193272877796494,
+      "learning_rate": 8.10960951572785e-06,
+      "loss": 0.1925,
+      "step": 12650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5800760007450267,
+      "learning_rate": 8.109300965665348e-06,
+      "loss": 0.2006,
+      "step": 12651
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.108460015956529,
+      "learning_rate": 8.10899239629493e-06,
+      "loss": 0.6621,
+      "step": 12652
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6199031985135106,
+      "learning_rate": 8.108683807618511e-06,
+      "loss": 0.2401,
+      "step": 12653
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.486864494460595,
+      "learning_rate": 8.10837519963801e-06,
+      "loss": 0.3622,
+      "step": 12654
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4844548530181083,
+      "learning_rate": 8.108066572355343e-06,
+      "loss": 0.3591,
+      "step": 12655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.096215496775557,
+      "learning_rate": 8.107757925772423e-06,
+      "loss": 0.1824,
+      "step": 12656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.918728349762335,
+      "learning_rate": 8.10744925989117e-06,
+      "loss": 0.3873,
+      "step": 12657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9256915058016197,
+      "learning_rate": 8.107140574713503e-06,
+      "loss": 0.4307,
+      "step": 12658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8299094603828732,
+      "learning_rate": 8.106831870241331e-06,
+      "loss": 0.4096,
+      "step": 12659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7731200534857783,
+      "learning_rate": 8.10652314647658e-06,
+      "loss": 0.3823,
+      "step": 12660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7284542697857839,
+      "learning_rate": 8.10621440342116e-06,
+      "loss": 0.3679,
+      "step": 12661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4774476582152003,
+      "learning_rate": 8.105905641076991e-06,
+      "loss": 0.3071,
+      "step": 12662
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4162728095806119,
+      "learning_rate": 8.10559685944599e-06,
+      "loss": 0.2488,
+      "step": 12663
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5313870942772632,
+      "learning_rate": 8.105288058530076e-06,
+      "loss": 0.2992,
+      "step": 12664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7442917539386664,
+      "learning_rate": 8.104979238331165e-06,
+      "loss": 0.2181,
+      "step": 12665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6770884588313315,
+      "learning_rate": 8.104670398851174e-06,
+      "loss": 0.3822,
+      "step": 12666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5845261716107681,
+      "learning_rate": 8.104361540092022e-06,
+      "loss": 0.2219,
+      "step": 12667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6708636790260407,
+      "learning_rate": 8.104052662055629e-06,
+      "loss": 0.2807,
+      "step": 12668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4405907130350368,
+      "learning_rate": 8.10374376474391e-06,
+      "loss": 0.242,
+      "step": 12669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.657532159686057,
+      "learning_rate": 8.103434848158782e-06,
+      "loss": 0.3271,
+      "step": 12670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7340763883159993,
+      "learning_rate": 8.103125912302168e-06,
+      "loss": 0.2964,
+      "step": 12671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6640418369281886,
+      "learning_rate": 8.10281695717598e-06,
+      "loss": 0.3677,
+      "step": 12672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.092560401714696,
+      "learning_rate": 8.102507982782144e-06,
+      "loss": 0.4975,
+      "step": 12673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6259577569647303,
+      "learning_rate": 8.102198989122573e-06,
+      "loss": 0.3037,
+      "step": 12674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8322756590007505,
+      "learning_rate": 8.10188997619919e-06,
+      "loss": 0.4023,
+      "step": 12675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.539124654589549,
+      "learning_rate": 8.101580944013907e-06,
+      "loss": 0.3307,
+      "step": 12676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7117085678967752,
+      "learning_rate": 8.10127189256865e-06,
+      "loss": 0.3767,
+      "step": 12677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5221205814792722,
+      "learning_rate": 8.100962821865335e-06,
+      "loss": 0.2457,
+      "step": 12678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3118801696951725,
+      "learning_rate": 8.100653731905882e-06,
+      "loss": 0.1469,
+      "step": 12679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4411573634521053,
+      "learning_rate": 8.100344622692212e-06,
+      "loss": 0.3935,
+      "step": 12680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.819205135572026,
+      "learning_rate": 8.100035494226241e-06,
+      "loss": 0.5097,
+      "step": 12681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.937459727991629,
+      "learning_rate": 8.09972634650989e-06,
+      "loss": 0.4453,
+      "step": 12682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5594966252355578,
+      "learning_rate": 8.099417179545081e-06,
+      "loss": 0.3479,
+      "step": 12683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6681728818816843,
+      "learning_rate": 8.09910799333373e-06,
+      "loss": 0.5647,
+      "step": 12684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.570606382113922,
+      "learning_rate": 8.09879878787776e-06,
+      "loss": 0.4962,
+      "step": 12685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5081382631800788,
+      "learning_rate": 8.09848956317909e-06,
+      "loss": 0.1684,
+      "step": 12686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.230521197942085,
+      "learning_rate": 8.098180319239638e-06,
+      "loss": 0.2994,
+      "step": 12687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.526171783416019,
+      "learning_rate": 8.097871056061331e-06,
+      "loss": 0.2089,
+      "step": 12688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1329276396345513,
+      "learning_rate": 8.09756177364608e-06,
+      "loss": 0.2021,
+      "step": 12689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7701318618383348,
+      "learning_rate": 8.097252471995813e-06,
+      "loss": 0.2805,
+      "step": 12690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7886048256277978,
+      "learning_rate": 8.096943151112448e-06,
+      "loss": 0.3324,
+      "step": 12691
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0312765853315824,
+      "learning_rate": 8.096633810997906e-06,
+      "loss": 0.5802,
+      "step": 12692
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.351692959657712,
+      "learning_rate": 8.096324451654107e-06,
+      "loss": 0.3891,
+      "step": 12693
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4182939550577307,
+      "learning_rate": 8.096015073082973e-06,
+      "loss": 0.3513,
+      "step": 12694
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.018196816133848,
+      "learning_rate": 8.095705675286427e-06,
+      "loss": 0.3734,
+      "step": 12695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.783917266305156,
+      "learning_rate": 8.095396258266386e-06,
+      "loss": 0.3719,
+      "step": 12696
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9927337832469991,
+      "learning_rate": 8.095086822024775e-06,
+      "loss": 0.4117,
+      "step": 12697
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6644134531975938,
+      "learning_rate": 8.094777366563514e-06,
+      "loss": 0.2437,
+      "step": 12698
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8064803243092706,
+      "learning_rate": 8.094467891884524e-06,
+      "loss": 0.3133,
+      "step": 12699
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.316307449029011,
+      "learning_rate": 8.09415839798973e-06,
+      "loss": 0.2092,
+      "step": 12700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.70106699606014,
+      "learning_rate": 8.093848884881047e-06,
+      "loss": 0.193,
+      "step": 12701
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6132996029306357,
+      "learning_rate": 8.093539352560406e-06,
+      "loss": 0.5312,
+      "step": 12702
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3125674281511184,
+      "learning_rate": 8.093229801029722e-06,
+      "loss": 0.2604,
+      "step": 12703
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6183885684500492,
+      "learning_rate": 8.092920230290921e-06,
+      "loss": 0.5437,
+      "step": 12704
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1232063300254578,
+      "learning_rate": 8.092610640345924e-06,
+      "loss": 0.3003,
+      "step": 12705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1079846487813017,
+      "learning_rate": 8.092301031196654e-06,
+      "loss": 0.1466,
+      "step": 12706
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1460640151282264,
+      "learning_rate": 8.091991402845033e-06,
+      "loss": 0.1902,
+      "step": 12707
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8180127986461645,
+      "learning_rate": 8.091681755292984e-06,
+      "loss": 0.1971,
+      "step": 12708
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2813547050910756,
+      "learning_rate": 8.091372088542431e-06,
+      "loss": 0.1268,
+      "step": 12709
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7489231183085414,
+      "learning_rate": 8.091062402595294e-06,
+      "loss": 0.2701,
+      "step": 12710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5854826601459038,
+      "learning_rate": 8.090752697453498e-06,
+      "loss": 0.1595,
+      "step": 12711
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5634528163652996,
+      "learning_rate": 8.090442973118966e-06,
+      "loss": 0.2823,
+      "step": 12712
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7432566343031084,
+      "learning_rate": 8.090133229593623e-06,
+      "loss": 0.3047,
+      "step": 12713
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.540830841637746,
+      "learning_rate": 8.089823466879388e-06,
+      "loss": 0.2558,
+      "step": 12714
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6817309347574498,
+      "learning_rate": 8.089513684978187e-06,
+      "loss": 0.3922,
+      "step": 12715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.413940180709568,
+      "learning_rate": 8.089203883891945e-06,
+      "loss": 0.1618,
+      "step": 12716
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2331139786948444,
+      "learning_rate": 8.088894063622586e-06,
+      "loss": 0.2709,
+      "step": 12717
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4621964909993936,
+      "learning_rate": 8.088584224172032e-06,
+      "loss": 0.5253,
+      "step": 12718
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6207795861321255,
+      "learning_rate": 8.088274365542207e-06,
+      "loss": 0.3687,
+      "step": 12719
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5397953767112011,
+      "learning_rate": 8.087964487735036e-06,
+      "loss": 0.2293,
+      "step": 12720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.709114556940061,
+      "learning_rate": 8.087654590752443e-06,
+      "loss": 0.2308,
+      "step": 12721
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8728132708863152,
+      "learning_rate": 8.087344674596353e-06,
+      "loss": 0.3347,
+      "step": 12722
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8799897705960849,
+      "learning_rate": 8.08703473926869e-06,
+      "loss": 0.2633,
+      "step": 12723
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3068184410865813,
+      "learning_rate": 8.086724784771378e-06,
+      "loss": 0.3021,
+      "step": 12724
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5848970476718962,
+      "learning_rate": 8.08641481110634e-06,
+      "loss": 0.2043,
+      "step": 12725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9845077965824236,
+      "learning_rate": 8.086104818275506e-06,
+      "loss": 0.3936,
+      "step": 12726
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9137253250223156,
+      "learning_rate": 8.085794806280798e-06,
+      "loss": 0.4979,
+      "step": 12727
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7080738890196758,
+      "learning_rate": 8.08548477512414e-06,
+      "loss": 0.4632,
+      "step": 12728
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7096506465767116,
+      "learning_rate": 8.085174724807459e-06,
+      "loss": 0.1839,
+      "step": 12729
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.002659318625013,
+      "learning_rate": 8.084864655332679e-06,
+      "loss": 0.1957,
+      "step": 12730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.127987903290575,
+      "learning_rate": 8.084554566701726e-06,
+      "loss": 0.2441,
+      "step": 12731
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5149388160995723,
+      "learning_rate": 8.084244458916527e-06,
+      "loss": 0.3266,
+      "step": 12732
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0448362796699513,
+      "learning_rate": 8.083934331979006e-06,
+      "loss": 0.1899,
+      "step": 12733
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6127954010074579,
+      "learning_rate": 8.083624185891088e-06,
+      "loss": 0.2243,
+      "step": 12734
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7791866264146812,
+      "learning_rate": 8.083314020654701e-06,
+      "loss": 0.2634,
+      "step": 12735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4450265820679182,
+      "learning_rate": 8.08300383627177e-06,
+      "loss": 0.2488,
+      "step": 12736
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4503514112724019,
+      "learning_rate": 8.082693632744222e-06,
+      "loss": 0.3303,
+      "step": 12737
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4932854916762894,
+      "learning_rate": 8.082383410073982e-06,
+      "loss": 0.3634,
+      "step": 12738
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6224170523205492,
+      "learning_rate": 8.082073168262976e-06,
+      "loss": 0.2531,
+      "step": 12739
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.053454409843023,
+      "learning_rate": 8.081762907313133e-06,
+      "loss": 0.412,
+      "step": 12740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5203997794735682,
+      "learning_rate": 8.081452627226379e-06,
+      "loss": 0.2977,
+      "step": 12741
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8229419458997966,
+      "learning_rate": 8.081142328004638e-06,
+      "loss": 0.4015,
+      "step": 12742
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7101855084254616,
+      "learning_rate": 8.080832009649838e-06,
+      "loss": 0.4826,
+      "step": 12743
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6520178362966487,
+      "learning_rate": 8.08052167216391e-06,
+      "loss": 0.386,
+      "step": 12744
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.519942838337492,
+      "learning_rate": 8.080211315548773e-06,
+      "loss": 0.3965,
+      "step": 12745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8411507631170236,
+      "learning_rate": 8.079900939806364e-06,
+      "loss": 0.4928,
+      "step": 12746
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7734712605896956,
+      "learning_rate": 8.079590544938603e-06,
+      "loss": 0.6256,
+      "step": 12747
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3005310797475484,
+      "learning_rate": 8.07928013094742e-06,
+      "loss": 0.2414,
+      "step": 12748
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4869943070740421,
+      "learning_rate": 8.078969697834743e-06,
+      "loss": 0.4003,
+      "step": 12749
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.436048657439113,
+      "learning_rate": 8.078659245602501e-06,
+      "loss": 0.2029,
+      "step": 12750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2282346379867262,
+      "learning_rate": 8.078348774252617e-06,
+      "loss": 0.1968,
+      "step": 12751
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6711285341841635,
+      "learning_rate": 8.078038283787024e-06,
+      "loss": 0.1825,
+      "step": 12752
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6410039146276107,
+      "learning_rate": 8.077727774207646e-06,
+      "loss": 0.5372,
+      "step": 12753
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.528675084481365,
+      "learning_rate": 8.077417245516415e-06,
+      "loss": 0.3122,
+      "step": 12754
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4393785190787718,
+      "learning_rate": 8.077106697715256e-06,
+      "loss": 0.2075,
+      "step": 12755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2929644267652443,
+      "learning_rate": 8.0767961308061e-06,
+      "loss": 0.1846,
+      "step": 12756
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.73909754285097,
+      "learning_rate": 8.076485544790875e-06,
+      "loss": 0.336,
+      "step": 12757
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5798455386622807,
+      "learning_rate": 8.076174939671509e-06,
+      "loss": 0.6019,
+      "step": 12758
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7391274781953516,
+      "learning_rate": 8.07586431544993e-06,
+      "loss": 0.4503,
+      "step": 12759
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7905914944281835,
+      "learning_rate": 8.075553672128068e-06,
+      "loss": 0.3734,
+      "step": 12760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4344303154979656,
+      "learning_rate": 8.075243009707853e-06,
+      "loss": 0.1704,
+      "step": 12761
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8548635559914104,
+      "learning_rate": 8.074932328191212e-06,
+      "loss": 0.4637,
+      "step": 12762
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.800423299005033,
+      "learning_rate": 8.074621627580075e-06,
+      "loss": 0.5323,
+      "step": 12763
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8761946369214448,
+      "learning_rate": 8.074310907876373e-06,
+      "loss": 0.3695,
+      "step": 12764
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4842407165823008,
+      "learning_rate": 8.074000169082031e-06,
+      "loss": 0.3918,
+      "step": 12765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.867821724369311,
+      "learning_rate": 8.073689411198983e-06,
+      "loss": 0.2955,
+      "step": 12766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6881638430928845,
+      "learning_rate": 8.073378634229158e-06,
+      "loss": 0.5564,
+      "step": 12767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.029095956440473,
+      "learning_rate": 8.073067838174486e-06,
+      "loss": 0.3898,
+      "step": 12768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.482750209187345,
+      "learning_rate": 8.072757023036895e-06,
+      "loss": 0.3054,
+      "step": 12769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8287407457569924,
+      "learning_rate": 8.072446188818316e-06,
+      "loss": 0.6015,
+      "step": 12770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7450338214586096,
+      "learning_rate": 8.07213533552068e-06,
+      "loss": 0.429,
+      "step": 12771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0513262514155939,
+      "learning_rate": 8.071824463145917e-06,
+      "loss": 0.1302,
+      "step": 12772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9079018689842675,
+      "learning_rate": 8.071513571695958e-06,
+      "loss": 0.6374,
+      "step": 12773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.615106939710959,
+      "learning_rate": 8.07120266117273e-06,
+      "loss": 0.3704,
+      "step": 12774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5482065750363132,
+      "learning_rate": 8.070891731578168e-06,
+      "loss": 0.4548,
+      "step": 12775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8419082624100471,
+      "learning_rate": 8.070580782914202e-06,
+      "loss": 0.4871,
+      "step": 12776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1417574427869317,
+      "learning_rate": 8.07026981518276e-06,
+      "loss": 0.3765,
+      "step": 12777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3380336128174979,
+      "learning_rate": 8.069958828385777e-06,
+      "loss": 0.2873,
+      "step": 12778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6165509330192627,
+      "learning_rate": 8.069647822525182e-06,
+      "loss": 0.2802,
+      "step": 12779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7991987378841534,
+      "learning_rate": 8.069336797602906e-06,
+      "loss": 0.3264,
+      "step": 12780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6989227601658334,
+      "learning_rate": 8.069025753620882e-06,
+      "loss": 0.2376,
+      "step": 12781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1738024104216727,
+      "learning_rate": 8.068714690581038e-06,
+      "loss": 0.1658,
+      "step": 12782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2058226821699662,
+      "learning_rate": 8.068403608485312e-06,
+      "loss": 0.1955,
+      "step": 12783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6980882852409138,
+      "learning_rate": 8.068092507335628e-06,
+      "loss": 0.3691,
+      "step": 12784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5766542587571648,
+      "learning_rate": 8.067781387133922e-06,
+      "loss": 0.3453,
+      "step": 12785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6562618579080142,
+      "learning_rate": 8.067470247882127e-06,
+      "loss": 0.3032,
+      "step": 12786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7394593436393198,
+      "learning_rate": 8.067159089582173e-06,
+      "loss": 0.2456,
+      "step": 12787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.963053932798184,
+      "learning_rate": 8.066847912235994e-06,
+      "loss": 0.5442,
+      "step": 12788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2906443181681895,
+      "learning_rate": 8.066536715845521e-06,
+      "loss": 0.2382,
+      "step": 12789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.746244837904163,
+      "learning_rate": 8.066225500412685e-06,
+      "loss": 0.5127,
+      "step": 12790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8642553566246582,
+      "learning_rate": 8.065914265939421e-06,
+      "loss": 0.4037,
+      "step": 12791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.193939364602718,
+      "learning_rate": 8.065603012427662e-06,
+      "loss": 0.1738,
+      "step": 12792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4084626407084957,
+      "learning_rate": 8.065291739879338e-06,
+      "loss": 0.2636,
+      "step": 12793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7354052937292581,
+      "learning_rate": 8.064980448296384e-06,
+      "loss": 0.2291,
+      "step": 12794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9355465208605644,
+      "learning_rate": 8.064669137680734e-06,
+      "loss": 0.2556,
+      "step": 12795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2958351197401634,
+      "learning_rate": 8.06435780803432e-06,
+      "loss": 0.2061,
+      "step": 12796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6827208785366774,
+      "learning_rate": 8.064046459359074e-06,
+      "loss": 0.3851,
+      "step": 12797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4571144576257864,
+      "learning_rate": 8.06373509165693e-06,
+      "loss": 0.2362,
+      "step": 12798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0064106882814325,
+      "learning_rate": 8.063423704929822e-06,
+      "loss": 0.4302,
+      "step": 12799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4289285020884075,
+      "learning_rate": 8.063112299179684e-06,
+      "loss": 0.333,
+      "step": 12800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8500524330602408,
+      "learning_rate": 8.06280087440845e-06,
+      "loss": 0.3906,
+      "step": 12801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8200908083845684,
+      "learning_rate": 8.062489430618053e-06,
+      "loss": 0.3493,
+      "step": 12802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.651812188339536,
+      "learning_rate": 8.062177967810427e-06,
+      "loss": 0.3565,
+      "step": 12803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2351916665797023,
+      "learning_rate": 8.061866485987507e-06,
+      "loss": 0.2526,
+      "step": 12804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6206361262590157,
+      "learning_rate": 8.061554985151226e-06,
+      "loss": 0.363,
+      "step": 12805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.943155843533507,
+      "learning_rate": 8.061243465303517e-06,
+      "loss": 0.4429,
+      "step": 12806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.95234091469634,
+      "learning_rate": 8.060931926446319e-06,
+      "loss": 0.5529,
+      "step": 12807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4719910168937316,
+      "learning_rate": 8.060620368581564e-06,
+      "loss": 0.3512,
+      "step": 12808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6679723860189575,
+      "learning_rate": 8.060308791711184e-06,
+      "loss": 0.4308,
+      "step": 12809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6654667492403652,
+      "learning_rate": 8.059997195837118e-06,
+      "loss": 0.2103,
+      "step": 12810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3946419973769364,
+      "learning_rate": 8.0596855809613e-06,
+      "loss": 0.2493,
+      "step": 12811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2905504395651441,
+      "learning_rate": 8.059373947085664e-06,
+      "loss": 0.1277,
+      "step": 12812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3483367749694044,
+      "learning_rate": 8.059062294212144e-06,
+      "loss": 0.3048,
+      "step": 12813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.436174179087336,
+      "learning_rate": 8.058750622342679e-06,
+      "loss": 0.1793,
+      "step": 12814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4959281525958346,
+      "learning_rate": 8.0584389314792e-06,
+      "loss": 0.4245,
+      "step": 12815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9381211634852902,
+      "learning_rate": 8.058127221623645e-06,
+      "loss": 0.475,
+      "step": 12816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.759930377492104,
+      "learning_rate": 8.05781549277795e-06,
+      "loss": 0.4348,
+      "step": 12817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.085689654967108,
+      "learning_rate": 8.05750374494405e-06,
+      "loss": 0.1669,
+      "step": 12818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.624286446805065,
+      "learning_rate": 8.057191978123883e-06,
+      "loss": 0.3597,
+      "step": 12819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9806759498298905,
+      "learning_rate": 8.05688019231938e-06,
+      "loss": 0.1979,
+      "step": 12820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6255258158793098,
+      "learning_rate": 8.056568387532482e-06,
+      "loss": 0.3346,
+      "step": 12821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9509497335637913,
+      "learning_rate": 8.056256563765122e-06,
+      "loss": 0.2632,
+      "step": 12822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3692377582492186,
+      "learning_rate": 8.055944721019238e-06,
+      "loss": 0.1947,
+      "step": 12823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1976733596575104,
+      "learning_rate": 8.055632859296765e-06,
+      "loss": 0.4574,
+      "step": 12824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7240730604137349,
+      "learning_rate": 8.055320978599643e-06,
+      "loss": 0.2318,
+      "step": 12825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9239783659230971,
+      "learning_rate": 8.055009078929806e-06,
+      "loss": 0.5012,
+      "step": 12826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.862373958873605,
+      "learning_rate": 8.054697160289189e-06,
+      "loss": 0.5495,
+      "step": 12827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4160113420092824,
+      "learning_rate": 8.054385222679733e-06,
+      "loss": 0.2601,
+      "step": 12828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1252125416757437,
+      "learning_rate": 8.054073266103373e-06,
+      "loss": 0.128,
+      "step": 12829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.61872799726762,
+      "learning_rate": 8.053761290562046e-06,
+      "loss": 0.1883,
+      "step": 12830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4719364838302056,
+      "learning_rate": 8.053449296057689e-06,
+      "loss": 0.1774,
+      "step": 12831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4048975109666286,
+      "learning_rate": 8.05313728259224e-06,
+      "loss": 0.3622,
+      "step": 12832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5339183548372644,
+      "learning_rate": 8.052825250167637e-06,
+      "loss": 0.2813,
+      "step": 12833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5215574744003075,
+      "learning_rate": 8.052513198785817e-06,
+      "loss": 0.432,
+      "step": 12834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5867707783495704,
+      "learning_rate": 8.052201128448718e-06,
+      "loss": 0.2824,
+      "step": 12835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.540201012919796,
+      "learning_rate": 8.051889039158279e-06,
+      "loss": 0.4139,
+      "step": 12836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3532358736421062,
+      "learning_rate": 8.051576930916433e-06,
+      "loss": 0.2482,
+      "step": 12837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5154082865973226,
+      "learning_rate": 8.051264803725127e-06,
+      "loss": 0.3419,
+      "step": 12838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.91852791980762,
+      "learning_rate": 8.05095265758629e-06,
+      "loss": 0.2568,
+      "step": 12839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3658720977383538,
+      "learning_rate": 8.050640492501867e-06,
+      "loss": 0.2379,
+      "step": 12840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.952194572082968,
+      "learning_rate": 8.050328308473793e-06,
+      "loss": 0.5874,
+      "step": 12841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0469722142320754,
+      "learning_rate": 8.050016105504006e-06,
+      "loss": 0.3548,
+      "step": 12842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.782360928839678,
+      "learning_rate": 8.049703883594449e-06,
+      "loss": 0.3171,
+      "step": 12843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.774445270522122,
+      "learning_rate": 8.049391642747055e-06,
+      "loss": 0.3043,
+      "step": 12844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.962074275142366,
+      "learning_rate": 8.049079382963768e-06,
+      "loss": 0.4134,
+      "step": 12845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4203999683884185,
+      "learning_rate": 8.048767104246525e-06,
+      "loss": 0.1771,
+      "step": 12846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.539079315074326,
+      "learning_rate": 8.048454806597263e-06,
+      "loss": 0.2492,
+      "step": 12847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7461469202672204,
+      "learning_rate": 8.048142490017925e-06,
+      "loss": 0.3435,
+      "step": 12848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4785594190413132,
+      "learning_rate": 8.04783015451045e-06,
+      "loss": 0.1551,
+      "step": 12849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4438260269996084,
+      "learning_rate": 8.047517800076773e-06,
+      "loss": 0.1279,
+      "step": 12850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3323007049588262,
+      "learning_rate": 8.047205426718841e-06,
+      "loss": 0.2358,
+      "step": 12851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7131672704297694,
+      "learning_rate": 8.046893034438588e-06,
+      "loss": 0.3966,
+      "step": 12852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2219630933430243,
+      "learning_rate": 8.046580623237954e-06,
+      "loss": 0.2034,
+      "step": 12853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.289489305955888,
+      "learning_rate": 8.046268193118882e-06,
+      "loss": 0.3741,
+      "step": 12854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.601632279527407,
+      "learning_rate": 8.045955744083312e-06,
+      "loss": 0.3396,
+      "step": 12855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.472860726850423,
+      "learning_rate": 8.045643276133182e-06,
+      "loss": 0.1613,
+      "step": 12856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3052318431089678,
+      "learning_rate": 8.045330789270431e-06,
+      "loss": 0.1794,
+      "step": 12857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7431848989571714,
+      "learning_rate": 8.045018283497005e-06,
+      "loss": 0.5134,
+      "step": 12858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4455224490045393,
+      "learning_rate": 8.04470575881484e-06,
+      "loss": 0.4045,
+      "step": 12859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5074086809730063,
+      "learning_rate": 8.044393215225877e-06,
+      "loss": 0.3488,
+      "step": 12860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5736719890526831,
+      "learning_rate": 8.044080652732058e-06,
+      "loss": 0.3264,
+      "step": 12861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5524383070723373,
+      "learning_rate": 8.043768071335326e-06,
+      "loss": 0.2279,
+      "step": 12862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1430354645789005,
+      "learning_rate": 8.043455471037618e-06,
+      "loss": 0.1688,
+      "step": 12863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8947516598251912,
+      "learning_rate": 8.043142851840877e-06,
+      "loss": 0.5842,
+      "step": 12864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4116251149445205,
+      "learning_rate": 8.042830213747044e-06,
+      "loss": 0.1965,
+      "step": 12865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4940227287502843,
+      "learning_rate": 8.04251755675806e-06,
+      "loss": 0.1657,
+      "step": 12866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7730179429031168,
+      "learning_rate": 8.042204880875869e-06,
+      "loss": 0.5693,
+      "step": 12867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9663639728637552,
+      "learning_rate": 8.041892186102408e-06,
+      "loss": 0.1982,
+      "step": 12868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3544850558950614,
+      "learning_rate": 8.041579472439625e-06,
+      "loss": 0.383,
+      "step": 12869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4664423980281538,
+      "learning_rate": 8.041266739889457e-06,
+      "loss": 0.1809,
+      "step": 12870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7214882366084308,
+      "learning_rate": 8.040953988453847e-06,
+      "loss": 0.3709,
+      "step": 12871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1306064081429925,
+      "learning_rate": 8.040641218134737e-06,
+      "loss": 0.3314,
+      "step": 12872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.68036470955701,
+      "learning_rate": 8.04032842893407e-06,
+      "loss": 0.2181,
+      "step": 12873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9205335164743182,
+      "learning_rate": 8.04001562085379e-06,
+      "loss": 0.1479,
+      "step": 12874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2153284571692027,
+      "learning_rate": 8.039702793895834e-06,
+      "loss": 0.2175,
+      "step": 12875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5798134434731868,
+      "learning_rate": 8.039389948062151e-06,
+      "loss": 0.3238,
+      "step": 12876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.811340059525309,
+      "learning_rate": 8.039077083354679e-06,
+      "loss": 0.2436,
+      "step": 12877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.810465742966168,
+      "learning_rate": 8.038764199775363e-06,
+      "loss": 0.2377,
+      "step": 12878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5682219686208902,
+      "learning_rate": 8.038451297326146e-06,
+      "loss": 0.3388,
+      "step": 12879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4175113118778295,
+      "learning_rate": 8.03813837600897e-06,
+      "loss": 0.3221,
+      "step": 12880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2295953254374525,
+      "learning_rate": 8.037825435825776e-06,
+      "loss": 0.1735,
+      "step": 12881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6691868196036752,
+      "learning_rate": 8.037512476778515e-06,
+      "loss": 0.3287,
+      "step": 12882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9410236449833096,
+      "learning_rate": 8.037199498869121e-06,
+      "loss": 0.4441,
+      "step": 12883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5281399142038736,
+      "learning_rate": 8.036886502099546e-06,
+      "loss": 0.3856,
+      "step": 12884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6319017220094982,
+      "learning_rate": 8.036573486471726e-06,
+      "loss": 0.1739,
+      "step": 12885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6858621304681884,
+      "learning_rate": 8.03626045198761e-06,
+      "loss": 0.4642,
+      "step": 12886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3318038337979086,
+      "learning_rate": 8.03594739864914e-06,
+      "loss": 0.1763,
+      "step": 12887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5308048467325261,
+      "learning_rate": 8.035634326458258e-06,
+      "loss": 0.3615,
+      "step": 12888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.539197671347266,
+      "learning_rate": 8.035321235416912e-06,
+      "loss": 0.2962,
+      "step": 12889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6669593164189807,
+      "learning_rate": 8.035008125527044e-06,
+      "loss": 0.3775,
+      "step": 12890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7196677997780572,
+      "learning_rate": 8.034694996790598e-06,
+      "loss": 0.1717,
+      "step": 12891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2491013992471227,
+      "learning_rate": 8.03438184920952e-06,
+      "loss": 0.1724,
+      "step": 12892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2721780409809154,
+      "learning_rate": 8.034068682785755e-06,
+      "loss": 0.1651,
+      "step": 12893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.854858723797633,
+      "learning_rate": 8.033755497521244e-06,
+      "loss": 0.2838,
+      "step": 12894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5875079361274174,
+      "learning_rate": 8.033442293417934e-06,
+      "loss": 0.254,
+      "step": 12895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1151000250561545,
+      "learning_rate": 8.033129070477771e-06,
+      "loss": 0.7323,
+      "step": 12896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.895810554615575,
+      "learning_rate": 8.032815828702698e-06,
+      "loss": 0.541,
+      "step": 12897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.67062826553037,
+      "learning_rate": 8.032502568094662e-06,
+      "loss": 0.3241,
+      "step": 12898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3567639711930117,
+      "learning_rate": 8.032189288655607e-06,
+      "loss": 0.2725,
+      "step": 12899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7644474050139045,
+      "learning_rate": 8.031875990387482e-06,
+      "loss": 0.3592,
+      "step": 12900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4559351221496264,
+      "learning_rate": 8.031562673292226e-06,
+      "loss": 0.3019,
+      "step": 12901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4472648322017008,
+      "learning_rate": 8.031249337371789e-06,
+      "loss": 0.2513,
+      "step": 12902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4719379093460263,
+      "learning_rate": 8.030935982628115e-06,
+      "loss": 0.2551,
+      "step": 12903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.646373806156157,
+      "learning_rate": 8.030622609063151e-06,
+      "loss": 0.4296,
+      "step": 12904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8697109568044343,
+      "learning_rate": 8.030309216678844e-06,
+      "loss": 0.4485,
+      "step": 12905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7264599618627485,
+      "learning_rate": 8.029995805477137e-06,
+      "loss": 0.3805,
+      "step": 12906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.085349395390081,
+      "learning_rate": 8.029682375459978e-06,
+      "loss": 0.2944,
+      "step": 12907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.830797939356323,
+      "learning_rate": 8.029368926629313e-06,
+      "loss": 0.3006,
+      "step": 12908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3017540822186722,
+      "learning_rate": 8.029055458987089e-06,
+      "loss": 0.2501,
+      "step": 12909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4129569410644567,
+      "learning_rate": 8.028741972535251e-06,
+      "loss": 0.274,
+      "step": 12910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1758291006648534,
+      "learning_rate": 8.028428467275748e-06,
+      "loss": 0.1604,
+      "step": 12911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6229942143386187,
+      "learning_rate": 8.028114943210525e-06,
+      "loss": 0.2028,
+      "step": 12912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6924141310288516,
+      "learning_rate": 8.027801400341531e-06,
+      "loss": 0.3588,
+      "step": 12913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.739020416533739,
+      "learning_rate": 8.02748783867071e-06,
+      "loss": 0.2158,
+      "step": 12914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3070414730864737,
+      "learning_rate": 8.02717425820001e-06,
+      "loss": 0.1795,
+      "step": 12915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.365590604331488,
+      "learning_rate": 8.026860658931378e-06,
+      "loss": 0.3214,
+      "step": 12916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.783351436601834,
+      "learning_rate": 8.026547040866765e-06,
+      "loss": 0.5989,
+      "step": 12917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2177502906025026,
+      "learning_rate": 8.026233404008115e-06,
+      "loss": 0.1605,
+      "step": 12918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.953852272765727,
+      "learning_rate": 8.025919748357376e-06,
+      "loss": 0.2653,
+      "step": 12919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4375599273339195,
+      "learning_rate": 8.025606073916496e-06,
+      "loss": 0.4465,
+      "step": 12920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9466879780859667,
+      "learning_rate": 8.025292380687422e-06,
+      "loss": 0.4939,
+      "step": 12921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6203962755939263,
+      "learning_rate": 8.024978668672103e-06,
+      "loss": 0.329,
+      "step": 12922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3386167993490954,
+      "learning_rate": 8.024664937872489e-06,
+      "loss": 0.2511,
+      "step": 12923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6358496468390156,
+      "learning_rate": 8.024351188290523e-06,
+      "loss": 0.3031,
+      "step": 12924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7903350693553062,
+      "learning_rate": 8.024037419928158e-06,
+      "loss": 0.4818,
+      "step": 12925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6228304740517139,
+      "learning_rate": 8.023723632787341e-06,
+      "loss": 0.4509,
+      "step": 12926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.482080380053237,
+      "learning_rate": 8.023409826870019e-06,
+      "loss": 0.3837,
+      "step": 12927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.555935943083282,
+      "learning_rate": 8.023096002178144e-06,
+      "loss": 0.3273,
+      "step": 12928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3074030854009309,
+      "learning_rate": 8.02278215871366e-06,
+      "loss": 0.242,
+      "step": 12929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2140217471763664,
+      "learning_rate": 8.02246829647852e-06,
+      "loss": 0.4002,
+      "step": 12930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4655567646409748,
+      "learning_rate": 8.022154415474672e-06,
+      "loss": 0.281,
+      "step": 12931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7619070187220556,
+      "learning_rate": 8.021840515704064e-06,
+      "loss": 0.2734,
+      "step": 12932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3218255402395558,
+      "learning_rate": 8.021526597168647e-06,
+      "loss": 0.1987,
+      "step": 12933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7000398427838272,
+      "learning_rate": 8.021212659870369e-06,
+      "loss": 0.4233,
+      "step": 12934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3458230736125387,
+      "learning_rate": 8.02089870381118e-06,
+      "loss": 0.2371,
+      "step": 12935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.305264099991852,
+      "learning_rate": 8.020584728993028e-06,
+      "loss": 0.3169,
+      "step": 12936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3723444084531604,
+      "learning_rate": 8.020270735417865e-06,
+      "loss": 0.1522,
+      "step": 12937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4710697904976549,
+      "learning_rate": 8.01995672308764e-06,
+      "loss": 0.2244,
+      "step": 12938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7045225251978378,
+      "learning_rate": 8.019642692004302e-06,
+      "loss": 0.2311,
+      "step": 12939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4542749838215023,
+      "learning_rate": 8.019328642169802e-06,
+      "loss": 0.334,
+      "step": 12940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7322617976172572,
+      "learning_rate": 8.019014573586091e-06,
+      "loss": 0.3543,
+      "step": 12941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8955841009794705,
+      "learning_rate": 8.018700486255118e-06,
+      "loss": 0.3679,
+      "step": 12942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3266457959501103,
+      "learning_rate": 8.018386380178832e-06,
+      "loss": 0.1901,
+      "step": 12943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0040905256109656,
+      "learning_rate": 8.018072255359185e-06,
+      "loss": 0.4382,
+      "step": 12944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.743738228406106,
+      "learning_rate": 8.01775811179813e-06,
+      "loss": 0.3704,
+      "step": 12945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.762588129985301,
+      "learning_rate": 8.017443949497616e-06,
+      "loss": 0.2265,
+      "step": 12946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8274046281456775,
+      "learning_rate": 8.017129768459592e-06,
+      "loss": 0.5183,
+      "step": 12947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3261132740000694,
+      "learning_rate": 8.01681556868601e-06,
+      "loss": 0.2411,
+      "step": 12948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6015963969481477,
+      "learning_rate": 8.016501350178822e-06,
+      "loss": 0.3649,
+      "step": 12949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7770254583724747,
+      "learning_rate": 8.016187112939979e-06,
+      "loss": 0.206,
+      "step": 12950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.310475098737932,
+      "learning_rate": 8.015872856971433e-06,
+      "loss": 0.188,
+      "step": 12951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.930979063841825,
+      "learning_rate": 8.015558582275132e-06,
+      "loss": 0.336,
+      "step": 12952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2478845293720937,
+      "learning_rate": 8.015244288853032e-06,
+      "loss": 0.1511,
+      "step": 12953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7342525782101137,
+      "learning_rate": 8.014929976707082e-06,
+      "loss": 0.2675,
+      "step": 12954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1123829622972456,
+      "learning_rate": 8.014615645839235e-06,
+      "loss": 0.6528,
+      "step": 12955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.816499633849664,
+      "learning_rate": 8.014301296251442e-06,
+      "loss": 0.4025,
+      "step": 12956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2098216653246991,
+      "learning_rate": 8.013986927945656e-06,
+      "loss": 0.2023,
+      "step": 12957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9983195233510787,
+      "learning_rate": 8.013672540923827e-06,
+      "loss": 0.4751,
+      "step": 12958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7567830377304285,
+      "learning_rate": 8.01335813518791e-06,
+      "loss": 0.4605,
+      "step": 12959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4594789149436365,
+      "learning_rate": 8.013043710739858e-06,
+      "loss": 0.177,
+      "step": 12960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7505576275612609,
+      "learning_rate": 8.01272926758162e-06,
+      "loss": 0.5629,
+      "step": 12961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4208366813217033,
+      "learning_rate": 8.012414805715149e-06,
+      "loss": 0.1336,
+      "step": 12962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5979219905830064,
+      "learning_rate": 8.0121003251424e-06,
+      "loss": 0.3469,
+      "step": 12963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6465122813329462,
+      "learning_rate": 8.011785825865325e-06,
+      "loss": 0.4739,
+      "step": 12964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9895777344416852,
+      "learning_rate": 8.011471307885878e-06,
+      "loss": 0.4633,
+      "step": 12965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4387878319450196,
+      "learning_rate": 8.01115677120601e-06,
+      "loss": 0.2979,
+      "step": 12966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6241169407685763,
+      "learning_rate": 8.010842215827674e-06,
+      "loss": 0.2434,
+      "step": 12967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0097369977630308,
+      "learning_rate": 8.010527641752826e-06,
+      "loss": 0.2071,
+      "step": 12968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9494728866827267,
+      "learning_rate": 8.010213048983418e-06,
+      "loss": 0.2252,
+      "step": 12969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7287467284802578,
+      "learning_rate": 8.009898437521402e-06,
+      "loss": 0.3517,
+      "step": 12970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8815166675679385,
+      "learning_rate": 8.009583807368734e-06,
+      "loss": 0.4021,
+      "step": 12971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7187144600921795,
+      "learning_rate": 8.009269158527366e-06,
+      "loss": 0.4858,
+      "step": 12972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.173233746981658,
+      "learning_rate": 8.008954490999255e-06,
+      "loss": 0.3228,
+      "step": 12973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7469651140980689,
+      "learning_rate": 8.008639804786349e-06,
+      "loss": 0.5045,
+      "step": 12974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.736008841098464,
+      "learning_rate": 8.008325099890608e-06,
+      "loss": 0.3215,
+      "step": 12975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5588621850088442,
+      "learning_rate": 8.008010376313983e-06,
+      "loss": 0.3816,
+      "step": 12976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2450439102451412,
+      "learning_rate": 8.007695634058431e-06,
+      "loss": 0.1571,
+      "step": 12977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5860238392352561,
+      "learning_rate": 8.007380873125903e-06,
+      "loss": 0.2391,
+      "step": 12978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4521121833139117,
+      "learning_rate": 8.007066093518357e-06,
+      "loss": 0.3467,
+      "step": 12979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7929655004401255,
+      "learning_rate": 8.006751295237745e-06,
+      "loss": 0.5198,
+      "step": 12980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.640897937051454,
+      "learning_rate": 8.006436478286024e-06,
+      "loss": 0.3645,
+      "step": 12981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4492759076386097,
+      "learning_rate": 8.006121642665146e-06,
+      "loss": 0.3526,
+      "step": 12982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.617120847388116,
+      "learning_rate": 8.005806788377068e-06,
+      "loss": 0.3655,
+      "step": 12983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8702594294623287,
+      "learning_rate": 8.005491915423748e-06,
+      "loss": 0.4413,
+      "step": 12984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.530589506205833,
+      "learning_rate": 8.005177023807136e-06,
+      "loss": 0.1958,
+      "step": 12985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.177275592018266,
+      "learning_rate": 8.004862113529191e-06,
+      "loss": 0.1971,
+      "step": 12986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.187732520688264,
+      "learning_rate": 8.004547184591864e-06,
+      "loss": 0.5023,
+      "step": 12987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5888479568077318,
+      "learning_rate": 8.004232236997118e-06,
+      "loss": 0.3773,
+      "step": 12988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8208060352743736,
+      "learning_rate": 8.003917270746902e-06,
+      "loss": 0.3728,
+      "step": 12989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.550951897867348,
+      "learning_rate": 8.003602285843175e-06,
+      "loss": 0.3739,
+      "step": 12990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4028360411244438,
+      "learning_rate": 8.003287282287893e-06,
+      "loss": 0.2371,
+      "step": 12991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0191572445284915,
+      "learning_rate": 8.002972260083011e-06,
+      "loss": 0.304,
+      "step": 12992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6624689710560951,
+      "learning_rate": 8.002657219230485e-06,
+      "loss": 0.3556,
+      "step": 12993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9742125688825185,
+      "learning_rate": 8.002342159732273e-06,
+      "loss": 0.4497,
+      "step": 12994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.748615597533975,
+      "learning_rate": 8.00202708159033e-06,
+      "loss": 0.2387,
+      "step": 12995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4936310553678185,
+      "learning_rate": 8.001711984806612e-06,
+      "loss": 0.1768,
+      "step": 12996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1920272912612127,
+      "learning_rate": 8.001396869383079e-06,
+      "loss": 0.1527,
+      "step": 12997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4917868231039917,
+      "learning_rate": 8.001081735321683e-06,
+      "loss": 0.3094,
+      "step": 12998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9653990474671639,
+      "learning_rate": 8.000766582624386e-06,
+      "loss": 0.6083,
+      "step": 12999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.878021799756718,
+      "learning_rate": 8.00045141129314e-06,
+      "loss": 0.1585,
+      "step": 13000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.483566302306238,
+      "learning_rate": 8.000136221329906e-06,
+      "loss": 0.2167,
+      "step": 13001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3753354140307696,
+      "learning_rate": 7.99982101273664e-06,
+      "loss": 0.2613,
+      "step": 13002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9474869050435304,
+      "learning_rate": 7.999505785515298e-06,
+      "loss": 0.4256,
+      "step": 13003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3875911601079167,
+      "learning_rate": 7.999190539667839e-06,
+      "loss": 0.3012,
+      "step": 13004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4359320662253494,
+      "learning_rate": 7.998875275196218e-06,
+      "loss": 0.3219,
+      "step": 13005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4383734402593813,
+      "learning_rate": 7.998559992102397e-06,
+      "loss": 0.4105,
+      "step": 13006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8793613651924048,
+      "learning_rate": 7.998244690388333e-06,
+      "loss": 0.6204,
+      "step": 13007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6724516284188615,
+      "learning_rate": 7.997929370055982e-06,
+      "loss": 0.2884,
+      "step": 13008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5169375312156135,
+      "learning_rate": 7.9976140311073e-06,
+      "loss": 0.2564,
+      "step": 13009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4210164863340113,
+      "learning_rate": 7.99729867354425e-06,
+      "loss": 0.384,
+      "step": 13010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4545560962850912,
+      "learning_rate": 7.996983297368789e-06,
+      "loss": 0.3345,
+      "step": 13011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9811080412443611,
+      "learning_rate": 7.996667902582871e-06,
+      "loss": 0.422,
+      "step": 13012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2135608191336342,
+      "learning_rate": 7.996352489188462e-06,
+      "loss": 0.2416,
+      "step": 13013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8827075137550648,
+      "learning_rate": 7.996037057187515e-06,
+      "loss": 0.3531,
+      "step": 13014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.076659625133288,
+      "learning_rate": 7.99572160658199e-06,
+      "loss": 0.1364,
+      "step": 13015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.184765924591419,
+      "learning_rate": 7.995406137373848e-06,
+      "loss": 0.21,
+      "step": 13016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7318576336860343,
+      "learning_rate": 7.995090649565044e-06,
+      "loss": 0.6151,
+      "step": 13017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1495834394323325,
+      "learning_rate": 7.99477514315754e-06,
+      "loss": 0.1972,
+      "step": 13018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.347768835534145,
+      "learning_rate": 7.994459618153295e-06,
+      "loss": 0.1555,
+      "step": 13019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0296922597763247,
+      "learning_rate": 7.994144074554267e-06,
+      "loss": 0.3934,
+      "step": 13020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8442322690538966,
+      "learning_rate": 7.993828512362418e-06,
+      "loss": 0.1803,
+      "step": 13021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9814225106692271,
+      "learning_rate": 7.993512931579705e-06,
+      "loss": 0.382,
+      "step": 13022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7568965856867436,
+      "learning_rate": 7.993197332208089e-06,
+      "loss": 0.4231,
+      "step": 13023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7281708583645137,
+      "learning_rate": 7.992881714249529e-06,
+      "loss": 0.4162,
+      "step": 13024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7787594407542135,
+      "learning_rate": 7.992566077705985e-06,
+      "loss": 0.3635,
+      "step": 13025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8174599780263558,
+      "learning_rate": 7.992250422579418e-06,
+      "loss": 0.3438,
+      "step": 13026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7154912920825063,
+      "learning_rate": 7.991934748871787e-06,
+      "loss": 0.4656,
+      "step": 13027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.342279821352662,
+      "learning_rate": 7.991619056585052e-06,
+      "loss": 0.2455,
+      "step": 13028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7326856840215172,
+      "learning_rate": 7.991303345721175e-06,
+      "loss": 0.5738,
+      "step": 13029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.847137503356343,
+      "learning_rate": 7.990987616282116e-06,
+      "loss": 0.6019,
+      "step": 13030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6284810032576167,
+      "learning_rate": 7.990671868269832e-06,
+      "loss": 0.3013,
+      "step": 13031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6201634690731563,
+      "learning_rate": 7.99035610168629e-06,
+      "loss": 0.4785,
+      "step": 13032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.542793307410046,
+      "learning_rate": 7.990040316533445e-06,
+      "loss": 0.3625,
+      "step": 13033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3801167283758264,
+      "learning_rate": 7.989724512813263e-06,
+      "loss": 0.2175,
+      "step": 13034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7481270414755792,
+      "learning_rate": 7.989408690527702e-06,
+      "loss": 0.2951,
+      "step": 13035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.011090641666915,
+      "learning_rate": 7.989092849678723e-06,
+      "loss": 0.5971,
+      "step": 13036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.386641868706622,
+      "learning_rate": 7.98877699026829e-06,
+      "loss": 0.1175,
+      "step": 13037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8151639238004715,
+      "learning_rate": 7.98846111229836e-06,
+      "loss": 0.5602,
+      "step": 13038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8481220642094907,
+      "learning_rate": 7.988145215770898e-06,
+      "loss": 0.5272,
+      "step": 13039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.131415083995185,
+      "learning_rate": 7.987829300687864e-06,
+      "loss": 0.1675,
+      "step": 13040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.639714614541038,
+      "learning_rate": 7.987513367051221e-06,
+      "loss": 0.4008,
+      "step": 13041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.800107484413096,
+      "learning_rate": 7.98719741486293e-06,
+      "loss": 0.3725,
+      "step": 13042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4570126545105506,
+      "learning_rate": 7.986881444124954e-06,
+      "loss": 0.3347,
+      "step": 13043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7360151135341872,
+      "learning_rate": 7.986565454839254e-06,
+      "loss": 0.4184,
+      "step": 13044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9844837534381936,
+      "learning_rate": 7.986249447007793e-06,
+      "loss": 0.18,
+      "step": 13045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7517463387683385,
+      "learning_rate": 7.985933420632531e-06,
+      "loss": 0.3806,
+      "step": 13046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6572709445708846,
+      "learning_rate": 7.985617375715433e-06,
+      "loss": 0.5444,
+      "step": 13047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7998742490128787,
+      "learning_rate": 7.985301312258463e-06,
+      "loss": 0.3962,
+      "step": 13048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.941020382272499,
+      "learning_rate": 7.984985230263579e-06,
+      "loss": 0.467,
+      "step": 13049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9012129279728938,
+      "learning_rate": 7.984669129732747e-06,
+      "loss": 0.4965,
+      "step": 13050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2139120917694473,
+      "learning_rate": 7.984353010667928e-06,
+      "loss": 0.3444,
+      "step": 13051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.943949302194756,
+      "learning_rate": 7.984036873071088e-06,
+      "loss": 0.4945,
+      "step": 13052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8196643424423553,
+      "learning_rate": 7.983720716944188e-06,
+      "loss": 0.2997,
+      "step": 13053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4508150311441665,
+      "learning_rate": 7.98340454228919e-06,
+      "loss": 0.3306,
+      "step": 13054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.019992137536006,
+      "learning_rate": 7.983088349108062e-06,
+      "loss": 0.5579,
+      "step": 13055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8128971680036297,
+      "learning_rate": 7.982772137402762e-06,
+      "loss": 0.1937,
+      "step": 13056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5605986566271848,
+      "learning_rate": 7.982455907175255e-06,
+      "loss": 0.4308,
+      "step": 13057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4231230873078715,
+      "learning_rate": 7.982139658427508e-06,
+      "loss": 0.3309,
+      "step": 13058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9876221649243744,
+      "learning_rate": 7.981823391161484e-06,
+      "loss": 0.3555,
+      "step": 13059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8822293822942395,
+      "learning_rate": 7.98150710537914e-06,
+      "loss": 0.37,
+      "step": 13060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9291901420470492,
+      "learning_rate": 7.98119080108245e-06,
+      "loss": 0.3862,
+      "step": 13061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5391458752602958,
+      "learning_rate": 7.980874478273371e-06,
+      "loss": 0.3454,
+      "step": 13062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.460596080399059,
+      "learning_rate": 7.980558136953872e-06,
+      "loss": 0.2881,
+      "step": 13063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.354177816662979,
+      "learning_rate": 7.980241777125915e-06,
+      "loss": 0.1754,
+      "step": 13064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5088595179431195,
+      "learning_rate": 7.979925398791463e-06,
+      "loss": 0.2743,
+      "step": 13065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9055208586151995,
+      "learning_rate": 7.979609001952484e-06,
+      "loss": 0.4118,
+      "step": 13066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.513172165617328,
+      "learning_rate": 7.979292586610941e-06,
+      "loss": 0.3212,
+      "step": 13067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5675596309762003,
+      "learning_rate": 7.9789761527688e-06,
+      "loss": 0.2869,
+      "step": 13068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5569150962704608,
+      "learning_rate": 7.978659700428024e-06,
+      "loss": 0.3586,
+      "step": 13069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.739801275902207,
+      "learning_rate": 7.97834322959058e-06,
+      "loss": 0.2683,
+      "step": 13070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.175805845668721,
+      "learning_rate": 7.978026740258429e-06,
+      "loss": 0.1982,
+      "step": 13071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8085572587156888,
+      "learning_rate": 7.977710232433543e-06,
+      "loss": 0.2683,
+      "step": 13072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.794144290776643,
+      "learning_rate": 7.977393706117881e-06,
+      "loss": 0.186,
+      "step": 13073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7960066417894394,
+      "learning_rate": 7.977077161313416e-06,
+      "loss": 0.4754,
+      "step": 13074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7782791526761283,
+      "learning_rate": 7.976760598022107e-06,
+      "loss": 0.4214,
+      "step": 13075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6254520039794502,
+      "learning_rate": 7.976444016245922e-06,
+      "loss": 0.1985,
+      "step": 13076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4835959498272135,
+      "learning_rate": 7.976127415986825e-06,
+      "loss": 0.1721,
+      "step": 13077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.58841298134504,
+      "learning_rate": 7.975810797246786e-06,
+      "loss": 0.2742,
+      "step": 13078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9401936265089097,
+      "learning_rate": 7.975494160027768e-06,
+      "loss": 0.554,
+      "step": 13079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6574440955525258,
+      "learning_rate": 7.97517750433174e-06,
+      "loss": 0.2179,
+      "step": 13080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4380858305662034,
+      "learning_rate": 7.974860830160667e-06,
+      "loss": 0.3869,
+      "step": 13081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1577343949139602,
+      "learning_rate": 7.974544137516512e-06,
+      "loss": 0.2301,
+      "step": 13082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6877330193083506,
+      "learning_rate": 7.974227426401247e-06,
+      "loss": 0.237,
+      "step": 13083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9135209621485372,
+      "learning_rate": 7.973910696816835e-06,
+      "loss": 0.4794,
+      "step": 13084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.467018809081435,
+      "learning_rate": 7.973593948765245e-06,
+      "loss": 0.1989,
+      "step": 13085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.660703851481003,
+      "learning_rate": 7.973277182248442e-06,
+      "loss": 0.2129,
+      "step": 13086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7344802876037135,
+      "learning_rate": 7.972960397268395e-06,
+      "loss": 0.4655,
+      "step": 13087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8319958722047758,
+      "learning_rate": 7.972643593827069e-06,
+      "loss": 0.4964,
+      "step": 13088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9909654309743456,
+      "learning_rate": 7.972326771926435e-06,
+      "loss": 0.2557,
+      "step": 13089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8365068729655263,
+      "learning_rate": 7.972009931568455e-06,
+      "loss": 0.4603,
+      "step": 13090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8074776025231243,
+      "learning_rate": 7.971693072755102e-06,
+      "loss": 0.3717,
+      "step": 13091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1453660859404615,
+      "learning_rate": 7.971376195488339e-06,
+      "loss": 0.1496,
+      "step": 13092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7001946551712228,
+      "learning_rate": 7.971059299770136e-06,
+      "loss": 0.309,
+      "step": 13093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.649142153129898,
+      "learning_rate": 7.970742385602462e-06,
+      "loss": 0.1456,
+      "step": 13094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.814036926233839,
+      "learning_rate": 7.970425452987282e-06,
+      "loss": 0.5378,
+      "step": 13095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.379684355679274,
+      "learning_rate": 7.970108501926566e-06,
+      "loss": 0.177,
+      "step": 13096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6792575108381607,
+      "learning_rate": 7.969791532422281e-06,
+      "loss": 0.2628,
+      "step": 13097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4924062701452192,
+      "learning_rate": 7.969474544476396e-06,
+      "loss": 0.2102,
+      "step": 13098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.201112182809718,
+      "learning_rate": 7.969157538090882e-06,
+      "loss": 0.2426,
+      "step": 13099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9433751052909904,
+      "learning_rate": 7.968840513267702e-06,
+      "loss": 0.5406,
+      "step": 13100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4782232035683582,
+      "learning_rate": 7.96852347000883e-06,
+      "loss": 0.3002,
+      "step": 13101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3260896879187665,
+      "learning_rate": 7.968206408316231e-06,
+      "loss": 0.3317,
+      "step": 13102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7367512173520028,
+      "learning_rate": 7.967889328191874e-06,
+      "loss": 0.5307,
+      "step": 13103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.70932453366949,
+      "learning_rate": 7.96757222963773e-06,
+      "loss": 0.4816,
+      "step": 13104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2517601257198088,
+      "learning_rate": 7.967255112655768e-06,
+      "loss": 0.2545,
+      "step": 13105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6069460427593998,
+      "learning_rate": 7.966937977247955e-06,
+      "loss": 0.3767,
+      "step": 13106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.356642299470033,
+      "learning_rate": 7.966620823416265e-06,
+      "loss": 0.2246,
+      "step": 13107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5149102135430206,
+      "learning_rate": 7.966303651162662e-06,
+      "loss": 0.3265,
+      "step": 13108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1828267536287154,
+      "learning_rate": 7.965986460489118e-06,
+      "loss": 0.2152,
+      "step": 13109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3824298862722058,
+      "learning_rate": 7.9656692513976e-06,
+      "loss": 0.2511,
+      "step": 13110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.25525687457668,
+      "learning_rate": 7.965352023890085e-06,
+      "loss": 0.3189,
+      "step": 13111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3378536483778238,
+      "learning_rate": 7.965034777968534e-06,
+      "loss": 0.3869,
+      "step": 13112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5452134572179417,
+      "learning_rate": 7.964717513634925e-06,
+      "loss": 0.2842,
+      "step": 13113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0625033920433586,
+      "learning_rate": 7.964400230891223e-06,
+      "loss": 0.1866,
+      "step": 13114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4142131830514075,
+      "learning_rate": 7.964082929739397e-06,
+      "loss": 0.3101,
+      "step": 13115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8537477683557413,
+      "learning_rate": 7.963765610181421e-06,
+      "loss": 0.325,
+      "step": 13116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2585688251504226,
+      "learning_rate": 7.963448272219266e-06,
+      "loss": 0.3525,
+      "step": 13117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6633703377276297,
+      "learning_rate": 7.963130915854902e-06,
+      "loss": 0.2443,
+      "step": 13118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4910084075393266,
+      "learning_rate": 7.962813541090295e-06,
+      "loss": 0.3628,
+      "step": 13119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3552382626853983,
+      "learning_rate": 7.96249614792742e-06,
+      "loss": 0.7264,
+      "step": 13120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3671687315606287,
+      "learning_rate": 7.96217873636825e-06,
+      "loss": 0.1735,
+      "step": 13121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.099967842551224,
+      "learning_rate": 7.961861306414752e-06,
+      "loss": 0.3195,
+      "step": 13122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3869598501447762,
+      "learning_rate": 7.961543858068898e-06,
+      "loss": 0.4113,
+      "step": 13123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.699839905956181,
+      "learning_rate": 7.961226391332661e-06,
+      "loss": 0.4353,
+      "step": 13124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8417790048972194,
+      "learning_rate": 7.960908906208011e-06,
+      "loss": 0.3925,
+      "step": 13125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7423965752028898,
+      "learning_rate": 7.960591402696919e-06,
+      "loss": 0.383,
+      "step": 13126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5086251870450036,
+      "learning_rate": 7.960273880801357e-06,
+      "loss": 0.3216,
+      "step": 13127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5258533567196912,
+      "learning_rate": 7.959956340523297e-06,
+      "loss": 0.2129,
+      "step": 13128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.627600620011598,
+      "learning_rate": 7.959638781864714e-06,
+      "loss": 0.3849,
+      "step": 13129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7009387727053196,
+      "learning_rate": 7.959321204827574e-06,
+      "loss": 0.3564,
+      "step": 13130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3317124762807482,
+      "learning_rate": 7.95900360941385e-06,
+      "loss": 0.2661,
+      "step": 13131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7512861277102776,
+      "learning_rate": 7.95868599562552e-06,
+      "loss": 0.4857,
+      "step": 13132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6334774816979227,
+      "learning_rate": 7.95836836346455e-06,
+      "loss": 0.3016,
+      "step": 13133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4173168951326305,
+      "learning_rate": 7.958050712932916e-06,
+      "loss": 0.2054,
+      "step": 13134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3811936088633023,
+      "learning_rate": 7.95773304403259e-06,
+      "loss": 0.2821,
+      "step": 13135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5099457192927734,
+      "learning_rate": 7.957415356765543e-06,
+      "loss": 0.3346,
+      "step": 13136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8487216406195262,
+      "learning_rate": 7.957097651133748e-06,
+      "loss": 0.4166,
+      "step": 13137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.56787737592961,
+      "learning_rate": 7.95677992713918e-06,
+      "loss": 0.3475,
+      "step": 13138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8266309232598132,
+      "learning_rate": 7.956462184783809e-06,
+      "loss": 0.477,
+      "step": 13139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.721540252829513,
+      "learning_rate": 7.95614442406961e-06,
+      "loss": 0.2287,
+      "step": 13140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7628460172439908,
+      "learning_rate": 7.955826644998556e-06,
+      "loss": 0.4105,
+      "step": 13141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6184038987090261,
+      "learning_rate": 7.955508847572621e-06,
+      "loss": 0.2319,
+      "step": 13142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5319958957520343,
+      "learning_rate": 7.955191031793776e-06,
+      "loss": 0.2657,
+      "step": 13143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5550544679476532,
+      "learning_rate": 7.954873197663998e-06,
+      "loss": 0.4815,
+      "step": 13144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1563571025675516,
+      "learning_rate": 7.954555345185259e-06,
+      "loss": 0.2138,
+      "step": 13145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4131806901379065,
+      "learning_rate": 7.954237474359532e-06,
+      "loss": 0.3156,
+      "step": 13146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4012537468376054,
+      "learning_rate": 7.95391958518879e-06,
+      "loss": 0.2358,
+      "step": 13147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2754553134874866,
+      "learning_rate": 7.953601677675009e-06,
+      "loss": 0.5843,
+      "step": 13148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5083265728373583,
+      "learning_rate": 7.953283751820165e-06,
+      "loss": 0.2731,
+      "step": 13149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5138674898803077,
+      "learning_rate": 7.952965807626228e-06,
+      "loss": 0.3266,
+      "step": 13150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3279455821077308,
+      "learning_rate": 7.952647845095174e-06,
+      "loss": 0.2433,
+      "step": 13151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.183962799304109,
+      "learning_rate": 7.952329864228978e-06,
+      "loss": 0.3046,
+      "step": 13152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2903359509732402,
+      "learning_rate": 7.952011865029614e-06,
+      "loss": 0.2739,
+      "step": 13153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.550992617387713,
+      "learning_rate": 7.951693847499057e-06,
+      "loss": 0.3008,
+      "step": 13154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.425219490480834,
+      "learning_rate": 7.95137581163928e-06,
+      "loss": 0.2676,
+      "step": 13155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6021532225360076,
+      "learning_rate": 7.951057757452264e-06,
+      "loss": 0.5348,
+      "step": 13156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8961916893120285,
+      "learning_rate": 7.950739684939976e-06,
+      "loss": 0.4159,
+      "step": 13157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.418280543576895,
+      "learning_rate": 7.950421594104398e-06,
+      "loss": 0.3302,
+      "step": 13158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7695577674209344,
+      "learning_rate": 7.9501034849475e-06,
+      "loss": 0.3695,
+      "step": 13159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.652702454935142,
+      "learning_rate": 7.949785357471258e-06,
+      "loss": 0.4604,
+      "step": 13160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2633887995223763,
+      "learning_rate": 7.94946721167765e-06,
+      "loss": 0.5198,
+      "step": 13161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7840623362867367,
+      "learning_rate": 7.949149047568651e-06,
+      "loss": 0.2676,
+      "step": 13162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2898789103548813,
+      "learning_rate": 7.948830865146238e-06,
+      "loss": 0.2112,
+      "step": 13163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5356621430422663,
+      "learning_rate": 7.948512664412383e-06,
+      "loss": 0.243,
+      "step": 13164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6924995650638825,
+      "learning_rate": 7.948194445369065e-06,
+      "loss": 0.4799,
+      "step": 13165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4394468594625647,
+      "learning_rate": 7.947876208018258e-06,
+      "loss": 0.2178,
+      "step": 13166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7609824674123196,
+      "learning_rate": 7.947557952361939e-06,
+      "loss": 0.4003,
+      "step": 13167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7721927156137864,
+      "learning_rate": 7.947239678402085e-06,
+      "loss": 0.3467,
+      "step": 13168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9845697690341484,
+      "learning_rate": 7.946921386140673e-06,
+      "loss": 0.329,
+      "step": 13169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.068559725156205,
+      "learning_rate": 7.946603075579677e-06,
+      "loss": 0.2675,
+      "step": 13170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8210041426677497,
+      "learning_rate": 7.946284746721077e-06,
+      "loss": 0.2416,
+      "step": 13171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.154454350420184,
+      "learning_rate": 7.945966399566844e-06,
+      "loss": 0.1772,
+      "step": 13172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.049070273054889,
+      "learning_rate": 7.945648034118962e-06,
+      "loss": 0.68,
+      "step": 13173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.570841088221032,
+      "learning_rate": 7.945329650379403e-06,
+      "loss": 0.2453,
+      "step": 13174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0544673018440176,
+      "learning_rate": 7.945011248350148e-06,
+      "loss": 0.3641,
+      "step": 13175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1675821305528826,
+      "learning_rate": 7.94469282803317e-06,
+      "loss": 0.2558,
+      "step": 13176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2596737934467597,
+      "learning_rate": 7.944374389430448e-06,
+      "loss": 0.3536,
+      "step": 13177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.827134736344608,
+      "learning_rate": 7.94405593254396e-06,
+      "loss": 0.3987,
+      "step": 13178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.063192005820131,
+      "learning_rate": 7.943737457375682e-06,
+      "loss": 0.2807,
+      "step": 13179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5423323127524247,
+      "learning_rate": 7.943418963927594e-06,
+      "loss": 0.3477,
+      "step": 13180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6588629722860546,
+      "learning_rate": 7.943100452201673e-06,
+      "loss": 0.3592,
+      "step": 13181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1860202365908965,
+      "learning_rate": 7.942781922199895e-06,
+      "loss": 0.3442,
+      "step": 13182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3324841990527077,
+      "learning_rate": 7.94246337392424e-06,
+      "loss": 0.2127,
+      "step": 13183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.581428646184978,
+      "learning_rate": 7.942144807376686e-06,
+      "loss": 0.2229,
+      "step": 13184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5075903615234025,
+      "learning_rate": 7.94182622255921e-06,
+      "loss": 0.2134,
+      "step": 13185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.130517045730176,
+      "learning_rate": 7.941507619473791e-06,
+      "loss": 0.3407,
+      "step": 13186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5491017151686846,
+      "learning_rate": 7.941188998122409e-06,
+      "loss": 0.1555,
+      "step": 13187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3610685038006975,
+      "learning_rate": 7.940870358507037e-06,
+      "loss": 0.3349,
+      "step": 13188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9689929448955994,
+      "learning_rate": 7.940551700629661e-06,
+      "loss": 0.325,
+      "step": 13189
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6085344475944543,
+      "learning_rate": 7.940233024492257e-06,
+      "loss": 0.3628,
+      "step": 13190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9297715913928914,
+      "learning_rate": 7.939914330096802e-06,
+      "loss": 0.3701,
+      "step": 13191
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6332000883850173,
+      "learning_rate": 7.939595617445276e-06,
+      "loss": 0.4641,
+      "step": 13192
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3096661628617965,
+      "learning_rate": 7.93927688653966e-06,
+      "loss": 0.2705,
+      "step": 13193
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9395267390007627,
+      "learning_rate": 7.938958137381929e-06,
+      "loss": 0.2137,
+      "step": 13194
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8562756580769402,
+      "learning_rate": 7.938639369974066e-06,
+      "loss": 0.1859,
+      "step": 13195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4849600266485545,
+      "learning_rate": 7.938320584318052e-06,
+      "loss": 0.389,
+      "step": 13196
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8102848023795262,
+      "learning_rate": 7.938001780415861e-06,
+      "loss": 0.3524,
+      "step": 13197
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3356189264285239,
+      "learning_rate": 7.937682958269477e-06,
+      "loss": 0.1696,
+      "step": 13198
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0216531397451365,
+      "learning_rate": 7.937364117880878e-06,
+      "loss": 0.3826,
+      "step": 13199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.763739767278082,
+      "learning_rate": 7.937045259252045e-06,
+      "loss": 0.3706,
+      "step": 13200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5016447601451948,
+      "learning_rate": 7.936726382384958e-06,
+      "loss": 0.2142,
+      "step": 13201
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.60580019766436,
+      "learning_rate": 7.936407487281597e-06,
+      "loss": 0.2247,
+      "step": 13202
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6189105764301222,
+      "learning_rate": 7.93608857394394e-06,
+      "loss": 0.3938,
+      "step": 13203
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5624930929984973,
+      "learning_rate": 7.93576964237397e-06,
+      "loss": 0.2388,
+      "step": 13204
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0209311829470145,
+      "learning_rate": 7.935450692573667e-06,
+      "loss": 0.4228,
+      "step": 13205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8502976890065248,
+      "learning_rate": 7.935131724545011e-06,
+      "loss": 0.3816,
+      "step": 13206
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.601036739114274,
+      "learning_rate": 7.934812738289984e-06,
+      "loss": 0.2112,
+      "step": 13207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.579455593718738,
+      "learning_rate": 7.934493733810564e-06,
+      "loss": 0.3607,
+      "step": 13208
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6417100310019213,
+      "learning_rate": 7.934174711108735e-06,
+      "loss": 0.2051,
+      "step": 13209
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3872400714252067,
+      "learning_rate": 7.933855670186477e-06,
+      "loss": 0.2504,
+      "step": 13210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3514242322018453,
+      "learning_rate": 7.933536611045771e-06,
+      "loss": 0.2332,
+      "step": 13211
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.062695403668409,
+      "learning_rate": 7.933217533688598e-06,
+      "loss": 0.3495,
+      "step": 13212
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7517290216555979,
+      "learning_rate": 7.93289843811694e-06,
+      "loss": 0.3622,
+      "step": 13213
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6481171608580396,
+      "learning_rate": 7.932579324332776e-06,
+      "loss": 0.4023,
+      "step": 13214
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.830659893962738,
+      "learning_rate": 7.932260192338093e-06,
+      "loss": 0.5046,
+      "step": 13215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.508459164993551,
+      "learning_rate": 7.931941042134868e-06,
+      "loss": 0.32,
+      "step": 13216
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.271338136528417,
+      "learning_rate": 7.931621873725084e-06,
+      "loss": 0.2054,
+      "step": 13217
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.753328325853099,
+      "learning_rate": 7.931302687110723e-06,
+      "loss": 0.2576,
+      "step": 13218
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0738802220378871,
+      "learning_rate": 7.930983482293766e-06,
+      "loss": 0.1946,
+      "step": 13219
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1583155814168644,
+      "learning_rate": 7.930664259276199e-06,
+      "loss": 0.6777,
+      "step": 13220
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.51189117286188,
+      "learning_rate": 7.930345018060002e-06,
+      "loss": 0.3543,
+      "step": 13221
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8733840495701863,
+      "learning_rate": 7.930025758647155e-06,
+      "loss": 0.3247,
+      "step": 13222
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0307892790398867,
+      "learning_rate": 7.929706481039644e-06,
+      "loss": 0.3839,
+      "step": 13223
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1556404468472903,
+      "learning_rate": 7.92938718523945e-06,
+      "loss": 0.1695,
+      "step": 13224
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.780986297897415,
+      "learning_rate": 7.929067871248557e-06,
+      "loss": 0.4825,
+      "step": 13225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8002815091022133,
+      "learning_rate": 7.928748539068946e-06,
+      "loss": 0.227,
+      "step": 13226
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2132099119916737,
+      "learning_rate": 7.9284291887026e-06,
+      "loss": 0.3784,
+      "step": 13227
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.294109239091149,
+      "learning_rate": 7.928109820151505e-06,
+      "loss": 0.2488,
+      "step": 13228
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7002635506088706,
+      "learning_rate": 7.92779043341764e-06,
+      "loss": 0.3861,
+      "step": 13229
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4010734928219892,
+      "learning_rate": 7.927471028502993e-06,
+      "loss": 0.1787,
+      "step": 13230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7406737247133042,
+      "learning_rate": 7.927151605409542e-06,
+      "loss": 0.2748,
+      "step": 13231
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7648584803438905,
+      "learning_rate": 7.926832164139275e-06,
+      "loss": 0.4171,
+      "step": 13232
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2785322510512096,
+      "learning_rate": 7.926512704694174e-06,
+      "loss": 0.3396,
+      "step": 13233
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.277868115628908,
+      "learning_rate": 7.926193227076223e-06,
+      "loss": 0.4172,
+      "step": 13234
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9061648553062125,
+      "learning_rate": 7.925873731287405e-06,
+      "loss": 0.3153,
+      "step": 13235
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7142417748326784,
+      "learning_rate": 7.925554217329704e-06,
+      "loss": 0.2232,
+      "step": 13236
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4435820814219127,
+      "learning_rate": 7.925234685205108e-06,
+      "loss": 0.2404,
+      "step": 13237
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6563101253752321,
+      "learning_rate": 7.924915134915595e-06,
+      "loss": 0.4756,
+      "step": 13238
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5799014649343517,
+      "learning_rate": 7.924595566463153e-06,
+      "loss": 0.2155,
+      "step": 13239
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1256966387675493,
+      "learning_rate": 7.924275979849765e-06,
+      "loss": 0.186,
+      "step": 13240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6014379906521292,
+      "learning_rate": 7.923956375077418e-06,
+      "loss": 0.2234,
+      "step": 13241
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7364491140566711,
+      "learning_rate": 7.923636752148091e-06,
+      "loss": 0.3758,
+      "step": 13242
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5334242509421534,
+      "learning_rate": 7.923317111063777e-06,
+      "loss": 0.2655,
+      "step": 13243
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3631459007733462,
+      "learning_rate": 7.922997451826455e-06,
+      "loss": 0.2637,
+      "step": 13244
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8527206790360746,
+      "learning_rate": 7.92267777443811e-06,
+      "loss": 0.4567,
+      "step": 13245
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4077550886939834,
+      "learning_rate": 7.922358078900732e-06,
+      "loss": 0.2008,
+      "step": 13246
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7463877556271998,
+      "learning_rate": 7.9220383652163e-06,
+      "loss": 0.2358,
+      "step": 13247
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4293428526160405,
+      "learning_rate": 7.921718633386804e-06,
+      "loss": 0.4715,
+      "step": 13248
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5704911211064,
+      "learning_rate": 7.921398883414227e-06,
+      "loss": 0.3342,
+      "step": 13249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9195573369490424,
+      "learning_rate": 7.921079115300556e-06,
+      "loss": 0.5321,
+      "step": 13250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8303239752924436,
+      "learning_rate": 7.920759329047775e-06,
+      "loss": 0.6645,
+      "step": 13251
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1906045025681764,
+      "learning_rate": 7.920439524657872e-06,
+      "loss": 0.4403,
+      "step": 13252
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.408641984889137,
+      "learning_rate": 7.92011970213283e-06,
+      "loss": 0.3875,
+      "step": 13253
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.331867258263316,
+      "learning_rate": 7.919799861474638e-06,
+      "loss": 0.3901,
+      "step": 13254
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7429398635597628,
+      "learning_rate": 7.919480002685279e-06,
+      "loss": 0.4468,
+      "step": 13255
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7349366575957967,
+      "learning_rate": 7.919160125766741e-06,
+      "loss": 0.2872,
+      "step": 13256
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2483590404255325,
+      "learning_rate": 7.918840230721013e-06,
+      "loss": 0.1798,
+      "step": 13257
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4083966983228438,
+      "learning_rate": 7.918520317550075e-06,
+      "loss": 0.2875,
+      "step": 13258
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1097185986285567,
+      "learning_rate": 7.91820038625592e-06,
+      "loss": 0.1357,
+      "step": 13259
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5125864567262919,
+      "learning_rate": 7.91788043684053e-06,
+      "loss": 0.3292,
+      "step": 13260
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7874286345594113,
+      "learning_rate": 7.917560469305896e-06,
+      "loss": 0.4009,
+      "step": 13261
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.480449374431955,
+      "learning_rate": 7.917240483654002e-06,
+      "loss": 0.1884,
+      "step": 13262
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3476983906888302,
+      "learning_rate": 7.916920479886837e-06,
+      "loss": 0.2361,
+      "step": 13263
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2160555631093972,
+      "learning_rate": 7.916600458006384e-06,
+      "loss": 0.195,
+      "step": 13264
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7136328447330569,
+      "learning_rate": 7.916280418014637e-06,
+      "loss": 0.3569,
+      "step": 13265
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.66123944791897,
+      "learning_rate": 7.915960359913577e-06,
+      "loss": 0.3087,
+      "step": 13266
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0103528394762678,
+      "learning_rate": 7.915640283705192e-06,
+      "loss": 0.5125,
+      "step": 13267
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.038565739596887,
+      "learning_rate": 7.915320189391476e-06,
+      "loss": 0.4424,
+      "step": 13268
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4344677145088085,
+      "learning_rate": 7.91500007697441e-06,
+      "loss": 0.3374,
+      "step": 13269
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0669294652740777,
+      "learning_rate": 7.914679946455984e-06,
+      "loss": 0.1753,
+      "step": 13270
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2248315594389265,
+      "learning_rate": 7.914359797838185e-06,
+      "loss": 0.2503,
+      "step": 13271
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9627279147753303,
+      "learning_rate": 7.914039631123005e-06,
+      "loss": 0.2452,
+      "step": 13272
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1833832832136064,
+      "learning_rate": 7.913719446312426e-06,
+      "loss": 0.7191,
+      "step": 13273
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5979395570901784,
+      "learning_rate": 7.913399243408441e-06,
+      "loss": 0.36,
+      "step": 13274
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5117215001280853,
+      "learning_rate": 7.913079022413037e-06,
+      "loss": 0.4048,
+      "step": 13275
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4524590289005364,
+      "learning_rate": 7.912758783328202e-06,
+      "loss": 0.2749,
+      "step": 13276
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.01380798219888,
+      "learning_rate": 7.912438526155926e-06,
+      "loss": 0.1649,
+      "step": 13277
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4031292195246063,
+      "learning_rate": 7.912118250898195e-06,
+      "loss": 0.3064,
+      "step": 13278
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1856754977919934,
+      "learning_rate": 7.911797957557e-06,
+      "loss": 0.2351,
+      "step": 13279
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0435672707486097,
+      "learning_rate": 7.911477646134331e-06,
+      "loss": 0.2652,
+      "step": 13280
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.134507752642355,
+      "learning_rate": 7.911157316632175e-06,
+      "loss": 0.3544,
+      "step": 13281
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8960002316555755,
+      "learning_rate": 7.91083696905252e-06,
+      "loss": 0.4054,
+      "step": 13282
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7102106370638153,
+      "learning_rate": 7.910516603397358e-06,
+      "loss": 0.246,
+      "step": 13283
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7108634514657532,
+      "learning_rate": 7.910196219668676e-06,
+      "loss": 0.4521,
+      "step": 13284
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8267517311617358,
+      "learning_rate": 7.909875817868465e-06,
+      "loss": 0.3329,
+      "step": 13285
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.401437366868284,
+      "learning_rate": 7.909555397998713e-06,
+      "loss": 0.3402,
+      "step": 13286
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.733584395628795,
+      "learning_rate": 7.909234960061415e-06,
+      "loss": 0.4557,
+      "step": 13287
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1357016355272918,
+      "learning_rate": 7.908914504058554e-06,
+      "loss": 0.1355,
+      "step": 13288
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1290393752725376,
+      "learning_rate": 7.908594029992125e-06,
+      "loss": 0.3324,
+      "step": 13289
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.723571778424439,
+      "learning_rate": 7.908273537864114e-06,
+      "loss": 0.499,
+      "step": 13290
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6025309170196895,
+      "learning_rate": 7.907953027676515e-06,
+      "loss": 0.3595,
+      "step": 13291
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7429320707270464,
+      "learning_rate": 7.907632499431313e-06,
+      "loss": 0.2395,
+      "step": 13292
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1798532729483815,
+      "learning_rate": 7.907311953130505e-06,
+      "loss": 0.1616,
+      "step": 13293
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3965594158151355,
+      "learning_rate": 7.906991388776076e-06,
+      "loss": 0.3273,
+      "step": 13294
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4012997999836112,
+      "learning_rate": 7.90667080637002e-06,
+      "loss": 0.238,
+      "step": 13295
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.367345273587321,
+      "learning_rate": 7.906350205914325e-06,
+      "loss": 0.2613,
+      "step": 13296
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6176536044999403,
+      "learning_rate": 7.906029587410986e-06,
+      "loss": 0.3795,
+      "step": 13297
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7855206645576558,
+      "learning_rate": 7.90570895086199e-06,
+      "loss": 0.3431,
+      "step": 13298
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.505503772057027,
+      "learning_rate": 7.905388296269332e-06,
+      "loss": 0.2122,
+      "step": 13299
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7607123974878691,
+      "learning_rate": 7.905067623634997e-06,
+      "loss": 0.4342,
+      "step": 13300
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4791983460562081,
+      "learning_rate": 7.904746932960982e-06,
+      "loss": 0.1582,
+      "step": 13301
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.305375485995373,
+      "learning_rate": 7.904426224249275e-06,
+      "loss": 0.163,
+      "step": 13302
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9674670546480488,
+      "learning_rate": 7.90410549750187e-06,
+      "loss": 0.3323,
+      "step": 13303
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9721020335475736,
+      "learning_rate": 7.903784752720757e-06,
+      "loss": 0.2015,
+      "step": 13304
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.504174193751328,
+      "learning_rate": 7.903463989907929e-06,
+      "loss": 0.1932,
+      "step": 13305
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6663367843993133,
+      "learning_rate": 7.903143209065377e-06,
+      "loss": 0.4815,
+      "step": 13306
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5159174536195634,
+      "learning_rate": 7.90282241019509e-06,
+      "loss": 0.1348,
+      "step": 13307
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6847938411409773,
+      "learning_rate": 7.902501593299066e-06,
+      "loss": 0.2942,
+      "step": 13308
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.685942984546441,
+      "learning_rate": 7.902180758379295e-06,
+      "loss": 0.3307,
+      "step": 13309
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1404286633653677,
+      "learning_rate": 7.901859905437767e-06,
+      "loss": 0.2255,
+      "step": 13310
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8106508564641812,
+      "learning_rate": 7.901539034476476e-06,
+      "loss": 0.3067,
+      "step": 13311
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1382957371592752,
+      "learning_rate": 7.901218145497415e-06,
+      "loss": 0.3909,
+      "step": 13312
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3971257917517632,
+      "learning_rate": 7.900897238502576e-06,
+      "loss": 0.1719,
+      "step": 13313
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7842647107351011,
+      "learning_rate": 7.900576313493953e-06,
+      "loss": 0.0819,
+      "step": 13314
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.203311091343998,
+      "learning_rate": 7.900255370473536e-06,
+      "loss": 0.5379,
+      "step": 13315
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2023844731246502,
+      "learning_rate": 7.89993440944332e-06,
+      "loss": 0.2524,
+      "step": 13316
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8212625946494385,
+      "learning_rate": 7.899613430405298e-06,
+      "loss": 0.5576,
+      "step": 13317
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3069401610335514,
+      "learning_rate": 7.899292433361464e-06,
+      "loss": 0.491,
+      "step": 13318
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6500985648958737,
+      "learning_rate": 7.898971418313808e-06,
+      "loss": 0.5156,
+      "step": 13319
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7237090593715458,
+      "learning_rate": 7.898650385264328e-06,
+      "loss": 0.3037,
+      "step": 13320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8209090667384917,
+      "learning_rate": 7.898329334215014e-06,
+      "loss": 0.5227,
+      "step": 13321
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.606676196403167,
+      "learning_rate": 7.898008265167862e-06,
+      "loss": 0.3359,
+      "step": 13322
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7037501675122206,
+      "learning_rate": 7.897687178124864e-06,
+      "loss": 0.3637,
+      "step": 13323
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.554816417404975,
+      "learning_rate": 7.897366073088014e-06,
+      "loss": 0.1979,
+      "step": 13324
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.292834810885565,
+      "learning_rate": 7.897044950059307e-06,
+      "loss": 0.1738,
+      "step": 13325
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9534508948232254,
+      "learning_rate": 7.896723809040736e-06,
+      "loss": 0.3906,
+      "step": 13326
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.638770213035213,
+      "learning_rate": 7.896402650034297e-06,
+      "loss": 0.3836,
+      "step": 13327
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5114459967180962,
+      "learning_rate": 7.896081473041982e-06,
+      "loss": 0.2744,
+      "step": 13328
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8316883587774844,
+      "learning_rate": 7.895760278065787e-06,
+      "loss": 0.3953,
+      "step": 13329
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3642241269692608,
+      "learning_rate": 7.895439065107707e-06,
+      "loss": 0.2664,
+      "step": 13330
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7713090542779646,
+      "learning_rate": 7.895117834169734e-06,
+      "loss": 0.3556,
+      "step": 13331
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8345378024464527,
+      "learning_rate": 7.894796585253866e-06,
+      "loss": 0.3315,
+      "step": 13332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.481671959092858,
+      "learning_rate": 7.894475318362094e-06,
+      "loss": 0.4557,
+      "step": 13333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.165563037071675,
+      "learning_rate": 7.894154033496417e-06,
+      "loss": 0.2008,
+      "step": 13334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3551466776315022,
+      "learning_rate": 7.893832730658828e-06,
+      "loss": 0.2591,
+      "step": 13335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9531653133046616,
+      "learning_rate": 7.893511409851322e-06,
+      "loss": 0.1843,
+      "step": 13336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9049533045169533,
+      "learning_rate": 7.893190071075895e-06,
+      "loss": 0.6543,
+      "step": 13337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8652805204120098,
+      "learning_rate": 7.892868714334543e-06,
+      "loss": 0.1501,
+      "step": 13338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.790242880181542,
+      "learning_rate": 7.892547339629261e-06,
+      "loss": 0.318,
+      "step": 13339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.114952379678085,
+      "learning_rate": 7.892225946962044e-06,
+      "loss": 0.5451,
+      "step": 13340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5539350126279021,
+      "learning_rate": 7.891904536334888e-06,
+      "loss": 0.387,
+      "step": 13341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5130172569136338,
+      "learning_rate": 7.89158310774979e-06,
+      "loss": 0.3814,
+      "step": 13342
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2416654481694491,
+      "learning_rate": 7.891261661208745e-06,
+      "loss": 0.2102,
+      "step": 13343
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.604199613247286,
+      "learning_rate": 7.890940196713748e-06,
+      "loss": 0.328,
+      "step": 13344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9671506440434667,
+      "learning_rate": 7.890618714266799e-06,
+      "loss": 0.5735,
+      "step": 13345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.304410805171941,
+      "learning_rate": 7.890297213869888e-06,
+      "loss": 0.2827,
+      "step": 13346
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6724748560308362,
+      "learning_rate": 7.889975695525018e-06,
+      "loss": 0.5763,
+      "step": 13347
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3873421302113151,
+      "learning_rate": 7.889654159234182e-06,
+      "loss": 0.1535,
+      "step": 13348
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8504967096805052,
+      "learning_rate": 7.889332604999376e-06,
+      "loss": 0.3351,
+      "step": 13349
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6779831711809836,
+      "learning_rate": 7.8890110328226e-06,
+      "loss": 0.3065,
+      "step": 13350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6758139991672454,
+      "learning_rate": 7.888689442705849e-06,
+      "loss": 0.622,
+      "step": 13351
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.46511843839715,
+      "learning_rate": 7.888367834651118e-06,
+      "loss": 0.2712,
+      "step": 13352
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4471330672084795,
+      "learning_rate": 7.88804620866041e-06,
+      "loss": 0.2271,
+      "step": 13353
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4318331477540083,
+      "learning_rate": 7.887724564735716e-06,
+      "loss": 0.7545,
+      "step": 13354
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0897470683613384,
+      "learning_rate": 7.887402902879033e-06,
+      "loss": 0.4339,
+      "step": 13355
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5809838583657263,
+      "learning_rate": 7.887081223092365e-06,
+      "loss": 0.4764,
+      "step": 13356
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5221546714115264,
+      "learning_rate": 7.886759525377703e-06,
+      "loss": 0.2733,
+      "step": 13357
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9388273637573896,
+      "learning_rate": 7.886437809737049e-06,
+      "loss": 0.6693,
+      "step": 13358
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4780383580644614,
+      "learning_rate": 7.886116076172398e-06,
+      "loss": 0.2744,
+      "step": 13359
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.001447409046453,
+      "learning_rate": 7.885794324685749e-06,
+      "loss": 0.3641,
+      "step": 13360
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.25128173976984,
+      "learning_rate": 7.8854725552791e-06,
+      "loss": 0.2441,
+      "step": 13361
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6063640817436355,
+      "learning_rate": 7.88515076795445e-06,
+      "loss": 0.3973,
+      "step": 13362
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3245368648561695,
+      "learning_rate": 7.884828962713795e-06,
+      "loss": 0.248,
+      "step": 13363
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3014774120677406,
+      "learning_rate": 7.884507139559135e-06,
+      "loss": 0.2471,
+      "step": 13364
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5665572371928072,
+      "learning_rate": 7.884185298492468e-06,
+      "loss": 0.2279,
+      "step": 13365
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.406763938415685,
+      "learning_rate": 7.883863439515792e-06,
+      "loss": 0.4136,
+      "step": 13366
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3833042391145727,
+      "learning_rate": 7.883541562631107e-06,
+      "loss": 0.169,
+      "step": 13367
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.085007499514387,
+      "learning_rate": 7.883219667840411e-06,
+      "loss": 0.3837,
+      "step": 13368
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3276296393039817,
+      "learning_rate": 7.882897755145702e-06,
+      "loss": 0.2141,
+      "step": 13369
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0206553763428716,
+      "learning_rate": 7.882575824548979e-06,
+      "loss": 0.4414,
+      "step": 13370
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6677706576247233,
+      "learning_rate": 7.882253876052244e-06,
+      "loss": 0.3889,
+      "step": 13371
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6036612134441457,
+      "learning_rate": 7.881931909657493e-06,
+      "loss": 0.3052,
+      "step": 13372
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5076351981848395,
+      "learning_rate": 7.881609925366726e-06,
+      "loss": 0.2375,
+      "step": 13373
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8088039271950058,
+      "learning_rate": 7.881287923181944e-06,
+      "loss": 0.3536,
+      "step": 13374
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5404763093889218,
+      "learning_rate": 7.880965903105145e-06,
+      "loss": 0.2214,
+      "step": 13375
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6623604491827768,
+      "learning_rate": 7.88064386513833e-06,
+      "loss": 0.3606,
+      "step": 13376
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6865047327891853,
+      "learning_rate": 7.880321809283495e-06,
+      "loss": 0.3997,
+      "step": 13377
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3877533700823554,
+      "learning_rate": 7.879999735542645e-06,
+      "loss": 0.5285,
+      "step": 13378
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3434548941083158,
+      "learning_rate": 7.879677643917778e-06,
+      "loss": 0.2942,
+      "step": 13379
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1229647241482825,
+      "learning_rate": 7.879355534410892e-06,
+      "loss": 0.1903,
+      "step": 13380
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6425834682890248,
+      "learning_rate": 7.87903340702399e-06,
+      "loss": 0.2809,
+      "step": 13381
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5438359901672256,
+      "learning_rate": 7.878711261759073e-06,
+      "loss": 0.3066,
+      "step": 13382
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7333788803501702,
+      "learning_rate": 7.878389098618137e-06,
+      "loss": 0.3365,
+      "step": 13383
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3516442792329426,
+      "learning_rate": 7.878066917603188e-06,
+      "loss": 0.3437,
+      "step": 13384
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3027357161484983,
+      "learning_rate": 7.877744718716222e-06,
+      "loss": 0.1994,
+      "step": 13385
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5457460204200781,
+      "learning_rate": 7.877422501959241e-06,
+      "loss": 0.1454,
+      "step": 13386
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.009418587337952,
+      "learning_rate": 7.87710026733425e-06,
+      "loss": 0.5172,
+      "step": 13387
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5566296212107766,
+      "learning_rate": 7.876778014843245e-06,
+      "loss": 0.2387,
+      "step": 13388
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.451590872284949,
+      "learning_rate": 7.876455744488228e-06,
+      "loss": 0.4234,
+      "step": 13389
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.609866041899036,
+      "learning_rate": 7.876133456271201e-06,
+      "loss": 0.3632,
+      "step": 13390
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5564254387050929,
+      "learning_rate": 7.875811150194166e-06,
+      "loss": 0.2883,
+      "step": 13391
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2705275983391322,
+      "learning_rate": 7.875488826259124e-06,
+      "loss": 0.2271,
+      "step": 13392
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.304704313701786,
+      "learning_rate": 7.875166484468076e-06,
+      "loss": 0.2286,
+      "step": 13393
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.050454271680079,
+      "learning_rate": 7.874844124823022e-06,
+      "loss": 0.6427,
+      "step": 13394
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.909027452570221,
+      "learning_rate": 7.874521747325968e-06,
+      "loss": 0.3728,
+      "step": 13395
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9086606873943275,
+      "learning_rate": 7.874199351978912e-06,
+      "loss": 0.2107,
+      "step": 13396
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7877497812794299,
+      "learning_rate": 7.873876938783858e-06,
+      "loss": 0.3797,
+      "step": 13397
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.554995120587467,
+      "learning_rate": 7.873554507742809e-06,
+      "loss": 0.3747,
+      "step": 13398
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.301171884218947,
+      "learning_rate": 7.873232058857765e-06,
+      "loss": 0.1975,
+      "step": 13399
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9772760950238649,
+      "learning_rate": 7.872909592130728e-06,
+      "loss": 0.1223,
+      "step": 13400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1891966966505607,
+      "learning_rate": 7.872587107563704e-06,
+      "loss": 0.3963,
+      "step": 13401
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.067765857844975,
+      "learning_rate": 7.87226460515869e-06,
+      "loss": 0.488,
+      "step": 13402
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4850925141915308,
+      "learning_rate": 7.871942084917694e-06,
+      "loss": 0.1714,
+      "step": 13403
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8665614255617538,
+      "learning_rate": 7.871619546842715e-06,
+      "loss": 0.5202,
+      "step": 13404
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7041647432086353,
+      "learning_rate": 7.87129699093576e-06,
+      "loss": 0.1695,
+      "step": 13405
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4182482725022296,
+      "learning_rate": 7.870974417198826e-06,
+      "loss": 0.3402,
+      "step": 13406
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0309461109549947,
+      "learning_rate": 7.870651825633922e-06,
+      "loss": 0.212,
+      "step": 13407
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9501321445219535,
+      "learning_rate": 7.870329216243048e-06,
+      "loss": 0.518,
+      "step": 13408
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1652345384132836,
+      "learning_rate": 7.870006589028208e-06,
+      "loss": 0.2319,
+      "step": 13409
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5777857529127277,
+      "learning_rate": 7.869683943991405e-06,
+      "loss": 0.4039,
+      "step": 13410
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3050752737361815,
+      "learning_rate": 7.869361281134644e-06,
+      "loss": 0.3105,
+      "step": 13411
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5067657372766161,
+      "learning_rate": 7.869038600459926e-06,
+      "loss": 0.3198,
+      "step": 13412
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.648422991431149,
+      "learning_rate": 7.868715901969259e-06,
+      "loss": 0.195,
+      "step": 13413
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7594307063178447,
+      "learning_rate": 7.868393185664643e-06,
+      "loss": 0.1489,
+      "step": 13414
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.372746986968201,
+      "learning_rate": 7.868070451548082e-06,
+      "loss": 0.2196,
+      "step": 13415
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9314307050305444,
+      "learning_rate": 7.867747699621582e-06,
+      "loss": 0.1657,
+      "step": 13416
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6696570727445847,
+      "learning_rate": 7.867424929887148e-06,
+      "loss": 0.2845,
+      "step": 13417
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.285700042024387,
+      "learning_rate": 7.86710214234678e-06,
+      "loss": 0.3445,
+      "step": 13418
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8012061166083497,
+      "learning_rate": 7.866779337002488e-06,
+      "loss": 0.4894,
+      "step": 13419
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7301251175734802,
+      "learning_rate": 7.866456513856272e-06,
+      "loss": 0.5001,
+      "step": 13420
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0037940925308795,
+      "learning_rate": 7.86613367291014e-06,
+      "loss": 0.3693,
+      "step": 13421
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7361892839968475,
+      "learning_rate": 7.865810814166094e-06,
+      "loss": 0.1448,
+      "step": 13422
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6112094627383053,
+      "learning_rate": 7.865487937626142e-06,
+      "loss": 0.2376,
+      "step": 13423
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4318816019919227,
+      "learning_rate": 7.865165043292284e-06,
+      "loss": 0.2508,
+      "step": 13424
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9211612399464124,
+      "learning_rate": 7.86484213116653e-06,
+      "loss": 0.3371,
+      "step": 13425
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8925333480340834,
+      "learning_rate": 7.864519201250884e-06,
+      "loss": 0.3621,
+      "step": 13426
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4612601097408222,
+      "learning_rate": 7.86419625354735e-06,
+      "loss": 0.2269,
+      "step": 13427
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4274672442288208,
+      "learning_rate": 7.863873288057934e-06,
+      "loss": 0.1761,
+      "step": 13428
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4272700607660522,
+      "learning_rate": 7.863550304784642e-06,
+      "loss": 0.2492,
+      "step": 13429
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9201720880069426,
+      "learning_rate": 7.863227303729477e-06,
+      "loss": 0.325,
+      "step": 13430
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2005783529803806,
+      "learning_rate": 7.86290428489445e-06,
+      "loss": 0.1483,
+      "step": 13431
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9737177155290142,
+      "learning_rate": 7.862581248281562e-06,
+      "loss": 0.5192,
+      "step": 13432
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7006463554132198,
+      "learning_rate": 7.86225819389282e-06,
+      "loss": 0.4722,
+      "step": 13433
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2695455168876082,
+      "learning_rate": 7.861935121730233e-06,
+      "loss": 0.2885,
+      "step": 13434
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5299460542275205,
+      "learning_rate": 7.861612031795805e-06,
+      "loss": 0.2226,
+      "step": 13435
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6120268352980902,
+      "learning_rate": 7.86128892409154e-06,
+      "loss": 0.4022,
+      "step": 13436
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0573323038570384,
+      "learning_rate": 7.860965798619448e-06,
+      "loss": 0.193,
+      "step": 13437
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5496995411895647,
+      "learning_rate": 7.860642655381535e-06,
+      "loss": 0.2753,
+      "step": 13438
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8526889094919474,
+      "learning_rate": 7.860319494379806e-06,
+      "loss": 0.3844,
+      "step": 13439
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5400095799538274,
+      "learning_rate": 7.85999631561627e-06,
+      "loss": 0.2615,
+      "step": 13440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8079463549806007,
+      "learning_rate": 7.859673119092932e-06,
+      "loss": 0.4166,
+      "step": 13441
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5803005170452693,
+      "learning_rate": 7.859349904811799e-06,
+      "loss": 0.287,
+      "step": 13442
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6988015243612469,
+      "learning_rate": 7.859026672774878e-06,
+      "loss": 0.48,
+      "step": 13443
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5277216241431726,
+      "learning_rate": 7.85870342298418e-06,
+      "loss": 0.23,
+      "step": 13444
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9623378430786458,
+      "learning_rate": 7.858380155441706e-06,
+      "loss": 0.3781,
+      "step": 13445
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3014705395353867,
+      "learning_rate": 7.858056870149466e-06,
+      "loss": 0.1873,
+      "step": 13446
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8880599252446544,
+      "learning_rate": 7.85773356710947e-06,
+      "loss": 0.3346,
+      "step": 13447
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4509914768678176,
+      "learning_rate": 7.857410246323723e-06,
+      "loss": 0.1888,
+      "step": 13448
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.928351847381977,
+      "learning_rate": 7.857086907794232e-06,
+      "loss": 0.3903,
+      "step": 13449
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1491749052860754,
+      "learning_rate": 7.856763551523008e-06,
+      "loss": 0.3051,
+      "step": 13450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3852572973866684,
+      "learning_rate": 7.856440177512057e-06,
+      "loss": 0.1783,
+      "step": 13451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6500187957209702,
+      "learning_rate": 7.856116785763386e-06,
+      "loss": 0.1946,
+      "step": 13452
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8599109358126744,
+      "learning_rate": 7.855793376279006e-06,
+      "loss": 0.5058,
+      "step": 13453
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.132712847439157,
+      "learning_rate": 7.855469949060922e-06,
+      "loss": 0.3207,
+      "step": 13454
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9125756775462606,
+      "learning_rate": 7.855146504111145e-06,
+      "loss": 0.3955,
+      "step": 13455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2813192905719903,
+      "learning_rate": 7.85482304143168e-06,
+      "loss": 0.2826,
+      "step": 13456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.805789379931055,
+      "learning_rate": 7.85449956102454e-06,
+      "loss": 0.4126,
+      "step": 13457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.14106681886411,
+      "learning_rate": 7.854176062891733e-06,
+      "loss": 0.2636,
+      "step": 13458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1542128255154904,
+      "learning_rate": 7.853852547035265e-06,
+      "loss": 0.3843,
+      "step": 13459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1169938589884991,
+      "learning_rate": 7.853529013457149e-06,
+      "loss": 0.1719,
+      "step": 13460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2955266602014324,
+      "learning_rate": 7.85320546215939e-06,
+      "loss": 0.5377,
+      "step": 13461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9371257543592526,
+      "learning_rate": 7.852881893144e-06,
+      "loss": 0.4338,
+      "step": 13462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6488255545734614,
+      "learning_rate": 7.852558306412985e-06,
+      "loss": 0.1207,
+      "step": 13463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3327360833410495,
+      "learning_rate": 7.852234701968359e-06,
+      "loss": 0.3003,
+      "step": 13464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0952402827063208,
+      "learning_rate": 7.851911079812127e-06,
+      "loss": 0.1817,
+      "step": 13465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4703976236607224,
+      "learning_rate": 7.851587439946303e-06,
+      "loss": 0.2242,
+      "step": 13466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.324182532623163,
+      "learning_rate": 7.85126378237289e-06,
+      "loss": 0.231,
+      "step": 13467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8898893276709499,
+      "learning_rate": 7.850940107093907e-06,
+      "loss": 0.226,
+      "step": 13468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5248794138635229,
+      "learning_rate": 7.850616414111358e-06,
+      "loss": 0.1699,
+      "step": 13469
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7085395713696863,
+      "learning_rate": 7.850292703427252e-06,
+      "loss": 0.319,
+      "step": 13470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1153781803239964,
+      "learning_rate": 7.849968975043604e-06,
+      "loss": 0.4257,
+      "step": 13471
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4780888262671368,
+      "learning_rate": 7.84964522896242e-06,
+      "loss": 0.1408,
+      "step": 13472
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1979883448952466,
+      "learning_rate": 7.849321465185713e-06,
+      "loss": 0.338,
+      "step": 13473
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1871158126243038,
+      "learning_rate": 7.848997683715492e-06,
+      "loss": 0.5373,
+      "step": 13474
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6911743943407165,
+      "learning_rate": 7.848673884553769e-06,
+      "loss": 0.3376,
+      "step": 13475
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5858040222598666,
+      "learning_rate": 7.84835006770255e-06,
+      "loss": 0.3724,
+      "step": 13476
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8004816679096751,
+      "learning_rate": 7.848026233163853e-06,
+      "loss": 0.4697,
+      "step": 13477
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3757905299617281,
+      "learning_rate": 7.847702380939685e-06,
+      "loss": 0.2989,
+      "step": 13478
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7759347633637423,
+      "learning_rate": 7.847378511032057e-06,
+      "loss": 0.5226,
+      "step": 13479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4359481256241753,
+      "learning_rate": 7.84705462344298e-06,
+      "loss": 0.4973,
+      "step": 13480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5753049912791164,
+      "learning_rate": 7.84673071817447e-06,
+      "loss": 0.2761,
+      "step": 13481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8976022782548174,
+      "learning_rate": 7.846406795228529e-06,
+      "loss": 0.3603,
+      "step": 13482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4901052371526466,
+      "learning_rate": 7.846082854607175e-06,
+      "loss": 0.1748,
+      "step": 13483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7629454881654334,
+      "learning_rate": 7.845758896312422e-06,
+      "loss": 0.3462,
+      "step": 13484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3025719800213973,
+      "learning_rate": 7.845434920346274e-06,
+      "loss": 0.2783,
+      "step": 13485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.748262615849489,
+      "learning_rate": 7.845110926710748e-06,
+      "loss": 0.4532,
+      "step": 13486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.468402598263191,
+      "learning_rate": 7.844786915407856e-06,
+      "loss": 0.3682,
+      "step": 13487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6149524012464276,
+      "learning_rate": 7.844462886439607e-06,
+      "loss": 0.2292,
+      "step": 13488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8016593761484792,
+      "learning_rate": 7.844138839808016e-06,
+      "loss": 0.3813,
+      "step": 13489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4999417651316214,
+      "learning_rate": 7.843814775515094e-06,
+      "loss": 0.4036,
+      "step": 13490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.75703493934607,
+      "learning_rate": 7.843490693562854e-06,
+      "loss": 0.243,
+      "step": 13491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4395186876456687,
+      "learning_rate": 7.843166593953306e-06,
+      "loss": 0.2284,
+      "step": 13492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.499689218989518,
+      "learning_rate": 7.842842476688466e-06,
+      "loss": 0.2063,
+      "step": 13493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9140211529061137,
+      "learning_rate": 7.842518341770346e-06,
+      "loss": 0.1989,
+      "step": 13494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.880244779541758,
+      "learning_rate": 7.842194189200956e-06,
+      "loss": 0.3281,
+      "step": 13495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5940118041705287,
+      "learning_rate": 7.841870018982311e-06,
+      "loss": 0.3936,
+      "step": 13496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2936326782846668,
+      "learning_rate": 7.841545831116427e-06,
+      "loss": 0.1801,
+      "step": 13497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5241595684471607,
+      "learning_rate": 7.84122162560531e-06,
+      "loss": 0.4024,
+      "step": 13498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4816049541718077,
+      "learning_rate": 7.840897402450981e-06,
+      "loss": 0.2521,
+      "step": 13499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7324654273716187,
+      "learning_rate": 7.840573161655447e-06,
+      "loss": 0.3608,
+      "step": 13500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9293411731751267,
+      "learning_rate": 7.840248903220727e-06,
+      "loss": 0.328,
+      "step": 13501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1797190633149754,
+      "learning_rate": 7.839924627148827e-06,
+      "loss": 0.5147,
+      "step": 13502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8113402117173114,
+      "learning_rate": 7.839600333441769e-06,
+      "loss": 0.3342,
+      "step": 13503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.659338516347935,
+      "learning_rate": 7.83927602210156e-06,
+      "loss": 0.3326,
+      "step": 13504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9417695716744805,
+      "learning_rate": 7.838951693130222e-06,
+      "loss": 0.6271,
+      "step": 13505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0022601362073753,
+      "learning_rate": 7.83862734652976e-06,
+      "loss": 0.2138,
+      "step": 13506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.59511582201809,
+      "learning_rate": 7.838302982302193e-06,
+      "loss": 0.1863,
+      "step": 13507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5509032122219006,
+      "learning_rate": 7.837978600449534e-06,
+      "loss": 0.1632,
+      "step": 13508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3139161179210883,
+      "learning_rate": 7.837654200973797e-06,
+      "loss": 0.1944,
+      "step": 13509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.705733121013183,
+      "learning_rate": 7.837329783876996e-06,
+      "loss": 0.2254,
+      "step": 13510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5223093611263376,
+      "learning_rate": 7.837005349161148e-06,
+      "loss": 0.2376,
+      "step": 13511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2998508413648489,
+      "learning_rate": 7.836680896828269e-06,
+      "loss": 0.2234,
+      "step": 13512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2334530402123662,
+      "learning_rate": 7.836356426880366e-06,
+      "loss": 0.3657,
+      "step": 13513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5904203897807063,
+      "learning_rate": 7.836031939319461e-06,
+      "loss": 0.4579,
+      "step": 13514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.268632701069351,
+      "learning_rate": 7.835707434147568e-06,
+      "loss": 0.4006,
+      "step": 13515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.01316926383452,
+      "learning_rate": 7.8353829113667e-06,
+      "loss": 0.2027,
+      "step": 13516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7701794445542933,
+      "learning_rate": 7.835058370978875e-06,
+      "loss": 0.3211,
+      "step": 13517
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7298602547735547,
+      "learning_rate": 7.834733812986102e-06,
+      "loss": 0.3991,
+      "step": 13518
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8935051213624072,
+      "learning_rate": 7.834409237390403e-06,
+      "loss": 0.3047,
+      "step": 13519
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.248461077833584,
+      "learning_rate": 7.834084644193793e-06,
+      "loss": 0.7244,
+      "step": 13520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.621493765873413,
+      "learning_rate": 7.833760033398284e-06,
+      "loss": 0.4034,
+      "step": 13521
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4335973503857673,
+      "learning_rate": 7.833435405005894e-06,
+      "loss": 0.342,
+      "step": 13522
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3274816301120087,
+      "learning_rate": 7.833110759018638e-06,
+      "loss": 0.2967,
+      "step": 13523
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7577957873079746,
+      "learning_rate": 7.832786095438533e-06,
+      "loss": 0.3452,
+      "step": 13524
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8657396960185444,
+      "learning_rate": 7.832461414267595e-06,
+      "loss": 0.3284,
+      "step": 13525
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.651134075222161,
+      "learning_rate": 7.83213671550784e-06,
+      "loss": 0.344,
+      "step": 13526
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9848652707098153,
+      "learning_rate": 7.831811999161285e-06,
+      "loss": 0.3797,
+      "step": 13527
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3939497133185268,
+      "learning_rate": 7.831487265229945e-06,
+      "loss": 0.2659,
+      "step": 13528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5758449919805777,
+      "learning_rate": 7.831162513715835e-06,
+      "loss": 0.3564,
+      "step": 13529
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8441532995123375,
+      "learning_rate": 7.830837744620976e-06,
+      "loss": 0.5259,
+      "step": 13530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4683783674825808,
+      "learning_rate": 7.830512957947382e-06,
+      "loss": 0.2298,
+      "step": 13531
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7224944637732822,
+      "learning_rate": 7.83018815369707e-06,
+      "loss": 0.4968,
+      "step": 13532
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7049879069892278,
+      "learning_rate": 7.829863331872057e-06,
+      "loss": 0.4219,
+      "step": 13533
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6789611487911917,
+      "learning_rate": 7.829538492474358e-06,
+      "loss": 0.5216,
+      "step": 13534
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4150845590758012,
+      "learning_rate": 7.829213635505995e-06,
+      "loss": 0.153,
+      "step": 13535
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.434515045067424,
+      "learning_rate": 7.828888760968982e-06,
+      "loss": 0.2269,
+      "step": 13536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.152102891353673,
+      "learning_rate": 7.828563868865337e-06,
+      "loss": 0.4903,
+      "step": 13537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7618044178977037,
+      "learning_rate": 7.828238959197077e-06,
+      "loss": 0.3377,
+      "step": 13538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2420919654504756,
+      "learning_rate": 7.827914031966222e-06,
+      "loss": 0.1768,
+      "step": 13539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4344111406847067,
+      "learning_rate": 7.827589087174785e-06,
+      "loss": 0.3211,
+      "step": 13540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5644301818803594,
+      "learning_rate": 7.82726412482479e-06,
+      "loss": 0.2688,
+      "step": 13541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9653162170153002,
+      "learning_rate": 7.826939144918248e-06,
+      "loss": 0.3936,
+      "step": 13542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0525117614002861,
+      "learning_rate": 7.826614147457183e-06,
+      "loss": 0.185,
+      "step": 13543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3451902106880376,
+      "learning_rate": 7.826289132443608e-06,
+      "loss": 0.2304,
+      "step": 13544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2342477910973593,
+      "learning_rate": 7.825964099879546e-06,
+      "loss": 0.2165,
+      "step": 13545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7875440975565196,
+      "learning_rate": 7.825639049767013e-06,
+      "loss": 0.2751,
+      "step": 13546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0631080836434896,
+      "learning_rate": 7.82531398210803e-06,
+      "loss": 0.6185,
+      "step": 13547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3470725135956698,
+      "learning_rate": 7.82498889690461e-06,
+      "loss": 0.1963,
+      "step": 13548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6367998763124219,
+      "learning_rate": 7.824663794158778e-06,
+      "loss": 0.3004,
+      "step": 13549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.797068730567026,
+      "learning_rate": 7.824338673872548e-06,
+      "loss": 0.3349,
+      "step": 13550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9572844122656972,
+      "learning_rate": 7.82401353604794e-06,
+      "loss": 0.3589,
+      "step": 13551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1541340162015152,
+      "learning_rate": 7.823688380686978e-06,
+      "loss": 0.486,
+      "step": 13552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9993333599575625,
+      "learning_rate": 7.823363207791672e-06,
+      "loss": 0.4024,
+      "step": 13553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.759911881443569,
+      "learning_rate": 7.823038017364048e-06,
+      "loss": 0.4625,
+      "step": 13554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1921118733840943,
+      "learning_rate": 7.822712809406126e-06,
+      "loss": 0.1803,
+      "step": 13555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1898551932453376,
+      "learning_rate": 7.82238758391992e-06,
+      "loss": 0.6935,
+      "step": 13556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8345600236319697,
+      "learning_rate": 7.822062340907453e-06,
+      "loss": 0.4027,
+      "step": 13557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.962061567443669,
+      "learning_rate": 7.821737080370746e-06,
+      "loss": 0.3376,
+      "step": 13558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7053284630632117,
+      "learning_rate": 7.821411802311814e-06,
+      "loss": 0.4144,
+      "step": 13559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.930592583359154,
+      "learning_rate": 7.821086506732682e-06,
+      "loss": 0.6408,
+      "step": 13560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.531170428164581,
+      "learning_rate": 7.820761193635369e-06,
+      "loss": 0.1908,
+      "step": 13561
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.889253686486121,
+      "learning_rate": 7.820435863021891e-06,
+      "loss": 0.3079,
+      "step": 13562
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0258964488165705,
+      "learning_rate": 7.820110514894273e-06,
+      "loss": 0.6089,
+      "step": 13563
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6564196373778421,
+      "learning_rate": 7.819785149254534e-06,
+      "loss": 0.3269,
+      "step": 13564
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3393641316240146,
+      "learning_rate": 7.819459766104693e-06,
+      "loss": 0.584,
+      "step": 13565
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6737424033970476,
+      "learning_rate": 7.81913436544677e-06,
+      "loss": 0.3052,
+      "step": 13566
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4797686101441172,
+      "learning_rate": 7.81880894728279e-06,
+      "loss": 0.3602,
+      "step": 13567
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.522343867182617,
+      "learning_rate": 7.81848351161477e-06,
+      "loss": 0.2006,
+      "step": 13568
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.625491638351715,
+      "learning_rate": 7.818158058444732e-06,
+      "loss": 0.4138,
+      "step": 13569
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5038149099974745,
+      "learning_rate": 7.817832587774696e-06,
+      "loss": 0.2982,
+      "step": 13570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.524994702310428,
+      "learning_rate": 7.817507099606685e-06,
+      "loss": 0.3959,
+      "step": 13571
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.609612389528225,
+      "learning_rate": 7.817181593942718e-06,
+      "loss": 0.3616,
+      "step": 13572
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8587771704829366,
+      "learning_rate": 7.816856070784818e-06,
+      "loss": 0.4944,
+      "step": 13573
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3727184421061245,
+      "learning_rate": 7.816530530135005e-06,
+      "loss": 0.3442,
+      "step": 13574
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4463169988990252,
+      "learning_rate": 7.8162049719953e-06,
+      "loss": 0.2027,
+      "step": 13575
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.702162899584019,
+      "learning_rate": 7.81587939636773e-06,
+      "loss": 0.4345,
+      "step": 13576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.08056505481551,
+      "learning_rate": 7.815553803254309e-06,
+      "loss": 0.1142,
+      "step": 13577
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9273412703678678,
+      "learning_rate": 7.815228192657064e-06,
+      "loss": 0.3728,
+      "step": 13578
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.540808195940989,
+      "learning_rate": 7.814902564578016e-06,
+      "loss": 0.2357,
+      "step": 13579
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5684486164247071,
+      "learning_rate": 7.814576919019185e-06,
+      "loss": 0.2496,
+      "step": 13580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6641466853736715,
+      "learning_rate": 7.814251255982594e-06,
+      "loss": 0.3406,
+      "step": 13581
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4263699818426678,
+      "learning_rate": 7.813925575470268e-06,
+      "loss": 0.3215,
+      "step": 13582
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.549474256705404,
+      "learning_rate": 7.813599877484227e-06,
+      "loss": 0.4522,
+      "step": 13583
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7175881435379052,
+      "learning_rate": 7.813274162026494e-06,
+      "loss": 0.224,
+      "step": 13584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3127454743849436,
+      "learning_rate": 7.81294842909909e-06,
+      "loss": 0.2372,
+      "step": 13585
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.342569867048486,
+      "learning_rate": 7.81262267870404e-06,
+      "loss": 0.3184,
+      "step": 13586
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7205261633088456,
+      "learning_rate": 7.812296910843367e-06,
+      "loss": 0.495,
+      "step": 13587
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3405880931851477,
+      "learning_rate": 7.811971125519092e-06,
+      "loss": 0.2065,
+      "step": 13588
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.850788864242459,
+      "learning_rate": 7.811645322733237e-06,
+      "loss": 0.5112,
+      "step": 13589
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6094023646648519,
+      "learning_rate": 7.811319502487828e-06,
+      "loss": 0.3812,
+      "step": 13590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.702973949663488,
+      "learning_rate": 7.810993664784889e-06,
+      "loss": 0.4702,
+      "step": 13591
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.667456714410662,
+      "learning_rate": 7.810667809626442e-06,
+      "loss": 0.1643,
+      "step": 13592
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.79849713124727,
+      "learning_rate": 7.81034193701451e-06,
+      "loss": 0.3097,
+      "step": 13593
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7782611784721072,
+      "learning_rate": 7.810016046951116e-06,
+      "loss": 0.2927,
+      "step": 13594
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.672302919955243,
+      "learning_rate": 7.809690139438283e-06,
+      "loss": 0.3003,
+      "step": 13595
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8987137628805386,
+      "learning_rate": 7.809364214478038e-06,
+      "loss": 0.4541,
+      "step": 13596
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0764131773627077,
+      "learning_rate": 7.809038272072402e-06,
+      "loss": 0.5308,
+      "step": 13597
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5562213155868212,
+      "learning_rate": 7.808712312223402e-06,
+      "loss": 0.3796,
+      "step": 13598
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5277009726735065,
+      "learning_rate": 7.808386334933058e-06,
+      "loss": 0.3798,
+      "step": 13599
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4973661226282817,
+      "learning_rate": 7.808060340203398e-06,
+      "loss": 0.32,
+      "step": 13600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2015951810729995,
+      "learning_rate": 7.807734328036444e-06,
+      "loss": 0.6993,
+      "step": 13601
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7918920504580744,
+      "learning_rate": 7.807408298434223e-06,
+      "loss": 0.3454,
+      "step": 13602
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4476288405333606,
+      "learning_rate": 7.807082251398756e-06,
+      "loss": 0.1749,
+      "step": 13603
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7875530463463836,
+      "learning_rate": 7.806756186932071e-06,
+      "loss": 0.462,
+      "step": 13604
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7423488559328228,
+      "learning_rate": 7.806430105036191e-06,
+      "loss": 0.2432,
+      "step": 13605
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6978146528732043,
+      "learning_rate": 7.80610400571314e-06,
+      "loss": 0.2049,
+      "step": 13606
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5114715137428423,
+      "learning_rate": 7.805777888964945e-06,
+      "loss": 0.3466,
+      "step": 13607
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3495217376277453,
+      "learning_rate": 7.805451754793629e-06,
+      "loss": 0.1677,
+      "step": 13608
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.760330985785027,
+      "learning_rate": 7.805125603201221e-06,
+      "loss": 0.3714,
+      "step": 13609
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5576708793354939,
+      "learning_rate": 7.804799434189742e-06,
+      "loss": 0.2883,
+      "step": 13610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.55533507772395,
+      "learning_rate": 7.804473247761219e-06,
+      "loss": 0.2773,
+      "step": 13611
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6615994822228617,
+      "learning_rate": 7.804147043917678e-06,
+      "loss": 0.2435,
+      "step": 13612
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2281612612648063,
+      "learning_rate": 7.803820822661146e-06,
+      "loss": 0.2352,
+      "step": 13613
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9871339504371281,
+      "learning_rate": 7.803494583993644e-06,
+      "loss": 0.2378,
+      "step": 13614
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6294208738251283,
+      "learning_rate": 7.803168327917204e-06,
+      "loss": 0.3519,
+      "step": 13615
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7671882958245952,
+      "learning_rate": 7.802842054433848e-06,
+      "loss": 0.1962,
+      "step": 13616
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.289593780736659,
+      "learning_rate": 7.802515763545604e-06,
+      "loss": 0.3219,
+      "step": 13617
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.451566348306858,
+      "learning_rate": 7.802189455254494e-06,
+      "loss": 0.429,
+      "step": 13618
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4172614503616783,
+      "learning_rate": 7.80186312956255e-06,
+      "loss": 0.1798,
+      "step": 13619
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1159889554477902,
+      "learning_rate": 7.801536786471796e-06,
+      "loss": 0.1615,
+      "step": 13620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6169670710321884,
+      "learning_rate": 7.801210425984256e-06,
+      "loss": 0.4488,
+      "step": 13621
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9515216521515313,
+      "learning_rate": 7.80088404810196e-06,
+      "loss": 0.3494,
+      "step": 13622
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.823313038209611,
+      "learning_rate": 7.800557652826934e-06,
+      "loss": 0.207,
+      "step": 13623
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.361045528504149,
+      "learning_rate": 7.800231240161205e-06,
+      "loss": 0.3521,
+      "step": 13624
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4982029416482994,
+      "learning_rate": 7.7999048101068e-06,
+      "loss": 0.3585,
+      "step": 13625
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2517998195342868,
+      "learning_rate": 7.799578362665746e-06,
+      "loss": 0.2575,
+      "step": 13626
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5206253040664628,
+      "learning_rate": 7.799251897840068e-06,
+      "loss": 0.1988,
+      "step": 13627
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8105576548951071,
+      "learning_rate": 7.798925415631794e-06,
+      "loss": 0.1502,
+      "step": 13628
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6862722407472697,
+      "learning_rate": 7.798598916042954e-06,
+      "loss": 0.3073,
+      "step": 13629
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.668873418751661,
+      "learning_rate": 7.798272399075573e-06,
+      "loss": 0.3422,
+      "step": 13630
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.909945579722183,
+      "learning_rate": 7.79794586473168e-06,
+      "loss": 0.4773,
+      "step": 13631
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8755523503847644,
+      "learning_rate": 7.7976193130133e-06,
+      "loss": 0.3968,
+      "step": 13632
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3151082691909202,
+      "learning_rate": 7.797292743922465e-06,
+      "loss": 0.1743,
+      "step": 13633
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.091975392365102,
+      "learning_rate": 7.7969661574612e-06,
+      "loss": 0.5598,
+      "step": 13634
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5038681867662507,
+      "learning_rate": 7.796639553631533e-06,
+      "loss": 0.3735,
+      "step": 13635
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5985528521618946,
+      "learning_rate": 7.796312932435494e-06,
+      "loss": 0.1813,
+      "step": 13636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3460777865706999,
+      "learning_rate": 7.79598629387511e-06,
+      "loss": 0.2255,
+      "step": 13637
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7885252070812832,
+      "learning_rate": 7.795659637952409e-06,
+      "loss": 0.365,
+      "step": 13638
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4384149485106288,
+      "learning_rate": 7.79533296466942e-06,
+      "loss": 0.2196,
+      "step": 13639
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0368059321840772,
+      "learning_rate": 7.795006274028171e-06,
+      "loss": 0.3088,
+      "step": 13640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1387716400706003,
+      "learning_rate": 7.794679566030691e-06,
+      "loss": 0.2241,
+      "step": 13641
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.030640231821236,
+      "learning_rate": 7.794352840679009e-06,
+      "loss": 0.308,
+      "step": 13642
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4929426988437784,
+      "learning_rate": 7.794026097975154e-06,
+      "loss": 0.2318,
+      "step": 13643
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.52723188551599,
+      "learning_rate": 7.793699337921155e-06,
+      "loss": 0.2646,
+      "step": 13644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5059395309865176,
+      "learning_rate": 7.79337256051904e-06,
+      "loss": 0.2166,
+      "step": 13645
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5203815424925062,
+      "learning_rate": 7.79304576577084e-06,
+      "loss": 0.2675,
+      "step": 13646
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.295183447682461,
+      "learning_rate": 7.792718953678583e-06,
+      "loss": 0.3805,
+      "step": 13647
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6657585491297104,
+      "learning_rate": 7.792392124244299e-06,
+      "loss": 0.3813,
+      "step": 13648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.391226217948331,
+      "learning_rate": 7.792065277470015e-06,
+      "loss": 0.1645,
+      "step": 13649
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4322007099805911,
+      "learning_rate": 7.791738413357766e-06,
+      "loss": 0.2745,
+      "step": 13650
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5982259000628665,
+      "learning_rate": 7.791411531909575e-06,
+      "loss": 0.3599,
+      "step": 13651
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2493731118856326,
+      "learning_rate": 7.791084633127479e-06,
+      "loss": 0.1465,
+      "step": 13652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.485793073797482,
+      "learning_rate": 7.790757717013502e-06,
+      "loss": 0.2113,
+      "step": 13653
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.43955910042461,
+      "learning_rate": 7.790430783569677e-06,
+      "loss": 0.5739,
+      "step": 13654
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.413427998255509,
+      "learning_rate": 7.790103832798035e-06,
+      "loss": 0.181,
+      "step": 13655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.128381433455596,
+      "learning_rate": 7.789776864700603e-06,
+      "loss": 0.3933,
+      "step": 13656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6981190290756747,
+      "learning_rate": 7.789449879279416e-06,
+      "loss": 0.4849,
+      "step": 13657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.671584393744257,
+      "learning_rate": 7.789122876536499e-06,
+      "loss": 0.3969,
+      "step": 13658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.244392680890837,
+      "learning_rate": 7.788795856473887e-06,
+      "loss": 0.2952,
+      "step": 13659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5335401176833832,
+      "learning_rate": 7.788468819093609e-06,
+      "loss": 0.3314,
+      "step": 13660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7771655381114981,
+      "learning_rate": 7.788141764397695e-06,
+      "loss": 0.2587,
+      "step": 13661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.429873998912071,
+      "learning_rate": 7.787814692388175e-06,
+      "loss": 0.2588,
+      "step": 13662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.384899504179559,
+      "learning_rate": 7.787487603067085e-06,
+      "loss": 0.3433,
+      "step": 13663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4388824616645837,
+      "learning_rate": 7.787160496436452e-06,
+      "loss": 0.1801,
+      "step": 13664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5127256647737983,
+      "learning_rate": 7.786833372498309e-06,
+      "loss": 0.3548,
+      "step": 13665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4787068035535023,
+      "learning_rate": 7.786506231254684e-06,
+      "loss": 0.1591,
+      "step": 13666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6374085828737681,
+      "learning_rate": 7.786179072707613e-06,
+      "loss": 0.2539,
+      "step": 13667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1652589375745617,
+      "learning_rate": 7.785851896859125e-06,
+      "loss": 0.5356,
+      "step": 13668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2220206265269007,
+      "learning_rate": 7.785524703711253e-06,
+      "loss": 0.5308,
+      "step": 13669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2970621249769527,
+      "learning_rate": 7.785197493266028e-06,
+      "loss": 0.1366,
+      "step": 13670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.80826388454974,
+      "learning_rate": 7.784870265525481e-06,
+      "loss": 0.3733,
+      "step": 13671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6824720899086474,
+      "learning_rate": 7.784543020491645e-06,
+      "loss": 0.5385,
+      "step": 13672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6616814103838697,
+      "learning_rate": 7.784215758166554e-06,
+      "loss": 0.4711,
+      "step": 13673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5212548833296924,
+      "learning_rate": 7.783888478552235e-06,
+      "loss": 0.4528,
+      "step": 13674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.668197507827828,
+      "learning_rate": 7.783561181650728e-06,
+      "loss": 0.2444,
+      "step": 13675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6054274050175819,
+      "learning_rate": 7.783233867464057e-06,
+      "loss": 0.4035,
+      "step": 13676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.682354309305403,
+      "learning_rate": 7.78290653599426e-06,
+      "loss": 0.2809,
+      "step": 13677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1794536481948241,
+      "learning_rate": 7.782579187243367e-06,
+      "loss": 0.2446,
+      "step": 13678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6342050851530887,
+      "learning_rate": 7.782251821213414e-06,
+      "loss": 0.2546,
+      "step": 13679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0354343304864915,
+      "learning_rate": 7.78192443790643e-06,
+      "loss": 0.31,
+      "step": 13680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5574006531022744,
+      "learning_rate": 7.781597037324451e-06,
+      "loss": 0.1721,
+      "step": 13681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.464469088642643,
+      "learning_rate": 7.781269619469509e-06,
+      "loss": 0.329,
+      "step": 13682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5233474655933992,
+      "learning_rate": 7.780942184343635e-06,
+      "loss": 0.2106,
+      "step": 13683
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6415488905799127,
+      "learning_rate": 7.780614731948866e-06,
+      "loss": 0.1395,
+      "step": 13684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5705719823575697,
+      "learning_rate": 7.780287262287234e-06,
+      "loss": 0.3172,
+      "step": 13685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.66271170114376,
+      "learning_rate": 7.77995977536077e-06,
+      "loss": 0.3284,
+      "step": 13686
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9004461392079481,
+      "learning_rate": 7.77963227117151e-06,
+      "loss": 0.404,
+      "step": 13687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1978306020696976,
+      "learning_rate": 7.779304749721489e-06,
+      "loss": 0.3345,
+      "step": 13688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8950436032684956,
+      "learning_rate": 7.778977211012738e-06,
+      "loss": 0.1494,
+      "step": 13689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9777340413740685,
+      "learning_rate": 7.778649655047293e-06,
+      "loss": 0.2317,
+      "step": 13690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7573634951718433,
+      "learning_rate": 7.778322081827187e-06,
+      "loss": 0.3219,
+      "step": 13691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4847841025318995,
+      "learning_rate": 7.777994491354454e-06,
+      "loss": 0.2387,
+      "step": 13692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1707639150600757,
+      "learning_rate": 7.77766688363113e-06,
+      "loss": 0.3605,
+      "step": 13693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1831315841028116,
+      "learning_rate": 7.777339258659247e-06,
+      "loss": 0.2969,
+      "step": 13694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6833269700041813,
+      "learning_rate": 7.777011616440842e-06,
+      "loss": 0.3856,
+      "step": 13695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4525921495796996,
+      "learning_rate": 7.776683956977946e-06,
+      "loss": 0.3089,
+      "step": 13696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4370036538118447,
+      "learning_rate": 7.776356280272596e-06,
+      "loss": 0.3026,
+      "step": 13697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8656804215523541,
+      "learning_rate": 7.776028586326826e-06,
+      "loss": 0.3715,
+      "step": 13698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.570932763993466,
+      "learning_rate": 7.775700875142672e-06,
+      "loss": 0.3357,
+      "step": 13699
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3704187084521846,
+      "learning_rate": 7.775373146722167e-06,
+      "loss": 0.2457,
+      "step": 13700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7996104762274965,
+      "learning_rate": 7.77504540106735e-06,
+      "loss": 0.168,
+      "step": 13701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.050366715201291,
+      "learning_rate": 7.774717638180254e-06,
+      "loss": 0.1829,
+      "step": 13702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5233485489332494,
+      "learning_rate": 7.774389858062912e-06,
+      "loss": 0.3252,
+      "step": 13703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7102388411299316,
+      "learning_rate": 7.774062060717363e-06,
+      "loss": 0.1689,
+      "step": 13704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4443959087639524,
+      "learning_rate": 7.77373424614564e-06,
+      "loss": 0.2202,
+      "step": 13705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2990670550865053,
+      "learning_rate": 7.773406414349779e-06,
+      "loss": 0.2928,
+      "step": 13706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9744068498448577,
+      "learning_rate": 7.773078565331815e-06,
+      "loss": 0.4114,
+      "step": 13707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2772683433416525,
+      "learning_rate": 7.772750699093787e-06,
+      "loss": 0.1973,
+      "step": 13708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0253330248272547,
+      "learning_rate": 7.772422815637731e-06,
+      "loss": 0.5748,
+      "step": 13709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8258449016929574,
+      "learning_rate": 7.772094914965677e-06,
+      "loss": 0.3692,
+      "step": 13710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.362382384788336,
+      "learning_rate": 7.771766997079668e-06,
+      "loss": 0.3424,
+      "step": 13711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1083235203912922,
+      "learning_rate": 7.771439061981737e-06,
+      "loss": 0.2489,
+      "step": 13712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9241810876658496,
+      "learning_rate": 7.771111109673924e-06,
+      "loss": 0.3296,
+      "step": 13713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4384257611266655,
+      "learning_rate": 7.77078314015826e-06,
+      "loss": 0.2129,
+      "step": 13714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5194412047862855,
+      "learning_rate": 7.770455153436784e-06,
+      "loss": 0.2077,
+      "step": 13715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5971195466676276,
+      "learning_rate": 7.770127149511533e-06,
+      "loss": 0.3017,
+      "step": 13716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9370595350582827,
+      "learning_rate": 7.769799128384544e-06,
+      "loss": 0.2048,
+      "step": 13717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0627423364426183,
+      "learning_rate": 7.769471090057852e-06,
+      "loss": 0.2376,
+      "step": 13718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7068275530899801,
+      "learning_rate": 7.769143034533497e-06,
+      "loss": 0.5042,
+      "step": 13719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4663297211518789,
+      "learning_rate": 7.768814961813515e-06,
+      "loss": 0.3393,
+      "step": 13720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7722664129950367,
+      "learning_rate": 7.768486871899943e-06,
+      "loss": 0.495,
+      "step": 13721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2038604111209814,
+      "learning_rate": 7.768158764794818e-06,
+      "loss": 0.1976,
+      "step": 13722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1567778027064184,
+      "learning_rate": 7.767830640500178e-06,
+      "loss": 0.2586,
+      "step": 13723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9807243407962447,
+      "learning_rate": 7.767502499018062e-06,
+      "loss": 0.5209,
+      "step": 13724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6281269459703642,
+      "learning_rate": 7.767174340350503e-06,
+      "loss": 0.411,
+      "step": 13725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.37633552498611,
+      "learning_rate": 7.766846164499544e-06,
+      "loss": 0.3372,
+      "step": 13726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.350775533678393,
+      "learning_rate": 7.76651797146722e-06,
+      "loss": 0.3478,
+      "step": 13727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1488630648127813,
+      "learning_rate": 7.76618976125557e-06,
+      "loss": 0.2277,
+      "step": 13728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3713421827469494,
+      "learning_rate": 7.76586153386663e-06,
+      "loss": 0.2689,
+      "step": 13729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1570284792651906,
+      "learning_rate": 7.765533289302442e-06,
+      "loss": 0.8048,
+      "step": 13730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5534529948397042,
+      "learning_rate": 7.76520502756504e-06,
+      "loss": 0.2102,
+      "step": 13731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7130493125353652,
+      "learning_rate": 7.764876748656469e-06,
+      "loss": 0.3738,
+      "step": 13732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.38087023092748,
+      "learning_rate": 7.76454845257876e-06,
+      "loss": 0.3241,
+      "step": 13733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4707208786864248,
+      "learning_rate": 7.764220139333956e-06,
+      "loss": 0.321,
+      "step": 13734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8548049542484093,
+      "learning_rate": 7.763891808924094e-06,
+      "loss": 0.4584,
+      "step": 13735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6379687149861515,
+      "learning_rate": 7.763563461351213e-06,
+      "loss": 0.3216,
+      "step": 13736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5950120248759814,
+      "learning_rate": 7.763235096617355e-06,
+      "loss": 0.4308,
+      "step": 13737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.322988226348242,
+      "learning_rate": 7.762906714724554e-06,
+      "loss": 0.4411,
+      "step": 13738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.02069161760441,
+      "learning_rate": 7.762578315674852e-06,
+      "loss": 0.704,
+      "step": 13739
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0407822982076254,
+      "learning_rate": 7.762249899470288e-06,
+      "loss": 0.3677,
+      "step": 13740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8160386933920916,
+      "learning_rate": 7.761921466112902e-06,
+      "loss": 0.5674,
+      "step": 13741
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.649784319575741,
+      "learning_rate": 7.761593015604732e-06,
+      "loss": 0.1996,
+      "step": 13742
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7193742788750799,
+      "learning_rate": 7.761264547947817e-06,
+      "loss": 0.5169,
+      "step": 13743
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3376107918673006,
+      "learning_rate": 7.7609360631442e-06,
+      "loss": 0.2386,
+      "step": 13744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5977744033139445,
+      "learning_rate": 7.760607561195918e-06,
+      "loss": 0.1255,
+      "step": 13745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1402112360583412,
+      "learning_rate": 7.760279042105012e-06,
+      "loss": 0.2186,
+      "step": 13746
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0693256293306055,
+      "learning_rate": 7.759950505873523e-06,
+      "loss": 0.6026,
+      "step": 13747
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2782681471261859,
+      "learning_rate": 7.75962195250349e-06,
+      "loss": 0.1861,
+      "step": 13748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5806878308123946,
+      "learning_rate": 7.75929338199695e-06,
+      "loss": 0.3372,
+      "step": 13749
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.269503736931438,
+      "learning_rate": 7.75896479435595e-06,
+      "loss": 0.1527,
+      "step": 13750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.816170551020442,
+      "learning_rate": 7.758636189582522e-06,
+      "loss": 0.2252,
+      "step": 13751
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.180328840498226,
+      "learning_rate": 7.758307567678716e-06,
+      "loss": 0.2641,
+      "step": 13752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.028139164144844,
+      "learning_rate": 7.757978928646566e-06,
+      "loss": 0.2654,
+      "step": 13753
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7256354328649777,
+      "learning_rate": 7.757650272488116e-06,
+      "loss": 0.4359,
+      "step": 13754
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5505381328739758,
+      "learning_rate": 7.757321599205405e-06,
+      "loss": 0.2297,
+      "step": 13755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.100442227919255,
+      "learning_rate": 7.756992908800475e-06,
+      "loss": 0.2164,
+      "step": 13756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4144313052058082,
+      "learning_rate": 7.756664201275369e-06,
+      "loss": 0.1478,
+      "step": 13757
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3054213583950944,
+      "learning_rate": 7.756335476632122e-06,
+      "loss": 0.3504,
+      "step": 13758
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7749886052343482,
+      "learning_rate": 7.756006734872781e-06,
+      "loss": 0.3026,
+      "step": 13759
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5253486267296439,
+      "learning_rate": 7.755677975999385e-06,
+      "loss": 0.184,
+      "step": 13760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.073593727692543,
+      "learning_rate": 7.755349200013977e-06,
+      "loss": 0.1834,
+      "step": 13761
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.840863379624781,
+      "learning_rate": 7.755020406918596e-06,
+      "loss": 0.2539,
+      "step": 13762
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5997889258439923,
+      "learning_rate": 7.754691596715286e-06,
+      "loss": 0.2129,
+      "step": 13763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5859836092480173,
+      "learning_rate": 7.75436276940609e-06,
+      "loss": 0.4174,
+      "step": 13764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6124525135278325,
+      "learning_rate": 7.754033924993047e-06,
+      "loss": 0.4407,
+      "step": 13765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0662694355719258,
+      "learning_rate": 7.753705063478201e-06,
+      "loss": 0.7099,
+      "step": 13766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5459622498268606,
+      "learning_rate": 7.753376184863593e-06,
+      "loss": 0.1778,
+      "step": 13767
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.493639476729431,
+      "learning_rate": 7.753047289151265e-06,
+      "loss": 0.2341,
+      "step": 13768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2975541687163674,
+      "learning_rate": 7.752718376343261e-06,
+      "loss": 0.3443,
+      "step": 13769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.869313418414447,
+      "learning_rate": 7.752389446441622e-06,
+      "loss": 0.5529,
+      "step": 13770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6231149046374254,
+      "learning_rate": 7.75206049944839e-06,
+      "loss": 0.3213,
+      "step": 13771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5223839540330222,
+      "learning_rate": 7.751731535365611e-06,
+      "loss": 0.4377,
+      "step": 13772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6567020609050227,
+      "learning_rate": 7.751402554195325e-06,
+      "loss": 0.1814,
+      "step": 13773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2420342323208415,
+      "learning_rate": 7.751073555939574e-06,
+      "loss": 0.2678,
+      "step": 13774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5637346825432654,
+      "learning_rate": 7.750744540600402e-06,
+      "loss": 0.3992,
+      "step": 13775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2484389899511124,
+      "learning_rate": 7.750415508179853e-06,
+      "loss": 0.3553,
+      "step": 13776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4060776816860632,
+      "learning_rate": 7.75008645867997e-06,
+      "loss": 0.2369,
+      "step": 13777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6035178877969032,
+      "learning_rate": 7.749757392102795e-06,
+      "loss": 0.3388,
+      "step": 13778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.394917060739055,
+      "learning_rate": 7.749428308450374e-06,
+      "loss": 0.3318,
+      "step": 13779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3421278022128613,
+      "learning_rate": 7.749099207724748e-06,
+      "loss": 0.2556,
+      "step": 13780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1669185303941647,
+      "learning_rate": 7.748770089927962e-06,
+      "loss": 0.133,
+      "step": 13781
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9372620205700113,
+      "learning_rate": 7.748440955062057e-06,
+      "loss": 0.5616,
+      "step": 13782
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.115756293636221,
+      "learning_rate": 7.748111803129083e-06,
+      "loss": 0.4251,
+      "step": 13783
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3015674128742187,
+      "learning_rate": 7.747782634131075e-06,
+      "loss": 0.1907,
+      "step": 13784
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7702364242786297,
+      "learning_rate": 7.747453448070087e-06,
+      "loss": 0.2968,
+      "step": 13785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7236184354620163,
+      "learning_rate": 7.747124244948155e-06,
+      "loss": 0.4371,
+      "step": 13786
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.204534904665672,
+      "learning_rate": 7.746795024767328e-06,
+      "loss": 0.5584,
+      "step": 13787
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6574810734052308,
+      "learning_rate": 7.746465787529647e-06,
+      "loss": 0.3599,
+      "step": 13788
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.745142282776223,
+      "learning_rate": 7.746136533237161e-06,
+      "loss": 0.3199,
+      "step": 13789
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.780720422891946,
+      "learning_rate": 7.745807261891912e-06,
+      "loss": 0.377,
+      "step": 13790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8216878623120498,
+      "learning_rate": 7.745477973495943e-06,
+      "loss": 0.5962,
+      "step": 13791
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4871090781008898,
+      "learning_rate": 7.7451486680513e-06,
+      "loss": 0.36,
+      "step": 13792
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9151318890893556,
+      "learning_rate": 7.744819345560028e-06,
+      "loss": 0.4694,
+      "step": 13793
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4323928113828333,
+      "learning_rate": 7.744490006024174e-06,
+      "loss": 0.1547,
+      "step": 13794
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.211286699196441,
+      "learning_rate": 7.74416064944578e-06,
+      "loss": 0.2379,
+      "step": 13795
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4498463682884257,
+      "learning_rate": 7.743831275826894e-06,
+      "loss": 0.2828,
+      "step": 13796
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.047675536762388,
+      "learning_rate": 7.74350188516956e-06,
+      "loss": 0.7018,
+      "step": 13797
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2535967935189745,
+      "learning_rate": 7.74317247747582e-06,
+      "loss": 0.1563,
+      "step": 13798
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3812007751798718,
+      "learning_rate": 7.742843052747726e-06,
+      "loss": 0.21,
+      "step": 13799
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3531246198517086,
+      "learning_rate": 7.742513610987321e-06,
+      "loss": 0.1907,
+      "step": 13800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5901412216935231,
+      "learning_rate": 7.742184152196648e-06,
+      "loss": 0.2261,
+      "step": 13801
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.036695433658437,
+      "learning_rate": 7.741854676377757e-06,
+      "loss": 0.2022,
+      "step": 13802
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4077876373534282,
+      "learning_rate": 7.741525183532692e-06,
+      "loss": 0.3115,
+      "step": 13803
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8654739585514268,
+      "learning_rate": 7.741195673663498e-06,
+      "loss": 0.1878,
+      "step": 13804
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3379077687522507,
+      "learning_rate": 7.740866146772224e-06,
+      "loss": 0.2831,
+      "step": 13805
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6622935354454598,
+      "learning_rate": 7.740536602860911e-06,
+      "loss": 0.2186,
+      "step": 13806
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.434936712955412,
+      "learning_rate": 7.740207041931612e-06,
+      "loss": 0.1608,
+      "step": 13807
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5172695892000925,
+      "learning_rate": 7.73987746398637e-06,
+      "loss": 0.3442,
+      "step": 13808
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8111582097395833,
+      "learning_rate": 7.739547869027231e-06,
+      "loss": 0.3351,
+      "step": 13809
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.608477703427309,
+      "learning_rate": 7.739218257056244e-06,
+      "loss": 0.4422,
+      "step": 13810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5228402927872144,
+      "learning_rate": 7.738888628075454e-06,
+      "loss": 0.3215,
+      "step": 13811
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.675956436817021,
+      "learning_rate": 7.738558982086909e-06,
+      "loss": 0.3965,
+      "step": 13812
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8880640667317206,
+      "learning_rate": 7.738229319092654e-06,
+      "loss": 0.2754,
+      "step": 13813
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9944340810330146,
+      "learning_rate": 7.737899639094739e-06,
+      "loss": 0.246,
+      "step": 13814
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.701134708668636,
+      "learning_rate": 7.737569942095208e-06,
+      "loss": 0.4105,
+      "step": 13815
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.51823247207971,
+      "learning_rate": 7.737240228096112e-06,
+      "loss": 0.167,
+      "step": 13816
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.880706693764516,
+      "learning_rate": 7.736910497099497e-06,
+      "loss": 0.5242,
+      "step": 13817
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.213817957670236,
+      "learning_rate": 7.736580749107409e-06,
+      "loss": 0.6613,
+      "step": 13818
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0386051454311094,
+      "learning_rate": 7.736250984121897e-06,
+      "loss": 0.7395,
+      "step": 13819
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3908023613970044,
+      "learning_rate": 7.735921202145006e-06,
+      "loss": 0.4107,
+      "step": 13820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5523275166216135,
+      "learning_rate": 7.735591403178789e-06,
+      "loss": 0.3043,
+      "step": 13821
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3634486493810174,
+      "learning_rate": 7.735261587225291e-06,
+      "loss": 0.5034,
+      "step": 13822
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4464755353418,
+      "learning_rate": 7.73493175428656e-06,
+      "loss": 0.3348,
+      "step": 13823
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7819253551594034,
+      "learning_rate": 7.734601904364645e-06,
+      "loss": 0.2573,
+      "step": 13824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6913871411660322,
+      "learning_rate": 7.734272037461594e-06,
+      "loss": 0.2649,
+      "step": 13825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6489371044345125,
+      "learning_rate": 7.733942153579454e-06,
+      "loss": 0.4398,
+      "step": 13826
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4936942911134554,
+      "learning_rate": 7.733612252720277e-06,
+      "loss": 0.4111,
+      "step": 13827
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4123157176348153,
+      "learning_rate": 7.733282334886106e-06,
+      "loss": 0.2239,
+      "step": 13828
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6559105538758256,
+      "learning_rate": 7.732952400078996e-06,
+      "loss": 0.3683,
+      "step": 13829
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8432173383814172,
+      "learning_rate": 7.732622448300991e-06,
+      "loss": 0.4183,
+      "step": 13830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8727573513101365,
+      "learning_rate": 7.732292479554143e-06,
+      "loss": 0.4369,
+      "step": 13831
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.087563929651839,
+      "learning_rate": 7.731962493840497e-06,
+      "loss": 0.1712,
+      "step": 13832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9331229667517213,
+      "learning_rate": 7.731632491162109e-06,
+      "loss": 0.2211,
+      "step": 13833
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8257465295587574,
+      "learning_rate": 7.731302471521021e-06,
+      "loss": 0.5694,
+      "step": 13834
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5731662563799047,
+      "learning_rate": 7.730972434919286e-06,
+      "loss": 0.27,
+      "step": 13835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7856294559867787,
+      "learning_rate": 7.730642381358954e-06,
+      "loss": 0.4873,
+      "step": 13836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.115465188576025,
+      "learning_rate": 7.730312310842071e-06,
+      "loss": 0.528,
+      "step": 13837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4865192385704313,
+      "learning_rate": 7.72998222337069e-06,
+      "loss": 0.2393,
+      "step": 13838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2475810359586057,
+      "learning_rate": 7.729652118946862e-06,
+      "loss": 0.2466,
+      "step": 13839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3444167269657212,
+      "learning_rate": 7.729321997572633e-06,
+      "loss": 0.2831,
+      "step": 13840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3501622162157219,
+      "learning_rate": 7.728991859250055e-06,
+      "loss": 0.384,
+      "step": 13841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5271398561877874,
+      "learning_rate": 7.728661703981177e-06,
+      "loss": 0.201,
+      "step": 13842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3259107279226863,
+      "learning_rate": 7.72833153176805e-06,
+      "loss": 0.2315,
+      "step": 13843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6682874879776683,
+      "learning_rate": 7.728001342612724e-06,
+      "loss": 0.4955,
+      "step": 13844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.548155289703239,
+      "learning_rate": 7.72767113651725e-06,
+      "loss": 0.2866,
+      "step": 13845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6724712788035936,
+      "learning_rate": 7.727340913483677e-06,
+      "loss": 0.3839,
+      "step": 13846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5519818047677894,
+      "learning_rate": 7.727010673514058e-06,
+      "loss": 0.1854,
+      "step": 13847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8809455582470973,
+      "learning_rate": 7.726680416610442e-06,
+      "loss": 0.3885,
+      "step": 13848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0773500638549836,
+      "learning_rate": 7.726350142774878e-06,
+      "loss": 0.5409,
+      "step": 13849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.640919241177956,
+      "learning_rate": 7.726019852009422e-06,
+      "loss": 0.319,
+      "step": 13850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6611103771742168,
+      "learning_rate": 7.725689544316121e-06,
+      "loss": 0.3348,
+      "step": 13851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2213393270110233,
+      "learning_rate": 7.725359219697026e-06,
+      "loss": 0.2052,
+      "step": 13852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.191675968537215,
+      "learning_rate": 7.725028878154191e-06,
+      "loss": 0.3855,
+      "step": 13853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7247180839656584,
+      "learning_rate": 7.724698519689665e-06,
+      "loss": 0.381,
+      "step": 13854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5957405268045646,
+      "learning_rate": 7.7243681443055e-06,
+      "loss": 0.3848,
+      "step": 13855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.880383685873935,
+      "learning_rate": 7.72403775200375e-06,
+      "loss": 0.4906,
+      "step": 13856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.007651426481856,
+      "learning_rate": 7.723707342786462e-06,
+      "loss": 0.3426,
+      "step": 13857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7021533180293786,
+      "learning_rate": 7.72337691665569e-06,
+      "loss": 0.493,
+      "step": 13858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9686878096899482,
+      "learning_rate": 7.723046473613486e-06,
+      "loss": 0.1552,
+      "step": 13859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7172264108596813,
+      "learning_rate": 7.722716013661901e-06,
+      "loss": 0.3793,
+      "step": 13860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8417208545647221,
+      "learning_rate": 7.72238553680299e-06,
+      "loss": 0.4676,
+      "step": 13861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.42860074899609,
+      "learning_rate": 7.722055043038803e-06,
+      "loss": 0.7274,
+      "step": 13862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6559944945413183,
+      "learning_rate": 7.721724532371392e-06,
+      "loss": 0.4911,
+      "step": 13863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5918460580915452,
+      "learning_rate": 7.72139400480281e-06,
+      "loss": 0.2588,
+      "step": 13864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.056685763277625,
+      "learning_rate": 7.721063460335109e-06,
+      "loss": 0.3447,
+      "step": 13865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8527776406657421,
+      "learning_rate": 7.72073289897034e-06,
+      "loss": 0.2081,
+      "step": 13866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8094234829008866,
+      "learning_rate": 7.72040232071056e-06,
+      "loss": 0.4781,
+      "step": 13867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6313952313736009,
+      "learning_rate": 7.72007172555782e-06,
+      "loss": 0.3838,
+      "step": 13868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6695423810424466,
+      "learning_rate": 7.71974111351417e-06,
+      "loss": 0.3823,
+      "step": 13869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4806022684486735,
+      "learning_rate": 7.719410484581666e-06,
+      "loss": 0.4004,
+      "step": 13870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6985810102529446,
+      "learning_rate": 7.719079838762363e-06,
+      "loss": 0.1867,
+      "step": 13871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5152944223966265,
+      "learning_rate": 7.718749176058307e-06,
+      "loss": 0.2656,
+      "step": 13872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4489522904924343,
+      "learning_rate": 7.718418496471558e-06,
+      "loss": 0.3176,
+      "step": 13873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0587271352186143,
+      "learning_rate": 7.71808780000417e-06,
+      "loss": 0.1954,
+      "step": 13874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9389102325779746,
+      "learning_rate": 7.71775708665819e-06,
+      "loss": 0.5066,
+      "step": 13875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6249145269839902,
+      "learning_rate": 7.717426356435678e-06,
+      "loss": 0.3558,
+      "step": 13876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.804198668483707,
+      "learning_rate": 7.717095609338684e-06,
+      "loss": 0.1629,
+      "step": 13877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.987654746527111,
+      "learning_rate": 7.716764845369265e-06,
+      "loss": 0.3551,
+      "step": 13878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3529891989929526,
+      "learning_rate": 7.716434064529473e-06,
+      "loss": 0.3148,
+      "step": 13879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4828796509188502,
+      "learning_rate": 7.71610326682136e-06,
+      "loss": 0.3062,
+      "step": 13880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7740720420326788,
+      "learning_rate": 7.715772452246984e-06,
+      "loss": 0.4422,
+      "step": 13881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6427150583059005,
+      "learning_rate": 7.715441620808397e-06,
+      "loss": 0.3428,
+      "step": 13882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3372874745079748,
+      "learning_rate": 7.715110772507652e-06,
+      "loss": 0.3532,
+      "step": 13883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4516844915819604,
+      "learning_rate": 7.71477990734681e-06,
+      "loss": 0.4013,
+      "step": 13884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.744450395335875,
+      "learning_rate": 7.714449025327917e-06,
+      "loss": 0.2977,
+      "step": 13885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4690614278838519,
+      "learning_rate": 7.714118126453034e-06,
+      "loss": 0.2861,
+      "step": 13886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1663692519598505,
+      "learning_rate": 7.713787210724213e-06,
+      "loss": 0.1901,
+      "step": 13887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.53496089206729,
+      "learning_rate": 7.713456278143508e-06,
+      "loss": 0.2002,
+      "step": 13888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3996101902143023,
+      "learning_rate": 7.713125328712978e-06,
+      "loss": 0.243,
+      "step": 13889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8912925258762976,
+      "learning_rate": 7.712794362434672e-06,
+      "loss": 0.2922,
+      "step": 13890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4438709906046099,
+      "learning_rate": 7.712463379310652e-06,
+      "loss": 0.3866,
+      "step": 13891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.800741054557974,
+      "learning_rate": 7.71213237934297e-06,
+      "loss": 0.4456,
+      "step": 13892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2377551191513414,
+      "learning_rate": 7.71180136253368e-06,
+      "loss": 0.1684,
+      "step": 13893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7740136651619693,
+      "learning_rate": 7.71147032888484e-06,
+      "loss": 0.4232,
+      "step": 13894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4992129936028238,
+      "learning_rate": 7.711139278398503e-06,
+      "loss": 0.2294,
+      "step": 13895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.035963001076876,
+      "learning_rate": 7.710808211076729e-06,
+      "loss": 0.411,
+      "step": 13896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4664423345191746,
+      "learning_rate": 7.710477126921569e-06,
+      "loss": 0.1883,
+      "step": 13897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3782322887256224,
+      "learning_rate": 7.710146025935082e-06,
+      "loss": 0.1711,
+      "step": 13898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.504739792926227,
+      "learning_rate": 7.709814908119322e-06,
+      "loss": 0.2516,
+      "step": 13899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8729350043690813,
+      "learning_rate": 7.709483773476347e-06,
+      "loss": 0.3863,
+      "step": 13900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9560433862124482,
+      "learning_rate": 7.709152622008214e-06,
+      "loss": 0.507,
+      "step": 13901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4484526622106197,
+      "learning_rate": 7.708821453716977e-06,
+      "loss": 0.362,
+      "step": 13902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.726582475261873,
+      "learning_rate": 7.708490268604693e-06,
+      "loss": 0.4364,
+      "step": 13903
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7076218203242126,
+      "learning_rate": 7.708159066673419e-06,
+      "loss": 0.3833,
+      "step": 13904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7924048548882427,
+      "learning_rate": 7.707827847925212e-06,
+      "loss": 0.3497,
+      "step": 13905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4212593459230525,
+      "learning_rate": 7.707496612362128e-06,
+      "loss": 0.2169,
+      "step": 13906
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6023970685743028,
+      "learning_rate": 7.707165359986225e-06,
+      "loss": 0.427,
+      "step": 13907
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.941817408140414,
+      "learning_rate": 7.70683409079956e-06,
+      "loss": 0.3614,
+      "step": 13908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4258531082021457,
+      "learning_rate": 7.706502804804188e-06,
+      "loss": 0.2304,
+      "step": 13909
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6297136679548154,
+      "learning_rate": 7.706171502002169e-06,
+      "loss": 0.3622,
+      "step": 13910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6430120042520833,
+      "learning_rate": 7.705840182395556e-06,
+      "loss": 0.3964,
+      "step": 13911
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1544628201529537,
+      "learning_rate": 7.705508845986412e-06,
+      "loss": 0.1883,
+      "step": 13912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.625746482173065,
+      "learning_rate": 7.705177492776791e-06,
+      "loss": 0.3799,
+      "step": 13913
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1944459990269216,
+      "learning_rate": 7.704846122768752e-06,
+      "loss": 0.2232,
+      "step": 13914
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2228633736407255,
+      "learning_rate": 7.704514735964353e-06,
+      "loss": 0.1912,
+      "step": 13915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6607647263173906,
+      "learning_rate": 7.704183332365649e-06,
+      "loss": 0.3435,
+      "step": 13916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9752629610538546,
+      "learning_rate": 7.703851911974701e-06,
+      "loss": 0.4109,
+      "step": 13917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5067161146522836,
+      "learning_rate": 7.703520474793564e-06,
+      "loss": 0.376,
+      "step": 13918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.531399028688267,
+      "learning_rate": 7.7031890208243e-06,
+      "loss": 0.2404,
+      "step": 13919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6057681897330018,
+      "learning_rate": 7.702857550068965e-06,
+      "loss": 0.3277,
+      "step": 13920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.440675131366181,
+      "learning_rate": 7.702526062529618e-06,
+      "loss": 0.2203,
+      "step": 13921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5940372124098088,
+      "learning_rate": 7.702194558208318e-06,
+      "loss": 0.4411,
+      "step": 13922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5330769359388412,
+      "learning_rate": 7.70186303710712e-06,
+      "loss": 0.206,
+      "step": 13923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7976875251062066,
+      "learning_rate": 7.701531499228087e-06,
+      "loss": 0.3977,
+      "step": 13924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.990140881836579,
+      "learning_rate": 7.701199944573276e-06,
+      "loss": 0.3214,
+      "step": 13925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4422369013433018,
+      "learning_rate": 7.700868373144746e-06,
+      "loss": 0.3265,
+      "step": 13926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3457417586363203,
+      "learning_rate": 7.700536784944555e-06,
+      "loss": 0.2397,
+      "step": 13927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9680251384136065,
+      "learning_rate": 7.700205179974765e-06,
+      "loss": 0.3601,
+      "step": 13928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.386493086922659,
+      "learning_rate": 7.69987355823743e-06,
+      "loss": 0.2007,
+      "step": 13929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3658053468909586,
+      "learning_rate": 7.699541919734614e-06,
+      "loss": 0.317,
+      "step": 13930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2227525307524438,
+      "learning_rate": 7.699210264468375e-06,
+      "loss": 0.162,
+      "step": 13931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4217811097596245,
+      "learning_rate": 7.698878592440771e-06,
+      "loss": 0.2061,
+      "step": 13932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5678535942780896,
+      "learning_rate": 7.698546903653866e-06,
+      "loss": 0.2941,
+      "step": 13933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5260399182627518,
+      "learning_rate": 7.698215198109713e-06,
+      "loss": 0.2505,
+      "step": 13934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7948704069809256,
+      "learning_rate": 7.697883475810378e-06,
+      "loss": 0.2966,
+      "step": 13935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2069411421600809,
+      "learning_rate": 7.697551736757918e-06,
+      "loss": 0.186,
+      "step": 13936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4089873427826984,
+      "learning_rate": 7.697219980954392e-06,
+      "loss": 0.3206,
+      "step": 13937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6856923885899016,
+      "learning_rate": 7.696888208401863e-06,
+      "loss": 0.3787,
+      "step": 13938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0327331043663,
+      "learning_rate": 7.696556419102388e-06,
+      "loss": 0.1775,
+      "step": 13939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8494420685874937,
+      "learning_rate": 7.69622461305803e-06,
+      "loss": 0.5002,
+      "step": 13940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.375189858812893,
+      "learning_rate": 7.695892790270849e-06,
+      "loss": 0.3478,
+      "step": 13941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7285648015672428,
+      "learning_rate": 7.695560950742904e-06,
+      "loss": 0.3863,
+      "step": 13942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0713185187317635,
+      "learning_rate": 7.695229094476256e-06,
+      "loss": 0.7251,
+      "step": 13943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6834934673845365,
+      "learning_rate": 7.694897221472968e-06,
+      "loss": 0.5212,
+      "step": 13944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6480817230077416,
+      "learning_rate": 7.694565331735097e-06,
+      "loss": 0.4479,
+      "step": 13945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.519417248546095,
+      "learning_rate": 7.694233425264708e-06,
+      "loss": 0.2811,
+      "step": 13946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.638005579274466,
+      "learning_rate": 7.693901502063859e-06,
+      "loss": 0.3367,
+      "step": 13947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5060844772651798,
+      "learning_rate": 7.693569562134614e-06,
+      "loss": 0.2474,
+      "step": 13948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.929004617858054,
+      "learning_rate": 7.69323760547903e-06,
+      "loss": 0.3649,
+      "step": 13949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.602642418633508,
+      "learning_rate": 7.692905632099173e-06,
+      "loss": 0.2723,
+      "step": 13950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4298246218894555,
+      "learning_rate": 7.6925736419971e-06,
+      "loss": 0.2187,
+      "step": 13951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6474389265271259,
+      "learning_rate": 7.692241635174877e-06,
+      "loss": 0.4613,
+      "step": 13952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.840261752030832,
+      "learning_rate": 7.691909611634563e-06,
+      "loss": 0.1888,
+      "step": 13953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.746168019808175,
+      "learning_rate": 7.691577571378219e-06,
+      "loss": 0.366,
+      "step": 13954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.711394775806316,
+      "learning_rate": 7.69124551440791e-06,
+      "loss": 0.5439,
+      "step": 13955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5815612612894512,
+      "learning_rate": 7.690913440725696e-06,
+      "loss": 0.2806,
+      "step": 13956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5770038880990886,
+      "learning_rate": 7.690581350333639e-06,
+      "loss": 0.2204,
+      "step": 13957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5040793727956325,
+      "learning_rate": 7.6902492432338e-06,
+      "loss": 0.3005,
+      "step": 13958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7190018664396238,
+      "learning_rate": 7.689917119428246e-06,
+      "loss": 0.217,
+      "step": 13959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3592910219977796,
+      "learning_rate": 7.689584978919033e-06,
+      "loss": 0.33,
+      "step": 13960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4704770070149085,
+      "learning_rate": 7.68925282170823e-06,
+      "loss": 0.3639,
+      "step": 13961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4477187449928135,
+      "learning_rate": 7.688920647797894e-06,
+      "loss": 0.2676,
+      "step": 13962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.476487639714386,
+      "learning_rate": 7.688588457190092e-06,
+      "loss": 0.1976,
+      "step": 13963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9673657546359076,
+      "learning_rate": 7.688256249886883e-06,
+      "loss": 0.1798,
+      "step": 13964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6922856570981697,
+      "learning_rate": 7.687924025890331e-06,
+      "loss": 0.4462,
+      "step": 13965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1328320135544385,
+      "learning_rate": 7.687591785202504e-06,
+      "loss": 0.2435,
+      "step": 13966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3933854618523323,
+      "learning_rate": 7.687259527825457e-06,
+      "loss": 0.1947,
+      "step": 13967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4383567661765022,
+      "learning_rate": 7.686927253761259e-06,
+      "loss": 0.3018,
+      "step": 13968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8181453419068985,
+      "learning_rate": 7.68659496301197e-06,
+      "loss": 0.1435,
+      "step": 13969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4749206517584899,
+      "learning_rate": 7.686262655579656e-06,
+      "loss": 0.4076,
+      "step": 13970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4114650480408617,
+      "learning_rate": 7.685930331466378e-06,
+      "loss": 0.1448,
+      "step": 13971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.282355577889109,
+      "learning_rate": 7.685597990674202e-06,
+      "loss": 0.353,
+      "step": 13972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.660493538196664,
+      "learning_rate": 7.685265633205192e-06,
+      "loss": 0.2709,
+      "step": 13973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5278773534239052,
+      "learning_rate": 7.684933259061409e-06,
+      "loss": 0.3779,
+      "step": 13974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.165115155711196,
+      "learning_rate": 7.68460086824492e-06,
+      "loss": 0.1842,
+      "step": 13975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4801496605991706,
+      "learning_rate": 7.684268460757786e-06,
+      "loss": 0.2223,
+      "step": 13976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8729696247929064,
+      "learning_rate": 7.683936036602075e-06,
+      "loss": 0.3748,
+      "step": 13977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2163858193927837,
+      "learning_rate": 7.683603595779846e-06,
+      "loss": 0.2289,
+      "step": 13978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9547553886925237,
+      "learning_rate": 7.68327113829317e-06,
+      "loss": 0.3283,
+      "step": 13979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.006612111810407,
+      "learning_rate": 7.682938664144107e-06,
+      "loss": 0.3317,
+      "step": 13980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7659424099831138,
+      "learning_rate": 7.682606173334721e-06,
+      "loss": 0.3635,
+      "step": 13981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8312949956241769,
+      "learning_rate": 7.68227366586708e-06,
+      "loss": 0.4742,
+      "step": 13982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5016453332114206,
+      "learning_rate": 7.681941141743246e-06,
+      "loss": 0.3848,
+      "step": 13983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5857657565063725,
+      "learning_rate": 7.681608600965285e-06,
+      "loss": 0.216,
+      "step": 13984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4221810415000176,
+      "learning_rate": 7.681276043535262e-06,
+      "loss": 0.2544,
+      "step": 13985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.00030481727673,
+      "learning_rate": 7.680943469455242e-06,
+      "loss": 0.4906,
+      "step": 13986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4446031427726058,
+      "learning_rate": 7.68061087872729e-06,
+      "loss": 0.2021,
+      "step": 13987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6106124028196924,
+      "learning_rate": 7.68027827135347e-06,
+      "loss": 0.3607,
+      "step": 13988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7120342092792225,
+      "learning_rate": 7.679945647335852e-06,
+      "loss": 0.5292,
+      "step": 13989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3155826082094513,
+      "learning_rate": 7.679613006676496e-06,
+      "loss": 0.3555,
+      "step": 13990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.818545298182876,
+      "learning_rate": 7.679280349377468e-06,
+      "loss": 0.597,
+      "step": 13991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7676548556539222,
+      "learning_rate": 7.67894767544084e-06,
+      "loss": 0.1953,
+      "step": 13992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.482560132481296,
+      "learning_rate": 7.678614984868672e-06,
+      "loss": 0.2656,
+      "step": 13993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.492252722907526,
+      "learning_rate": 7.678282277663031e-06,
+      "loss": 0.3342,
+      "step": 13994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0112050928421414,
+      "learning_rate": 7.677949553825983e-06,
+      "loss": 0.6602,
+      "step": 13995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7022952204155475,
+      "learning_rate": 7.677616813359596e-06,
+      "loss": 0.2345,
+      "step": 13996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.523323602575752,
+      "learning_rate": 7.677284056265935e-06,
+      "loss": 0.2258,
+      "step": 13997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0236318275854996,
+      "learning_rate": 7.676951282547064e-06,
+      "loss": 0.1855,
+      "step": 13998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6968267354019957,
+      "learning_rate": 7.676618492205053e-06,
+      "loss": 0.2516,
+      "step": 13999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.379656781344604,
+      "learning_rate": 7.676285685241967e-06,
+      "loss": 0.54,
+      "step": 14000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8104970384012307,
+      "learning_rate": 7.675952861659872e-06,
+      "loss": 0.3492,
+      "step": 14001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2270950215969374,
+      "learning_rate": 7.675620021460836e-06,
+      "loss": 0.5564,
+      "step": 14002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6622194419939023,
+      "learning_rate": 7.675287164646925e-06,
+      "loss": 0.1661,
+      "step": 14003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1008301796472113,
+      "learning_rate": 7.674954291220206e-06,
+      "loss": 0.1312,
+      "step": 14004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2573377545563305,
+      "learning_rate": 7.674621401182748e-06,
+      "loss": 0.2014,
+      "step": 14005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4496130214183753,
+      "learning_rate": 7.674288494536614e-06,
+      "loss": 0.2572,
+      "step": 14006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8394326846221827,
+      "learning_rate": 7.673955571283875e-06,
+      "loss": 0.2773,
+      "step": 14007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5363304753095204,
+      "learning_rate": 7.673622631426598e-06,
+      "loss": 0.3409,
+      "step": 14008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.322871206158278,
+      "learning_rate": 7.673289674966849e-06,
+      "loss": 0.3393,
+      "step": 14009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0459652764593765,
+      "learning_rate": 7.672956701906695e-06,
+      "loss": 0.3676,
+      "step": 14010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7165120260647004,
+      "learning_rate": 7.672623712248206e-06,
+      "loss": 0.2294,
+      "step": 14011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9032006752145902,
+      "learning_rate": 7.672290705993448e-06,
+      "loss": 0.3658,
+      "step": 14012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5348275902973225,
+      "learning_rate": 7.671957683144489e-06,
+      "loss": 0.3387,
+      "step": 14013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6301700802866004,
+      "learning_rate": 7.6716246437034e-06,
+      "loss": 0.5035,
+      "step": 14014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6550985530864752,
+      "learning_rate": 7.671291587672242e-06,
+      "loss": 0.3008,
+      "step": 14015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8813308350726614,
+      "learning_rate": 7.670958515053091e-06,
+      "loss": 0.3915,
+      "step": 14016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3270668596987394,
+      "learning_rate": 7.67062542584801e-06,
+      "loss": 0.1828,
+      "step": 14017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1376532510842785,
+      "learning_rate": 7.670292320059071e-06,
+      "loss": 0.7986,
+      "step": 14018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8645177333274399,
+      "learning_rate": 7.669959197688343e-06,
+      "loss": 0.3051,
+      "step": 14019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.649321070704069,
+      "learning_rate": 7.669626058737889e-06,
+      "loss": 0.4664,
+      "step": 14020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7347940248064335,
+      "learning_rate": 7.669292903209784e-06,
+      "loss": 0.5289,
+      "step": 14021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6015157367112973,
+      "learning_rate": 7.668959731106092e-06,
+      "loss": 0.3811,
+      "step": 14022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6029927055062925,
+      "learning_rate": 7.668626542428884e-06,
+      "loss": 0.2765,
+      "step": 14023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5497752993897518,
+      "learning_rate": 7.668293337180228e-06,
+      "loss": 0.2185,
+      "step": 14024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9691392158090082,
+      "learning_rate": 7.667960115362197e-06,
+      "loss": 0.4658,
+      "step": 14025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4646586530089831,
+      "learning_rate": 7.667626876976855e-06,
+      "loss": 0.2741,
+      "step": 14026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7369621050586264,
+      "learning_rate": 7.667293622026274e-06,
+      "loss": 0.3463,
+      "step": 14027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7905654070958847,
+      "learning_rate": 7.666960350512522e-06,
+      "loss": 0.3599,
+      "step": 14028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5813911744375262,
+      "learning_rate": 7.666627062437671e-06,
+      "loss": 0.1703,
+      "step": 14029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6756286885101062,
+      "learning_rate": 7.666293757803788e-06,
+      "loss": 0.3918,
+      "step": 14030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4676363151899718,
+      "learning_rate": 7.665960436612946e-06,
+      "loss": 0.3815,
+      "step": 14031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5897322411762291,
+      "learning_rate": 7.66562709886721e-06,
+      "loss": 0.1895,
+      "step": 14032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.729664576946285,
+      "learning_rate": 7.665293744568654e-06,
+      "loss": 0.4237,
+      "step": 14033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.483636456518207,
+      "learning_rate": 7.664960373719348e-06,
+      "loss": 0.2407,
+      "step": 14034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0486234788484599,
+      "learning_rate": 7.664626986321358e-06,
+      "loss": 0.1104,
+      "step": 14035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8848939367704467,
+      "learning_rate": 7.664293582376759e-06,
+      "loss": 0.5418,
+      "step": 14036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.017748420227183,
+      "learning_rate": 7.663960161887619e-06,
+      "loss": 0.5797,
+      "step": 14037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5221026025198605,
+      "learning_rate": 7.66362672485601e-06,
+      "loss": 0.1408,
+      "step": 14038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3903693194492364,
+      "learning_rate": 7.663293271284e-06,
+      "loss": 0.1932,
+      "step": 14039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9205123831482076,
+      "learning_rate": 7.662959801173662e-06,
+      "loss": 0.3351,
+      "step": 14040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2789746701365925,
+      "learning_rate": 7.662626314527065e-06,
+      "loss": 0.2291,
+      "step": 14041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6127160331445511,
+      "learning_rate": 7.662292811346281e-06,
+      "loss": 0.1521,
+      "step": 14042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5163155806220716,
+      "learning_rate": 7.661959291633381e-06,
+      "loss": 0.3526,
+      "step": 14043
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0164567050215547,
+      "learning_rate": 7.661625755390435e-06,
+      "loss": 0.2429,
+      "step": 14044
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3473849341348967,
+      "learning_rate": 7.661292202619516e-06,
+      "loss": 0.3214,
+      "step": 14045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6004883176070508,
+      "learning_rate": 7.660958633322693e-06,
+      "loss": 0.2168,
+      "step": 14046
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4601072716496803,
+      "learning_rate": 7.660625047502039e-06,
+      "loss": 0.1946,
+      "step": 14047
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9986219186559818,
+      "learning_rate": 7.660291445159626e-06,
+      "loss": 0.1899,
+      "step": 14048
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4880318748756696,
+      "learning_rate": 7.659957826297523e-06,
+      "loss": 0.6524,
+      "step": 14049
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9538912721968608,
+      "learning_rate": 7.659624190917805e-06,
+      "loss": 0.1753,
+      "step": 14050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.875221940654019,
+      "learning_rate": 7.65929053902254e-06,
+      "loss": 0.2308,
+      "step": 14051
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2972580332592774,
+      "learning_rate": 7.658956870613804e-06,
+      "loss": 0.1998,
+      "step": 14052
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8134603997795704,
+      "learning_rate": 7.658623185693665e-06,
+      "loss": 0.3634,
+      "step": 14053
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8351164869123096,
+      "learning_rate": 7.6582894842642e-06,
+      "loss": 0.5587,
+      "step": 14054
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6049647131721863,
+      "learning_rate": 7.657955766327476e-06,
+      "loss": 0.3559,
+      "step": 14055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3791608980534549,
+      "learning_rate": 7.657622031885567e-06,
+      "loss": 0.2806,
+      "step": 14056
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7931676143490154,
+      "learning_rate": 7.657288280940548e-06,
+      "loss": 0.3487,
+      "step": 14057
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6373870027116604,
+      "learning_rate": 7.656954513494487e-06,
+      "loss": 0.4044,
+      "step": 14058
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2085058254273966,
+      "learning_rate": 7.656620729549459e-06,
+      "loss": 0.2371,
+      "step": 14059
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8161895458612007,
+      "learning_rate": 7.65628692910754e-06,
+      "loss": 0.4525,
+      "step": 14060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.578114177883924,
+      "learning_rate": 7.655953112170797e-06,
+      "loss": 0.2605,
+      "step": 14061
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7663610704828527,
+      "learning_rate": 7.655619278741308e-06,
+      "loss": 0.4629,
+      "step": 14062
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8657399555872858,
+      "learning_rate": 7.655285428821143e-06,
+      "loss": 0.329,
+      "step": 14063
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.075965470957235,
+      "learning_rate": 7.654951562412372e-06,
+      "loss": 0.2605,
+      "step": 14064
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8294799625089686,
+      "learning_rate": 7.654617679517075e-06,
+      "loss": 0.4788,
+      "step": 14065
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.697389944898753,
+      "learning_rate": 7.654283780137322e-06,
+      "loss": 0.3002,
+      "step": 14066
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7917698090735554,
+      "learning_rate": 7.653949864275186e-06,
+      "loss": 0.2714,
+      "step": 14067
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.82422614249257,
+      "learning_rate": 7.65361593193274e-06,
+      "loss": 0.4813,
+      "step": 14068
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9867287197155403,
+      "learning_rate": 7.653281983112061e-06,
+      "loss": 0.4451,
+      "step": 14069
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.380384236648651,
+      "learning_rate": 7.652948017815218e-06,
+      "loss": 0.2321,
+      "step": 14070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8618263866638272,
+      "learning_rate": 7.65261403604429e-06,
+      "loss": 0.4509,
+      "step": 14071
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6395658752848146,
+      "learning_rate": 7.652280037801346e-06,
+      "loss": 0.4926,
+      "step": 14072
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5496626333936678,
+      "learning_rate": 7.651946023088464e-06,
+      "loss": 0.374,
+      "step": 14073
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.892810695046504,
+      "learning_rate": 7.651611991907716e-06,
+      "loss": 0.3248,
+      "step": 14074
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0775381851588928,
+      "learning_rate": 7.651277944261175e-06,
+      "loss": 0.4129,
+      "step": 14075
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.742536347364509,
+      "learning_rate": 7.65094388015092e-06,
+      "loss": 0.6336,
+      "step": 14076
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.761469965113745,
+      "learning_rate": 7.65060979957902e-06,
+      "loss": 0.511,
+      "step": 14077
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4849259759373978,
+      "learning_rate": 7.650275702547553e-06,
+      "loss": 0.2651,
+      "step": 14078
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4721605493665395,
+      "learning_rate": 7.649941589058592e-06,
+      "loss": 0.365,
+      "step": 14079
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8230693655545522,
+      "learning_rate": 7.649607459114214e-06,
+      "loss": 0.2262,
+      "step": 14080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2521515761868487,
+      "learning_rate": 7.64927331271649e-06,
+      "loss": 0.2922,
+      "step": 14081
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9739840014322942,
+      "learning_rate": 7.648939149867498e-06,
+      "loss": 0.4027,
+      "step": 14082
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.59702092761828,
+      "learning_rate": 7.648604970569314e-06,
+      "loss": 0.2625,
+      "step": 14083
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8520017738606709,
+      "learning_rate": 7.648270774824009e-06,
+      "loss": 0.4862,
+      "step": 14084
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9337609806377825,
+      "learning_rate": 7.647936562633661e-06,
+      "loss": 0.4778,
+      "step": 14085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7949129650239348,
+      "learning_rate": 7.647602334000346e-06,
+      "loss": 0.2188,
+      "step": 14086
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6919105341871465,
+      "learning_rate": 7.64726808892614e-06,
+      "loss": 0.368,
+      "step": 14087
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3807137073911453,
+      "learning_rate": 7.646933827413114e-06,
+      "loss": 0.2797,
+      "step": 14088
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8601801091972642,
+      "learning_rate": 7.646599549463349e-06,
+      "loss": 0.1836,
+      "step": 14089
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.327490316935334,
+      "learning_rate": 7.646265255078918e-06,
+      "loss": 0.1663,
+      "step": 14090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2308739827462654,
+      "learning_rate": 7.645930944261897e-06,
+      "loss": 0.2526,
+      "step": 14091
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2760763792908756,
+      "learning_rate": 7.645596617014363e-06,
+      "loss": 0.319,
+      "step": 14092
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4317142837123364,
+      "learning_rate": 7.645262273338391e-06,
+      "loss": 0.2715,
+      "step": 14093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8552417375326355,
+      "learning_rate": 7.64492791323606e-06,
+      "loss": 0.2721,
+      "step": 14094
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9259607026402954,
+      "learning_rate": 7.64459353670944e-06,
+      "loss": 0.5464,
+      "step": 14095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8783825027374113,
+      "learning_rate": 7.644259143760616e-06,
+      "loss": 0.2634,
+      "step": 14096
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9368528854373046,
+      "learning_rate": 7.643924734391657e-06,
+      "loss": 0.3389,
+      "step": 14097
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0378140504419013,
+      "learning_rate": 7.643590308604643e-06,
+      "loss": 0.5652,
+      "step": 14098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8772539181594103,
+      "learning_rate": 7.643255866401649e-06,
+      "loss": 0.5546,
+      "step": 14099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8309710450459893,
+      "learning_rate": 7.642921407784754e-06,
+      "loss": 0.2712,
+      "step": 14100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3932567296614278,
+      "learning_rate": 7.642586932756033e-06,
+      "loss": 0.3294,
+      "step": 14101
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.654439323264328,
+      "learning_rate": 7.642252441317566e-06,
+      "loss": 0.5285,
+      "step": 14102
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7828726634640317,
+      "learning_rate": 7.641917933471425e-06,
+      "loss": 0.3505,
+      "step": 14103
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4955458745936634,
+      "learning_rate": 7.641583409219692e-06,
+      "loss": 0.1729,
+      "step": 14104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5303144735495342,
+      "learning_rate": 7.641248868564442e-06,
+      "loss": 0.3152,
+      "step": 14105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3765877840987035,
+      "learning_rate": 7.640914311507754e-06,
+      "loss": 0.2102,
+      "step": 14106
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4080753189056077,
+      "learning_rate": 7.640579738051703e-06,
+      "loss": 0.3286,
+      "step": 14107
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.039635006967139,
+      "learning_rate": 7.640245148198367e-06,
+      "loss": 0.2437,
+      "step": 14108
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1128381449629563,
+      "learning_rate": 7.639910541949827e-06,
+      "loss": 0.1584,
+      "step": 14109
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5018950625979408,
+      "learning_rate": 7.639575919308158e-06,
+      "loss": 0.377,
+      "step": 14110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6805653385156967,
+      "learning_rate": 7.639241280275438e-06,
+      "loss": 0.449,
+      "step": 14111
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4108053968509515,
+      "learning_rate": 7.638906624853744e-06,
+      "loss": 0.6991,
+      "step": 14112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.52394354558416,
+      "learning_rate": 7.638571953045157e-06,
+      "loss": 0.258,
+      "step": 14113
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5439402455313707,
+      "learning_rate": 7.638237264851753e-06,
+      "loss": 0.3924,
+      "step": 14114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6231518446377302,
+      "learning_rate": 7.637902560275614e-06,
+      "loss": 0.3931,
+      "step": 14115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6331648149903542,
+      "learning_rate": 7.637567839318812e-06,
+      "loss": 0.3565,
+      "step": 14116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3864443158203719,
+      "learning_rate": 7.63723310198343e-06,
+      "loss": 0.2276,
+      "step": 14117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6302352807577232,
+      "learning_rate": 7.636898348271546e-06,
+      "loss": 0.3387,
+      "step": 14118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7255620301711043,
+      "learning_rate": 7.636563578185238e-06,
+      "loss": 0.4103,
+      "step": 14119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.767959944441406,
+      "learning_rate": 7.636228791726588e-06,
+      "loss": 0.1984,
+      "step": 14120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5449967878579882,
+      "learning_rate": 7.63589398889767e-06,
+      "loss": 0.2575,
+      "step": 14121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.262117321097281,
+      "learning_rate": 7.635559169700564e-06,
+      "loss": 0.298,
+      "step": 14122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.478505856419976,
+      "learning_rate": 7.63522433413735e-06,
+      "loss": 0.3442,
+      "step": 14123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.814240798394041,
+      "learning_rate": 7.634889482210111e-06,
+      "loss": 0.277,
+      "step": 14124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.322379931378944,
+      "learning_rate": 7.634554613920919e-06,
+      "loss": 0.1682,
+      "step": 14125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3091577367864526,
+      "learning_rate": 7.63421972927186e-06,
+      "loss": 0.219,
+      "step": 14126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.71737981713792,
+      "learning_rate": 7.63388482826501e-06,
+      "loss": 0.3769,
+      "step": 14127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6153340868172592,
+      "learning_rate": 7.63354991090245e-06,
+      "loss": 0.1913,
+      "step": 14128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1337451647031858,
+      "learning_rate": 7.63321497718626e-06,
+      "loss": 0.1638,
+      "step": 14129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.818300912083167,
+      "learning_rate": 7.632880027118517e-06,
+      "loss": 0.5354,
+      "step": 14130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6055843515845134,
+      "learning_rate": 7.632545060701305e-06,
+      "loss": 0.2638,
+      "step": 14131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6204361811366315,
+      "learning_rate": 7.632210077936701e-06,
+      "loss": 0.2037,
+      "step": 14132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1922906641614597,
+      "learning_rate": 7.631875078826786e-06,
+      "loss": 0.2585,
+      "step": 14133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.794962316976885,
+      "learning_rate": 7.63154006337364e-06,
+      "loss": 0.3632,
+      "step": 14134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7383488245484404,
+      "learning_rate": 7.631205031579344e-06,
+      "loss": 0.393,
+      "step": 14135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.047092047678685,
+      "learning_rate": 7.63086998344598e-06,
+      "loss": 0.3511,
+      "step": 14136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.942638482038866,
+      "learning_rate": 7.630534918975627e-06,
+      "loss": 0.4792,
+      "step": 14137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3418425367601234,
+      "learning_rate": 7.630199838170366e-06,
+      "loss": 0.2338,
+      "step": 14138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2235897936493363,
+      "learning_rate": 7.629864741032276e-06,
+      "loss": 0.3722,
+      "step": 14139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7297650536068068,
+      "learning_rate": 7.62952962756344e-06,
+      "loss": 0.4441,
+      "step": 14140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2721890578451343,
+      "learning_rate": 7.629194497765937e-06,
+      "loss": 0.2345,
+      "step": 14141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2330629991731907,
+      "learning_rate": 7.628859351641851e-06,
+      "loss": 0.3139,
+      "step": 14142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2380429273200464,
+      "learning_rate": 7.628524189193259e-06,
+      "loss": 0.239,
+      "step": 14143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0247855721961665,
+      "learning_rate": 7.628189010422247e-06,
+      "loss": 0.223,
+      "step": 14144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3404904075675614,
+      "learning_rate": 7.627853815330893e-06,
+      "loss": 0.2301,
+      "step": 14145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.586044073047497,
+      "learning_rate": 7.627518603921279e-06,
+      "loss": 0.2083,
+      "step": 14146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4183305562505233,
+      "learning_rate": 7.627183376195487e-06,
+      "loss": 0.3112,
+      "step": 14147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6882366298973395,
+      "learning_rate": 7.6268481321556e-06,
+      "loss": 0.3695,
+      "step": 14148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6907080162805712,
+      "learning_rate": 7.626512871803697e-06,
+      "loss": 0.3509,
+      "step": 14149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8093789375670812,
+      "learning_rate": 7.626177595141863e-06,
+      "loss": 0.4773,
+      "step": 14150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.345381073970706,
+      "learning_rate": 7.6258423021721775e-06,
+      "loss": 0.2945,
+      "step": 14151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.580234611787692,
+      "learning_rate": 7.625506992896723e-06,
+      "loss": 0.3373,
+      "step": 14152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.319041266745734,
+      "learning_rate": 7.625171667317583e-06,
+      "loss": 0.5535,
+      "step": 14153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.593230080882277,
+      "learning_rate": 7.624836325436838e-06,
+      "loss": 0.4139,
+      "step": 14154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9620189724144859,
+      "learning_rate": 7.6245009672565724e-06,
+      "loss": 0.5386,
+      "step": 14155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6336480699481022,
+      "learning_rate": 7.6241655927788675e-06,
+      "loss": 0.2283,
+      "step": 14156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1523801830157399,
+      "learning_rate": 7.6238302020058045e-06,
+      "loss": 0.2033,
+      "step": 14157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9085259803798038,
+      "learning_rate": 7.62349479493947e-06,
+      "loss": 0.3174,
+      "step": 14158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6414327199908654,
+      "learning_rate": 7.6231593715819415e-06,
+      "loss": 0.2493,
+      "step": 14159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4695846894691222,
+      "learning_rate": 7.622823931935307e-06,
+      "loss": 0.2671,
+      "step": 14160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9594599830109243,
+      "learning_rate": 7.622488476001646e-06,
+      "loss": 0.3437,
+      "step": 14161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.821163183367211,
+      "learning_rate": 7.622153003783044e-06,
+      "loss": 0.3842,
+      "step": 14162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9951163696645888,
+      "learning_rate": 7.6218175152815806e-06,
+      "loss": 0.2486,
+      "step": 14163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4751809686028816,
+      "learning_rate": 7.621482010499343e-06,
+      "loss": 0.3729,
+      "step": 14164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2770737237190186,
+      "learning_rate": 7.621146489438413e-06,
+      "loss": 0.1837,
+      "step": 14165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.612363392565795,
+      "learning_rate": 7.6208109521008756e-06,
+      "loss": 0.4365,
+      "step": 14166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8280263531796974,
+      "learning_rate": 7.620475398488812e-06,
+      "loss": 0.3862,
+      "step": 14167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8162317089595876,
+      "learning_rate": 7.620139828604306e-06,
+      "loss": 0.3903,
+      "step": 14168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.497313089710649,
+      "learning_rate": 7.619804242449443e-06,
+      "loss": 0.359,
+      "step": 14169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4261697422785864,
+      "learning_rate": 7.6194686400263065e-06,
+      "loss": 0.2778,
+      "step": 14170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4922191906105307,
+      "learning_rate": 7.619133021336979e-06,
+      "loss": 0.3663,
+      "step": 14171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3825601277417034,
+      "learning_rate": 7.618797386383546e-06,
+      "loss": 0.2463,
+      "step": 14172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9283859210858318,
+      "learning_rate": 7.618461735168093e-06,
+      "loss": 0.6074,
+      "step": 14173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8165117294695137,
+      "learning_rate": 7.618126067692702e-06,
+      "loss": 0.2749,
+      "step": 14174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3012528953419262,
+      "learning_rate": 7.617790383959458e-06,
+      "loss": 0.2461,
+      "step": 14175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6189406746609087,
+      "learning_rate": 7.617454683970444e-06,
+      "loss": 0.3602,
+      "step": 14176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7951990521660937,
+      "learning_rate": 7.617118967727748e-06,
+      "loss": 0.5092,
+      "step": 14177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5266787844892578,
+      "learning_rate": 7.616783235233454e-06,
+      "loss": 0.3436,
+      "step": 14178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3108822241440674,
+      "learning_rate": 7.6164474864896445e-06,
+      "loss": 0.1949,
+      "step": 14179
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4541826500966084,
+      "learning_rate": 7.616111721498405e-06,
+      "loss": 0.3765,
+      "step": 14180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5463035568361685,
+      "learning_rate": 7.6157759402618225e-06,
+      "loss": 0.3043,
+      "step": 14181
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9256654853536195,
+      "learning_rate": 7.61544014278198e-06,
+      "loss": 0.3361,
+      "step": 14182
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7770898574748102,
+      "learning_rate": 7.615104329060965e-06,
+      "loss": 0.5245,
+      "step": 14183
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0089748780650796,
+      "learning_rate": 7.614768499100859e-06,
+      "loss": 0.257,
+      "step": 14184
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5839827733907823,
+      "learning_rate": 7.61443265290375e-06,
+      "loss": 0.3095,
+      "step": 14185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7959449392814812,
+      "learning_rate": 7.6140967904717254e-06,
+      "loss": 0.4045,
+      "step": 14186
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4746173454220841,
+      "learning_rate": 7.613760911806867e-06,
+      "loss": 0.3635,
+      "step": 14187
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.380074943238647,
+      "learning_rate": 7.613425016911263e-06,
+      "loss": 0.2305,
+      "step": 14188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3849007469291885,
+      "learning_rate": 7.613089105786998e-06,
+      "loss": 0.3186,
+      "step": 14189
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6679945235751377,
+      "learning_rate": 7.612753178436159e-06,
+      "loss": 0.619,
+      "step": 14190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4073839384202218,
+      "learning_rate": 7.612417234860831e-06,
+      "loss": 0.2432,
+      "step": 14191
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1183888638333026,
+      "learning_rate": 7.612081275063101e-06,
+      "loss": 0.1464,
+      "step": 14192
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0684807445201756,
+      "learning_rate": 7.611745299045055e-06,
+      "loss": 0.1298,
+      "step": 14193
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8863790415553157,
+      "learning_rate": 7.611409306808777e-06,
+      "loss": 0.5555,
+      "step": 14194
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8606588114932916,
+      "learning_rate": 7.611073298356357e-06,
+      "loss": 0.488,
+      "step": 14195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9648577118940642,
+      "learning_rate": 7.6107372736898785e-06,
+      "loss": 0.6622,
+      "step": 14196
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.230366802533579,
+      "learning_rate": 7.6104012328114316e-06,
+      "loss": 0.3114,
+      "step": 14197
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6238053543002897,
+      "learning_rate": 7.6100651757231e-06,
+      "loss": 0.1367,
+      "step": 14198
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4815699198653107,
+      "learning_rate": 7.6097291024269724e-06,
+      "loss": 0.4624,
+      "step": 14199
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.623862644574451,
+      "learning_rate": 7.609393012925133e-06,
+      "loss": 0.4065,
+      "step": 14200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4125985528947693,
+      "learning_rate": 7.609056907219671e-06,
+      "loss": 0.1539,
+      "step": 14201
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4262493358045498,
+      "learning_rate": 7.6087207853126754e-06,
+      "loss": 0.2162,
+      "step": 14202
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5962714907118827,
+      "learning_rate": 7.608384647206229e-06,
+      "loss": 0.359,
+      "step": 14203
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5365337308202125,
+      "learning_rate": 7.608048492902422e-06,
+      "loss": 0.2272,
+      "step": 14204
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3104496255822116,
+      "learning_rate": 7.607712322403342e-06,
+      "loss": 0.3709,
+      "step": 14205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6873944501186082,
+      "learning_rate": 7.6073761357110746e-06,
+      "loss": 0.3115,
+      "step": 14206
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.512232444286055,
+      "learning_rate": 7.6070399328277085e-06,
+      "loss": 0.4102,
+      "step": 14207
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4350318763429,
+      "learning_rate": 7.606703713755332e-06,
+      "loss": 0.2294,
+      "step": 14208
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1840390069754725,
+      "learning_rate": 7.606367478496033e-06,
+      "loss": 0.2061,
+      "step": 14209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4855410651234542,
+      "learning_rate": 7.606031227051897e-06,
+      "loss": 0.2647,
+      "step": 14210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.754264038906648,
+      "learning_rate": 7.605694959425015e-06,
+      "loss": 0.3888,
+      "step": 14211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.267516923089999,
+      "learning_rate": 7.6053586756174745e-06,
+      "loss": 0.2072,
+      "step": 14212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6171160742056403,
+      "learning_rate": 7.605022375631363e-06,
+      "loss": 0.3309,
+      "step": 14213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1446030125562547,
+      "learning_rate": 7.604686059468769e-06,
+      "loss": 0.1714,
+      "step": 14214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1164561148564056,
+      "learning_rate": 7.604349727131781e-06,
+      "loss": 0.3781,
+      "step": 14215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5563658180955566,
+      "learning_rate": 7.604013378622487e-06,
+      "loss": 0.3798,
+      "step": 14216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8177648775073971,
+      "learning_rate": 7.6036770139429765e-06,
+      "loss": 0.4661,
+      "step": 14217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3904659630348541,
+      "learning_rate": 7.603340633095338e-06,
+      "loss": 0.2109,
+      "step": 14218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7009531837509835,
+      "learning_rate": 7.6030042360816595e-06,
+      "loss": 0.5059,
+      "step": 14219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6484446050843344,
+      "learning_rate": 7.602667822904031e-06,
+      "loss": 0.3241,
+      "step": 14220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6720606709707944,
+      "learning_rate": 7.602331393564541e-06,
+      "loss": 0.3078,
+      "step": 14221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3290842014499022,
+      "learning_rate": 7.60199494806528e-06,
+      "loss": 0.2527,
+      "step": 14222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2706882396914243,
+      "learning_rate": 7.601658486408335e-06,
+      "loss": 0.604,
+      "step": 14223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3073399757740192,
+      "learning_rate": 7.601322008595798e-06,
+      "loss": 0.2528,
+      "step": 14224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.832444711647375,
+      "learning_rate": 7.600985514629755e-06,
+      "loss": 0.4176,
+      "step": 14225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4184851533659404,
+      "learning_rate": 7.600649004512297e-06,
+      "loss": 0.1921,
+      "step": 14226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3828529793800566,
+      "learning_rate": 7.6003124782455155e-06,
+      "loss": 0.1661,
+      "step": 14227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8018945326790075,
+      "learning_rate": 7.599975935831498e-06,
+      "loss": 0.4149,
+      "step": 14228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4552775969527143,
+      "learning_rate": 7.599639377272334e-06,
+      "loss": 0.3887,
+      "step": 14229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5053656692728854,
+      "learning_rate": 7.599302802570117e-06,
+      "loss": 0.3043,
+      "step": 14230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9522023921073062,
+      "learning_rate": 7.598966211726932e-06,
+      "loss": 0.4429,
+      "step": 14231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8557337762384785,
+      "learning_rate": 7.598629604744874e-06,
+      "loss": 0.3845,
+      "step": 14232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9284104395618138,
+      "learning_rate": 7.59829298162603e-06,
+      "loss": 0.6592,
+      "step": 14233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7121523502499485,
+      "learning_rate": 7.59795634237249e-06,
+      "loss": 0.2125,
+      "step": 14234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.580862319864368,
+      "learning_rate": 7.597619686986347e-06,
+      "loss": 0.5101,
+      "step": 14235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7640402272983358,
+      "learning_rate": 7.59728301546969e-06,
+      "loss": 0.491,
+      "step": 14236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6385451735026315,
+      "learning_rate": 7.59694632782461e-06,
+      "loss": 0.2978,
+      "step": 14237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.212734771396204,
+      "learning_rate": 7.5966096240531975e-06,
+      "loss": 0.1519,
+      "step": 14238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.851498323528353,
+      "learning_rate": 7.5962729041575434e-06,
+      "loss": 0.4446,
+      "step": 14239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.17789241747593,
+      "learning_rate": 7.595936168139738e-06,
+      "loss": 0.571,
+      "step": 14240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3101924226461212,
+      "learning_rate": 7.595599416001874e-06,
+      "loss": 0.1343,
+      "step": 14241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.413069531621551,
+      "learning_rate": 7.595262647746041e-06,
+      "loss": 0.2813,
+      "step": 14242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2672676799539504,
+      "learning_rate": 7.594925863374332e-06,
+      "loss": 0.1933,
+      "step": 14243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6283659782288595,
+      "learning_rate": 7.594589062888835e-06,
+      "loss": 0.2313,
+      "step": 14244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5136884134681177,
+      "learning_rate": 7.594252246291647e-06,
+      "loss": 0.346,
+      "step": 14245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8946103895509565,
+      "learning_rate": 7.5939154135848535e-06,
+      "loss": 0.4498,
+      "step": 14246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5060865352104398,
+      "learning_rate": 7.5935785647705495e-06,
+      "loss": 0.1431,
+      "step": 14247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3970910335173652,
+      "learning_rate": 7.593241699850826e-06,
+      "loss": 0.193,
+      "step": 14248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.505722098630299,
+      "learning_rate": 7.592904818827774e-06,
+      "loss": 0.2562,
+      "step": 14249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1179145852497157,
+      "learning_rate": 7.592567921703488e-06,
+      "loss": 0.3791,
+      "step": 14250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.645809864654491,
+      "learning_rate": 7.5922310084800565e-06,
+      "loss": 0.2092,
+      "step": 14251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5547003434000553,
+      "learning_rate": 7.591894079159574e-06,
+      "loss": 0.4063,
+      "step": 14252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0410992157603831,
+      "learning_rate": 7.5915571337441325e-06,
+      "loss": 0.1255,
+      "step": 14253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.602952957170342,
+      "learning_rate": 7.591220172235825e-06,
+      "loss": 0.2257,
+      "step": 14254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7763225178609727,
+      "learning_rate": 7.590883194636741e-06,
+      "loss": 0.2257,
+      "step": 14255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8676260149938757,
+      "learning_rate": 7.590546200948977e-06,
+      "loss": 0.2202,
+      "step": 14256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4333959437804848,
+      "learning_rate": 7.590209191174622e-06,
+      "loss": 0.2571,
+      "step": 14257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4749339674862043,
+      "learning_rate": 7.589872165315773e-06,
+      "loss": 0.3956,
+      "step": 14258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7117749955770438,
+      "learning_rate": 7.589535123374516e-06,
+      "loss": 0.2488,
+      "step": 14259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5269895225748147,
+      "learning_rate": 7.58919806535295e-06,
+      "loss": 0.3676,
+      "step": 14260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.261551134292572,
+      "learning_rate": 7.588860991253169e-06,
+      "loss": 0.3391,
+      "step": 14261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.360150103908459,
+      "learning_rate": 7.58852390107726e-06,
+      "loss": 0.1175,
+      "step": 14262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.941767607126652,
+      "learning_rate": 7.5881867948273214e-06,
+      "loss": 0.3214,
+      "step": 14263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7715666897963585,
+      "learning_rate": 7.5878496725054425e-06,
+      "loss": 0.4974,
+      "step": 14264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9021415289645573,
+      "learning_rate": 7.5875125341137215e-06,
+      "loss": 0.352,
+      "step": 14265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8259191572791473,
+      "learning_rate": 7.587175379654248e-06,
+      "loss": 0.4928,
+      "step": 14266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5735434865816853,
+      "learning_rate": 7.586838209129119e-06,
+      "loss": 0.3454,
+      "step": 14267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0003971361211863,
+      "learning_rate": 7.586501022540425e-06,
+      "loss": 0.4342,
+      "step": 14268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3445282972082844,
+      "learning_rate": 7.58616381989026e-06,
+      "loss": 0.3078,
+      "step": 14269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.41564327211356,
+      "learning_rate": 7.585826601180721e-06,
+      "loss": 0.2076,
+      "step": 14270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.457839036231863,
+      "learning_rate": 7.585489366413899e-06,
+      "loss": 0.3662,
+      "step": 14271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6743692048770995,
+      "learning_rate": 7.5851521155918895e-06,
+      "loss": 0.2336,
+      "step": 14272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1330354093055546,
+      "learning_rate": 7.584814848716786e-06,
+      "loss": 0.1423,
+      "step": 14273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9090282331330217,
+      "learning_rate": 7.584477565790685e-06,
+      "loss": 0.2001,
+      "step": 14274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5000712146152737,
+      "learning_rate": 7.584140266815676e-06,
+      "loss": 0.2486,
+      "step": 14275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0327372754756317,
+      "learning_rate": 7.583802951793859e-06,
+      "loss": 0.4051,
+      "step": 14276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7653523905625133,
+      "learning_rate": 7.583465620727326e-06,
+      "loss": 0.3447,
+      "step": 14277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7732042839872406,
+      "learning_rate": 7.583128273618175e-06,
+      "loss": 0.3326,
+      "step": 14278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9512212611488517,
+      "learning_rate": 7.582790910468494e-06,
+      "loss": 0.2772,
+      "step": 14279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4172732405081792,
+      "learning_rate": 7.582453531280382e-06,
+      "loss": 0.3378,
+      "step": 14280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0954303368382754,
+      "learning_rate": 7.582116136055937e-06,
+      "loss": 0.6709,
+      "step": 14281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5282126318684284,
+      "learning_rate": 7.581778724797249e-06,
+      "loss": 0.2845,
+      "step": 14282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3639228835650121,
+      "learning_rate": 7.581441297506414e-06,
+      "loss": 0.2225,
+      "step": 14283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9321528605108234,
+      "learning_rate": 7.581103854185532e-06,
+      "loss": 0.296,
+      "step": 14284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.747336280497738,
+      "learning_rate": 7.580766394836692e-06,
+      "loss": 0.3825,
+      "step": 14285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3938661854047638,
+      "learning_rate": 7.580428919461993e-06,
+      "loss": 0.2402,
+      "step": 14286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5211961262601492,
+      "learning_rate": 7.580091428063532e-06,
+      "loss": 0.2498,
+      "step": 14287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2420908182534587,
+      "learning_rate": 7.579753920643402e-06,
+      "loss": 0.1667,
+      "step": 14288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2435516313647412,
+      "learning_rate": 7.5794163972037e-06,
+      "loss": 0.2559,
+      "step": 14289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5926212539802451,
+      "learning_rate": 7.579078857746522e-06,
+      "loss": 0.1817,
+      "step": 14290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6523692803180354,
+      "learning_rate": 7.578741302273964e-06,
+      "loss": 0.2391,
+      "step": 14291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.662813839742504,
+      "learning_rate": 7.578403730788122e-06,
+      "loss": 0.4981,
+      "step": 14292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1600180248158034,
+      "learning_rate": 7.5780661432910905e-06,
+      "loss": 0.2082,
+      "step": 14293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6557360348724521,
+      "learning_rate": 7.577728539784969e-06,
+      "loss": 0.1354,
+      "step": 14294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9159932733101122,
+      "learning_rate": 7.577390920271851e-06,
+      "loss": 0.3068,
+      "step": 14295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6604038816779072,
+      "learning_rate": 7.577053284753836e-06,
+      "loss": 0.3553,
+      "step": 14296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9501254394422678,
+      "learning_rate": 7.576715633233018e-06,
+      "loss": 0.286,
+      "step": 14297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1803922190369187,
+      "learning_rate": 7.5763779657114946e-06,
+      "loss": 0.169,
+      "step": 14298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0595162092186803,
+      "learning_rate": 7.5760402821913636e-06,
+      "loss": 0.0997,
+      "step": 14299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9977031253893974,
+      "learning_rate": 7.57570258267472e-06,
+      "loss": 0.2029,
+      "step": 14300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0429918769271387,
+      "learning_rate": 7.575364867163663e-06,
+      "loss": 0.3276,
+      "step": 14301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7031426822549383,
+      "learning_rate": 7.575027135660288e-06,
+      "loss": 0.1375,
+      "step": 14302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0589679978627067,
+      "learning_rate": 7.574689388166694e-06,
+      "loss": 0.3652,
+      "step": 14303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6452001629306463,
+      "learning_rate": 7.574351624684975e-06,
+      "loss": 0.3777,
+      "step": 14304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.414791301363903,
+      "learning_rate": 7.574013845217232e-06,
+      "loss": 0.4141,
+      "step": 14305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8171107002882279,
+      "learning_rate": 7.573676049765562e-06,
+      "loss": 0.6198,
+      "step": 14306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.183738658708706,
+      "learning_rate": 7.57333823833206e-06,
+      "loss": 0.1474,
+      "step": 14307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1173355664655171,
+      "learning_rate": 7.573000410918825e-06,
+      "loss": 0.1384,
+      "step": 14308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3095966751779076,
+      "learning_rate": 7.572662567527956e-06,
+      "loss": 0.2428,
+      "step": 14309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1658539332448352,
+      "learning_rate": 7.572324708161552e-06,
+      "loss": 0.3096,
+      "step": 14310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3994961998653845,
+      "learning_rate": 7.571986832821707e-06,
+      "loss": 0.3448,
+      "step": 14311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3675058321244395,
+      "learning_rate": 7.571648941510521e-06,
+      "loss": 0.2139,
+      "step": 14312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5408214064749057,
+      "learning_rate": 7.571311034230095e-06,
+      "loss": 0.1979,
+      "step": 14313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.622837806025401,
+      "learning_rate": 7.570973110982523e-06,
+      "loss": 0.3825,
+      "step": 14314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7622761531812479,
+      "learning_rate": 7.570635171769905e-06,
+      "loss": 0.1944,
+      "step": 14315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.374168274490838,
+      "learning_rate": 7.570297216594341e-06,
+      "loss": 0.1756,
+      "step": 14316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7349696385780484,
+      "learning_rate": 7.569959245457927e-06,
+      "loss": 0.4248,
+      "step": 14317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.94083211073121,
+      "learning_rate": 7.569621258362765e-06,
+      "loss": 0.5342,
+      "step": 14318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.073015135193937,
+      "learning_rate": 7.5692832553109505e-06,
+      "loss": 0.1633,
+      "step": 14319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5473970967453152,
+      "learning_rate": 7.5689452363045836e-06,
+      "loss": 0.2087,
+      "step": 14320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.563589828219832,
+      "learning_rate": 7.568607201345763e-06,
+      "loss": 0.313,
+      "step": 14321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.070830946012025,
+      "learning_rate": 7.568269150436591e-06,
+      "loss": 0.5249,
+      "step": 14322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.430589575016447,
+      "learning_rate": 7.567931083579162e-06,
+      "loss": 0.1882,
+      "step": 14323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4651815177599439,
+      "learning_rate": 7.567593000775578e-06,
+      "loss": 0.1515,
+      "step": 14324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1102497335753982,
+      "learning_rate": 7.567254902027938e-06,
+      "loss": 0.4929,
+      "step": 14325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2877714107135414,
+      "learning_rate": 7.566916787338341e-06,
+      "loss": 0.3286,
+      "step": 14326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.016312502263561,
+      "learning_rate": 7.566578656708887e-06,
+      "loss": 0.4028,
+      "step": 14327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5373644804930768,
+      "learning_rate": 7.566240510141675e-06,
+      "loss": 0.1661,
+      "step": 14328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6774839900488243,
+      "learning_rate": 7.565902347638806e-06,
+      "loss": 0.375,
+      "step": 14329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4757561626493059,
+      "learning_rate": 7.56556416920238e-06,
+      "loss": 0.1753,
+      "step": 14330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.767033179988173,
+      "learning_rate": 7.565225974834495e-06,
+      "loss": 0.4298,
+      "step": 14331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8869469704032285,
+      "learning_rate": 7.564887764537252e-06,
+      "loss": 0.6329,
+      "step": 14332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.990608425385776,
+      "learning_rate": 7.5645495383127534e-06,
+      "loss": 0.3348,
+      "step": 14333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2477821567833716,
+      "learning_rate": 7.564211296163097e-06,
+      "loss": 0.149,
+      "step": 14334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4677043220861383,
+      "learning_rate": 7.563873038090382e-06,
+      "loss": 0.361,
+      "step": 14335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7595305608645229,
+      "learning_rate": 7.563534764096714e-06,
+      "loss": 0.2665,
+      "step": 14336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6288648961441552,
+      "learning_rate": 7.563196474184188e-06,
+      "loss": 0.4091,
+      "step": 14337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7174840339601516,
+      "learning_rate": 7.562858168354908e-06,
+      "loss": 0.4296,
+      "step": 14338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7209742675602473,
+      "learning_rate": 7.5625198466109736e-06,
+      "loss": 0.2207,
+      "step": 14339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6032757225346566,
+      "learning_rate": 7.562181508954487e-06,
+      "loss": 0.3221,
+      "step": 14340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8979856322230075,
+      "learning_rate": 7.561843155387546e-06,
+      "loss": 0.2896,
+      "step": 14341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2428626064317372,
+      "learning_rate": 7.561504785912256e-06,
+      "loss": 0.2349,
+      "step": 14342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.820279771895338,
+      "learning_rate": 7.561166400530714e-06,
+      "loss": 0.46,
+      "step": 14343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6829371001533218,
+      "learning_rate": 7.560827999245025e-06,
+      "loss": 0.4548,
+      "step": 14344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.383364772894929,
+      "learning_rate": 7.560489582057289e-06,
+      "loss": 0.1988,
+      "step": 14345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.612162992187988,
+      "learning_rate": 7.560151148969606e-06,
+      "loss": 0.3522,
+      "step": 14346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7728191864900253,
+      "learning_rate": 7.559812699984078e-06,
+      "loss": 0.3112,
+      "step": 14347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9398012396263795,
+      "learning_rate": 7.559474235102809e-06,
+      "loss": 0.4121,
+      "step": 14348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7957407221946122,
+      "learning_rate": 7.5591357543278975e-06,
+      "loss": 0.2524,
+      "step": 14349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9669467315860831,
+      "learning_rate": 7.558797257661448e-06,
+      "loss": 0.2988,
+      "step": 14350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6111207192769783,
+      "learning_rate": 7.558458745105561e-06,
+      "loss": 0.4262,
+      "step": 14351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3535431753109748,
+      "learning_rate": 7.558120216662339e-06,
+      "loss": 0.1769,
+      "step": 14352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6012841773685054,
+      "learning_rate": 7.557781672333885e-06,
+      "loss": 0.3483,
+      "step": 14353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2551589555525235,
+      "learning_rate": 7.557443112122299e-06,
+      "loss": 0.4653,
+      "step": 14354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6317266842003464,
+      "learning_rate": 7.557104536029687e-06,
+      "loss": 0.2812,
+      "step": 14355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2714149101607461,
+      "learning_rate": 7.556765944058148e-06,
+      "loss": 0.2083,
+      "step": 14356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4944564684969504,
+      "learning_rate": 7.556427336209787e-06,
+      "loss": 0.2483,
+      "step": 14357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6360875508119288,
+      "learning_rate": 7.5560887124867035e-06,
+      "loss": 0.3501,
+      "step": 14358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4335958288525885,
+      "learning_rate": 7.555750072891004e-06,
+      "loss": 0.2248,
+      "step": 14359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2550508497962705,
+      "learning_rate": 7.5554114174247895e-06,
+      "loss": 0.1742,
+      "step": 14360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.716669700694721,
+      "learning_rate": 7.5550727460901615e-06,
+      "loss": 0.1884,
+      "step": 14361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5975791847277347,
+      "learning_rate": 7.554734058889226e-06,
+      "loss": 0.3932,
+      "step": 14362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7159722075862651,
+      "learning_rate": 7.5543953558240845e-06,
+      "loss": 0.3046,
+      "step": 14363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1292600209395558,
+      "learning_rate": 7.554056636896843e-06,
+      "loss": 0.1207,
+      "step": 14364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.361652442359499,
+      "learning_rate": 7.553717902109599e-06,
+      "loss": 0.2717,
+      "step": 14365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5283158079818302,
+      "learning_rate": 7.553379151464462e-06,
+      "loss": 0.4792,
+      "step": 14366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5256995402390747,
+      "learning_rate": 7.5530403849635325e-06,
+      "loss": 0.2048,
+      "step": 14367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2375806818294464,
+      "learning_rate": 7.5527016026089136e-06,
+      "loss": 0.2694,
+      "step": 14368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2507245258809552,
+      "learning_rate": 7.552362804402711e-06,
+      "loss": 0.2151,
+      "step": 14369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6576716275730927,
+      "learning_rate": 7.552023990347026e-06,
+      "loss": 0.313,
+      "step": 14370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8096999705060681,
+      "learning_rate": 7.551685160443967e-06,
+      "loss": 0.6224,
+      "step": 14371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0884959072386184,
+      "learning_rate": 7.551346314695633e-06,
+      "loss": 0.3857,
+      "step": 14372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6423895701572058,
+      "learning_rate": 7.551007453104131e-06,
+      "loss": 0.3869,
+      "step": 14373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8509739290407676,
+      "learning_rate": 7.550668575671563e-06,
+      "loss": 0.4754,
+      "step": 14374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6820234379141792,
+      "learning_rate": 7.550329682400037e-06,
+      "loss": 0.3826,
+      "step": 14375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9825937732151593,
+      "learning_rate": 7.549990773291655e-06,
+      "loss": 0.4039,
+      "step": 14376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3732533306461308,
+      "learning_rate": 7.549651848348521e-06,
+      "loss": 0.3444,
+      "step": 14377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3631673235140565,
+      "learning_rate": 7.549312907572741e-06,
+      "loss": 0.2932,
+      "step": 14378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5593648427620985,
+      "learning_rate": 7.5489739509664195e-06,
+      "loss": 0.4042,
+      "step": 14379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8012120048207239,
+      "learning_rate": 7.548634978531662e-06,
+      "loss": 0.4668,
+      "step": 14380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1090824560119144,
+      "learning_rate": 7.54829599027057e-06,
+      "loss": 0.5816,
+      "step": 14381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2169178264259586,
+      "learning_rate": 7.547956986185253e-06,
+      "loss": 0.1802,
+      "step": 14382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7793462727595686,
+      "learning_rate": 7.547617966277814e-06,
+      "loss": 0.3935,
+      "step": 14383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.108478489476309,
+      "learning_rate": 7.547278930550357e-06,
+      "loss": 0.1759,
+      "step": 14384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2638688252752224,
+      "learning_rate": 7.546939879004988e-06,
+      "loss": 0.4888,
+      "step": 14385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3278837770295673,
+      "learning_rate": 7.546600811643816e-06,
+      "loss": 0.1899,
+      "step": 14386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4983881865591573,
+      "learning_rate": 7.546261728468941e-06,
+      "loss": 0.2283,
+      "step": 14387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6215136064740183,
+      "learning_rate": 7.545922629482474e-06,
+      "loss": 0.303,
+      "step": 14388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.207114499542645,
+      "learning_rate": 7.545583514686516e-06,
+      "loss": 0.5364,
+      "step": 14389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1587305848458476,
+      "learning_rate": 7.545244384083175e-06,
+      "loss": 0.4169,
+      "step": 14390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2984265057870679,
+      "learning_rate": 7.5449052376745565e-06,
+      "loss": 0.186,
+      "step": 14391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5552616005877868,
+      "learning_rate": 7.544566075462767e-06,
+      "loss": 0.2124,
+      "step": 14392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6243211492834895,
+      "learning_rate": 7.544226897449914e-06,
+      "loss": 0.439,
+      "step": 14393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6889058103057037,
+      "learning_rate": 7.5438877036380986e-06,
+      "loss": 0.4474,
+      "step": 14394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.147203876744924,
+      "learning_rate": 7.543548494029433e-06,
+      "loss": 0.4508,
+      "step": 14395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0853582662416899,
+      "learning_rate": 7.543209268626019e-06,
+      "loss": 0.2544,
+      "step": 14396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7651949565560991,
+      "learning_rate": 7.5428700274299675e-06,
+      "loss": 0.3605,
+      "step": 14397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6422968112123462,
+      "learning_rate": 7.542530770443382e-06,
+      "loss": 0.2917,
+      "step": 14398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3515032799376128,
+      "learning_rate": 7.5421914976683695e-06,
+      "loss": 0.2061,
+      "step": 14399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4916952616987174,
+      "learning_rate": 7.541852209107038e-06,
+      "loss": 0.2183,
+      "step": 14400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.34264529223628,
+      "learning_rate": 7.541512904761492e-06,
+      "loss": 0.278,
+      "step": 14401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0232536129261085,
+      "learning_rate": 7.541173584633843e-06,
+      "loss": 0.3649,
+      "step": 14402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.304745353055647,
+      "learning_rate": 7.540834248726192e-06,
+      "loss": 0.3121,
+      "step": 14403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4809142074153827,
+      "learning_rate": 7.540494897040651e-06,
+      "loss": 0.1937,
+      "step": 14404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5053597696368999,
+      "learning_rate": 7.540155529579325e-06,
+      "loss": 0.2957,
+      "step": 14405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.370682643868907,
+      "learning_rate": 7.539816146344323e-06,
+      "loss": 0.2728,
+      "step": 14406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9938173396299117,
+      "learning_rate": 7.539476747337752e-06,
+      "loss": 0.3233,
+      "step": 14407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7541704637285518,
+      "learning_rate": 7.539137332561719e-06,
+      "loss": 0.4706,
+      "step": 14408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6901121056070965,
+      "learning_rate": 7.53879790201833e-06,
+      "loss": 0.3309,
+      "step": 14409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7148044816302506,
+      "learning_rate": 7.5384584557096955e-06,
+      "loss": 0.5269,
+      "step": 14410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5435473166165115,
+      "learning_rate": 7.538118993637923e-06,
+      "loss": 0.3525,
+      "step": 14411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6136216902426301,
+      "learning_rate": 7.53777951580512e-06,
+      "loss": 0.19,
+      "step": 14412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8574480762900034,
+      "learning_rate": 7.537440022213394e-06,
+      "loss": 0.4847,
+      "step": 14413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9227064164122125,
+      "learning_rate": 7.537100512864852e-06,
+      "loss": 0.5673,
+      "step": 14414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6109254021657455,
+      "learning_rate": 7.536760987761607e-06,
+      "loss": 0.2987,
+      "step": 14415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.572138540688099,
+      "learning_rate": 7.536421446905762e-06,
+      "loss": 0.2565,
+      "step": 14416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.033336052410726,
+      "learning_rate": 7.5360818902994294e-06,
+      "loss": 0.7248,
+      "step": 14417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8978147623471704,
+      "learning_rate": 7.535742317944714e-06,
+      "loss": 0.3301,
+      "step": 14418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2929284501494092,
+      "learning_rate": 7.535402729843729e-06,
+      "loss": 0.2148,
+      "step": 14419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.662724122378456,
+      "learning_rate": 7.535063125998578e-06,
+      "loss": 0.3547,
+      "step": 14420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.738805383249964,
+      "learning_rate": 7.534723506411375e-06,
+      "loss": 0.3879,
+      "step": 14421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2948058013780845,
+      "learning_rate": 7.534383871084225e-06,
+      "loss": 0.4204,
+      "step": 14422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4391827892349458,
+      "learning_rate": 7.534044220019239e-06,
+      "loss": 0.2163,
+      "step": 14423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6210734906137108,
+      "learning_rate": 7.533704553218526e-06,
+      "loss": 0.328,
+      "step": 14424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5308414590560557,
+      "learning_rate": 7.533364870684195e-06,
+      "loss": 0.1987,
+      "step": 14425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.576122297336858,
+      "learning_rate": 7.533025172418354e-06,
+      "loss": 0.3158,
+      "step": 14426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.86498301145471,
+      "learning_rate": 7.5326854584231144e-06,
+      "loss": 0.516,
+      "step": 14427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5943533746321223,
+      "learning_rate": 7.532345728700585e-06,
+      "loss": 0.3215,
+      "step": 14428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7947671510177405,
+      "learning_rate": 7.532005983252874e-06,
+      "loss": 0.1787,
+      "step": 14429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.0409060021304875,
+      "learning_rate": 7.531666222082094e-06,
+      "loss": 0.4867,
+      "step": 14430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3574836607120084,
+      "learning_rate": 7.531326445190355e-06,
+      "loss": 0.3914,
+      "step": 14431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6403737148215054,
+      "learning_rate": 7.5309866525797626e-06,
+      "loss": 0.3664,
+      "step": 14432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.680233083876929,
+      "learning_rate": 7.53064684425243e-06,
+      "loss": 0.3126,
+      "step": 14433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7801417784938451,
+      "learning_rate": 7.5303070202104675e-06,
+      "loss": 0.4215,
+      "step": 14434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7797082285222314,
+      "learning_rate": 7.529967180455984e-06,
+      "loss": 0.5601,
+      "step": 14435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6887984843180202,
+      "learning_rate": 7.5296273249910915e-06,
+      "loss": 0.3464,
+      "step": 14436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7706003085219735,
+      "learning_rate": 7.529287453817899e-06,
+      "loss": 0.132,
+      "step": 14437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.217933940212963,
+      "learning_rate": 7.528947566938516e-06,
+      "loss": 0.5443,
+      "step": 14438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.228873128074805,
+      "learning_rate": 7.528607664355057e-06,
+      "loss": 0.172,
+      "step": 14439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2360680790269098,
+      "learning_rate": 7.528267746069627e-06,
+      "loss": 0.1784,
+      "step": 14440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.543779174822443,
+      "learning_rate": 7.5279278120843425e-06,
+      "loss": 0.4343,
+      "step": 14441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7002584718457543,
+      "learning_rate": 7.527587862401311e-06,
+      "loss": 0.3893,
+      "step": 14442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.648610209958713,
+      "learning_rate": 7.527247897022645e-06,
+      "loss": 0.2352,
+      "step": 14443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9536857601070037,
+      "learning_rate": 7.526907915950455e-06,
+      "loss": 0.5996,
+      "step": 14444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6249728475649707,
+      "learning_rate": 7.5265679191868515e-06,
+      "loss": 0.3009,
+      "step": 14445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.018427486315107,
+      "learning_rate": 7.526227906733946e-06,
+      "loss": 0.1956,
+      "step": 14446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5850101728895065,
+      "learning_rate": 7.52588787859385e-06,
+      "loss": 0.2573,
+      "step": 14447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5109671269286375,
+      "learning_rate": 7.525547834768677e-06,
+      "loss": 0.2066,
+      "step": 14448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4065607840123502,
+      "learning_rate": 7.525207775260537e-06,
+      "loss": 0.2132,
+      "step": 14449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.446072322000426,
+      "learning_rate": 7.524867700071541e-06,
+      "loss": 0.2693,
+      "step": 14450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.994116152502048,
+      "learning_rate": 7.5245276092038e-06,
+      "loss": 0.3123,
+      "step": 14451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.600433605362755,
+      "learning_rate": 7.524187502659428e-06,
+      "loss": 0.4804,
+      "step": 14452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7724044496556806,
+      "learning_rate": 7.523847380440536e-06,
+      "loss": 0.3034,
+      "step": 14453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.856474300762207,
+      "learning_rate": 7.523507242549237e-06,
+      "loss": 0.3392,
+      "step": 14454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2986247053967608,
+      "learning_rate": 7.523167088987644e-06,
+      "loss": 0.2458,
+      "step": 14455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.04707827541483,
+      "learning_rate": 7.522826919757864e-06,
+      "loss": 0.2399,
+      "step": 14456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5273748721491,
+      "learning_rate": 7.522486734862016e-06,
+      "loss": 0.2071,
+      "step": 14457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4556905315392663,
+      "learning_rate": 7.522146534302209e-06,
+      "loss": 0.1956,
+      "step": 14458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9804849851259596,
+      "learning_rate": 7.521806318080554e-06,
+      "loss": 0.1927,
+      "step": 14459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.564511992148281,
+      "learning_rate": 7.521466086199167e-06,
+      "loss": 0.1913,
+      "step": 14460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.076154305068824,
+      "learning_rate": 7.5211258386601595e-06,
+      "loss": 0.5281,
+      "step": 14461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8793880747293853,
+      "learning_rate": 7.520785575465644e-06,
+      "loss": 0.1508,
+      "step": 14462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2900072306099526,
+      "learning_rate": 7.520445296617734e-06,
+      "loss": 0.2881,
+      "step": 14463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7196761886956702,
+      "learning_rate": 7.520105002118543e-06,
+      "loss": 0.2181,
+      "step": 14464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6752692732084438,
+      "learning_rate": 7.5197646919701826e-06,
+      "loss": 0.6799,
+      "step": 14465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9532625455107995,
+      "learning_rate": 7.5194243661747675e-06,
+      "loss": 0.3787,
+      "step": 14466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9590818584372276,
+      "learning_rate": 7.519084024734409e-06,
+      "loss": 0.6293,
+      "step": 14467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4391091943883643,
+      "learning_rate": 7.518743667651224e-06,
+      "loss": 0.4792,
+      "step": 14468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.924722438469746,
+      "learning_rate": 7.5184032949273225e-06,
+      "loss": 0.4154,
+      "step": 14469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8182558015853802,
+      "learning_rate": 7.51806290656482e-06,
+      "loss": 0.5544,
+      "step": 14470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9300872344307944,
+      "learning_rate": 7.51772250256583e-06,
+      "loss": 0.3413,
+      "step": 14471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7121745083896156,
+      "learning_rate": 7.517382082932465e-06,
+      "loss": 0.342,
+      "step": 14472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.312722119201104,
+      "learning_rate": 7.51704164766684e-06,
+      "loss": 0.3048,
+      "step": 14473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7196483648437195,
+      "learning_rate": 7.51670119677107e-06,
+      "loss": 0.4301,
+      "step": 14474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4029276868669704,
+      "learning_rate": 7.516360730247268e-06,
+      "loss": 0.1867,
+      "step": 14475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7414040664310038,
+      "learning_rate": 7.516020248097549e-06,
+      "loss": 0.3729,
+      "step": 14476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0539369031493515,
+      "learning_rate": 7.515679750324025e-06,
+      "loss": 0.2127,
+      "step": 14477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5563263617781755,
+      "learning_rate": 7.515339236928813e-06,
+      "loss": 0.4327,
+      "step": 14478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.821225342645817,
+      "learning_rate": 7.514998707914026e-06,
+      "loss": 0.3519,
+      "step": 14479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4586049547873512,
+      "learning_rate": 7.514658163281779e-06,
+      "loss": 0.1907,
+      "step": 14480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7207923285822961,
+      "learning_rate": 7.514317603034188e-06,
+      "loss": 0.4922,
+      "step": 14481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2884364760984859,
+      "learning_rate": 7.513977027173365e-06,
+      "loss": 0.1506,
+      "step": 14482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5932885149316602,
+      "learning_rate": 7.513636435701428e-06,
+      "loss": 0.2215,
+      "step": 14483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9119970115245224,
+      "learning_rate": 7.513295828620488e-06,
+      "loss": 0.5311,
+      "step": 14484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0319252652369624,
+      "learning_rate": 7.5129552059326645e-06,
+      "loss": 0.6541,
+      "step": 14485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2807310779147982,
+      "learning_rate": 7.5126145676400705e-06,
+      "loss": 0.2091,
+      "step": 14486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5756778557350506,
+      "learning_rate": 7.512273913744821e-06,
+      "loss": 0.2166,
+      "step": 14487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.427934224073314,
+      "learning_rate": 7.5119332442490335e-06,
+      "loss": 0.2242,
+      "step": 14488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5461657588780529,
+      "learning_rate": 7.51159255915482e-06,
+      "loss": 0.4503,
+      "step": 14489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2143607237256955,
+      "learning_rate": 7.5112518584642995e-06,
+      "loss": 0.2265,
+      "step": 14490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9528832553029043,
+      "learning_rate": 7.510911142179584e-06,
+      "loss": 0.5782,
+      "step": 14491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.540610974636808,
+      "learning_rate": 7.510570410302793e-06,
+      "loss": 0.3725,
+      "step": 14492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9377682023382647,
+      "learning_rate": 7.510229662836039e-06,
+      "loss": 0.3186,
+      "step": 14493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.523514129472056,
+      "learning_rate": 7.509888899781441e-06,
+      "loss": 0.2162,
+      "step": 14494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.714919640919032,
+      "learning_rate": 7.5095481211411125e-06,
+      "loss": 0.2153,
+      "step": 14495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8041209698176588,
+      "learning_rate": 7.509207326917172e-06,
+      "loss": 0.4676,
+      "step": 14496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.703280192487623,
+      "learning_rate": 7.508866517111733e-06,
+      "loss": 0.3993,
+      "step": 14497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7615041940468887,
+      "learning_rate": 7.5085256917269155e-06,
+      "loss": 0.4177,
+      "step": 14498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8503309250313527,
+      "learning_rate": 7.508184850764833e-06,
+      "loss": 0.6745,
+      "step": 14499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4354513694497413,
+      "learning_rate": 7.507843994227602e-06,
+      "loss": 0.2649,
+      "step": 14500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4701110007544989,
+      "learning_rate": 7.507503122117341e-06,
+      "loss": 0.2118,
+      "step": 14501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1244549332720686,
+      "learning_rate": 7.507162234436164e-06,
+      "loss": 0.4107,
+      "step": 14502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.530581279627122,
+      "learning_rate": 7.506821331186192e-06,
+      "loss": 0.3636,
+      "step": 14503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5048576342972313,
+      "learning_rate": 7.5064804123695365e-06,
+      "loss": 0.3434,
+      "step": 14504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7617220149041686,
+      "learning_rate": 7.506139477988319e-06,
+      "loss": 0.4119,
+      "step": 14505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6640933038430459,
+      "learning_rate": 7.505798528044655e-06,
+      "loss": 0.3221,
+      "step": 14506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0223407998137923,
+      "learning_rate": 7.5054575625406615e-06,
+      "loss": 0.2387,
+      "step": 14507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.549249940366807,
+      "learning_rate": 7.505116581478455e-06,
+      "loss": 0.1569,
+      "step": 14508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9364093010105072,
+      "learning_rate": 7.504775584860155e-06,
+      "loss": 0.3864,
+      "step": 14509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5313434961541128,
+      "learning_rate": 7.504434572687879e-06,
+      "loss": 0.1482,
+      "step": 14510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2697920953021637,
+      "learning_rate": 7.504093544963742e-06,
+      "loss": 0.4066,
+      "step": 14511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8119411017869718,
+      "learning_rate": 7.503752501689864e-06,
+      "loss": 0.3535,
+      "step": 14512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2047431084431213,
+      "learning_rate": 7.50341144286836e-06,
+      "loss": 0.2401,
+      "step": 14513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5525208044424226,
+      "learning_rate": 7.503070368501353e-06,
+      "loss": 0.5517,
+      "step": 14514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7499505955670596,
+      "learning_rate": 7.502729278590957e-06,
+      "loss": 0.4589,
+      "step": 14515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.636472426294518,
+      "learning_rate": 7.50238817313929e-06,
+      "loss": 0.3431,
+      "step": 14516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.300863091583377,
+      "learning_rate": 7.502047052148471e-06,
+      "loss": 0.2678,
+      "step": 14517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8549974666177154,
+      "learning_rate": 7.5017059156206186e-06,
+      "loss": 0.499,
+      "step": 14518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6798226656977489,
+      "learning_rate": 7.501364763557851e-06,
+      "loss": 0.36,
+      "step": 14519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7138727639045093,
+      "learning_rate": 7.501023595962288e-06,
+      "loss": 0.2368,
+      "step": 14520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.687019844974097,
+      "learning_rate": 7.500682412836045e-06,
+      "loss": 0.1903,
+      "step": 14521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2484482805366084,
+      "learning_rate": 7.500341214181244e-06,
+      "loss": 0.4276,
+      "step": 14522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7959221925974236,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.4396,
+      "step": 14523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7108041850716356,
+      "learning_rate": 7.499658770294436e-06,
+      "loss": 0.3654,
+      "step": 14524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.317180589355638,
+      "learning_rate": 7.4993175250666686e-06,
+      "loss": 0.4019,
+      "step": 14525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3359777628536669,
+      "learning_rate": 7.4989762643188156e-06,
+      "loss": 0.3227,
+      "step": 14526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5335777796857712,
+      "learning_rate": 7.498634988053e-06,
+      "loss": 0.3343,
+      "step": 14527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8590595194102888,
+      "learning_rate": 7.498293696271337e-06,
+      "loss": 0.2653,
+      "step": 14528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6124768133492473,
+      "learning_rate": 7.4979523889759485e-06,
+      "loss": 0.3982,
+      "step": 14529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.724517152335826,
+      "learning_rate": 7.497611066168953e-06,
+      "loss": 0.3453,
+      "step": 14530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7876534806229991,
+      "learning_rate": 7.497269727852471e-06,
+      "loss": 0.4084,
+      "step": 14531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5119711790838588,
+      "learning_rate": 7.496928374028621e-06,
+      "loss": 0.4583,
+      "step": 14532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.245108648800801,
+      "learning_rate": 7.496587004699522e-06,
+      "loss": 0.3755,
+      "step": 14533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7322510631866044,
+      "learning_rate": 7.496245619867295e-06,
+      "loss": 0.2204,
+      "step": 14534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.143618078315509,
+      "learning_rate": 7.49590421953406e-06,
+      "loss": 0.4092,
+      "step": 14535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.97160762449229,
+      "learning_rate": 7.495562803701937e-06,
+      "loss": 0.3306,
+      "step": 14536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9667269174356665,
+      "learning_rate": 7.495221372373045e-06,
+      "loss": 0.5795,
+      "step": 14537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6842498784135276,
+      "learning_rate": 7.494879925549505e-06,
+      "loss": 0.4579,
+      "step": 14538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6915180288297544,
+      "learning_rate": 7.494538463233438e-06,
+      "loss": 0.1683,
+      "step": 14539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5933550831558547,
+      "learning_rate": 7.494196985426964e-06,
+      "loss": 0.2788,
+      "step": 14540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.162893284136955,
+      "learning_rate": 7.493855492132202e-06,
+      "loss": 0.1941,
+      "step": 14541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.82030629905987,
+      "learning_rate": 7.493513983351275e-06,
+      "loss": 0.2922,
+      "step": 14542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.60907172382487,
+      "learning_rate": 7.493172459086302e-06,
+      "loss": 0.3173,
+      "step": 14543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.797480912259745,
+      "learning_rate": 7.492830919339403e-06,
+      "loss": 0.1507,
+      "step": 14544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3416923485012289,
+      "learning_rate": 7.492489364112703e-06,
+      "loss": 0.2939,
+      "step": 14545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5965969099233157,
+      "learning_rate": 7.492147793408316e-06,
+      "loss": 0.3676,
+      "step": 14546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.438672824738122,
+      "learning_rate": 7.49180620722837e-06,
+      "loss": 0.2158,
+      "step": 14547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3608270267448284,
+      "learning_rate": 7.4914646055749825e-06,
+      "loss": 0.3234,
+      "step": 14548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0027764055734867,
+      "learning_rate": 7.491122988450276e-06,
+      "loss": 0.2788,
+      "step": 14549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6412444852473398,
+      "learning_rate": 7.49078135585637e-06,
+      "loss": 0.3972,
+      "step": 14550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3970660832086557,
+      "learning_rate": 7.49043970779539e-06,
+      "loss": 0.3383,
+      "step": 14551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8623124689060626,
+      "learning_rate": 7.490098044269451e-06,
+      "loss": 0.4191,
+      "step": 14552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4166640645124808,
+      "learning_rate": 7.489756365280681e-06,
+      "loss": 0.2446,
+      "step": 14553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7636301303963025,
+      "learning_rate": 7.489414670831198e-06,
+      "loss": 0.2428,
+      "step": 14554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.464971002773479,
+      "learning_rate": 7.489072960923127e-06,
+      "loss": 0.3275,
+      "step": 14555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.06189515274328,
+      "learning_rate": 7.488731235558586e-06,
+      "loss": 0.2159,
+      "step": 14556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7822088872999817,
+      "learning_rate": 7.488389494739701e-06,
+      "loss": 0.4865,
+      "step": 14557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2799771676807148,
+      "learning_rate": 7.48804773846859e-06,
+      "loss": 0.2444,
+      "step": 14558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1449272785560332,
+      "learning_rate": 7.487705966747378e-06,
+      "loss": 0.2902,
+      "step": 14559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7153399190658913,
+      "learning_rate": 7.487364179578187e-06,
+      "loss": 0.356,
+      "step": 14560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2555132191585727,
+      "learning_rate": 7.487022376963138e-06,
+      "loss": 0.325,
+      "step": 14561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5926093830356138,
+      "learning_rate": 7.486680558904356e-06,
+      "loss": 0.2102,
+      "step": 14562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4107151205293642,
+      "learning_rate": 7.486338725403961e-06,
+      "loss": 0.3049,
+      "step": 14563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3192727210022193,
+      "learning_rate": 7.485996876464077e-06,
+      "loss": 0.3043,
+      "step": 14564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0807488585213927,
+      "learning_rate": 7.485655012086828e-06,
+      "loss": 0.1821,
+      "step": 14565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1238477985621116,
+      "learning_rate": 7.4853131322743336e-06,
+      "loss": 0.2936,
+      "step": 14566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4674083232756538,
+      "learning_rate": 7.4849712370287196e-06,
+      "loss": 0.1906,
+      "step": 14567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4979583559902183,
+      "learning_rate": 7.4846293263521075e-06,
+      "loss": 0.2606,
+      "step": 14568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6436750443850148,
+      "learning_rate": 7.484287400246623e-06,
+      "loss": 0.3869,
+      "step": 14569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5566835289837597,
+      "learning_rate": 7.483945458714386e-06,
+      "loss": 0.2151,
+      "step": 14570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5714319897899354,
+      "learning_rate": 7.483603501757522e-06,
+      "loss": 0.3205,
+      "step": 14571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.162586063122246,
+      "learning_rate": 7.483261529378152e-06,
+      "loss": 0.4678,
+      "step": 14572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6445305047294034,
+      "learning_rate": 7.482919541578404e-06,
+      "loss": 0.2801,
+      "step": 14573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3414657772855136,
+      "learning_rate": 7.482577538360398e-06,
+      "loss": 0.1386,
+      "step": 14574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.609081195197421,
+      "learning_rate": 7.482235519726258e-06,
+      "loss": 0.3262,
+      "step": 14575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0657944269281567,
+      "learning_rate": 7.481893485678109e-06,
+      "loss": 0.4377,
+      "step": 14576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7963490172419554,
+      "learning_rate": 7.4815514362180755e-06,
+      "loss": 0.4262,
+      "step": 14577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.284897818858274,
+      "learning_rate": 7.4812093713482795e-06,
+      "loss": 0.1866,
+      "step": 14578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.547423509922494,
+      "learning_rate": 7.480867291070846e-06,
+      "loss": 0.2325,
+      "step": 14579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7976852704772293,
+      "learning_rate": 7.480525195387901e-06,
+      "loss": 0.3404,
+      "step": 14580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6659283095238497,
+      "learning_rate": 7.480183084301566e-06,
+      "loss": 0.4054,
+      "step": 14581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2224445845611367,
+      "learning_rate": 7.4798409578139675e-06,
+      "loss": 0.2056,
+      "step": 14582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.718292713021247,
+      "learning_rate": 7.479498815927229e-06,
+      "loss": 0.391,
+      "step": 14583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0830177846992677,
+      "learning_rate": 7.479156658643474e-06,
+      "loss": 0.6584,
+      "step": 14584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5883617124614877,
+      "learning_rate": 7.478814485964829e-06,
+      "loss": 0.4184,
+      "step": 14585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.629882236524038,
+      "learning_rate": 7.47847229789342e-06,
+      "loss": 0.402,
+      "step": 14586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6173540135594993,
+      "learning_rate": 7.478130094431368e-06,
+      "loss": 0.2262,
+      "step": 14587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.637285166399591,
+      "learning_rate": 7.477787875580802e-06,
+      "loss": 0.1871,
+      "step": 14588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4468497375029548,
+      "learning_rate": 7.477445641343844e-06,
+      "loss": 0.1793,
+      "step": 14589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6675285812742326,
+      "learning_rate": 7.477103391722621e-06,
+      "loss": 0.3488,
+      "step": 14590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3359118397797054,
+      "learning_rate": 7.476761126719259e-06,
+      "loss": 0.2133,
+      "step": 14591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7252114563718561,
+      "learning_rate": 7.47641884633588e-06,
+      "loss": 0.2407,
+      "step": 14592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9740679948653899,
+      "learning_rate": 7.476076550574614e-06,
+      "loss": 0.2241,
+      "step": 14593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.288899844918237,
+      "learning_rate": 7.475734239437581e-06,
+      "loss": 0.2045,
+      "step": 14594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5844504567899669,
+      "learning_rate": 7.475391912926912e-06,
+      "loss": 0.4281,
+      "step": 14595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1019587054672133,
+      "learning_rate": 7.47504957104473e-06,
+      "loss": 0.5147,
+      "step": 14596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7742187001956073,
+      "learning_rate": 7.474707213793162e-06,
+      "loss": 0.4944,
+      "step": 14597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1600243207626375,
+      "learning_rate": 7.474364841174333e-06,
+      "loss": 0.2049,
+      "step": 14598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2459489833961819,
+      "learning_rate": 7.474022453190368e-06,
+      "loss": 0.1313,
+      "step": 14599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.671252326343147,
+      "learning_rate": 7.473680049843397e-06,
+      "loss": 0.4477,
+      "step": 14600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6456227021902161,
+      "learning_rate": 7.473337631135542e-06,
+      "loss": 0.4255,
+      "step": 14601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8424121601602559,
+      "learning_rate": 7.472995197068933e-06,
+      "loss": 0.6107,
+      "step": 14602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.857948858231459,
+      "learning_rate": 7.472652747645693e-06,
+      "loss": 0.3094,
+      "step": 14603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.806606972870704,
+      "learning_rate": 7.47231028286795e-06,
+      "loss": 0.4086,
+      "step": 14604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7297030858282714,
+      "learning_rate": 7.47196780273783e-06,
+      "loss": 0.239,
+      "step": 14605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8159373890034056,
+      "learning_rate": 7.471625307257462e-06,
+      "loss": 0.1294,
+      "step": 14606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8156154572405165,
+      "learning_rate": 7.471282796428971e-06,
+      "loss": 0.2735,
+      "step": 14607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5274585055146614,
+      "learning_rate": 7.470940270254483e-06,
+      "loss": 0.2173,
+      "step": 14608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.438837579717961,
+      "learning_rate": 7.470597728736126e-06,
+      "loss": 0.3561,
+      "step": 14609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3720029929808928,
+      "learning_rate": 7.470255171876028e-06,
+      "loss": 0.214,
+      "step": 14610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5680390787436855,
+      "learning_rate": 7.469912599676314e-06,
+      "loss": 0.3059,
+      "step": 14611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8107643243767408,
+      "learning_rate": 7.469570012139115e-06,
+      "loss": 0.2972,
+      "step": 14612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.620856951458335,
+      "learning_rate": 7.469227409266554e-06,
+      "loss": 0.1905,
+      "step": 14613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2432878991452958,
+      "learning_rate": 7.468884791060761e-06,
+      "loss": 0.2464,
+      "step": 14614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5687461770341653,
+      "learning_rate": 7.468542157523864e-06,
+      "loss": 0.334,
+      "step": 14615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3537837281785052,
+      "learning_rate": 7.468199508657987e-06,
+      "loss": 0.252,
+      "step": 14616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.622502046162795,
+      "learning_rate": 7.467856844465263e-06,
+      "loss": 0.3873,
+      "step": 14617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7352457781714086,
+      "learning_rate": 7.467514164947817e-06,
+      "loss": 0.2523,
+      "step": 14618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6162968938276254,
+      "learning_rate": 7.467171470107777e-06,
+      "loss": 0.2842,
+      "step": 14619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2695753926158826,
+      "learning_rate": 7.466828759947271e-06,
+      "loss": 0.2064,
+      "step": 14620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7616292588199995,
+      "learning_rate": 7.466486034468427e-06,
+      "loss": 0.3995,
+      "step": 14621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3534929155258388,
+      "learning_rate": 7.466143293673376e-06,
+      "loss": 0.2245,
+      "step": 14622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0375046241450345,
+      "learning_rate": 7.46580053756424e-06,
+      "loss": 0.281,
+      "step": 14623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.136523743744301,
+      "learning_rate": 7.465457766143155e-06,
+      "loss": 0.2019,
+      "step": 14624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5765079916320244,
+      "learning_rate": 7.465114979412244e-06,
+      "loss": 0.257,
+      "step": 14625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.926838408619817,
+      "learning_rate": 7.464772177373638e-06,
+      "loss": 0.4164,
+      "step": 14626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3487618251362585,
+      "learning_rate": 7.4644293600294656e-06,
+      "loss": 0.2167,
+      "step": 14627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6267458118459113,
+      "learning_rate": 7.464086527381855e-06,
+      "loss": 0.3046,
+      "step": 14628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9415624256370403,
+      "learning_rate": 7.4637436794329354e-06,
+      "loss": 0.3288,
+      "step": 14629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5980471899852382,
+      "learning_rate": 7.463400816184836e-06,
+      "loss": 0.4128,
+      "step": 14630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.07269716525134,
+      "learning_rate": 7.463057937639686e-06,
+      "loss": 0.2465,
+      "step": 14631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.673440227274805,
+      "learning_rate": 7.462715043799612e-06,
+      "loss": 0.2009,
+      "step": 14632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5334982715487497,
+      "learning_rate": 7.462372134666748e-06,
+      "loss": 0.3779,
+      "step": 14633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6651234043272054,
+      "learning_rate": 7.462029210243219e-06,
+      "loss": 0.5792,
+      "step": 14634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9534377381759886,
+      "learning_rate": 7.461686270531158e-06,
+      "loss": 0.4952,
+      "step": 14635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4910580844645078,
+      "learning_rate": 7.461343315532692e-06,
+      "loss": 0.3927,
+      "step": 14636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4551649362154713,
+      "learning_rate": 7.46100034524995e-06,
+      "loss": 0.4144,
+      "step": 14637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7149077231915604,
+      "learning_rate": 7.460657359685066e-06,
+      "loss": 0.2566,
+      "step": 14638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5250987727392027,
+      "learning_rate": 7.460314358840166e-06,
+      "loss": 0.3645,
+      "step": 14639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0972254468792704,
+      "learning_rate": 7.459971342717381e-06,
+      "loss": 0.8815,
+      "step": 14640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9627455472709807,
+      "learning_rate": 7.4596283113188406e-06,
+      "loss": 0.3751,
+      "step": 14641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5979449983552843,
+      "learning_rate": 7.459285264646676e-06,
+      "loss": 0.4778,
+      "step": 14642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4527265222823909,
+      "learning_rate": 7.4589422027030164e-06,
+      "loss": 0.3493,
+      "step": 14643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.727135291389769,
+      "learning_rate": 7.458599125489995e-06,
+      "loss": 0.4401,
+      "step": 14644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.504776795580845,
+      "learning_rate": 7.458256033009738e-06,
+      "loss": 0.235,
+      "step": 14645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9527238204924164,
+      "learning_rate": 7.457912925264378e-06,
+      "loss": 0.4149,
+      "step": 14646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0464168610989097,
+      "learning_rate": 7.4575698022560436e-06,
+      "loss": 0.4752,
+      "step": 14647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.599749812807015,
+      "learning_rate": 7.45722666398687e-06,
+      "loss": 0.2625,
+      "step": 14648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2018199753923335,
+      "learning_rate": 7.456883510458983e-06,
+      "loss": 0.1959,
+      "step": 14649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6678200863198738,
+      "learning_rate": 7.456540341674517e-06,
+      "loss": 0.2413,
+      "step": 14650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1699109705829165,
+      "learning_rate": 7.456197157635603e-06,
+      "loss": 0.1805,
+      "step": 14651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8256349763724984,
+      "learning_rate": 7.455853958344368e-06,
+      "loss": 0.572,
+      "step": 14652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3934953779873471,
+      "learning_rate": 7.455510743802947e-06,
+      "loss": 0.2371,
+      "step": 14653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8342219189743951,
+      "learning_rate": 7.45516751401347e-06,
+      "loss": 0.2528,
+      "step": 14654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4936204490984966,
+      "learning_rate": 7.45482426897807e-06,
+      "loss": 0.407,
+      "step": 14655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4587729388924258,
+      "learning_rate": 7.454481008698875e-06,
+      "loss": 0.2239,
+      "step": 14656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2907485339363103,
+      "learning_rate": 7.454137733178021e-06,
+      "loss": 0.33,
+      "step": 14657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8267219979617302,
+      "learning_rate": 7.453794442417636e-06,
+      "loss": 0.3603,
+      "step": 14658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7426915572872048,
+      "learning_rate": 7.453451136419853e-06,
+      "loss": 0.3337,
+      "step": 14659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8641041297173084,
+      "learning_rate": 7.453107815186803e-06,
+      "loss": 0.2167,
+      "step": 14660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.934593242911826,
+      "learning_rate": 7.45276447872062e-06,
+      "loss": 0.3611,
+      "step": 14661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5957435033164853,
+      "learning_rate": 7.452421127023433e-06,
+      "loss": 0.1788,
+      "step": 14662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0744861902763048,
+      "learning_rate": 7.452077760097378e-06,
+      "loss": 0.1711,
+      "step": 14663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8945352161012015,
+      "learning_rate": 7.451734377944585e-06,
+      "loss": 0.3476,
+      "step": 14664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6756949836027732,
+      "learning_rate": 7.4513909805671845e-06,
+      "loss": 0.382,
+      "step": 14665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.632657751023724,
+      "learning_rate": 7.451047567967313e-06,
+      "loss": 0.3003,
+      "step": 14666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.288524122229223,
+      "learning_rate": 7.450704140147101e-06,
+      "loss": 0.2956,
+      "step": 14667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.384169070449181,
+      "learning_rate": 7.450360697108679e-06,
+      "loss": 0.2479,
+      "step": 14668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9479507266516072,
+      "learning_rate": 7.450017238854183e-06,
+      "loss": 0.4577,
+      "step": 14669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2590103931567866,
+      "learning_rate": 7.449673765385744e-06,
+      "loss": 0.5486,
+      "step": 14670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.617440198767711,
+      "learning_rate": 7.4493302767054956e-06,
+      "loss": 0.2794,
+      "step": 14671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7365324867741279,
+      "learning_rate": 7.44898677281557e-06,
+      "loss": 0.4191,
+      "step": 14672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8135425635731288,
+      "learning_rate": 7.448643253718102e-06,
+      "loss": 0.4097,
+      "step": 14673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.609182374037315,
+      "learning_rate": 7.448299719415222e-06,
+      "loss": 0.2227,
+      "step": 14674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.820499856663609,
+      "learning_rate": 7.447956169909066e-06,
+      "loss": 0.4043,
+      "step": 14675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0540005873718274,
+      "learning_rate": 7.447612605201764e-06,
+      "loss": 0.4446,
+      "step": 14676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9616678709103699,
+      "learning_rate": 7.447269025295452e-06,
+      "loss": 0.6145,
+      "step": 14677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.503860834600196,
+      "learning_rate": 7.4469254301922646e-06,
+      "loss": 0.2601,
+      "step": 14678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5164301640915985,
+      "learning_rate": 7.446581819894333e-06,
+      "loss": 0.2933,
+      "step": 14679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.665585460123624,
+      "learning_rate": 7.446238194403792e-06,
+      "loss": 0.21,
+      "step": 14680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6630303266616502,
+      "learning_rate": 7.445894553722776e-06,
+      "loss": 0.2492,
+      "step": 14681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8707971278031132,
+      "learning_rate": 7.445550897853417e-06,
+      "loss": 0.2067,
+      "step": 14682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8252172657568395,
+      "learning_rate": 7.44520722679785e-06,
+      "loss": 0.5034,
+      "step": 14683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8710769541072185,
+      "learning_rate": 7.444863540558212e-06,
+      "loss": 0.4093,
+      "step": 14684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6865843036847417,
+      "learning_rate": 7.44451983913663e-06,
+      "loss": 0.3178,
+      "step": 14685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.956432189927863,
+      "learning_rate": 7.444176122535245e-06,
+      "loss": 0.4104,
+      "step": 14686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4249654761532589,
+      "learning_rate": 7.443832390756189e-06,
+      "loss": 0.3073,
+      "step": 14687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8498906784684623,
+      "learning_rate": 7.443488643801598e-06,
+      "loss": 0.6316,
+      "step": 14688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.103711727426579,
+      "learning_rate": 7.4431448816736025e-06,
+      "loss": 0.2353,
+      "step": 14689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6274783052735908,
+      "learning_rate": 7.442801104374342e-06,
+      "loss": 0.2036,
+      "step": 14690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8157543541106399,
+      "learning_rate": 7.4424573119059485e-06,
+      "loss": 0.1517,
+      "step": 14691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.894373590260544,
+      "learning_rate": 7.4421135042705586e-06,
+      "loss": 0.2591,
+      "step": 14692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8063383519696672,
+      "learning_rate": 7.441769681470304e-06,
+      "loss": 0.449,
+      "step": 14693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4531460934820528,
+      "learning_rate": 7.441425843507324e-06,
+      "loss": 0.2038,
+      "step": 14694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1937098066065897,
+      "learning_rate": 7.4410819903837516e-06,
+      "loss": 0.5686,
+      "step": 14695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6279358980990584,
+      "learning_rate": 7.440738122101721e-06,
+      "loss": 0.2548,
+      "step": 14696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5211071482079574,
+      "learning_rate": 7.440394238663369e-06,
+      "loss": 0.229,
+      "step": 14697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4047694412535174,
+      "learning_rate": 7.440050340070831e-06,
+      "loss": 0.3916,
+      "step": 14698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4860562102114954,
+      "learning_rate": 7.439706426326241e-06,
+      "loss": 0.2819,
+      "step": 14699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8641107300356894,
+      "learning_rate": 7.439362497431737e-06,
+      "loss": 0.3378,
+      "step": 14700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7561560052446294,
+      "learning_rate": 7.439018553389452e-06,
+      "loss": 0.372,
+      "step": 14701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4329725005608533,
+      "learning_rate": 7.438674594201525e-06,
+      "loss": 0.3313,
+      "step": 14702
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7665451783080557,
+      "learning_rate": 7.438330619870089e-06,
+      "loss": 0.4034,
+      "step": 14703
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5456256589645172,
+      "learning_rate": 7.437986630397281e-06,
+      "loss": 0.2258,
+      "step": 14704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7528236505851196,
+      "learning_rate": 7.437642625785238e-06,
+      "loss": 0.3133,
+      "step": 14705
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5067177192758043,
+      "learning_rate": 7.4372986060360965e-06,
+      "loss": 0.4424,
+      "step": 14706
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.567069493526113,
+      "learning_rate": 7.436954571151989e-06,
+      "loss": 0.3304,
+      "step": 14707
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.395527040684119,
+      "learning_rate": 7.436610521135057e-06,
+      "loss": 0.2526,
+      "step": 14708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2179723254476142,
+      "learning_rate": 7.436266455987432e-06,
+      "loss": 0.1348,
+      "step": 14709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8841782914792458,
+      "learning_rate": 7.435922375711255e-06,
+      "loss": 0.4463,
+      "step": 14710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6319158684030022,
+      "learning_rate": 7.435578280308659e-06,
+      "loss": 0.4908,
+      "step": 14711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1477639402913116,
+      "learning_rate": 7.4352341697817845e-06,
+      "loss": 0.1606,
+      "step": 14712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5235833269407648,
+      "learning_rate": 7.434890044132765e-06,
+      "loss": 0.2831,
+      "step": 14713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.080141523227842,
+      "learning_rate": 7.434545903363738e-06,
+      "loss": 0.423,
+      "step": 14714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2240781765122166,
+      "learning_rate": 7.434201747476841e-06,
+      "loss": 0.2437,
+      "step": 14715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.406064260985123,
+      "learning_rate": 7.4338575764742135e-06,
+      "loss": 0.1442,
+      "step": 14716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8427314288552095,
+      "learning_rate": 7.4335133903579906e-06,
+      "loss": 0.467,
+      "step": 14717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7944071553676548,
+      "learning_rate": 7.433169189130307e-06,
+      "loss": 0.3095,
+      "step": 14718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3057028736859944,
+      "learning_rate": 7.432824972793305e-06,
+      "loss": 0.285,
+      "step": 14719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4645424393072077,
+      "learning_rate": 7.432480741349118e-06,
+      "loss": 0.4126,
+      "step": 14720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9080541898144163,
+      "learning_rate": 7.432136494799888e-06,
+      "loss": 0.6616,
+      "step": 14721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.633100221918947,
+      "learning_rate": 7.431792233147747e-06,
+      "loss": 0.4854,
+      "step": 14722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4105424629486814,
+      "learning_rate": 7.431447956394837e-06,
+      "loss": 0.2774,
+      "step": 14723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7373571425480185,
+      "learning_rate": 7.431103664543294e-06,
+      "loss": 0.1962,
+      "step": 14724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5533214251076461,
+      "learning_rate": 7.430759357595257e-06,
+      "loss": 0.3432,
+      "step": 14725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2714765639292251,
+      "learning_rate": 7.430415035552864e-06,
+      "loss": 0.2159,
+      "step": 14726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5957099629585991,
+      "learning_rate": 7.430070698418253e-06,
+      "loss": 0.4694,
+      "step": 14727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6323531476605926,
+      "learning_rate": 7.429726346193562e-06,
+      "loss": 0.3049,
+      "step": 14728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6477832761502949,
+      "learning_rate": 7.429381978880927e-06,
+      "loss": 0.3686,
+      "step": 14729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8170770411487054,
+      "learning_rate": 7.429037596482491e-06,
+      "loss": 0.5194,
+      "step": 14730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4709342835672585,
+      "learning_rate": 7.42869319900039e-06,
+      "loss": 0.3922,
+      "step": 14731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4668526569025249,
+      "learning_rate": 7.4283487864367635e-06,
+      "loss": 0.3168,
+      "step": 14732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.443395979369385,
+      "learning_rate": 7.428004358793748e-06,
+      "loss": 0.2008,
+      "step": 14733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6329912699251716,
+      "learning_rate": 7.427659916073486e-06,
+      "loss": 0.2894,
+      "step": 14734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4707173933170397,
+      "learning_rate": 7.427315458278112e-06,
+      "loss": 0.2888,
+      "step": 14735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.102166438271535,
+      "learning_rate": 7.4269709854097694e-06,
+      "loss": 0.2619,
+      "step": 14736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1216212686360567,
+      "learning_rate": 7.426626497470594e-06,
+      "loss": 0.0887,
+      "step": 14737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7950098578606082,
+      "learning_rate": 7.426281994462725e-06,
+      "loss": 0.3391,
+      "step": 14738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6997510449004687,
+      "learning_rate": 7.425937476388305e-06,
+      "loss": 0.354,
+      "step": 14739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4976833951954553,
+      "learning_rate": 7.425592943249471e-06,
+      "loss": 0.3095,
+      "step": 14740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0236802489434345,
+      "learning_rate": 7.425248395048362e-06,
+      "loss": 0.4428,
+      "step": 14741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6200662711836058,
+      "learning_rate": 7.424903831787118e-06,
+      "loss": 0.3855,
+      "step": 14742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6815669611836401,
+      "learning_rate": 7.42455925346788e-06,
+      "loss": 0.2963,
+      "step": 14743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6925506992790083,
+      "learning_rate": 7.424214660092786e-06,
+      "loss": 0.3409,
+      "step": 14744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6440917037026226,
+      "learning_rate": 7.4238700516639765e-06,
+      "loss": 0.4015,
+      "step": 14745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3066838874554392,
+      "learning_rate": 7.42352542818359e-06,
+      "loss": 0.1635,
+      "step": 14746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.166148873004754,
+      "learning_rate": 7.42318078965377e-06,
+      "loss": 0.1501,
+      "step": 14747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8019644393913004,
+      "learning_rate": 7.422836136076652e-06,
+      "loss": 0.6233,
+      "step": 14748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.576902747135018,
+      "learning_rate": 7.422491467454382e-06,
+      "loss": 0.4597,
+      "step": 14749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.844234669555327,
+      "learning_rate": 7.422146783789095e-06,
+      "loss": 0.4674,
+      "step": 14750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4672118424111633,
+      "learning_rate": 7.4218020850829335e-06,
+      "loss": 0.266,
+      "step": 14751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5298546009148386,
+      "learning_rate": 7.421457371338038e-06,
+      "loss": 0.1796,
+      "step": 14752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7529332067937775,
+      "learning_rate": 7.421112642556548e-06,
+      "loss": 0.3941,
+      "step": 14753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5957366655016116,
+      "learning_rate": 7.420767898740606e-06,
+      "loss": 0.348,
+      "step": 14754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8182089007855515,
+      "learning_rate": 7.420423139892352e-06,
+      "loss": 0.1995,
+      "step": 14755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4099137278479796,
+      "learning_rate": 7.420078366013928e-06,
+      "loss": 0.4694,
+      "step": 14756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.549596050138811,
+      "learning_rate": 7.419733577107471e-06,
+      "loss": 0.2819,
+      "step": 14757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3594513301685156,
+      "learning_rate": 7.419388773175127e-06,
+      "loss": 0.2208,
+      "step": 14758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.077449058022181,
+      "learning_rate": 7.419043954219033e-06,
+      "loss": 0.1956,
+      "step": 14759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7722614691023248,
+      "learning_rate": 7.418699120241332e-06,
+      "loss": 0.4146,
+      "step": 14760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8013806018001064,
+      "learning_rate": 7.418354271244167e-06,
+      "loss": 0.405,
+      "step": 14761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0507352987318432,
+      "learning_rate": 7.418009407229678e-06,
+      "loss": 0.4709,
+      "step": 14762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.19993290465257,
+      "learning_rate": 7.417664528200005e-06,
+      "loss": 0.2259,
+      "step": 14763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4588208138783136,
+      "learning_rate": 7.4173196341572915e-06,
+      "loss": 0.3319,
+      "step": 14764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.515174476775115,
+      "learning_rate": 7.4169747251036785e-06,
+      "loss": 0.4664,
+      "step": 14765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6079472300926456,
+      "learning_rate": 7.416629801041308e-06,
+      "loss": 0.5209,
+      "step": 14766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5283660881816596,
+      "learning_rate": 7.416284861972322e-06,
+      "loss": 0.278,
+      "step": 14767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8655517028582538,
+      "learning_rate": 7.4159399078988625e-06,
+      "loss": 0.6653,
+      "step": 14768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.618052384937802,
+      "learning_rate": 7.415594938823072e-06,
+      "loss": 0.31,
+      "step": 14769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2510386532718896,
+      "learning_rate": 7.415249954747091e-06,
+      "loss": 0.2776,
+      "step": 14770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6843047797971598,
+      "learning_rate": 7.414904955673064e-06,
+      "loss": 0.2124,
+      "step": 14771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4131927766199084,
+      "learning_rate": 7.414559941603131e-06,
+      "loss": 0.3205,
+      "step": 14772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5294681492137843,
+      "learning_rate": 7.4142149125394355e-06,
+      "loss": 0.3759,
+      "step": 14773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6058339448444976,
+      "learning_rate": 7.4138698684841215e-06,
+      "loss": 0.2109,
+      "step": 14774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5643220994256173,
+      "learning_rate": 7.413524809439329e-06,
+      "loss": 0.2127,
+      "step": 14775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5680942548844792,
+      "learning_rate": 7.413179735407203e-06,
+      "loss": 0.4075,
+      "step": 14776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7275266587592173,
+      "learning_rate": 7.4128346463898834e-06,
+      "loss": 0.6355,
+      "step": 14777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4744965289054999,
+      "learning_rate": 7.412489542389517e-06,
+      "loss": 0.3172,
+      "step": 14778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.783633897691274,
+      "learning_rate": 7.412144423408244e-06,
+      "loss": 0.2832,
+      "step": 14779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.749011411897094,
+      "learning_rate": 7.411799289448209e-06,
+      "loss": 0.1805,
+      "step": 14780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3044665971928344,
+      "learning_rate": 7.411454140511554e-06,
+      "loss": 0.2791,
+      "step": 14781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7399505786276088,
+      "learning_rate": 7.411108976600423e-06,
+      "loss": 0.3036,
+      "step": 14782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3504348264880404,
+      "learning_rate": 7.41076379771696e-06,
+      "loss": 0.1948,
+      "step": 14783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3620207541944136,
+      "learning_rate": 7.410418603863305e-06,
+      "loss": 0.1922,
+      "step": 14784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8403317316611216,
+      "learning_rate": 7.410073395041607e-06,
+      "loss": 0.3939,
+      "step": 14785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7059848292896,
+      "learning_rate": 7.409728171254005e-06,
+      "loss": 0.3055,
+      "step": 14786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.516266313529566,
+      "learning_rate": 7.409382932502645e-06,
+      "loss": 0.3049,
+      "step": 14787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0445485616846777,
+      "learning_rate": 7.409037678789671e-06,
+      "loss": 0.2115,
+      "step": 14788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5831707088710543,
+      "learning_rate": 7.408692410117225e-06,
+      "loss": 0.3343,
+      "step": 14789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.942750129855192,
+      "learning_rate": 7.408347126487454e-06,
+      "loss": 0.223,
+      "step": 14790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5363957058271138,
+      "learning_rate": 7.4080018279025e-06,
+      "loss": 0.348,
+      "step": 14791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8733013783683699,
+      "learning_rate": 7.407656514364508e-06,
+      "loss": 0.2506,
+      "step": 14792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5558372659762076,
+      "learning_rate": 7.407311185875621e-06,
+      "loss": 0.3732,
+      "step": 14793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.232935581859158,
+      "learning_rate": 7.4069658424379854e-06,
+      "loss": 0.4676,
+      "step": 14794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5610368983378637,
+      "learning_rate": 7.406620484053745e-06,
+      "loss": 0.1866,
+      "step": 14795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.371407606061839,
+      "learning_rate": 7.406275110725043e-06,
+      "loss": 0.2483,
+      "step": 14796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7000636576206158,
+      "learning_rate": 7.405929722454026e-06,
+      "loss": 0.4875,
+      "step": 14797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.14508470574768,
+      "learning_rate": 7.405584319242838e-06,
+      "loss": 0.228,
+      "step": 14798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8792651838835084,
+      "learning_rate": 7.405238901093623e-06,
+      "loss": 0.2705,
+      "step": 14799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5528908692676535,
+      "learning_rate": 7.4048934680085276e-06,
+      "loss": 0.29,
+      "step": 14800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8427310245325124,
+      "learning_rate": 7.404548019989694e-06,
+      "loss": 0.2472,
+      "step": 14801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3057318700503238,
+      "learning_rate": 7.404202557039272e-06,
+      "loss": 0.1974,
+      "step": 14802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.181517458369998,
+      "learning_rate": 7.403857079159403e-06,
+      "loss": 0.7624,
+      "step": 14803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4045331063885844,
+      "learning_rate": 7.403511586352233e-06,
+      "loss": 0.286,
+      "step": 14804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0756695032819286,
+      "learning_rate": 7.40316607861991e-06,
+      "loss": 0.3062,
+      "step": 14805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.18399418825426,
+      "learning_rate": 7.402820555964575e-06,
+      "loss": 0.3099,
+      "step": 14806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8571994820840034,
+      "learning_rate": 7.402475018388378e-06,
+      "loss": 0.4429,
+      "step": 14807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.081009196156111,
+      "learning_rate": 7.402129465893461e-06,
+      "loss": 0.2114,
+      "step": 14808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4970172790569807,
+      "learning_rate": 7.401783898481971e-06,
+      "loss": 0.3945,
+      "step": 14809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9367109952196808,
+      "learning_rate": 7.401438316156055e-06,
+      "loss": 0.0964,
+      "step": 14810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5481395393377246,
+      "learning_rate": 7.4010927189178595e-06,
+      "loss": 0.1764,
+      "step": 14811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5326777876317663,
+      "learning_rate": 7.400747106769528e-06,
+      "loss": 0.4746,
+      "step": 14812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6371843181141579,
+      "learning_rate": 7.400401479713209e-06,
+      "loss": 0.3674,
+      "step": 14813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4823610434031569,
+      "learning_rate": 7.4000558377510475e-06,
+      "loss": 0.195,
+      "step": 14814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2364540216787996,
+      "learning_rate": 7.39971018088519e-06,
+      "loss": 0.1639,
+      "step": 14815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8597698533719798,
+      "learning_rate": 7.399364509117784e-06,
+      "loss": 0.5714,
+      "step": 14816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0456305062103057,
+      "learning_rate": 7.399018822450974e-06,
+      "loss": 0.5976,
+      "step": 14817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.253304015943253,
+      "learning_rate": 7.398673120886909e-06,
+      "loss": 0.3167,
+      "step": 14818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.662200512836007,
+      "learning_rate": 7.398327404427733e-06,
+      "loss": 0.3187,
+      "step": 14819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5109814908666284,
+      "learning_rate": 7.397981673075595e-06,
+      "loss": 0.3833,
+      "step": 14820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4846822872323275,
+      "learning_rate": 7.39763592683264e-06,
+      "loss": 0.3468,
+      "step": 14821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5564768476495183,
+      "learning_rate": 7.397290165701019e-06,
+      "loss": 0.2284,
+      "step": 14822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9211288259685186,
+      "learning_rate": 7.396944389682874e-06,
+      "loss": 0.4373,
+      "step": 14823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5215925929865473,
+      "learning_rate": 7.396598598780354e-06,
+      "loss": 0.3495,
+      "step": 14824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8703178557061912,
+      "learning_rate": 7.396252792995608e-06,
+      "loss": 0.2797,
+      "step": 14825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7694948794772396,
+      "learning_rate": 7.395906972330782e-06,
+      "loss": 0.4488,
+      "step": 14826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9512053172746202,
+      "learning_rate": 7.395561136788023e-06,
+      "loss": 0.5886,
+      "step": 14827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4375988262869597,
+      "learning_rate": 7.395215286369479e-06,
+      "loss": 0.387,
+      "step": 14828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4996857512635362,
+      "learning_rate": 7.394869421077299e-06,
+      "loss": 0.1775,
+      "step": 14829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7038202616504758,
+      "learning_rate": 7.394523540913627e-06,
+      "loss": 0.4437,
+      "step": 14830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4420814535098796,
+      "learning_rate": 7.394177645880616e-06,
+      "loss": 0.2398,
+      "step": 14831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1068455400333845,
+      "learning_rate": 7.393831735980408e-06,
+      "loss": 0.5776,
+      "step": 14832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9535034881074422,
+      "learning_rate": 7.3934858112151564e-06,
+      "loss": 0.684,
+      "step": 14833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0991528604698804,
+      "learning_rate": 7.3931398715870064e-06,
+      "loss": 0.2097,
+      "step": 14834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0813208060348716,
+      "learning_rate": 7.392793917098106e-06,
+      "loss": 0.1952,
+      "step": 14835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1778231479786425,
+      "learning_rate": 7.392447947750605e-06,
+      "loss": 0.1484,
+      "step": 14836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9134873674427755,
+      "learning_rate": 7.392101963546652e-06,
+      "loss": 0.2769,
+      "step": 14837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4992639193407122,
+      "learning_rate": 7.391755964488394e-06,
+      "loss": 0.2804,
+      "step": 14838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.562241134657622,
+      "learning_rate": 7.3914099505779805e-06,
+      "loss": 0.2242,
+      "step": 14839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8491879983205657,
+      "learning_rate": 7.391063921817558e-06,
+      "loss": 0.2067,
+      "step": 14840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4032608051465119,
+      "learning_rate": 7.390717878209278e-06,
+      "loss": 0.2996,
+      "step": 14841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.607924014442144,
+      "learning_rate": 7.390371819755289e-06,
+      "loss": 0.305,
+      "step": 14842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2790228397307521,
+      "learning_rate": 7.390025746457739e-06,
+      "loss": 0.2583,
+      "step": 14843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.502597441209213,
+      "learning_rate": 7.389679658318777e-06,
+      "loss": 0.2273,
+      "step": 14844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4973228251585315,
+      "learning_rate": 7.389333555340552e-06,
+      "loss": 0.3078,
+      "step": 14845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.480238209592266,
+      "learning_rate": 7.388987437525215e-06,
+      "loss": 0.2194,
+      "step": 14846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4672028161428599,
+      "learning_rate": 7.388641304874914e-06,
+      "loss": 0.204,
+      "step": 14847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4697775163404176,
+      "learning_rate": 7.388295157391796e-06,
+      "loss": 0.2228,
+      "step": 14848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.024972811857835,
+      "learning_rate": 7.387948995078016e-06,
+      "loss": 0.1716,
+      "step": 14849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0549295023630223,
+      "learning_rate": 7.387602817935717e-06,
+      "loss": 0.2284,
+      "step": 14850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8425073921839485,
+      "learning_rate": 7.387256625967054e-06,
+      "loss": 0.2693,
+      "step": 14851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8009994978644557,
+      "learning_rate": 7.386910419174175e-06,
+      "loss": 0.5368,
+      "step": 14852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4270856470839206,
+      "learning_rate": 7.386564197559229e-06,
+      "loss": 0.228,
+      "step": 14853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.979212742612134,
+      "learning_rate": 7.386217961124366e-06,
+      "loss": 0.3201,
+      "step": 14854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4606785213195648,
+      "learning_rate": 7.385871709871738e-06,
+      "loss": 0.3053,
+      "step": 14855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7353721020652437,
+      "learning_rate": 7.385525443803493e-06,
+      "loss": 0.4471,
+      "step": 14856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.968286599768751,
+      "learning_rate": 7.385179162921783e-06,
+      "loss": 0.5864,
+      "step": 14857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9351116426306405,
+      "learning_rate": 7.384832867228756e-06,
+      "loss": 0.4971,
+      "step": 14858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.845281074346778,
+      "learning_rate": 7.384486556726563e-06,
+      "loss": 0.4266,
+      "step": 14859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1903860523841812,
+      "learning_rate": 7.384140231417357e-06,
+      "loss": 0.216,
+      "step": 14860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8600596281461232,
+      "learning_rate": 7.3837938913032856e-06,
+      "loss": 0.4279,
+      "step": 14861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7387470521360264,
+      "learning_rate": 7.383447536386501e-06,
+      "loss": 0.2929,
+      "step": 14862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3810693373487775,
+      "learning_rate": 7.383101166669153e-06,
+      "loss": 0.323,
+      "step": 14863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9043274158082806,
+      "learning_rate": 7.382754782153394e-06,
+      "loss": 0.2981,
+      "step": 14864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5722587526064593,
+      "learning_rate": 7.382408382841374e-06,
+      "loss": 0.2109,
+      "step": 14865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.127330277197487,
+      "learning_rate": 7.382061968735243e-06,
+      "loss": 0.4531,
+      "step": 14866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2934797262675466,
+      "learning_rate": 7.3817155398371546e-06,
+      "loss": 0.2261,
+      "step": 14867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5098837476862705,
+      "learning_rate": 7.381369096149258e-06,
+      "loss": 0.2808,
+      "step": 14868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.077480259169194,
+      "learning_rate": 7.381022637673706e-06,
+      "loss": 0.6409,
+      "step": 14869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7777616659277071,
+      "learning_rate": 7.380676164412648e-06,
+      "loss": 0.2601,
+      "step": 14870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3203217968810925,
+      "learning_rate": 7.3803296763682365e-06,
+      "loss": 0.2476,
+      "step": 14871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7344267377045868,
+      "learning_rate": 7.379983173542624e-06,
+      "loss": 0.2415,
+      "step": 14872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4583330705052093,
+      "learning_rate": 7.379636655937962e-06,
+      "loss": 0.3672,
+      "step": 14873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1647520791024162,
+      "learning_rate": 7.3792901235563995e-06,
+      "loss": 0.2054,
+      "step": 14874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8333914044965502,
+      "learning_rate": 7.378943576400093e-06,
+      "loss": 0.4607,
+      "step": 14875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.684295008180717,
+      "learning_rate": 7.378597014471191e-06,
+      "loss": 0.5021,
+      "step": 14876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3787308816457537,
+      "learning_rate": 7.378250437771848e-06,
+      "loss": 0.3159,
+      "step": 14877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6230151796516885,
+      "learning_rate": 7.377903846304213e-06,
+      "loss": 0.3249,
+      "step": 14878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3217870136269436,
+      "learning_rate": 7.377557240070441e-06,
+      "loss": 0.1729,
+      "step": 14879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6062367358940264,
+      "learning_rate": 7.377210619072683e-06,
+      "loss": 0.3728,
+      "step": 14880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2924679913006805,
+      "learning_rate": 7.376863983313092e-06,
+      "loss": 0.3556,
+      "step": 14881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8465399767447328,
+      "learning_rate": 7.37651733279382e-06,
+      "loss": 0.1786,
+      "step": 14882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8305608485257816,
+      "learning_rate": 7.37617066751702e-06,
+      "loss": 0.2671,
+      "step": 14883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.532509845053272,
+      "learning_rate": 7.375823987484845e-06,
+      "loss": 0.4865,
+      "step": 14884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5109621786609098,
+      "learning_rate": 7.375477292699447e-06,
+      "loss": 0.304,
+      "step": 14885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.705106094847566,
+      "learning_rate": 7.37513058316298e-06,
+      "loss": 0.5286,
+      "step": 14886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7722654481803228,
+      "learning_rate": 7.374783858877594e-06,
+      "loss": 0.2911,
+      "step": 14887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6491462982541332,
+      "learning_rate": 7.374437119845445e-06,
+      "loss": 0.3825,
+      "step": 14888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6908710857680795,
+      "learning_rate": 7.374090366068687e-06,
+      "loss": 0.3202,
+      "step": 14889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4156038803773574,
+      "learning_rate": 7.37374359754947e-06,
+      "loss": 0.3194,
+      "step": 14890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8477031548576925,
+      "learning_rate": 7.3733968142899505e-06,
+      "loss": 0.2786,
+      "step": 14891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4050816132642954,
+      "learning_rate": 7.3730500162922794e-06,
+      "loss": 0.2103,
+      "step": 14892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0280549413456765,
+      "learning_rate": 7.372703203558612e-06,
+      "loss": 0.4342,
+      "step": 14893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9447205420348497,
+      "learning_rate": 7.372356376091101e-06,
+      "loss": 0.5187,
+      "step": 14894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.047694850122632,
+      "learning_rate": 7.372009533891901e-06,
+      "loss": 0.4638,
+      "step": 14895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2588719265981443,
+      "learning_rate": 7.371662676963163e-06,
+      "loss": 0.2025,
+      "step": 14896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.666014951949526,
+      "learning_rate": 7.371315805307046e-06,
+      "loss": 0.1923,
+      "step": 14897
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8683051670703756,
+      "learning_rate": 7.370968918925698e-06,
+      "loss": 0.4048,
+      "step": 14898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.19319530364359,
+      "learning_rate": 7.370622017821278e-06,
+      "loss": 0.151,
+      "step": 14899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9155988891371571,
+      "learning_rate": 7.370275101995938e-06,
+      "loss": 0.3468,
+      "step": 14900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6972534820261207,
+      "learning_rate": 7.369928171451832e-06,
+      "loss": 0.3101,
+      "step": 14901
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3287009224660717,
+      "learning_rate": 7.369581226191115e-06,
+      "loss": 0.2222,
+      "step": 14902
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.077920297708495,
+      "learning_rate": 7.369234266215941e-06,
+      "loss": 0.6711,
+      "step": 14903
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3218125323489895,
+      "learning_rate": 7.368887291528465e-06,
+      "loss": 0.2112,
+      "step": 14904
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8169108274488162,
+      "learning_rate": 7.368540302130841e-06,
+      "loss": 0.4738,
+      "step": 14905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7427638430589407,
+      "learning_rate": 7.368193298025225e-06,
+      "loss": 0.4727,
+      "step": 14906
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6238759972308878,
+      "learning_rate": 7.36784627921377e-06,
+      "loss": 0.3489,
+      "step": 14907
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0114935332217323,
+      "learning_rate": 7.367499245698632e-06,
+      "loss": 0.2002,
+      "step": 14908
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8251034390516456,
+      "learning_rate": 7.367152197481965e-06,
+      "loss": 0.3331,
+      "step": 14909
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0709992880107058,
+      "learning_rate": 7.366805134565926e-06,
+      "loss": 0.1968,
+      "step": 14910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1058694227941872,
+      "learning_rate": 7.366458056952668e-06,
+      "loss": 0.2676,
+      "step": 14911
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5379532723508862,
+      "learning_rate": 7.366110964644348e-06,
+      "loss": 0.2663,
+      "step": 14912
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.080350598322633,
+      "learning_rate": 7.365763857643121e-06,
+      "loss": 0.5482,
+      "step": 14913
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6063636248843836,
+      "learning_rate": 7.365416735951142e-06,
+      "loss": 0.3197,
+      "step": 14914
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5696952849933064,
+      "learning_rate": 7.365069599570566e-06,
+      "loss": 0.1831,
+      "step": 14915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2166117424752574,
+      "learning_rate": 7.3647224485035485e-06,
+      "loss": 0.2239,
+      "step": 14916
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3218254021432423,
+      "learning_rate": 7.3643752827522475e-06,
+      "loss": 0.2323,
+      "step": 14917
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.143913248663654,
+      "learning_rate": 7.364028102318816e-06,
+      "loss": 0.2335,
+      "step": 14918
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.358039435549686,
+      "learning_rate": 7.363680907205413e-06,
+      "loss": 0.3253,
+      "step": 14919
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.201704468543671,
+      "learning_rate": 7.36333369741419e-06,
+      "loss": 0.1981,
+      "step": 14920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3789846503363536,
+      "learning_rate": 7.362986472947308e-06,
+      "loss": 0.2354,
+      "step": 14921
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.372419563995516,
+      "learning_rate": 7.362639233806918e-06,
+      "loss": 0.1775,
+      "step": 14922
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.886657375222003,
+      "learning_rate": 7.362291979995182e-06,
+      "loss": 0.1722,
+      "step": 14923
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.596884842037919,
+      "learning_rate": 7.361944711514253e-06,
+      "loss": 0.283,
+      "step": 14924
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3470352995178536,
+      "learning_rate": 7.361597428366287e-06,
+      "loss": 0.2232,
+      "step": 14925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1667368908875821,
+      "learning_rate": 7.361250130553442e-06,
+      "loss": 0.2744,
+      "step": 14926
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4250723918713522,
+      "learning_rate": 7.3609028180778745e-06,
+      "loss": 0.1738,
+      "step": 14927
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9927946399703733,
+      "learning_rate": 7.36055549094174e-06,
+      "loss": 0.6446,
+      "step": 14928
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6255906778376683,
+      "learning_rate": 7.360208149147196e-06,
+      "loss": 0.3522,
+      "step": 14929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8087974170013648,
+      "learning_rate": 7.3598607926964005e-06,
+      "loss": 0.2033,
+      "step": 14930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.715873003146825,
+      "learning_rate": 7.359513421591508e-06,
+      "loss": 0.3746,
+      "step": 14931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.862763578850517,
+      "learning_rate": 7.3591660358346785e-06,
+      "loss": 0.2164,
+      "step": 14932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.810195351345844,
+      "learning_rate": 7.358818635428066e-06,
+      "loss": 0.3365,
+      "step": 14933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.145743987775796,
+      "learning_rate": 7.358471220373831e-06,
+      "loss": 0.1953,
+      "step": 14934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3471582543028728,
+      "learning_rate": 7.35812379067413e-06,
+      "loss": 0.2288,
+      "step": 14935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.578089685328204,
+      "learning_rate": 7.357776346331118e-06,
+      "loss": 0.4641,
+      "step": 14936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4699263832901048,
+      "learning_rate": 7.357428887346956e-06,
+      "loss": 0.1724,
+      "step": 14937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.62907194763055,
+      "learning_rate": 7.357081413723799e-06,
+      "loss": 0.5164,
+      "step": 14938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.58327036944688,
+      "learning_rate": 7.356733925463805e-06,
+      "loss": 0.2695,
+      "step": 14939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.24739486989737,
+      "learning_rate": 7.356386422569134e-06,
+      "loss": 0.3829,
+      "step": 14940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.455962553633653,
+      "learning_rate": 7.356038905041941e-06,
+      "loss": 0.1739,
+      "step": 14941
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1478404377219684,
+      "learning_rate": 7.355691372884386e-06,
+      "loss": 0.6554,
+      "step": 14942
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6783150677158973,
+      "learning_rate": 7.355343826098627e-06,
+      "loss": 0.436,
+      "step": 14943
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.431175994417326,
+      "learning_rate": 7.35499626468682e-06,
+      "loss": 0.3301,
+      "step": 14944
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.924598102137831,
+      "learning_rate": 7.354648688651127e-06,
+      "loss": 0.4275,
+      "step": 14945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.663817667523672,
+      "learning_rate": 7.354301097993704e-06,
+      "loss": 0.2971,
+      "step": 14946
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4206891485811082,
+      "learning_rate": 7.353953492716708e-06,
+      "loss": 0.1676,
+      "step": 14947
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.003690942273526,
+      "learning_rate": 7.3536058728223e-06,
+      "loss": 0.3986,
+      "step": 14948
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.939882878136666,
+      "learning_rate": 7.353258238312638e-06,
+      "loss": 0.3629,
+      "step": 14949
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5805445196117118,
+      "learning_rate": 7.352910589189881e-06,
+      "loss": 0.3402,
+      "step": 14950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.015044783095758,
+      "learning_rate": 7.3525629254561855e-06,
+      "loss": 0.4295,
+      "step": 14951
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8824307541440912,
+      "learning_rate": 7.3522152471137155e-06,
+      "loss": 0.6169,
+      "step": 14952
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.401675551244963,
+      "learning_rate": 7.351867554164625e-06,
+      "loss": 0.3409,
+      "step": 14953
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3383714292276334,
+      "learning_rate": 7.351519846611075e-06,
+      "loss": 0.1651,
+      "step": 14954
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4875076566226635,
+      "learning_rate": 7.351172124455224e-06,
+      "loss": 0.3549,
+      "step": 14955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.297184631429031,
+      "learning_rate": 7.350824387699231e-06,
+      "loss": 0.2364,
+      "step": 14956
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.935656508835023,
+      "learning_rate": 7.350476636345259e-06,
+      "loss": 0.5617,
+      "step": 14957
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.154608530170053,
+      "learning_rate": 7.350128870395462e-06,
+      "loss": 0.3162,
+      "step": 14958
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7275574048644335,
+      "learning_rate": 7.349781089852004e-06,
+      "loss": 0.4488,
+      "step": 14959
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3121383500825063,
+      "learning_rate": 7.34943329471704e-06,
+      "loss": 0.1583,
+      "step": 14960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6049884068287257,
+      "learning_rate": 7.349085484992735e-06,
+      "loss": 0.3147,
+      "step": 14961
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2114038845975494,
+      "learning_rate": 7.348737660681244e-06,
+      "loss": 0.1768,
+      "step": 14962
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9811647836066442,
+      "learning_rate": 7.348389821784731e-06,
+      "loss": 0.1845,
+      "step": 14963
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6424463969967014,
+      "learning_rate": 7.348041968305352e-06,
+      "loss": 0.4307,
+      "step": 14964
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6437338347782358,
+      "learning_rate": 7.347694100245272e-06,
+      "loss": 0.3584,
+      "step": 14965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4710978146829634,
+      "learning_rate": 7.347346217606647e-06,
+      "loss": 0.3284,
+      "step": 14966
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.348085078137778,
+      "learning_rate": 7.346998320391638e-06,
+      "loss": 0.1509,
+      "step": 14967
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.618783112964194,
+      "learning_rate": 7.346650408602407e-06,
+      "loss": 0.3493,
+      "step": 14968
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1815178887002977,
+      "learning_rate": 7.346302482241112e-06,
+      "loss": 0.1805,
+      "step": 14969
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6237454775768554,
+      "learning_rate": 7.3459545413099164e-06,
+      "loss": 0.2026,
+      "step": 14970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4985343843609116,
+      "learning_rate": 7.345606585810978e-06,
+      "loss": 0.1708,
+      "step": 14971
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7245613687644474,
+      "learning_rate": 7.3452586157464615e-06,
+      "loss": 0.25,
+      "step": 14972
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3046380736074343,
+      "learning_rate": 7.344910631118523e-06,
+      "loss": 0.1401,
+      "step": 14973
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8645757620851322,
+      "learning_rate": 7.344562631929325e-06,
+      "loss": 0.3758,
+      "step": 14974
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.691843388212435,
+      "learning_rate": 7.344214618181029e-06,
+      "loss": 0.196,
+      "step": 14975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.158269435847748,
+      "learning_rate": 7.343866589875798e-06,
+      "loss": 0.3204,
+      "step": 14976
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.843067669285665,
+      "learning_rate": 7.343518547015788e-06,
+      "loss": 0.4932,
+      "step": 14977
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7604571390484318,
+      "learning_rate": 7.3431704896031665e-06,
+      "loss": 0.4892,
+      "step": 14978
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.068501084888161,
+      "learning_rate": 7.342822417640091e-06,
+      "loss": 0.1994,
+      "step": 14979
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.424842070519869,
+      "learning_rate": 7.342474331128722e-06,
+      "loss": 0.2328,
+      "step": 14980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.600716800239367,
+      "learning_rate": 7.3421262300712245e-06,
+      "loss": 0.3029,
+      "step": 14981
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6750320769555072,
+      "learning_rate": 7.3417781144697575e-06,
+      "loss": 0.3183,
+      "step": 14982
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6725501208108287,
+      "learning_rate": 7.341429984326484e-06,
+      "loss": 0.2879,
+      "step": 14983
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3553948093661725,
+      "learning_rate": 7.341081839643564e-06,
+      "loss": 0.1679,
+      "step": 14984
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.895133359341351,
+      "learning_rate": 7.3407336804231625e-06,
+      "loss": 0.2294,
+      "step": 14985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2168000743205474,
+      "learning_rate": 7.340385506667438e-06,
+      "loss": 0.2416,
+      "step": 14986
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9348075043466257,
+      "learning_rate": 7.340037318378555e-06,
+      "loss": 0.2198,
+      "step": 14987
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3580485687987036,
+      "learning_rate": 7.339689115558675e-06,
+      "loss": 0.2249,
+      "step": 14988
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1672933852184053,
+      "learning_rate": 7.33934089820996e-06,
+      "loss": 0.1551,
+      "step": 14989
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.363501511845382,
+      "learning_rate": 7.338992666334572e-06,
+      "loss": 0.1539,
+      "step": 14990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7638173555548122,
+      "learning_rate": 7.338644419934674e-06,
+      "loss": 0.4808,
+      "step": 14991
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5855270521715605,
+      "learning_rate": 7.338296159012429e-06,
+      "loss": 0.3178,
+      "step": 14992
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8137622869050427,
+      "learning_rate": 7.337947883569999e-06,
+      "loss": 0.5532,
+      "step": 14993
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9122017827367916,
+      "learning_rate": 7.3375995936095465e-06,
+      "loss": 0.3987,
+      "step": 14994
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6706128481113645,
+      "learning_rate": 7.337251289133234e-06,
+      "loss": 0.1974,
+      "step": 14995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6002071111526495,
+      "learning_rate": 7.336902970143226e-06,
+      "loss": 0.371,
+      "step": 14996
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8594599992894185,
+      "learning_rate": 7.336554636641684e-06,
+      "loss": 0.265,
+      "step": 14997
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5362332305776678,
+      "learning_rate": 7.336206288630772e-06,
+      "loss": 0.4436,
+      "step": 14998
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3619175830721137,
+      "learning_rate": 7.335857926112651e-06,
+      "loss": 0.2794,
+      "step": 14999
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0092997831294412,
+      "learning_rate": 7.335509549089487e-06,
+      "loss": 0.3189,
+      "step": 15000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0514886362676625,
+      "learning_rate": 7.335161157563442e-06,
+      "loss": 0.1944,
+      "step": 15001
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.169785542576805,
+      "learning_rate": 7.334812751536678e-06,
+      "loss": 0.1891,
+      "step": 15002
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4597793517208695,
+      "learning_rate": 7.334464331011361e-06,
+      "loss": 0.2774,
+      "step": 15003
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2445686957310886,
+      "learning_rate": 7.334115895989654e-06,
+      "loss": 0.3181,
+      "step": 15004
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2093949614461992,
+      "learning_rate": 7.333767446473721e-06,
+      "loss": 0.1724,
+      "step": 15005
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8650294425988707,
+      "learning_rate": 7.333418982465723e-06,
+      "loss": 0.4108,
+      "step": 15006
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5947555427230695,
+      "learning_rate": 7.333070503967827e-06,
+      "loss": 0.3288,
+      "step": 15007
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9776964817263298,
+      "learning_rate": 7.332722010982196e-06,
+      "loss": 0.3113,
+      "step": 15008
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7779019328149726,
+      "learning_rate": 7.332373503510996e-06,
+      "loss": 0.3865,
+      "step": 15009
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4129586149532827,
+      "learning_rate": 7.332024981556386e-06,
+      "loss": 0.2651,
+      "step": 15010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2807594034608136,
+      "learning_rate": 7.331676445120535e-06,
+      "loss": 0.2366,
+      "step": 15011
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3419925680258529,
+      "learning_rate": 7.3313278942056056e-06,
+      "loss": 0.2623,
+      "step": 15012
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.443616422065909,
+      "learning_rate": 7.330979328813761e-06,
+      "loss": 0.35,
+      "step": 15013
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4393128620480036,
+      "learning_rate": 7.330630748947167e-06,
+      "loss": 0.196,
+      "step": 15014
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.723196031595872,
+      "learning_rate": 7.3302821546079894e-06,
+      "loss": 0.4134,
+      "step": 15015
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5978203236895498,
+      "learning_rate": 7.329933545798391e-06,
+      "loss": 0.2131,
+      "step": 15016
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5854283581986435,
+      "learning_rate": 7.329584922520536e-06,
+      "loss": 0.2333,
+      "step": 15017
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.53871277403484,
+      "learning_rate": 7.329236284776593e-06,
+      "loss": 0.4333,
+      "step": 15018
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9595502186416718,
+      "learning_rate": 7.328887632568722e-06,
+      "loss": 0.3052,
+      "step": 15019
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7302841535839886,
+      "learning_rate": 7.3285389658990924e-06,
+      "loss": 0.2579,
+      "step": 15020
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4097152295067066,
+      "learning_rate": 7.328190284769866e-06,
+      "loss": 0.2072,
+      "step": 15021
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3549641293751957,
+      "learning_rate": 7.32784158918321e-06,
+      "loss": 0.1979,
+      "step": 15022
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1049516233884855,
+      "learning_rate": 7.327492879141289e-06,
+      "loss": 0.3042,
+      "step": 15023
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6938789696876684,
+      "learning_rate": 7.327144154646269e-06,
+      "loss": 0.2733,
+      "step": 15024
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.714691170528092,
+      "learning_rate": 7.326795415700314e-06,
+      "loss": 0.3313,
+      "step": 15025
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8113955758513334,
+      "learning_rate": 7.326446662305591e-06,
+      "loss": 0.4877,
+      "step": 15026
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3942581758885293,
+      "learning_rate": 7.326097894464265e-06,
+      "loss": 0.2805,
+      "step": 15027
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6134315540781203,
+      "learning_rate": 7.3257491121785015e-06,
+      "loss": 0.2489,
+      "step": 15028
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1600826574033172,
+      "learning_rate": 7.325400315450467e-06,
+      "loss": 0.2013,
+      "step": 15029
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7574463356962924,
+      "learning_rate": 7.325051504282328e-06,
+      "loss": 0.3453,
+      "step": 15030
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.063842181388818,
+      "learning_rate": 7.324702678676248e-06,
+      "loss": 0.3294,
+      "step": 15031
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4968484912030209,
+      "learning_rate": 7.324353838634396e-06,
+      "loss": 0.3309,
+      "step": 15032
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4788232586875503,
+      "learning_rate": 7.324004984158938e-06,
+      "loss": 0.3033,
+      "step": 15033
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6581711603145897,
+      "learning_rate": 7.323656115252038e-06,
+      "loss": 0.4233,
+      "step": 15034
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5564680062834837,
+      "learning_rate": 7.323307231915864e-06,
+      "loss": 0.3574,
+      "step": 15035
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8350072654897311,
+      "learning_rate": 7.3229583341525835e-06,
+      "loss": 0.5302,
+      "step": 15036
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.782204447569877,
+      "learning_rate": 7.32260942196436e-06,
+      "loss": 0.48,
+      "step": 15037
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4221308865696523,
+      "learning_rate": 7.3222604953533615e-06,
+      "loss": 0.2107,
+      "step": 15038
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4106348234788681,
+      "learning_rate": 7.321911554321757e-06,
+      "loss": 0.3496,
+      "step": 15039
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.180249965383842,
+      "learning_rate": 7.32156259887171e-06,
+      "loss": 0.4897,
+      "step": 15040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1180232604922535,
+      "learning_rate": 7.321213629005389e-06,
+      "loss": 0.5184,
+      "step": 15041
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.165267278458079,
+      "learning_rate": 7.320864644724961e-06,
+      "loss": 0.2288,
+      "step": 15042
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.225239732235837,
+      "learning_rate": 7.320515646032593e-06,
+      "loss": 0.356,
+      "step": 15043
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5891106857975998,
+      "learning_rate": 7.320166632930452e-06,
+      "loss": 0.4078,
+      "step": 15044
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.299568723005321,
+      "learning_rate": 7.319817605420707e-06,
+      "loss": 0.281,
+      "step": 15045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5704120585447443,
+      "learning_rate": 7.319468563505523e-06,
+      "loss": 0.2198,
+      "step": 15046
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4404254667575873,
+      "learning_rate": 7.319119507187068e-06,
+      "loss": 0.3486,
+      "step": 15047
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2090610622380282,
+      "learning_rate": 7.31877043646751e-06,
+      "loss": 0.5013,
+      "step": 15048
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3979663401470261,
+      "learning_rate": 7.318421351349017e-06,
+      "loss": 0.3522,
+      "step": 15049
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6676094731350832,
+      "learning_rate": 7.3180722518337545e-06,
+      "loss": 0.3482,
+      "step": 15050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9550144450042453,
+      "learning_rate": 7.317723137923893e-06,
+      "loss": 0.2003,
+      "step": 15051
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5410485499820799,
+      "learning_rate": 7.317374009621598e-06,
+      "loss": 0.3689,
+      "step": 15052
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.484864324688294,
+      "learning_rate": 7.317024866929041e-06,
+      "loss": 0.2515,
+      "step": 15053
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.598684287946512,
+      "learning_rate": 7.316675709848389e-06,
+      "loss": 0.2865,
+      "step": 15054
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6915802766988832,
+      "learning_rate": 7.316326538381806e-06,
+      "loss": 0.3847,
+      "step": 15055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5468177014152114,
+      "learning_rate": 7.315977352531465e-06,
+      "loss": 0.3751,
+      "step": 15056
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4803198183380895,
+      "learning_rate": 7.315628152299532e-06,
+      "loss": 0.4332,
+      "step": 15057
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8586950862176466,
+      "learning_rate": 7.315278937688177e-06,
+      "loss": 0.3373,
+      "step": 15058
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7032397695245616,
+      "learning_rate": 7.314929708699568e-06,
+      "loss": 0.484,
+      "step": 15059
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4754678008619715,
+      "learning_rate": 7.314580465335873e-06,
+      "loss": 0.2307,
+      "step": 15060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.02699910801109,
+      "learning_rate": 7.31423120759926e-06,
+      "loss": 0.2141,
+      "step": 15061
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2116943570256917,
+      "learning_rate": 7.3138819354919e-06,
+      "loss": 0.3764,
+      "step": 15062
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6206904357722334,
+      "learning_rate": 7.313532649015961e-06,
+      "loss": 0.2511,
+      "step": 15063
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8109384954744354,
+      "learning_rate": 7.313183348173611e-06,
+      "loss": 0.4992,
+      "step": 15064
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1054899218172443,
+      "learning_rate": 7.31283403296702e-06,
+      "loss": 0.4226,
+      "step": 15065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3650548131953466,
+      "learning_rate": 7.312484703398356e-06,
+      "loss": 0.1979,
+      "step": 15066
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9408608789265014,
+      "learning_rate": 7.31213535946979e-06,
+      "loss": 0.4331,
+      "step": 15067
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8756806409769573,
+      "learning_rate": 7.31178600118349e-06,
+      "loss": 0.2772,
+      "step": 15068
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.796120891722715,
+      "learning_rate": 7.311436628541627e-06,
+      "loss": 0.2704,
+      "step": 15069
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.911373998425537,
+      "learning_rate": 7.311087241546368e-06,
+      "loss": 0.3705,
+      "step": 15070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.071102004294167,
+      "learning_rate": 7.310737840199886e-06,
+      "loss": 0.1834,
+      "step": 15071
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7001376594177031,
+      "learning_rate": 7.310388424504347e-06,
+      "loss": 0.3501,
+      "step": 15072
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.411948430979034,
+      "learning_rate": 7.310038994461924e-06,
+      "loss": 0.2416,
+      "step": 15073
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2496259964281395,
+      "learning_rate": 7.309689550074783e-06,
+      "loss": 0.1975,
+      "step": 15074
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9053230611575078,
+      "learning_rate": 7.3093400913450985e-06,
+      "loss": 0.3705,
+      "step": 15075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8745874865065426,
+      "learning_rate": 7.308990618275037e-06,
+      "loss": 0.5899,
+      "step": 15076
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9876370018585656,
+      "learning_rate": 7.30864113086677e-06,
+      "loss": 0.203,
+      "step": 15077
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4141565562944551,
+      "learning_rate": 7.308291629122468e-06,
+      "loss": 0.1883,
+      "step": 15078
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.344848452759608,
+      "learning_rate": 7.307942113044302e-06,
+      "loss": 0.2509,
+      "step": 15079
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8152264358854103,
+      "learning_rate": 7.307592582634441e-06,
+      "loss": 0.1244,
+      "step": 15080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2297928248592567,
+      "learning_rate": 7.307243037895054e-06,
+      "loss": 0.4113,
+      "step": 15081
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.632764886317002,
+      "learning_rate": 7.306893478828317e-06,
+      "loss": 0.2859,
+      "step": 15082
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9332919145297824,
+      "learning_rate": 7.306543905436394e-06,
+      "loss": 0.5169,
+      "step": 15083
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9730388270372388,
+      "learning_rate": 7.30619431772146e-06,
+      "loss": 0.5408,
+      "step": 15084
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2643739600604302,
+      "learning_rate": 7.3058447156856834e-06,
+      "loss": 0.4341,
+      "step": 15085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.022509866709272,
+      "learning_rate": 7.305495099331239e-06,
+      "loss": 0.7462,
+      "step": 15086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6722671369774338,
+      "learning_rate": 7.305145468660295e-06,
+      "loss": 0.3626,
+      "step": 15087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.674028159866909,
+      "learning_rate": 7.3047958236750214e-06,
+      "loss": 0.4764,
+      "step": 15088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8957734351774047,
+      "learning_rate": 7.3044461643775896e-06,
+      "loss": 0.4604,
+      "step": 15089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0413306378733413,
+      "learning_rate": 7.304096490770175e-06,
+      "loss": 0.2379,
+      "step": 15090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3352527062365975,
+      "learning_rate": 7.303746802854945e-06,
+      "loss": 0.1837,
+      "step": 15091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7961846020433359,
+      "learning_rate": 7.3033971006340705e-06,
+      "loss": 0.4498,
+      "step": 15092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8981340464067724,
+      "learning_rate": 7.303047384109727e-06,
+      "loss": 0.5613,
+      "step": 15093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.998622792641112,
+      "learning_rate": 7.302697653284083e-06,
+      "loss": 0.3111,
+      "step": 15094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4798168040277146,
+      "learning_rate": 7.302347908159312e-06,
+      "loss": 0.4017,
+      "step": 15095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4449199813661235,
+      "learning_rate": 7.301998148737584e-06,
+      "loss": 0.3194,
+      "step": 15096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7956509442949669,
+      "learning_rate": 7.301648375021072e-06,
+      "loss": 0.6006,
+      "step": 15097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2544149929469992,
+      "learning_rate": 7.301298587011948e-06,
+      "loss": 0.1874,
+      "step": 15098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6132939317403387,
+      "learning_rate": 7.300948784712384e-06,
+      "loss": 0.2237,
+      "step": 15099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.255077192636827,
+      "learning_rate": 7.300598968124552e-06,
+      "loss": 0.3219,
+      "step": 15100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.447036400684795,
+      "learning_rate": 7.300249137250623e-06,
+      "loss": 0.3988,
+      "step": 15101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8236012301938354,
+      "learning_rate": 7.299899292092773e-06,
+      "loss": 0.2675,
+      "step": 15102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0410821725212194,
+      "learning_rate": 7.29954943265317e-06,
+      "loss": 0.2724,
+      "step": 15103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0117236294733516,
+      "learning_rate": 7.29919955893399e-06,
+      "loss": 0.3428,
+      "step": 15104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2598080790559432,
+      "learning_rate": 7.298849670937403e-06,
+      "loss": 0.2021,
+      "step": 15105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5415434326285626,
+      "learning_rate": 7.2984997686655855e-06,
+      "loss": 0.3149,
+      "step": 15106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7083602245082898,
+      "learning_rate": 7.298149852120705e-06,
+      "loss": 0.4721,
+      "step": 15107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6303189735730108,
+      "learning_rate": 7.29779992130494e-06,
+      "loss": 0.3571,
+      "step": 15108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3743654596761765,
+      "learning_rate": 7.297449976220459e-06,
+      "loss": 0.2341,
+      "step": 15109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2962699823209607,
+      "learning_rate": 7.297100016869436e-06,
+      "loss": 0.3776,
+      "step": 15110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1629630871903587,
+      "learning_rate": 7.296750043254046e-06,
+      "loss": 0.2653,
+      "step": 15111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.070853863890703,
+      "learning_rate": 7.2964000553764604e-06,
+      "loss": 0.2171,
+      "step": 15112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8292981503938557,
+      "learning_rate": 7.296050053238854e-06,
+      "loss": 0.4142,
+      "step": 15113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2118096333527413,
+      "learning_rate": 7.295700036843399e-06,
+      "loss": 0.1531,
+      "step": 15114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4324431700485034,
+      "learning_rate": 7.29535000619227e-06,
+      "loss": 0.325,
+      "step": 15115
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0471909548515295,
+      "learning_rate": 7.2949999612876395e-06,
+      "loss": 0.1977,
+      "step": 15116
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2977143151224042,
+      "learning_rate": 7.294649902131681e-06,
+      "loss": 0.3532,
+      "step": 15117
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.582710143994565,
+      "learning_rate": 7.29429982872657e-06,
+      "loss": 0.2964,
+      "step": 15118
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1324383339700659,
+      "learning_rate": 7.293949741074478e-06,
+      "loss": 0.1693,
+      "step": 15119
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6418170390409965,
+      "learning_rate": 7.293599639177582e-06,
+      "loss": 0.3101,
+      "step": 15120
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.966805584006551,
+      "learning_rate": 7.293249523038054e-06,
+      "loss": 0.4691,
+      "step": 15121
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6309063887529647,
+      "learning_rate": 7.292899392658068e-06,
+      "loss": 0.5713,
+      "step": 15122
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2159595831037546,
+      "learning_rate": 7.292549248039799e-06,
+      "loss": 0.1034,
+      "step": 15123
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.650150685672573,
+      "learning_rate": 7.2921990891854195e-06,
+      "loss": 0.5316,
+      "step": 15124
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7189686657784562,
+      "learning_rate": 7.291848916097107e-06,
+      "loss": 0.3499,
+      "step": 15125
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5317254423139064,
+      "learning_rate": 7.291498728777033e-06,
+      "loss": 0.3823,
+      "step": 15126
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6896178479457773,
+      "learning_rate": 7.291148527227374e-06,
+      "loss": 0.3998,
+      "step": 15127
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0217401831015787,
+      "learning_rate": 7.290798311450305e-06,
+      "loss": 0.5877,
+      "step": 15128
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.67021882389551,
+      "learning_rate": 7.290448081447998e-06,
+      "loss": 0.2925,
+      "step": 15129
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4101105759216623,
+      "learning_rate": 7.29009783722263e-06,
+      "loss": 0.1763,
+      "step": 15130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.322379765169418,
+      "learning_rate": 7.289747578776377e-06,
+      "loss": 0.1876,
+      "step": 15131
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5126630189821213,
+      "learning_rate": 7.289397306111411e-06,
+      "loss": 0.2641,
+      "step": 15132
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8362248743661103,
+      "learning_rate": 7.289047019229909e-06,
+      "loss": 0.3318,
+      "step": 15133
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2159841802387168,
+      "learning_rate": 7.288696718134046e-06,
+      "loss": 0.2455,
+      "step": 15134
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6945568985007016,
+      "learning_rate": 7.2883464028259974e-06,
+      "loss": 0.1796,
+      "step": 15135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8879475984948915,
+      "learning_rate": 7.287996073307937e-06,
+      "loss": 0.3405,
+      "step": 15136
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8211012142557015,
+      "learning_rate": 7.287645729582043e-06,
+      "loss": 0.4022,
+      "step": 15137
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7230012422796277,
+      "learning_rate": 7.287295371650488e-06,
+      "loss": 0.2062,
+      "step": 15138
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.476903924498349,
+      "learning_rate": 7.286944999515451e-06,
+      "loss": 0.2156,
+      "step": 15139
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.135745089947159,
+      "learning_rate": 7.286594613179103e-06,
+      "loss": 0.3443,
+      "step": 15140
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5053176449701036,
+      "learning_rate": 7.286244212643625e-06,
+      "loss": 0.1577,
+      "step": 15141
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6767476903024456,
+      "learning_rate": 7.28589379791119e-06,
+      "loss": 0.498,
+      "step": 15142
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7634531880400335,
+      "learning_rate": 7.285543368983974e-06,
+      "loss": 0.343,
+      "step": 15143
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5185511181166915,
+      "learning_rate": 7.285192925864154e-06,
+      "loss": 0.3243,
+      "step": 15144
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7187232275524953,
+      "learning_rate": 7.284842468553904e-06,
+      "loss": 0.5663,
+      "step": 15145
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5550468714777974,
+      "learning_rate": 7.284491997055405e-06,
+      "loss": 0.5151,
+      "step": 15146
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.186418255639631,
+      "learning_rate": 7.284141511370828e-06,
+      "loss": 0.3672,
+      "step": 15147
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6228780371737361,
+      "learning_rate": 7.2837910115023525e-06,
+      "loss": 0.5002,
+      "step": 15148
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4043080298945014,
+      "learning_rate": 7.283440497452153e-06,
+      "loss": 0.2352,
+      "step": 15149
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6567064389587465,
+      "learning_rate": 7.28308996922241e-06,
+      "loss": 0.2389,
+      "step": 15150
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6958559578119403,
+      "learning_rate": 7.282739426815295e-06,
+      "loss": 0.2209,
+      "step": 15151
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8526680177036738,
+      "learning_rate": 7.282388870232988e-06,
+      "loss": 0.3202,
+      "step": 15152
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4611653237624513,
+      "learning_rate": 7.2820382994776665e-06,
+      "loss": 0.3615,
+      "step": 15153
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8747408528850666,
+      "learning_rate": 7.281687714551504e-06,
+      "loss": 0.2701,
+      "step": 15154
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.323263206174237,
+      "learning_rate": 7.2813371154566805e-06,
+      "loss": 0.4132,
+      "step": 15155
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9471622351591613,
+      "learning_rate": 7.280986502195372e-06,
+      "loss": 0.4746,
+      "step": 15156
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3193251399707357,
+      "learning_rate": 7.280635874769756e-06,
+      "loss": 0.2892,
+      "step": 15157
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7746123757862848,
+      "learning_rate": 7.280285233182009e-06,
+      "loss": 0.5287,
+      "step": 15158
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9604353432740307,
+      "learning_rate": 7.279934577434311e-06,
+      "loss": 0.4714,
+      "step": 15159
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9171221121135569,
+      "learning_rate": 7.279583907528836e-06,
+      "loss": 0.5771,
+      "step": 15160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4696352433848867,
+      "learning_rate": 7.279233223467764e-06,
+      "loss": 0.36,
+      "step": 15161
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.117726114953352,
+      "learning_rate": 7.278882525253271e-06,
+      "loss": 0.301,
+      "step": 15162
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0348660620118995,
+      "learning_rate": 7.278531812887537e-06,
+      "loss": 0.2079,
+      "step": 15163
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4959038297658362,
+      "learning_rate": 7.278181086372736e-06,
+      "loss": 0.2404,
+      "step": 15164
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7116204601362195,
+      "learning_rate": 7.277830345711051e-06,
+      "loss": 0.4965,
+      "step": 15165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5359018325544995,
+      "learning_rate": 7.277479590904656e-06,
+      "loss": 0.297,
+      "step": 15166
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8571645115150577,
+      "learning_rate": 7.277128821955731e-06,
+      "loss": 0.61,
+      "step": 15167
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1226493740567127,
+      "learning_rate": 7.2767780388664526e-06,
+      "loss": 0.276,
+      "step": 15168
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1987077761364844,
+      "learning_rate": 7.276427241639e-06,
+      "loss": 0.179,
+      "step": 15169
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5888929359117547,
+      "learning_rate": 7.2760764302755525e-06,
+      "loss": 0.3285,
+      "step": 15170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.299410029969408,
+      "learning_rate": 7.2757256047782875e-06,
+      "loss": 0.2415,
+      "step": 15171
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.930570211333929,
+      "learning_rate": 7.275374765149383e-06,
+      "loss": 0.5451,
+      "step": 15172
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0925200256295253,
+      "learning_rate": 7.275023911391019e-06,
+      "loss": 0.1497,
+      "step": 15173
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6540645404689938,
+      "learning_rate": 7.2746730435053735e-06,
+      "loss": 0.3694,
+      "step": 15174
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9057922869065056,
+      "learning_rate": 7.274322161494627e-06,
+      "loss": 0.4836,
+      "step": 15175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7057549355771087,
+      "learning_rate": 7.273971265360954e-06,
+      "loss": 0.5201,
+      "step": 15176
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.086105580666769,
+      "learning_rate": 7.273620355106538e-06,
+      "loss": 0.1603,
+      "step": 15177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.529144152814241,
+      "learning_rate": 7.273269430733554e-06,
+      "loss": 0.3382,
+      "step": 15178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6231250675246938,
+      "learning_rate": 7.272918492244185e-06,
+      "loss": 0.3245,
+      "step": 15179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.530490453312089,
+      "learning_rate": 7.272567539640608e-06,
+      "loss": 0.3561,
+      "step": 15180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.934286495454351,
+      "learning_rate": 7.272216572925004e-06,
+      "loss": 0.762,
+      "step": 15181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2378903698642574,
+      "learning_rate": 7.2718655920995496e-06,
+      "loss": 0.2297,
+      "step": 15182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.227036676048757,
+      "learning_rate": 7.271514597166427e-06,
+      "loss": 0.1898,
+      "step": 15183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.457527198963028,
+      "learning_rate": 7.271163588127814e-06,
+      "loss": 0.3212,
+      "step": 15184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.984954861901228,
+      "learning_rate": 7.270812564985893e-06,
+      "loss": 0.5249,
+      "step": 15185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.027265407540634,
+      "learning_rate": 7.270461527742841e-06,
+      "loss": 0.5889,
+      "step": 15186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6825794076248242,
+      "learning_rate": 7.2701104764008375e-06,
+      "loss": 0.353,
+      "step": 15187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4298857905764988,
+      "learning_rate": 7.269759410962065e-06,
+      "loss": 0.4766,
+      "step": 15188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3753937815171757,
+      "learning_rate": 7.269408331428701e-06,
+      "loss": 0.2064,
+      "step": 15189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.466705893426362,
+      "learning_rate": 7.269057237802928e-06,
+      "loss": 0.2822,
+      "step": 15190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3557884075744777,
+      "learning_rate": 7.268706130086925e-06,
+      "loss": 0.166,
+      "step": 15191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4916780985384388,
+      "learning_rate": 7.26835500828287e-06,
+      "loss": 0.3329,
+      "step": 15192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4348931465423778,
+      "learning_rate": 7.268003872392948e-06,
+      "loss": 0.3636,
+      "step": 15193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.64753982545013,
+      "learning_rate": 7.267652722419335e-06,
+      "loss": 0.3858,
+      "step": 15194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3302170657103032,
+      "learning_rate": 7.267301558364215e-06,
+      "loss": 0.3449,
+      "step": 15195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2095186442694148,
+      "learning_rate": 7.266950380229767e-06,
+      "loss": 0.3248,
+      "step": 15196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3847370306164546,
+      "learning_rate": 7.266599188018172e-06,
+      "loss": 0.3134,
+      "step": 15197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.13236434825386,
+      "learning_rate": 7.26624798173161e-06,
+      "loss": 0.4273,
+      "step": 15198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3642731367382395,
+      "learning_rate": 7.265896761372264e-06,
+      "loss": 0.1871,
+      "step": 15199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.038323785623882,
+      "learning_rate": 7.265545526942313e-06,
+      "loss": 0.3261,
+      "step": 15200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7668693090751333,
+      "learning_rate": 7.2651942784439386e-06,
+      "loss": 0.4676,
+      "step": 15201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5879686554602157,
+      "learning_rate": 7.264843015879321e-06,
+      "loss": 0.3738,
+      "step": 15202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9893786543939305,
+      "learning_rate": 7.264491739250645e-06,
+      "loss": 0.2651,
+      "step": 15203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.338523714956983,
+      "learning_rate": 7.264140448560088e-06,
+      "loss": 0.343,
+      "step": 15204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6320822282775547,
+      "learning_rate": 7.263789143809834e-06,
+      "loss": 0.4696,
+      "step": 15205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6504955190697703,
+      "learning_rate": 7.263437825002061e-06,
+      "loss": 0.4808,
+      "step": 15206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9771560619849006,
+      "learning_rate": 7.2630864921389554e-06,
+      "loss": 0.121,
+      "step": 15207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7943654739010093,
+      "learning_rate": 7.262735145222696e-06,
+      "loss": 0.402,
+      "step": 15208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7520923516653373,
+      "learning_rate": 7.262383784255463e-06,
+      "loss": 0.2209,
+      "step": 15209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6908190282724462,
+      "learning_rate": 7.262032409239442e-06,
+      "loss": 0.3685,
+      "step": 15210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7306256922730467,
+      "learning_rate": 7.261681020176813e-06,
+      "loss": 0.1976,
+      "step": 15211
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6046896689343675,
+      "learning_rate": 7.261329617069759e-06,
+      "loss": 0.2423,
+      "step": 15212
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4311681178462352,
+      "learning_rate": 7.2609781999204595e-06,
+      "loss": 0.2591,
+      "step": 15213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.202807451117651,
+      "learning_rate": 7.2606267687311005e-06,
+      "loss": 0.1667,
+      "step": 15214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1053606357883534,
+      "learning_rate": 7.26027532350386e-06,
+      "loss": 0.2012,
+      "step": 15215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7000760031638205,
+      "learning_rate": 7.259923864240926e-06,
+      "loss": 0.4708,
+      "step": 15216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7755857681443856,
+      "learning_rate": 7.259572390944475e-06,
+      "loss": 0.5004,
+      "step": 15217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.776474641692486,
+      "learning_rate": 7.2592209036166925e-06,
+      "loss": 0.5104,
+      "step": 15218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.147436317902942,
+      "learning_rate": 7.258869402259763e-06,
+      "loss": 0.5204,
+      "step": 15219
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.437445867597278,
+      "learning_rate": 7.258517886875864e-06,
+      "loss": 0.3951,
+      "step": 15220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8770952915173444,
+      "learning_rate": 7.2581663574671845e-06,
+      "loss": 0.5653,
+      "step": 15221
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5560903354304327,
+      "learning_rate": 7.257814814035903e-06,
+      "loss": 0.3514,
+      "step": 15222
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.562782426581748,
+      "learning_rate": 7.257463256584204e-06,
+      "loss": 0.4703,
+      "step": 15223
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7494694659161087,
+      "learning_rate": 7.25711168511427e-06,
+      "loss": 0.5182,
+      "step": 15224
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.162547395911748,
+      "learning_rate": 7.256760099628285e-06,
+      "loss": 0.2092,
+      "step": 15225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.428645841063322,
+      "learning_rate": 7.256408500128431e-06,
+      "loss": 0.2928,
+      "step": 15226
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7313932571288537,
+      "learning_rate": 7.256056886616894e-06,
+      "loss": 0.528,
+      "step": 15227
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.527985673288564,
+      "learning_rate": 7.255705259095853e-06,
+      "loss": 0.1881,
+      "step": 15228
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5322804584348255,
+      "learning_rate": 7.255353617567497e-06,
+      "loss": 0.2256,
+      "step": 15229
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.815919564788273,
+      "learning_rate": 7.255001962034005e-06,
+      "loss": 0.4814,
+      "step": 15230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9028388599080384,
+      "learning_rate": 7.254650292497562e-06,
+      "loss": 0.2319,
+      "step": 15231
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3577727214311228,
+      "learning_rate": 7.254298608960354e-06,
+      "loss": 0.165,
+      "step": 15232
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4705275599237624,
+      "learning_rate": 7.253946911424562e-06,
+      "loss": 0.2338,
+      "step": 15233
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0241020355969077,
+      "learning_rate": 7.253595199892372e-06,
+      "loss": 0.2976,
+      "step": 15234
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6990774337058943,
+      "learning_rate": 7.253243474365964e-06,
+      "loss": 0.4869,
+      "step": 15235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.13830674420714,
+      "learning_rate": 7.252891734847528e-06,
+      "loss": 0.5676,
+      "step": 15236
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6761666746109973,
+      "learning_rate": 7.252539981339244e-06,
+      "loss": 0.1973,
+      "step": 15237
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4361262391488714,
+      "learning_rate": 7.2521882138432985e-06,
+      "loss": 0.3023,
+      "step": 15238
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5523574738230064,
+      "learning_rate": 7.251836432361874e-06,
+      "loss": 0.3777,
+      "step": 15239
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7122931922865268,
+      "learning_rate": 7.251484636897157e-06,
+      "loss": 0.2154,
+      "step": 15240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.019361463009637,
+      "learning_rate": 7.251132827451329e-06,
+      "loss": 0.302,
+      "step": 15241
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.738285493309757,
+      "learning_rate": 7.250781004026579e-06,
+      "loss": 0.1952,
+      "step": 15242
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5164622637232754,
+      "learning_rate": 7.2504291666250885e-06,
+      "loss": 0.2917,
+      "step": 15243
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5192047248827747,
+      "learning_rate": 7.250077315249042e-06,
+      "loss": 0.3396,
+      "step": 15244
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1130043906900056,
+      "learning_rate": 7.249725449900628e-06,
+      "loss": 0.2063,
+      "step": 15245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3183069939273877,
+      "learning_rate": 7.249373570582027e-06,
+      "loss": 0.3832,
+      "step": 15246
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7690452177668865,
+      "learning_rate": 7.249021677295428e-06,
+      "loss": 0.223,
+      "step": 15247
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.876611116749886,
+      "learning_rate": 7.248669770043013e-06,
+      "loss": 0.4433,
+      "step": 15248
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9563054149292571,
+      "learning_rate": 7.248317848826969e-06,
+      "loss": 0.377,
+      "step": 15249
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8984424685189492,
+      "learning_rate": 7.2479659136494805e-06,
+      "loss": 0.3518,
+      "step": 15250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5815269655487976,
+      "learning_rate": 7.247613964512734e-06,
+      "loss": 0.3926,
+      "step": 15251
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2955616712481506,
+      "learning_rate": 7.247262001418915e-06,
+      "loss": 0.2676,
+      "step": 15252
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7293592430985107,
+      "learning_rate": 7.2469100243702076e-06,
+      "loss": 0.3896,
+      "step": 15253
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.626412994648573,
+      "learning_rate": 7.246558033368798e-06,
+      "loss": 0.3141,
+      "step": 15254
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6490060494732977,
+      "learning_rate": 7.246206028416872e-06,
+      "loss": 0.4264,
+      "step": 15255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6568484889441706,
+      "learning_rate": 7.245854009516617e-06,
+      "loss": 0.3502,
+      "step": 15256
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7990307357287199,
+      "learning_rate": 7.245501976670217e-06,
+      "loss": 0.4317,
+      "step": 15257
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.656463586782688,
+      "learning_rate": 7.24514992987986e-06,
+      "loss": 0.4295,
+      "step": 15258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6999039726273626,
+      "learning_rate": 7.244797869147728e-06,
+      "loss": 0.1818,
+      "step": 15259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1184511407530355,
+      "learning_rate": 7.2444457944760125e-06,
+      "loss": 0.1792,
+      "step": 15260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.416523122936555,
+      "learning_rate": 7.244093705866896e-06,
+      "loss": 0.2759,
+      "step": 15261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.978963687280041,
+      "learning_rate": 7.243741603322567e-06,
+      "loss": 0.1824,
+      "step": 15262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5389608843413765,
+      "learning_rate": 7.243389486845211e-06,
+      "loss": 0.3508,
+      "step": 15263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.060422893610454,
+      "learning_rate": 7.243037356437015e-06,
+      "loss": 0.3566,
+      "step": 15264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5362473219032626,
+      "learning_rate": 7.242685212100165e-06,
+      "loss": 0.4024,
+      "step": 15265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7190910932898256,
+      "learning_rate": 7.242333053836848e-06,
+      "loss": 0.3271,
+      "step": 15266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0954777524907664,
+      "learning_rate": 7.24198088164925e-06,
+      "loss": 0.2287,
+      "step": 15267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4062963610530417,
+      "learning_rate": 7.24162869553956e-06,
+      "loss": 0.3318,
+      "step": 15268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.087019752107492,
+      "learning_rate": 7.241276495509963e-06,
+      "loss": 0.267,
+      "step": 15269
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.748562826168312,
+      "learning_rate": 7.240924281562647e-06,
+      "loss": 0.2249,
+      "step": 15270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4643773391784727,
+      "learning_rate": 7.240572053699799e-06,
+      "loss": 0.2943,
+      "step": 15271
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9154529970372207,
+      "learning_rate": 7.240219811923608e-06,
+      "loss": 0.3859,
+      "step": 15272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.426660095722459,
+      "learning_rate": 7.239867556236257e-06,
+      "loss": 0.206,
+      "step": 15273
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2601979068462046,
+      "learning_rate": 7.239515286639937e-06,
+      "loss": 0.202,
+      "step": 15274
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5473332689576107,
+      "learning_rate": 7.239163003136835e-06,
+      "loss": 0.305,
+      "step": 15275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5789423893601147,
+      "learning_rate": 7.238810705729137e-06,
+      "loss": 0.3173,
+      "step": 15276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.443008624782144,
+      "learning_rate": 7.238458394419032e-06,
+      "loss": 0.2191,
+      "step": 15277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6528733579351522,
+      "learning_rate": 7.238106069208708e-06,
+      "loss": 0.2614,
+      "step": 15278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.900247425357142,
+      "learning_rate": 7.237753730100351e-06,
+      "loss": 0.3467,
+      "step": 15279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.339991967258112,
+      "learning_rate": 7.2374013770961515e-06,
+      "loss": 0.2078,
+      "step": 15280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3716930074260956,
+      "learning_rate": 7.237049010198296e-06,
+      "loss": 0.2904,
+      "step": 15281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.547099918332146,
+      "learning_rate": 7.236696629408972e-06,
+      "loss": 0.1299,
+      "step": 15282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9471362251745015,
+      "learning_rate": 7.23634423473037e-06,
+      "loss": 0.2464,
+      "step": 15283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7720354184188365,
+      "learning_rate": 7.2359918261646755e-06,
+      "loss": 0.2625,
+      "step": 15284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9850102406581127,
+      "learning_rate": 7.23563940371408e-06,
+      "loss": 0.2572,
+      "step": 15285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5077278874429323,
+      "learning_rate": 7.235286967380768e-06,
+      "loss": 0.1959,
+      "step": 15286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.422396291813962,
+      "learning_rate": 7.234934517166931e-06,
+      "loss": 0.1965,
+      "step": 15287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4252396992121477,
+      "learning_rate": 7.234582053074757e-06,
+      "loss": 0.2143,
+      "step": 15288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.890921879249193,
+      "learning_rate": 7.234229575106434e-06,
+      "loss": 0.5067,
+      "step": 15289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.605583172918214,
+      "learning_rate": 7.233877083264152e-06,
+      "loss": 0.1525,
+      "step": 15290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7264220322995736,
+      "learning_rate": 7.233524577550099e-06,
+      "loss": 0.4021,
+      "step": 15291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.66049369524052,
+      "learning_rate": 7.233172057966463e-06,
+      "loss": 0.3406,
+      "step": 15292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8001379779742273,
+      "learning_rate": 7.2328195245154345e-06,
+      "loss": 0.3917,
+      "step": 15293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3550546106897343,
+      "learning_rate": 7.232466977199203e-06,
+      "loss": 0.3371,
+      "step": 15294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4090197228507555,
+      "learning_rate": 7.2321144160199565e-06,
+      "loss": 0.2197,
+      "step": 15295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2636828296782046,
+      "learning_rate": 7.231761840979885e-06,
+      "loss": 0.262,
+      "step": 15296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6763216250042432,
+      "learning_rate": 7.231409252081179e-06,
+      "loss": 0.1735,
+      "step": 15297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4368671755434559,
+      "learning_rate": 7.2310566493260245e-06,
+      "loss": 0.2172,
+      "step": 15298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8296612943732116,
+      "learning_rate": 7.230704032716615e-06,
+      "loss": 0.6307,
+      "step": 15299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.539543749511611,
+      "learning_rate": 7.230351402255138e-06,
+      "loss": 0.2131,
+      "step": 15300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6426822706447792,
+      "learning_rate": 7.2299987579437834e-06,
+      "loss": 0.3174,
+      "step": 15301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6127851337608659,
+      "learning_rate": 7.229646099784742e-06,
+      "loss": 0.3646,
+      "step": 15302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.191345839594237,
+      "learning_rate": 7.2292934277802016e-06,
+      "loss": 0.2816,
+      "step": 15303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.878700558353858,
+      "learning_rate": 7.228940741932355e-06,
+      "loss": 0.5433,
+      "step": 15304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2756736851645913,
+      "learning_rate": 7.228588042243391e-06,
+      "loss": 0.1669,
+      "step": 15305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8231289077000266,
+      "learning_rate": 7.228235328715498e-06,
+      "loss": 0.4102,
+      "step": 15306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4604172417734331,
+      "learning_rate": 7.2278826013508704e-06,
+      "loss": 0.2981,
+      "step": 15307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6490711377401794,
+      "learning_rate": 7.2275298601516944e-06,
+      "loss": 0.2136,
+      "step": 15308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8821404907965784,
+      "learning_rate": 7.227177105120162e-06,
+      "loss": 0.4094,
+      "step": 15309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.504710004961063,
+      "learning_rate": 7.226824336258465e-06,
+      "loss": 0.1575,
+      "step": 15310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2776058552224527,
+      "learning_rate": 7.226471553568791e-06,
+      "loss": 0.2301,
+      "step": 15311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.697777033448923,
+      "learning_rate": 7.226118757053333e-06,
+      "loss": 0.4082,
+      "step": 15312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.652820869982856,
+      "learning_rate": 7.225765946714283e-06,
+      "loss": 0.4199,
+      "step": 15313
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7491853725286701,
+      "learning_rate": 7.2254131225538285e-06,
+      "loss": 0.2185,
+      "step": 15314
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9241170675877113,
+      "learning_rate": 7.225060284574162e-06,
+      "loss": 0.4073,
+      "step": 15315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5218925956789844,
+      "learning_rate": 7.224707432777475e-06,
+      "loss": 0.3388,
+      "step": 15316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.606510098435095,
+      "learning_rate": 7.224354567165959e-06,
+      "loss": 0.3038,
+      "step": 15317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.897309660955208,
+      "learning_rate": 7.224001687741803e-06,
+      "loss": 0.2134,
+      "step": 15318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3380941013504188,
+      "learning_rate": 7.2236487945071995e-06,
+      "loss": 0.3272,
+      "step": 15319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.166979562345278,
+      "learning_rate": 7.223295887464341e-06,
+      "loss": 0.3858,
+      "step": 15320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4568510698380768,
+      "learning_rate": 7.222942966615419e-06,
+      "loss": 0.1737,
+      "step": 15321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.90369640566703,
+      "learning_rate": 7.222590031962622e-06,
+      "loss": 0.4372,
+      "step": 15322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1389477953304994,
+      "learning_rate": 7.222237083508144e-06,
+      "loss": 0.239,
+      "step": 15323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5825788898158275,
+      "learning_rate": 7.221884121254178e-06,
+      "loss": 0.1716,
+      "step": 15324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5903243807549206,
+      "learning_rate": 7.221531145202913e-06,
+      "loss": 0.3027,
+      "step": 15325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1355453982473191,
+      "learning_rate": 7.221178155356542e-06,
+      "loss": 0.1637,
+      "step": 15326
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2630038227495755,
+      "learning_rate": 7.220825151717258e-06,
+      "loss": 0.31,
+      "step": 15327
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7308513786666695,
+      "learning_rate": 7.2204721342872506e-06,
+      "loss": 0.3604,
+      "step": 15328
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5909577948992248,
+      "learning_rate": 7.220119103068716e-06,
+      "loss": 0.3786,
+      "step": 15329
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1616924266363644,
+      "learning_rate": 7.219766058063841e-06,
+      "loss": 0.61,
+      "step": 15330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5246939734512743,
+      "learning_rate": 7.219412999274822e-06,
+      "loss": 0.1906,
+      "step": 15331
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0473153135765556,
+      "learning_rate": 7.219059926703851e-06,
+      "loss": 0.5721,
+      "step": 15332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.03914588085421,
+      "learning_rate": 7.218706840353118e-06,
+      "loss": 0.1399,
+      "step": 15333
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.705546329007715,
+      "learning_rate": 7.218353740224818e-06,
+      "loss": 0.5001,
+      "step": 15334
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.122167528288654,
+      "learning_rate": 7.218000626321145e-06,
+      "loss": 0.1592,
+      "step": 15335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.004147420279892,
+      "learning_rate": 7.217647498644286e-06,
+      "loss": 0.4099,
+      "step": 15336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.641094265537931,
+      "learning_rate": 7.21729435719644e-06,
+      "loss": 0.3791,
+      "step": 15337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7327728690043251,
+      "learning_rate": 7.216941201979797e-06,
+      "loss": 0.446,
+      "step": 15338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0955078780619192,
+      "learning_rate": 7.216588032996549e-06,
+      "loss": 0.1878,
+      "step": 15339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.129609717937884,
+      "learning_rate": 7.216234850248892e-06,
+      "loss": 0.2085,
+      "step": 15340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5377719619235748,
+      "learning_rate": 7.215881653739016e-06,
+      "loss": 0.202,
+      "step": 15341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5904609058740216,
+      "learning_rate": 7.215528443469118e-06,
+      "loss": 0.4366,
+      "step": 15342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5377993919241806,
+      "learning_rate": 7.215175219441388e-06,
+      "loss": 0.2371,
+      "step": 15343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7736442262676944,
+      "learning_rate": 7.21482198165802e-06,
+      "loss": 0.3139,
+      "step": 15344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.479108066864401,
+      "learning_rate": 7.214468730121209e-06,
+      "loss": 0.325,
+      "step": 15345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.195574222181678,
+      "learning_rate": 7.214115464833148e-06,
+      "loss": 0.358,
+      "step": 15346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.229702108216431,
+      "learning_rate": 7.2137621857960295e-06,
+      "loss": 0.1643,
+      "step": 15347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6913986205882618,
+      "learning_rate": 7.213408893012049e-06,
+      "loss": 0.2066,
+      "step": 15348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5432220202832079,
+      "learning_rate": 7.213055586483401e-06,
+      "loss": 0.2842,
+      "step": 15349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5392710113637154,
+      "learning_rate": 7.212702266212276e-06,
+      "loss": 0.2804,
+      "step": 15350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6303624201175717,
+      "learning_rate": 7.212348932200871e-06,
+      "loss": 0.4412,
+      "step": 15351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1002847315191384,
+      "learning_rate": 7.211995584451378e-06,
+      "loss": 0.1638,
+      "step": 15352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8453781225275854,
+      "learning_rate": 7.211642222965993e-06,
+      "loss": 0.2369,
+      "step": 15353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6677795990802056,
+      "learning_rate": 7.21128884774691e-06,
+      "loss": 0.4585,
+      "step": 15354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4190702731195364,
+      "learning_rate": 7.210935458796323e-06,
+      "loss": 0.1973,
+      "step": 15355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8972487141996535,
+      "learning_rate": 7.210582056116425e-06,
+      "loss": 0.2633,
+      "step": 15356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4226376039476092,
+      "learning_rate": 7.210228639709413e-06,
+      "loss": 0.2717,
+      "step": 15357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9991498738506919,
+      "learning_rate": 7.209875209577481e-06,
+      "loss": 0.4302,
+      "step": 15358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.397745216826086,
+      "learning_rate": 7.2095217657228234e-06,
+      "loss": 0.2325,
+      "step": 15359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.488627816446401,
+      "learning_rate": 7.209168308147635e-06,
+      "loss": 0.3417,
+      "step": 15360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7995751183281765,
+      "learning_rate": 7.20881483685411e-06,
+      "loss": 0.2186,
+      "step": 15361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4011061102878257,
+      "learning_rate": 7.208461351844443e-06,
+      "loss": 0.3019,
+      "step": 15362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4587012977670786,
+      "learning_rate": 7.208107853120832e-06,
+      "loss": 0.3049,
+      "step": 15363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.708455787656407,
+      "learning_rate": 7.207754340685469e-06,
+      "loss": 0.402,
+      "step": 15364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4371635312750277,
+      "learning_rate": 7.20740081454055e-06,
+      "loss": 0.3686,
+      "step": 15365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9803549980285617,
+      "learning_rate": 7.20704727468827e-06,
+      "loss": 0.2208,
+      "step": 15366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8002086130259056,
+      "learning_rate": 7.206693721130826e-06,
+      "loss": 0.3669,
+      "step": 15367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9800312905897608,
+      "learning_rate": 7.206340153870411e-06,
+      "loss": 0.2995,
+      "step": 15368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8760423862202429,
+      "learning_rate": 7.205986572909225e-06,
+      "loss": 0.1849,
+      "step": 15369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.685208623723935,
+      "learning_rate": 7.205632978249457e-06,
+      "loss": 0.2626,
+      "step": 15370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3527685693687013,
+      "learning_rate": 7.20527936989331e-06,
+      "loss": 0.3381,
+      "step": 15371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7084526673560028,
+      "learning_rate": 7.204925747842974e-06,
+      "loss": 0.3104,
+      "step": 15372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6202606082504827,
+      "learning_rate": 7.204572112100647e-06,
+      "loss": 0.2911,
+      "step": 15373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3360958089222645,
+      "learning_rate": 7.2042184626685265e-06,
+      "loss": 0.2451,
+      "step": 15374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4736221217026146,
+      "learning_rate": 7.203864799548807e-06,
+      "loss": 0.3878,
+      "step": 15375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4057221773130946,
+      "learning_rate": 7.203511122743683e-06,
+      "loss": 0.2997,
+      "step": 15376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.267253448347421,
+      "learning_rate": 7.203157432255355e-06,
+      "loss": 0.314,
+      "step": 15377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5143844059546956,
+      "learning_rate": 7.202803728086014e-06,
+      "loss": 0.2421,
+      "step": 15378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5952542209348206,
+      "learning_rate": 7.202450010237861e-06,
+      "loss": 0.1644,
+      "step": 15379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4386738982613076,
+      "learning_rate": 7.202096278713091e-06,
+      "loss": 0.2119,
+      "step": 15380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3283858309599752,
+      "learning_rate": 7.2017425335139e-06,
+      "loss": 0.303,
+      "step": 15381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8713559980318908,
+      "learning_rate": 7.201388774642486e-06,
+      "loss": 0.5594,
+      "step": 15382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.838318441544091,
+      "learning_rate": 7.2010350021010424e-06,
+      "loss": 0.5992,
+      "step": 15383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4961392095542356,
+      "learning_rate": 7.20068121589177e-06,
+      "loss": 0.4978,
+      "step": 15384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4404439375867555,
+      "learning_rate": 7.200327416016863e-06,
+      "loss": 0.1578,
+      "step": 15385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9840894028003364,
+      "learning_rate": 7.199973602478521e-06,
+      "loss": 0.4603,
+      "step": 15386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2728628107316915,
+      "learning_rate": 7.199619775278939e-06,
+      "loss": 0.3379,
+      "step": 15387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8946989201549636,
+      "learning_rate": 7.199265934420314e-06,
+      "loss": 0.5289,
+      "step": 15388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6903080612758357,
+      "learning_rate": 7.198912079904845e-06,
+      "loss": 0.2383,
+      "step": 15389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.518303559074509,
+      "learning_rate": 7.198558211734728e-06,
+      "loss": 0.3695,
+      "step": 15390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.801363399931371,
+      "learning_rate": 7.19820432991216e-06,
+      "loss": 0.2463,
+      "step": 15391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2902729427415642,
+      "learning_rate": 7.1978504344393405e-06,
+      "loss": 0.2746,
+      "step": 15392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7884932282793924,
+      "learning_rate": 7.197496525318466e-06,
+      "loss": 0.4751,
+      "step": 15393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5993401064388773,
+      "learning_rate": 7.197142602551732e-06,
+      "loss": 0.3558,
+      "step": 15394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3961915742919804,
+      "learning_rate": 7.1967886661413415e-06,
+      "loss": 0.2401,
+      "step": 15395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8717727225610767,
+      "learning_rate": 7.196434716089487e-06,
+      "loss": 0.4224,
+      "step": 15396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5660562766420725,
+      "learning_rate": 7.196080752398369e-06,
+      "loss": 0.1918,
+      "step": 15397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6168460508519045,
+      "learning_rate": 7.195726775070184e-06,
+      "loss": 0.321,
+      "step": 15398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9112502126837063,
+      "learning_rate": 7.195372784107133e-06,
+      "loss": 0.2209,
+      "step": 15399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5921461857123456,
+      "learning_rate": 7.1950187795114116e-06,
+      "loss": 0.3663,
+      "step": 15400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0879498108485395,
+      "learning_rate": 7.194664761285219e-06,
+      "loss": 0.548,
+      "step": 15401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.008657537372584,
+      "learning_rate": 7.194310729430754e-06,
+      "loss": 0.3586,
+      "step": 15402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9207490476029714,
+      "learning_rate": 7.193956683950214e-06,
+      "loss": 0.4581,
+      "step": 15403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1913448497709704,
+      "learning_rate": 7.193602624845798e-06,
+      "loss": 0.1352,
+      "step": 15404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.580413735564082,
+      "learning_rate": 7.193248552119705e-06,
+      "loss": 0.1868,
+      "step": 15405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.01930741742788,
+      "learning_rate": 7.192894465774133e-06,
+      "loss": 0.3999,
+      "step": 15406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0264386157579937,
+      "learning_rate": 7.1925403658112805e-06,
+      "loss": 0.3824,
+      "step": 15407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7164153107936586,
+      "learning_rate": 7.1921862522333475e-06,
+      "loss": 0.5104,
+      "step": 15408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.341310021309775,
+      "learning_rate": 7.191832125042533e-06,
+      "loss": 0.2222,
+      "step": 15409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7416910952546245,
+      "learning_rate": 7.1914779842410345e-06,
+      "loss": 0.433,
+      "step": 15410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4809810864123356,
+      "learning_rate": 7.191123829831052e-06,
+      "loss": 0.3325,
+      "step": 15411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5781469378976654,
+      "learning_rate": 7.190769661814786e-06,
+      "loss": 0.3246,
+      "step": 15412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9145717488361458,
+      "learning_rate": 7.190415480194432e-06,
+      "loss": 0.5097,
+      "step": 15413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3447603735347244,
+      "learning_rate": 7.190061284972195e-06,
+      "loss": 0.219,
+      "step": 15414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.910408237428973,
+      "learning_rate": 7.189707076150271e-06,
+      "loss": 0.4445,
+      "step": 15415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6297513132777617,
+      "learning_rate": 7.189352853730857e-06,
+      "loss": 0.4603,
+      "step": 15416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3030130890320486,
+      "learning_rate": 7.188998617716158e-06,
+      "loss": 0.2045,
+      "step": 15417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.834125486963733,
+      "learning_rate": 7.188644368108372e-06,
+      "loss": 0.3646,
+      "step": 15418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6953420878170835,
+      "learning_rate": 7.188290104909696e-06,
+      "loss": 0.2265,
+      "step": 15419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.195386463569651,
+      "learning_rate": 7.187935828122332e-06,
+      "loss": 0.2503,
+      "step": 15420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.860596019473047,
+      "learning_rate": 7.187581537748481e-06,
+      "loss": 0.5871,
+      "step": 15421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6293883765095543,
+      "learning_rate": 7.1872272337903414e-06,
+      "loss": 0.3571,
+      "step": 15422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2732601276274302,
+      "learning_rate": 7.186872916250115e-06,
+      "loss": 0.1748,
+      "step": 15423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8289974323536768,
+      "learning_rate": 7.186518585129999e-06,
+      "loss": 0.466,
+      "step": 15424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5473801818675839,
+      "learning_rate": 7.1861642404321975e-06,
+      "loss": 0.2517,
+      "step": 15425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5219317446935452,
+      "learning_rate": 7.185809882158908e-06,
+      "loss": 0.1824,
+      "step": 15426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6675137920038035,
+      "learning_rate": 7.185455510312331e-06,
+      "loss": 0.194,
+      "step": 15427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7698018054566962,
+      "learning_rate": 7.18510112489467e-06,
+      "loss": 0.2373,
+      "step": 15428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2783501915175506,
+      "learning_rate": 7.184746725908123e-06,
+      "loss": 0.1466,
+      "step": 15429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3911558568510762,
+      "learning_rate": 7.184392313354891e-06,
+      "loss": 0.1352,
+      "step": 15430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.909942209308186,
+      "learning_rate": 7.184037887237174e-06,
+      "loss": 0.4082,
+      "step": 15431
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7235689059505483,
+      "learning_rate": 7.183683447557176e-06,
+      "loss": 0.6113,
+      "step": 15432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8343604274086438,
+      "learning_rate": 7.183328994317095e-06,
+      "loss": 0.6063,
+      "step": 15433
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3763514780247192,
+      "learning_rate": 7.182974527519133e-06,
+      "loss": 0.2235,
+      "step": 15434
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7386120770707991,
+      "learning_rate": 7.182620047165492e-06,
+      "loss": 0.5456,
+      "step": 15435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4182294548119365,
+      "learning_rate": 7.182265553258372e-06,
+      "loss": 0.4174,
+      "step": 15436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.608336069348912,
+      "learning_rate": 7.181911045799975e-06,
+      "loss": 0.3049,
+      "step": 15437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.55957649435157,
+      "learning_rate": 7.181556524792501e-06,
+      "loss": 0.3643,
+      "step": 15438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9802374958704978,
+      "learning_rate": 7.181201990238155e-06,
+      "loss": 0.3356,
+      "step": 15439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.741224488929308,
+      "learning_rate": 7.1808474421391335e-06,
+      "loss": 0.3158,
+      "step": 15440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.459320105055264,
+      "learning_rate": 7.1804928804976426e-06,
+      "loss": 0.2326,
+      "step": 15441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7500147989192292,
+      "learning_rate": 7.180138305315883e-06,
+      "loss": 0.4676,
+      "step": 15442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4887149659413814,
+      "learning_rate": 7.179783716596053e-06,
+      "loss": 0.347,
+      "step": 15443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.749732184353053,
+      "learning_rate": 7.17942911434036e-06,
+      "loss": 0.3241,
+      "step": 15444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.868806336743409,
+      "learning_rate": 7.179074498551001e-06,
+      "loss": 0.3391,
+      "step": 15445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7558227870468581,
+      "learning_rate": 7.178719869230181e-06,
+      "loss": 0.2678,
+      "step": 15446
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4102297034642441,
+      "learning_rate": 7.178365226380101e-06,
+      "loss": 0.253,
+      "step": 15447
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6817850810707882,
+      "learning_rate": 7.178010570002967e-06,
+      "loss": 0.3494,
+      "step": 15448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0207527679592885,
+      "learning_rate": 7.177655900100975e-06,
+      "loss": 0.3741,
+      "step": 15449
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4087833129943186,
+      "learning_rate": 7.177301216676331e-06,
+      "loss": 0.3218,
+      "step": 15450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.019362433366981,
+      "learning_rate": 7.176946519731237e-06,
+      "loss": 0.3889,
+      "step": 15451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.239999148071481,
+      "learning_rate": 7.1765918092678965e-06,
+      "loss": 0.1919,
+      "step": 15452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4854870131892575,
+      "learning_rate": 7.17623708528851e-06,
+      "loss": 0.1965,
+      "step": 15453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5354447663952357,
+      "learning_rate": 7.175882347795282e-06,
+      "loss": 0.3402,
+      "step": 15454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.705852852527823,
+      "learning_rate": 7.175527596790414e-06,
+      "loss": 0.3953,
+      "step": 15455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7718935160703848,
+      "learning_rate": 7.175172832276111e-06,
+      "loss": 0.2316,
+      "step": 15456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4761521132419069,
+      "learning_rate": 7.174818054254573e-06,
+      "loss": 0.2281,
+      "step": 15457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7272314851252755,
+      "learning_rate": 7.174463262728007e-06,
+      "loss": 0.3683,
+      "step": 15458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1674783237873765,
+      "learning_rate": 7.174108457698612e-06,
+      "loss": 0.4336,
+      "step": 15459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0526004969007405,
+      "learning_rate": 7.1737536391685925e-06,
+      "loss": 0.5088,
+      "step": 15460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.920837525496249,
+      "learning_rate": 7.173398807140154e-06,
+      "loss": 0.2602,
+      "step": 15461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6606788867244549,
+      "learning_rate": 7.173043961615498e-06,
+      "loss": 0.5011,
+      "step": 15462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4257885353851263,
+      "learning_rate": 7.172689102596828e-06,
+      "loss": 0.8459,
+      "step": 15463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0321827654172604,
+      "learning_rate": 7.172334230086347e-06,
+      "loss": 0.3222,
+      "step": 15464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1803714193894552,
+      "learning_rate": 7.1719793440862616e-06,
+      "loss": 0.1528,
+      "step": 15465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.119954240424306,
+      "learning_rate": 7.171624444598773e-06,
+      "loss": 0.2941,
+      "step": 15466
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5744064399015107,
+      "learning_rate": 7.171269531626085e-06,
+      "loss": 0.33,
+      "step": 15467
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7902660215586736,
+      "learning_rate": 7.170914605170402e-06,
+      "loss": 0.2147,
+      "step": 15468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6037119235322723,
+      "learning_rate": 7.170559665233927e-06,
+      "loss": 0.3514,
+      "step": 15469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3370654297008735,
+      "learning_rate": 7.170204711818867e-06,
+      "loss": 0.1556,
+      "step": 15470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2935936773183085,
+      "learning_rate": 7.169849744927424e-06,
+      "loss": 0.4045,
+      "step": 15471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.445696952309839,
+      "learning_rate": 7.169494764561802e-06,
+      "loss": 0.1651,
+      "step": 15472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6193480341538522,
+      "learning_rate": 7.169139770724206e-06,
+      "loss": 0.212,
+      "step": 15473
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.405662056074322,
+      "learning_rate": 7.168784763416841e-06,
+      "loss": 0.3934,
+      "step": 15474
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.023018812068527,
+      "learning_rate": 7.16842974264191e-06,
+      "loss": 0.4345,
+      "step": 15475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3923143990477125,
+      "learning_rate": 7.168074708401619e-06,
+      "loss": 0.1358,
+      "step": 15476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8752381339192936,
+      "learning_rate": 7.167719660698172e-06,
+      "loss": 0.1982,
+      "step": 15477
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2894569255931168,
+      "learning_rate": 7.167364599533774e-06,
+      "loss": 0.2936,
+      "step": 15478
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.635225347598159,
+      "learning_rate": 7.167009524910629e-06,
+      "loss": 0.3278,
+      "step": 15479
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3186627243866265,
+      "learning_rate": 7.166654436830942e-06,
+      "loss": 0.1836,
+      "step": 15480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9128793368280095,
+      "learning_rate": 7.16629933529692e-06,
+      "loss": 0.6136,
+      "step": 15481
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4920954001799045,
+      "learning_rate": 7.165944220310766e-06,
+      "loss": 0.2951,
+      "step": 15482
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6722038337823717,
+      "learning_rate": 7.165589091874686e-06,
+      "loss": 0.4565,
+      "step": 15483
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.293298805959223,
+      "learning_rate": 7.165233949990885e-06,
+      "loss": 0.2229,
+      "step": 15484
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6930904768912483,
+      "learning_rate": 7.164878794661568e-06,
+      "loss": 0.3432,
+      "step": 15485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.16275638455163,
+      "learning_rate": 7.164523625888942e-06,
+      "loss": 0.3402,
+      "step": 15486
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2737273766859516,
+      "learning_rate": 7.1641684436752106e-06,
+      "loss": 0.2425,
+      "step": 15487
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6839143399138676,
+      "learning_rate": 7.16381324802258e-06,
+      "loss": 0.2147,
+      "step": 15488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.145029778719816,
+      "learning_rate": 7.1634580389332555e-06,
+      "loss": 0.1971,
+      "step": 15489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.034090539203037,
+      "learning_rate": 7.163102816409443e-06,
+      "loss": 0.3527,
+      "step": 15490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3058533645336996,
+      "learning_rate": 7.162747580453351e-06,
+      "loss": 0.2546,
+      "step": 15491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8270318541040111,
+      "learning_rate": 7.16239233106718e-06,
+      "loss": 0.225,
+      "step": 15492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.410055414655698,
+      "learning_rate": 7.162037068253141e-06,
+      "loss": 0.2834,
+      "step": 15493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9696763599146692,
+      "learning_rate": 7.161681792013438e-06,
+      "loss": 0.2468,
+      "step": 15494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.830439173991127,
+      "learning_rate": 7.161326502350276e-06,
+      "loss": 0.2323,
+      "step": 15495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8814512531697969,
+      "learning_rate": 7.160971199265865e-06,
+      "loss": 0.3534,
+      "step": 15496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6720474635504592,
+      "learning_rate": 7.160615882762406e-06,
+      "loss": 0.1578,
+      "step": 15497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2766311711082161,
+      "learning_rate": 7.160260552842111e-06,
+      "loss": 0.3988,
+      "step": 15498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.565047126329357,
+      "learning_rate": 7.159905209507182e-06,
+      "loss": 0.1677,
+      "step": 15499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7480624768290793,
+      "learning_rate": 7.159549852759828e-06,
+      "loss": 0.2113,
+      "step": 15500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5534886789227076,
+      "learning_rate": 7.159194482602257e-06,
+      "loss": 0.4207,
+      "step": 15501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7977766262767418,
+      "learning_rate": 7.158839099036672e-06,
+      "loss": 0.4941,
+      "step": 15502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0387090764010996,
+      "learning_rate": 7.158483702065282e-06,
+      "loss": 0.7041,
+      "step": 15503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9758784878455584,
+      "learning_rate": 7.158128291690292e-06,
+      "loss": 0.503,
+      "step": 15504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4070450416198261,
+      "learning_rate": 7.157772867913913e-06,
+      "loss": 0.1456,
+      "step": 15505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.516608921679188,
+      "learning_rate": 7.1574174307383495e-06,
+      "loss": 0.1955,
+      "step": 15506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5189886472744263,
+      "learning_rate": 7.157061980165808e-06,
+      "loss": 0.2609,
+      "step": 15507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6581355329381031,
+      "learning_rate": 7.156706516198496e-06,
+      "loss": 0.2712,
+      "step": 15508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9303493677428596,
+      "learning_rate": 7.156351038838622e-06,
+      "loss": 0.3742,
+      "step": 15509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9351772605978201,
+      "learning_rate": 7.155995548088394e-06,
+      "loss": 0.5654,
+      "step": 15510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.343260897847975,
+      "learning_rate": 7.155640043950017e-06,
+      "loss": 0.1631,
+      "step": 15511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9665511715561323,
+      "learning_rate": 7.155284526425701e-06,
+      "loss": 0.2362,
+      "step": 15512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0797427914532696,
+      "learning_rate": 7.154928995517652e-06,
+      "loss": 0.1726,
+      "step": 15513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6600899582137314,
+      "learning_rate": 7.1545734512280785e-06,
+      "loss": 0.4697,
+      "step": 15514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8479454507775737,
+      "learning_rate": 7.154217893559188e-06,
+      "loss": 0.379,
+      "step": 15515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4748312006453954,
+      "learning_rate": 7.15386232251319e-06,
+      "loss": 0.3277,
+      "step": 15516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6115669790157252,
+      "learning_rate": 7.153506738092288e-06,
+      "loss": 0.3426,
+      "step": 15517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4011128796296466,
+      "learning_rate": 7.153151140298695e-06,
+      "loss": 0.2252,
+      "step": 15518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9348042754377914,
+      "learning_rate": 7.152795529134618e-06,
+      "loss": 0.5297,
+      "step": 15519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0952075948347564,
+      "learning_rate": 7.152439904602264e-06,
+      "loss": 0.2318,
+      "step": 15520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.16332805589338,
+      "learning_rate": 7.152084266703843e-06,
+      "loss": 0.2555,
+      "step": 15521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3669295367335592,
+      "learning_rate": 7.15172861544156e-06,
+      "loss": 0.2536,
+      "step": 15522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.621916291723134,
+      "learning_rate": 7.1513729508176275e-06,
+      "loss": 0.1559,
+      "step": 15523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8236759958979327,
+      "learning_rate": 7.151017272834252e-06,
+      "loss": 0.2176,
+      "step": 15524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6236475167078424,
+      "learning_rate": 7.1506615814936435e-06,
+      "loss": 0.2086,
+      "step": 15525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8658100138671225,
+      "learning_rate": 7.150305876798008e-06,
+      "loss": 0.1653,
+      "step": 15526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6746421935325073,
+      "learning_rate": 7.149950158749558e-06,
+      "loss": 0.3693,
+      "step": 15527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6514129542094582,
+      "learning_rate": 7.149594427350499e-06,
+      "loss": 0.1923,
+      "step": 15528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7514085678192792,
+      "learning_rate": 7.149238682603044e-06,
+      "loss": 0.3656,
+      "step": 15529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0125815439437686,
+      "learning_rate": 7.148882924509397e-06,
+      "loss": 0.2501,
+      "step": 15530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4338737313136343,
+      "learning_rate": 7.148527153071772e-06,
+      "loss": 0.3412,
+      "step": 15531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.299933887827163,
+      "learning_rate": 7.1481713682923745e-06,
+      "loss": 0.3476,
+      "step": 15532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4872325847178072,
+      "learning_rate": 7.147815570173416e-06,
+      "loss": 0.2591,
+      "step": 15533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1415397976075967,
+      "learning_rate": 7.147459758717107e-06,
+      "loss": 0.4378,
+      "step": 15534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.997186693750349,
+      "learning_rate": 7.147103933925652e-06,
+      "loss": 0.1666,
+      "step": 15535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2537800868014635,
+      "learning_rate": 7.146748095801267e-06,
+      "loss": 0.2569,
+      "step": 15536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3450892062632787,
+      "learning_rate": 7.146392244346156e-06,
+      "loss": 0.2811,
+      "step": 15537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.478366313888408,
+      "learning_rate": 7.146036379562533e-06,
+      "loss": 0.2309,
+      "step": 15538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5464141260289375,
+      "learning_rate": 7.145680501452606e-06,
+      "loss": 0.4726,
+      "step": 15539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4875256555212075,
+      "learning_rate": 7.145324610018585e-06,
+      "loss": 0.3683,
+      "step": 15540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8021189175291648,
+      "learning_rate": 7.144968705262678e-06,
+      "loss": 0.503,
+      "step": 15541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6171668231029266,
+      "learning_rate": 7.1446127871870995e-06,
+      "loss": 0.2098,
+      "step": 15542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6975050407266004,
+      "learning_rate": 7.144256855794055e-06,
+      "loss": 0.3889,
+      "step": 15543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.129616556802365,
+      "learning_rate": 7.1439009110857595e-06,
+      "loss": 0.1796,
+      "step": 15544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8986451855623132,
+      "learning_rate": 7.143544953064419e-06,
+      "loss": 0.2298,
+      "step": 15545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.649305585113882,
+      "learning_rate": 7.143188981732245e-06,
+      "loss": 0.5936,
+      "step": 15546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5510577812877777,
+      "learning_rate": 7.142832997091451e-06,
+      "loss": 0.2302,
+      "step": 15547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.613067844906943,
+      "learning_rate": 7.142476999144244e-06,
+      "loss": 0.1884,
+      "step": 15548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1392884768775724,
+      "learning_rate": 7.142120987892835e-06,
+      "loss": 0.2216,
+      "step": 15549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3581134228381888,
+      "learning_rate": 7.141764963339437e-06,
+      "loss": 0.1812,
+      "step": 15550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.431862645565152,
+      "learning_rate": 7.14140892548626e-06,
+      "loss": 0.1579,
+      "step": 15551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.460946948739978,
+      "learning_rate": 7.141052874335512e-06,
+      "loss": 0.2327,
+      "step": 15552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8044789697464616,
+      "learning_rate": 7.140696809889409e-06,
+      "loss": 0.3042,
+      "step": 15553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3489963273967276,
+      "learning_rate": 7.140340732150157e-06,
+      "loss": 0.4304,
+      "step": 15554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7124614591159806,
+      "learning_rate": 7.139984641119973e-06,
+      "loss": 0.3604,
+      "step": 15555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0958988190763366,
+      "learning_rate": 7.139628536801062e-06,
+      "loss": 0.4136,
+      "step": 15556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7865559163888902,
+      "learning_rate": 7.139272419195638e-06,
+      "loss": 0.5116,
+      "step": 15557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5673007907975387,
+      "learning_rate": 7.138916288305914e-06,
+      "loss": 0.208,
+      "step": 15558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4749219664141182,
+      "learning_rate": 7.138560144134098e-06,
+      "loss": 0.383,
+      "step": 15559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2409867550184797,
+      "learning_rate": 7.138203986682405e-06,
+      "loss": 0.2055,
+      "step": 15560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9944800947702177,
+      "learning_rate": 7.137847815953044e-06,
+      "loss": 0.1875,
+      "step": 15561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.795358486433176,
+      "learning_rate": 7.137491631948229e-06,
+      "loss": 0.2401,
+      "step": 15562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6188198940733731,
+      "learning_rate": 7.137135434670169e-06,
+      "loss": 0.2238,
+      "step": 15563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2425454316789524,
+      "learning_rate": 7.13677922412108e-06,
+      "loss": 0.1518,
+      "step": 15564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8008657878304637,
+      "learning_rate": 7.13642300030317e-06,
+      "loss": 0.6859,
+      "step": 15565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4626589389169526,
+      "learning_rate": 7.136066763218652e-06,
+      "loss": 0.265,
+      "step": 15566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1435165645189795,
+      "learning_rate": 7.13571051286974e-06,
+      "loss": 0.5856,
+      "step": 15567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.644835754624034,
+      "learning_rate": 7.135354249258645e-06,
+      "loss": 0.4523,
+      "step": 15568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6124980239893023,
+      "learning_rate": 7.134997972387577e-06,
+      "loss": 0.4687,
+      "step": 15569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.090712885299473,
+      "learning_rate": 7.134641682258754e-06,
+      "loss": 0.1843,
+      "step": 15570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.665950572632158,
+      "learning_rate": 7.1342853788743825e-06,
+      "loss": 0.3136,
+      "step": 15571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.153157660121724,
+      "learning_rate": 7.133929062236677e-06,
+      "loss": 0.2352,
+      "step": 15572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5436241291156283,
+      "learning_rate": 7.133572732347853e-06,
+      "loss": 0.1978,
+      "step": 15573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.420235614930303,
+      "learning_rate": 7.13321638921012e-06,
+      "loss": 0.2114,
+      "step": 15574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7062271568145586,
+      "learning_rate": 7.132860032825693e-06,
+      "loss": 0.2687,
+      "step": 15575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4019425596624013,
+      "learning_rate": 7.132503663196781e-06,
+      "loss": 0.3223,
+      "step": 15576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1450659223971982,
+      "learning_rate": 7.132147280325603e-06,
+      "loss": 0.1033,
+      "step": 15577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9883856717851673,
+      "learning_rate": 7.131790884214367e-06,
+      "loss": 0.6354,
+      "step": 15578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6155223151581075,
+      "learning_rate": 7.131434474865288e-06,
+      "loss": 0.2675,
+      "step": 15579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9453208187798647,
+      "learning_rate": 7.131078052280579e-06,
+      "loss": 0.537,
+      "step": 15580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4415520023167974,
+      "learning_rate": 7.130721616462453e-06,
+      "loss": 0.204,
+      "step": 15581
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2608933334412926,
+      "learning_rate": 7.1303651674131224e-06,
+      "loss": 0.1611,
+      "step": 15582
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4879189548414304,
+      "learning_rate": 7.130008705134804e-06,
+      "loss": 0.3257,
+      "step": 15583
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3167324213907328,
+      "learning_rate": 7.129652229629708e-06,
+      "loss": 0.2948,
+      "step": 15584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7476747777431654,
+      "learning_rate": 7.129295740900049e-06,
+      "loss": 0.2747,
+      "step": 15585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5524717708664915,
+      "learning_rate": 7.1289392389480415e-06,
+      "loss": 0.2516,
+      "step": 15586
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3240519562050022,
+      "learning_rate": 7.128582723775899e-06,
+      "loss": 0.2899,
+      "step": 15587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5718973901416726,
+      "learning_rate": 7.1282261953858346e-06,
+      "loss": 0.3354,
+      "step": 15588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.203495021616222,
+      "learning_rate": 7.127869653780064e-06,
+      "loss": 0.4068,
+      "step": 15589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6384978673706867,
+      "learning_rate": 7.127513098960798e-06,
+      "loss": 0.3158,
+      "step": 15590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7768215088142416,
+      "learning_rate": 7.127156530930254e-06,
+      "loss": 0.2111,
+      "step": 15591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3451743755034418,
+      "learning_rate": 7.126799949690643e-06,
+      "loss": 0.3077,
+      "step": 15592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8346578277634127,
+      "learning_rate": 7.126443355244183e-06,
+      "loss": 0.6559,
+      "step": 15593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0848577147757097,
+      "learning_rate": 7.126086747593086e-06,
+      "loss": 0.6346,
+      "step": 15594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.521354219209626,
+      "learning_rate": 7.125730126739566e-06,
+      "loss": 0.3259,
+      "step": 15595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7205789047070807,
+      "learning_rate": 7.12537349268584e-06,
+      "loss": 0.5081,
+      "step": 15596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8676898671535307,
+      "learning_rate": 7.125016845434119e-06,
+      "loss": 0.6009,
+      "step": 15597
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.135318893814954,
+      "learning_rate": 7.124660184986621e-06,
+      "loss": 0.2717,
+      "step": 15598
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.059929112399863,
+      "learning_rate": 7.124303511345561e-06,
+      "loss": 0.3625,
+      "step": 15599
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1402931791534336,
+      "learning_rate": 7.1239468245131505e-06,
+      "loss": 0.5332,
+      "step": 15600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5244070147022128,
+      "learning_rate": 7.123590124491606e-06,
+      "loss": 0.1926,
+      "step": 15601
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.468297548664193,
+      "learning_rate": 7.123233411283144e-06,
+      "loss": 0.1936,
+      "step": 15602
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.984660623675067,
+      "learning_rate": 7.122876684889976e-06,
+      "loss": 0.2681,
+      "step": 15603
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.852182389161055,
+      "learning_rate": 7.122519945314323e-06,
+      "loss": 0.3371,
+      "step": 15604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8767449723017953,
+      "learning_rate": 7.122163192558393e-06,
+      "loss": 0.6582,
+      "step": 15605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2474290381122168,
+      "learning_rate": 7.121806426624408e-06,
+      "loss": 0.2051,
+      "step": 15606
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.434084481531744,
+      "learning_rate": 7.121449647514579e-06,
+      "loss": 0.2236,
+      "step": 15607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0533505865967836,
+      "learning_rate": 7.121092855231124e-06,
+      "loss": 0.2462,
+      "step": 15608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.00486364044224,
+      "learning_rate": 7.1207360497762575e-06,
+      "loss": 0.2496,
+      "step": 15609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4292742830353526,
+      "learning_rate": 7.120379231152194e-06,
+      "loss": 0.3352,
+      "step": 15610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1165993285231153,
+      "learning_rate": 7.120022399361153e-06,
+      "loss": 0.4876,
+      "step": 15611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8953271511483356,
+      "learning_rate": 7.119665554405346e-06,
+      "loss": 0.4795,
+      "step": 15612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7962061922734793,
+      "learning_rate": 7.1193086962869925e-06,
+      "loss": 0.3403,
+      "step": 15613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7128686936126045,
+      "learning_rate": 7.118951825008304e-06,
+      "loss": 0.3611,
+      "step": 15614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3305781848631109,
+      "learning_rate": 7.118594940571501e-06,
+      "loss": 0.1433,
+      "step": 15615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.049457964167453,
+      "learning_rate": 7.118238042978798e-06,
+      "loss": 0.2013,
+      "step": 15616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.655647762137793,
+      "learning_rate": 7.117881132232411e-06,
+      "loss": 0.391,
+      "step": 15617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0354350560273946,
+      "learning_rate": 7.1175242083345574e-06,
+      "loss": 0.1192,
+      "step": 15618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5191711254005615,
+      "learning_rate": 7.117167271287453e-06,
+      "loss": 0.2681,
+      "step": 15619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4598965927284908,
+      "learning_rate": 7.1168103210933115e-06,
+      "loss": 0.2179,
+      "step": 15620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.217727527196564,
+      "learning_rate": 7.116453357754354e-06,
+      "loss": 0.2569,
+      "step": 15621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.270920415992806,
+      "learning_rate": 7.116096381272794e-06,
+      "loss": 0.3,
+      "step": 15622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8827258838025356,
+      "learning_rate": 7.11573939165085e-06,
+      "loss": 0.5262,
+      "step": 15623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.571413946252058,
+      "learning_rate": 7.11538238889074e-06,
+      "loss": 0.6105,
+      "step": 15624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.346959967350522,
+      "learning_rate": 7.1150253729946765e-06,
+      "loss": 0.1848,
+      "step": 15625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0273495822888754,
+      "learning_rate": 7.11466834396488e-06,
+      "loss": 0.582,
+      "step": 15626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7139012064369645,
+      "learning_rate": 7.114311301803566e-06,
+      "loss": 0.4066,
+      "step": 15627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6629928724106948,
+      "learning_rate": 7.113954246512953e-06,
+      "loss": 0.3156,
+      "step": 15628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7063717681188941,
+      "learning_rate": 7.113597178095256e-06,
+      "loss": 0.2129,
+      "step": 15629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5232800989266975,
+      "learning_rate": 7.113240096552696e-06,
+      "loss": 0.3053,
+      "step": 15630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3404670911325796,
+      "learning_rate": 7.112883001887487e-06,
+      "loss": 0.1921,
+      "step": 15631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.750534559247329,
+      "learning_rate": 7.1125258941018475e-06,
+      "loss": 0.4066,
+      "step": 15632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2993896503870288,
+      "learning_rate": 7.112168773197996e-06,
+      "loss": 0.1982,
+      "step": 15633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7111725819876273,
+      "learning_rate": 7.111811639178148e-06,
+      "loss": 0.1967,
+      "step": 15634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.495753727712199,
+      "learning_rate": 7.1114544920445235e-06,
+      "loss": 0.3537,
+      "step": 15635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8623429100943034,
+      "learning_rate": 7.111097331799338e-06,
+      "loss": 0.4748,
+      "step": 15636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7026317345841897,
+      "learning_rate": 7.110740158444812e-06,
+      "loss": 0.249,
+      "step": 15637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8584077266555172,
+      "learning_rate": 7.11038297198316e-06,
+      "loss": 0.4958,
+      "step": 15638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0466743982455853,
+      "learning_rate": 7.110025772416604e-06,
+      "loss": 0.5075,
+      "step": 15639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8825321701107587,
+      "learning_rate": 7.10966855974736e-06,
+      "loss": 0.5482,
+      "step": 15640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.670334658997646,
+      "learning_rate": 7.1093113339776465e-06,
+      "loss": 0.5335,
+      "step": 15641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0069473978169,
+      "learning_rate": 7.1089540951096815e-06,
+      "loss": 0.3249,
+      "step": 15642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.699435367150368,
+      "learning_rate": 7.108596843145683e-06,
+      "loss": 0.1642,
+      "step": 15643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4478890521574526,
+      "learning_rate": 7.10823957808787e-06,
+      "loss": 0.323,
+      "step": 15644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6073611515668456,
+      "learning_rate": 7.107882299938462e-06,
+      "loss": 0.3115,
+      "step": 15645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.816636706673528,
+      "learning_rate": 7.107525008699676e-06,
+      "loss": 0.3408,
+      "step": 15646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2122726371047547,
+      "learning_rate": 7.10716770437373e-06,
+      "loss": 0.1572,
+      "step": 15647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4668374646459184,
+      "learning_rate": 7.106810386962846e-06,
+      "loss": 0.278,
+      "step": 15648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.655378188698139,
+      "learning_rate": 7.106453056469241e-06,
+      "loss": 0.3709,
+      "step": 15649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.731829500578971,
+      "learning_rate": 7.106095712895133e-06,
+      "loss": 0.1453,
+      "step": 15650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6698045432827162,
+      "learning_rate": 7.10573835624274e-06,
+      "loss": 0.3952,
+      "step": 15651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.053911757071836,
+      "learning_rate": 7.105380986514285e-06,
+      "loss": 0.5034,
+      "step": 15652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5691519049537457,
+      "learning_rate": 7.1050236037119845e-06,
+      "loss": 0.2742,
+      "step": 15653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.039675328180539,
+      "learning_rate": 7.104666207838059e-06,
+      "loss": 0.5178,
+      "step": 15654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3440131938806834,
+      "learning_rate": 7.1043087988947275e-06,
+      "loss": 0.1834,
+      "step": 15655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8803782496369412,
+      "learning_rate": 7.103951376884208e-06,
+      "loss": 0.5175,
+      "step": 15656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5654267092728202,
+      "learning_rate": 7.103593941808722e-06,
+      "loss": 0.2442,
+      "step": 15657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9869968260200734,
+      "learning_rate": 7.103236493670487e-06,
+      "loss": 0.2561,
+      "step": 15658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4019894178831387,
+      "learning_rate": 7.102879032471725e-06,
+      "loss": 0.2009,
+      "step": 15659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6416557911227816,
+      "learning_rate": 7.102521558214652e-06,
+      "loss": 0.3381,
+      "step": 15660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4093367237298327,
+      "learning_rate": 7.102164070901493e-06,
+      "loss": 0.3177,
+      "step": 15661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3927973378096308,
+      "learning_rate": 7.1018065705344645e-06,
+      "loss": 0.3629,
+      "step": 15662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.653235904837859,
+      "learning_rate": 7.101449057115787e-06,
+      "loss": 0.5242,
+      "step": 15663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8197346330827924,
+      "learning_rate": 7.10109153064768e-06,
+      "loss": 0.2344,
+      "step": 15664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.057898356921084,
+      "learning_rate": 7.100733991132366e-06,
+      "loss": 0.3678,
+      "step": 15665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3953920407162115,
+      "learning_rate": 7.100376438572063e-06,
+      "loss": 0.5045,
+      "step": 15666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3374218418779362,
+      "learning_rate": 7.100018872968993e-06,
+      "loss": 0.1712,
+      "step": 15667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6942568142140206,
+      "learning_rate": 7.099661294325373e-06,
+      "loss": 0.3262,
+      "step": 15668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7048595163792597,
+      "learning_rate": 7.0993037026434274e-06,
+      "loss": 0.227,
+      "step": 15669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.66877148360807,
+      "learning_rate": 7.098946097925375e-06,
+      "loss": 0.3881,
+      "step": 15670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.371868278860603,
+      "learning_rate": 7.098588480173436e-06,
+      "loss": 0.3576,
+      "step": 15671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.658027163314104,
+      "learning_rate": 7.098230849389831e-06,
+      "loss": 0.3964,
+      "step": 15672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9324124552117576,
+      "learning_rate": 7.0978732055767815e-06,
+      "loss": 0.3922,
+      "step": 15673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7032672052772841,
+      "learning_rate": 7.097515548736511e-06,
+      "loss": 0.3171,
+      "step": 15674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.762206430876374,
+      "learning_rate": 7.0971578788712355e-06,
+      "loss": 0.5073,
+      "step": 15675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2828495635928938,
+      "learning_rate": 7.09680019598318e-06,
+      "loss": 0.3379,
+      "step": 15676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6956064567086286,
+      "learning_rate": 7.0964425000745626e-06,
+      "loss": 0.4766,
+      "step": 15677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5961809750201807,
+      "learning_rate": 7.096084791147606e-06,
+      "loss": 0.3234,
+      "step": 15678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4540572726522405,
+      "learning_rate": 7.0957270692045325e-06,
+      "loss": 0.2338,
+      "step": 15679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5992773174822899,
+      "learning_rate": 7.09536933424756e-06,
+      "loss": 0.2459,
+      "step": 15680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.781343072835462,
+      "learning_rate": 7.0950115862789145e-06,
+      "loss": 0.3915,
+      "step": 15681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.681804948015569,
+      "learning_rate": 7.094653825300815e-06,
+      "loss": 0.3047,
+      "step": 15682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5047088661142554,
+      "learning_rate": 7.094296051315483e-06,
+      "loss": 0.3058,
+      "step": 15683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.584234441357243,
+      "learning_rate": 7.0939382643251385e-06,
+      "loss": 0.1834,
+      "step": 15684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3814056589267496,
+      "learning_rate": 7.0935804643320086e-06,
+      "loss": 0.3848,
+      "step": 15685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8683717754765206,
+      "learning_rate": 7.0932226513383094e-06,
+      "loss": 0.1897,
+      "step": 15686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6014066713399386,
+      "learning_rate": 7.092864825346266e-06,
+      "loss": 0.3324,
+      "step": 15687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8453790349860013,
+      "learning_rate": 7.0925069863581e-06,
+      "loss": 0.4048,
+      "step": 15688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.51817497829575,
+      "learning_rate": 7.092149134376032e-06,
+      "loss": 0.2222,
+      "step": 15689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6602767104034553,
+      "learning_rate": 7.091791269402287e-06,
+      "loss": 0.2248,
+      "step": 15690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.495019230720699,
+      "learning_rate": 7.091433391439084e-06,
+      "loss": 0.4839,
+      "step": 15691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9518867267631368,
+      "learning_rate": 7.091075500488648e-06,
+      "loss": 0.3574,
+      "step": 15692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4872576480071846,
+      "learning_rate": 7.0907175965531985e-06,
+      "loss": 0.2907,
+      "step": 15693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3559734987736343,
+      "learning_rate": 7.090359679634962e-06,
+      "loss": 0.2754,
+      "step": 15694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4583389943444411,
+      "learning_rate": 7.090001749736157e-06,
+      "loss": 0.305,
+      "step": 15695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5897874306678283,
+      "learning_rate": 7.089643806859008e-06,
+      "loss": 0.3421,
+      "step": 15696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.143779987842804,
+      "learning_rate": 7.089285851005739e-06,
+      "loss": 0.3752,
+      "step": 15697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4846450007977479,
+      "learning_rate": 7.08892788217857e-06,
+      "loss": 0.4226,
+      "step": 15698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8530643826327509,
+      "learning_rate": 7.088569900379727e-06,
+      "loss": 0.5374,
+      "step": 15699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9118952723834681,
+      "learning_rate": 7.08821190561143e-06,
+      "loss": 0.3574,
+      "step": 15700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7458969975454854,
+      "learning_rate": 7.087853897875905e-06,
+      "loss": 0.4039,
+      "step": 15701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3027418070614039,
+      "learning_rate": 7.087495877175372e-06,
+      "loss": 0.2283,
+      "step": 15702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5013754345084382,
+      "learning_rate": 7.0871378435120564e-06,
+      "loss": 0.2848,
+      "step": 15703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.51562659764943,
+      "learning_rate": 7.086779796888181e-06,
+      "loss": 0.2286,
+      "step": 15704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.473397141932404,
+      "learning_rate": 7.086421737305969e-06,
+      "loss": 0.2674,
+      "step": 15705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3400542716633954,
+      "learning_rate": 7.086063664767643e-06,
+      "loss": 0.1753,
+      "step": 15706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6233059140464188,
+      "learning_rate": 7.085705579275428e-06,
+      "loss": 0.3688,
+      "step": 15707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6845906278716476,
+      "learning_rate": 7.085347480831547e-06,
+      "loss": 0.3403,
+      "step": 15708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8989339854307679,
+      "learning_rate": 7.084989369438224e-06,
+      "loss": 0.2718,
+      "step": 15709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3649359118958357,
+      "learning_rate": 7.084631245097683e-06,
+      "loss": 0.3019,
+      "step": 15710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5720339397062,
+      "learning_rate": 7.084273107812146e-06,
+      "loss": 0.2152,
+      "step": 15711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2429732410726209,
+      "learning_rate": 7.0839149575838395e-06,
+      "loss": 0.3234,
+      "step": 15712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9504280276219932,
+      "learning_rate": 7.083556794414985e-06,
+      "loss": 0.525,
+      "step": 15713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5036944252551385,
+      "learning_rate": 7.083198618307809e-06,
+      "loss": 0.2441,
+      "step": 15714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5233095115303943,
+      "learning_rate": 7.082840429264534e-06,
+      "loss": 0.2489,
+      "step": 15715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1840071003594654,
+      "learning_rate": 7.082482227287385e-06,
+      "loss": 0.2517,
+      "step": 15716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.26702159718893,
+      "learning_rate": 7.082124012378585e-06,
+      "loss": 0.3854,
+      "step": 15717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9580328371346651,
+      "learning_rate": 7.081765784540362e-06,
+      "loss": 0.5147,
+      "step": 15718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8072203453531723,
+      "learning_rate": 7.0814075437749365e-06,
+      "loss": 0.358,
+      "step": 15719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7686911856237724,
+      "learning_rate": 7.081049290084534e-06,
+      "loss": 0.3784,
+      "step": 15720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.860186299670506,
+      "learning_rate": 7.0806910234713814e-06,
+      "loss": 0.4214,
+      "step": 15721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.00491981697717,
+      "learning_rate": 7.0803327439377e-06,
+      "loss": 0.1712,
+      "step": 15722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5937248489788873,
+      "learning_rate": 7.079974451485718e-06,
+      "loss": 0.4213,
+      "step": 15723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7037854226572855,
+      "learning_rate": 7.079616146117658e-06,
+      "loss": 0.344,
+      "step": 15724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2895797385359267,
+      "learning_rate": 7.079257827835746e-06,
+      "loss": 0.2588,
+      "step": 15725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5410625682686452,
+      "learning_rate": 7.078899496642207e-06,
+      "loss": 0.358,
+      "step": 15726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4643550464493416,
+      "learning_rate": 7.0785411525392655e-06,
+      "loss": 0.1996,
+      "step": 15727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7868973903313952,
+      "learning_rate": 7.078182795529147e-06,
+      "loss": 0.4849,
+      "step": 15728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3687752849821373,
+      "learning_rate": 7.077824425614078e-06,
+      "loss": 0.2832,
+      "step": 15729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1959472498820474,
+      "learning_rate": 7.077466042796282e-06,
+      "loss": 0.3187,
+      "step": 15730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8414377001867255,
+      "learning_rate": 7.077107647077986e-06,
+      "loss": 0.2156,
+      "step": 15731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5804847128152018,
+      "learning_rate": 7.076749238461416e-06,
+      "loss": 0.3399,
+      "step": 15732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8103319470549473,
+      "learning_rate": 7.076390816948794e-06,
+      "loss": 0.4584,
+      "step": 15733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7391677781736512,
+      "learning_rate": 7.076032382542348e-06,
+      "loss": 0.2165,
+      "step": 15734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.666351002860694,
+      "learning_rate": 7.0756739352443045e-06,
+      "loss": 0.3442,
+      "step": 15735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4179204835374677,
+      "learning_rate": 7.075315475056891e-06,
+      "loss": 0.1559,
+      "step": 15736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8008628752241953,
+      "learning_rate": 7.074957001982329e-06,
+      "loss": 0.3106,
+      "step": 15737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5698702103921194,
+      "learning_rate": 7.074598516022848e-06,
+      "loss": 0.227,
+      "step": 15738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6439515496931603,
+      "learning_rate": 7.074240017180671e-06,
+      "loss": 0.2102,
+      "step": 15739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.371216075689699,
+      "learning_rate": 7.073881505458028e-06,
+      "loss": 0.2626,
+      "step": 15740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6473846056066415,
+      "learning_rate": 7.0735229808571415e-06,
+      "loss": 0.2949,
+      "step": 15741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8781557705813334,
+      "learning_rate": 7.073164443380242e-06,
+      "loss": 0.3507,
+      "step": 15742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4997854948647646,
+      "learning_rate": 7.072805893029552e-06,
+      "loss": 0.2057,
+      "step": 15743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7428612454037211,
+      "learning_rate": 7.072447329807299e-06,
+      "loss": 0.2447,
+      "step": 15744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8167567171371837,
+      "learning_rate": 7.072088753715711e-06,
+      "loss": 0.1788,
+      "step": 15745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2205545731754481,
+      "learning_rate": 7.071730164757014e-06,
+      "loss": 0.1403,
+      "step": 15746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.97734758368414,
+      "learning_rate": 7.071371562933433e-06,
+      "loss": 0.3751,
+      "step": 15747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.625293249024675,
+      "learning_rate": 7.071012948247198e-06,
+      "loss": 0.3855,
+      "step": 15748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4554166208085735,
+      "learning_rate": 7.070654320700534e-06,
+      "loss": 0.2926,
+      "step": 15749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6259833556253847,
+      "learning_rate": 7.070295680295668e-06,
+      "loss": 0.4759,
+      "step": 15750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1523224177056295,
+      "learning_rate": 7.069937027034828e-06,
+      "loss": 0.3422,
+      "step": 15751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.028755823926559,
+      "learning_rate": 7.069578360920241e-06,
+      "loss": 0.2062,
+      "step": 15752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.056993057615028,
+      "learning_rate": 7.0692196819541315e-06,
+      "loss": 0.4897,
+      "step": 15753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3603168573491569,
+      "learning_rate": 7.068860990138731e-06,
+      "loss": 0.4915,
+      "step": 15754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8514663212579796,
+      "learning_rate": 7.068502285476263e-06,
+      "loss": 0.4573,
+      "step": 15755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.503220272911104,
+      "learning_rate": 7.068143567968958e-06,
+      "loss": 0.2345,
+      "step": 15756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7308428383676266,
+      "learning_rate": 7.067784837619042e-06,
+      "loss": 0.3196,
+      "step": 15757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5316304736492954,
+      "learning_rate": 7.067426094428743e-06,
+      "loss": 0.3642,
+      "step": 15758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0185642991750847,
+      "learning_rate": 7.067067338400289e-06,
+      "loss": 0.2029,
+      "step": 15759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8748772561110405,
+      "learning_rate": 7.066708569535907e-06,
+      "loss": 0.388,
+      "step": 15760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0055315739740927,
+      "learning_rate": 7.066349787837825e-06,
+      "loss": 0.381,
+      "step": 15761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.047054362992422,
+      "learning_rate": 7.065990993308271e-06,
+      "loss": 0.4526,
+      "step": 15762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7715561882696595,
+      "learning_rate": 7.065632185949475e-06,
+      "loss": 0.6253,
+      "step": 15763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2634589975515687,
+      "learning_rate": 7.065273365763662e-06,
+      "loss": 0.1863,
+      "step": 15764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0874361733712457,
+      "learning_rate": 7.064914532753062e-06,
+      "loss": 0.6654,
+      "step": 15765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.09848210790831,
+      "learning_rate": 7.064555686919902e-06,
+      "loss": 0.2099,
+      "step": 15766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6842607915939405,
+      "learning_rate": 7.064196828266412e-06,
+      "loss": 0.6511,
+      "step": 15767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3475149369196235,
+      "learning_rate": 7.063837956794819e-06,
+      "loss": 0.2862,
+      "step": 15768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5622954747333382,
+      "learning_rate": 7.063479072507352e-06,
+      "loss": 0.1963,
+      "step": 15769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1338358509511988,
+      "learning_rate": 7.06312017540624e-06,
+      "loss": 0.6264,
+      "step": 15770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3520593115970407,
+      "learning_rate": 7.062761265493709e-06,
+      "loss": 0.1821,
+      "step": 15771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1095534587426852,
+      "learning_rate": 7.062402342771993e-06,
+      "loss": 0.1726,
+      "step": 15772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.424508218858373,
+      "learning_rate": 7.062043407243316e-06,
+      "loss": 0.294,
+      "step": 15773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4339534142045105,
+      "learning_rate": 7.0616844589099106e-06,
+      "loss": 0.3011,
+      "step": 15774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2937847224934096,
+      "learning_rate": 7.061325497774002e-06,
+      "loss": 0.173,
+      "step": 15775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5337652451582182,
+      "learning_rate": 7.060966523837822e-06,
+      "loss": 0.1968,
+      "step": 15776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4298524605059788,
+      "learning_rate": 7.060607537103598e-06,
+      "loss": 0.1775,
+      "step": 15777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0085555568166376,
+      "learning_rate": 7.0602485375735615e-06,
+      "loss": 0.4595,
+      "step": 15778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.823748820058063,
+      "learning_rate": 7.059889525249939e-06,
+      "loss": 0.4665,
+      "step": 15779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4538470187003165,
+      "learning_rate": 7.059530500134962e-06,
+      "loss": 0.3594,
+      "step": 15780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8303816308566594,
+      "learning_rate": 7.059171462230858e-06,
+      "loss": 0.3231,
+      "step": 15781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.052363165221573,
+      "learning_rate": 7.058812411539859e-06,
+      "loss": 0.439,
+      "step": 15782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6338942973281654,
+      "learning_rate": 7.058453348064194e-06,
+      "loss": 0.3136,
+      "step": 15783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9021426061236715,
+      "learning_rate": 7.058094271806091e-06,
+      "loss": 0.3528,
+      "step": 15784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4848922129568733,
+      "learning_rate": 7.057735182767781e-06,
+      "loss": 0.3281,
+      "step": 15785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5532878632995157,
+      "learning_rate": 7.0573760809514925e-06,
+      "loss": 0.3876,
+      "step": 15786
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5606312305248156,
+      "learning_rate": 7.057016966359458e-06,
+      "loss": 0.3441,
+      "step": 15787
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6783175492949252,
+      "learning_rate": 7.0566578389939045e-06,
+      "loss": 0.2935,
+      "step": 15788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8696817633190004,
+      "learning_rate": 7.056298698857065e-06,
+      "loss": 0.5178,
+      "step": 15789
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1893055446681653,
+      "learning_rate": 7.0559395459511674e-06,
+      "loss": 0.4359,
+      "step": 15790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7686989618492996,
+      "learning_rate": 7.055580380278443e-06,
+      "loss": 0.4841,
+      "step": 15791
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6301861635555193,
+      "learning_rate": 7.0552212018411215e-06,
+      "loss": 0.3692,
+      "step": 15792
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.001976772261729,
+      "learning_rate": 7.054862010641434e-06,
+      "loss": 0.4545,
+      "step": 15793
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4022126794039542,
+      "learning_rate": 7.054502806681611e-06,
+      "loss": 0.2103,
+      "step": 15794
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8064177133318373,
+      "learning_rate": 7.054143589963882e-06,
+      "loss": 0.4239,
+      "step": 15795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8353695901170128,
+      "learning_rate": 7.053784360490479e-06,
+      "loss": 0.1557,
+      "step": 15796
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.597005401426776,
+      "learning_rate": 7.053425118263633e-06,
+      "loss": 0.3564,
+      "step": 15797
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8334296934495748,
+      "learning_rate": 7.053065863285573e-06,
+      "loss": 0.5179,
+      "step": 15798
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.558885984343535,
+      "learning_rate": 7.05270659555853e-06,
+      "loss": 0.2583,
+      "step": 15799
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8806636942226167,
+      "learning_rate": 7.052347315084738e-06,
+      "loss": 0.3278,
+      "step": 15800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3068088402326206,
+      "learning_rate": 7.0519880218664224e-06,
+      "loss": 0.1926,
+      "step": 15801
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3773478408943134,
+      "learning_rate": 7.051628715905821e-06,
+      "loss": 0.3453,
+      "step": 15802
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.285269442632179,
+      "learning_rate": 7.051269397205159e-06,
+      "loss": 0.4356,
+      "step": 15803
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8078612426098397,
+      "learning_rate": 7.050910065766673e-06,
+      "loss": 0.1739,
+      "step": 15804
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9829107198594642,
+      "learning_rate": 7.05055072159259e-06,
+      "loss": 0.3984,
+      "step": 15805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2155909984609161,
+      "learning_rate": 7.050191364685144e-06,
+      "loss": 0.1873,
+      "step": 15806
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3939213901822574,
+      "learning_rate": 7.049831995046566e-06,
+      "loss": 0.2384,
+      "step": 15807
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5438931774238023,
+      "learning_rate": 7.049472612679086e-06,
+      "loss": 0.3539,
+      "step": 15808
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.386959431138145,
+      "learning_rate": 7.049113217584937e-06,
+      "loss": 0.2058,
+      "step": 15809
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3924813792046946,
+      "learning_rate": 7.048753809766351e-06,
+      "loss": 0.2157,
+      "step": 15810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7874656072055515,
+      "learning_rate": 7.04839438922556e-06,
+      "loss": 0.4783,
+      "step": 15811
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.079556386283945,
+      "learning_rate": 7.048034955964795e-06,
+      "loss": 0.4482,
+      "step": 15812
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4707263980153786,
+      "learning_rate": 7.047675509986289e-06,
+      "loss": 0.3602,
+      "step": 15813
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8454634859040526,
+      "learning_rate": 7.047316051292272e-06,
+      "loss": 0.4175,
+      "step": 15814
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8445244956630045,
+      "learning_rate": 7.046956579884979e-06,
+      "loss": 0.366,
+      "step": 15815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8550762725133234,
+      "learning_rate": 7.04659709576664e-06,
+      "loss": 0.3257,
+      "step": 15816
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4406856865419881,
+      "learning_rate": 7.046237598939488e-06,
+      "loss": 0.2851,
+      "step": 15817
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.299012201215011,
+      "learning_rate": 7.045878089405757e-06,
+      "loss": 0.2496,
+      "step": 15818
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7231268467026737,
+      "learning_rate": 7.045518567167675e-06,
+      "loss": 0.2494,
+      "step": 15819
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1992008878887348,
+      "learning_rate": 7.045159032227479e-06,
+      "loss": 0.1906,
+      "step": 15820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.385180460959792,
+      "learning_rate": 7.044799484587401e-06,
+      "loss": 0.2928,
+      "step": 15821
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5206987240304082,
+      "learning_rate": 7.044439924249671e-06,
+      "loss": 0.2375,
+      "step": 15822
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7863853739046927,
+      "learning_rate": 7.044080351216525e-06,
+      "loss": 0.2498,
+      "step": 15823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.455154237740681,
+      "learning_rate": 7.043720765490194e-06,
+      "loss": 0.2989,
+      "step": 15824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.453510681877968,
+      "learning_rate": 7.04336116707291e-06,
+      "loss": 0.32,
+      "step": 15825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5895085250825973,
+      "learning_rate": 7.043001555966908e-06,
+      "loss": 0.2625,
+      "step": 15826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7569018527054776,
+      "learning_rate": 7.042641932174421e-06,
+      "loss": 0.5842,
+      "step": 15827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.724868138648912,
+      "learning_rate": 7.0422822956976826e-06,
+      "loss": 0.3634,
+      "step": 15828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3280956559585928,
+      "learning_rate": 7.041922646538924e-06,
+      "loss": 0.1803,
+      "step": 15829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7877347175175387,
+      "learning_rate": 7.0415629847003785e-06,
+      "loss": 0.5692,
+      "step": 15830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2826706639447814,
+      "learning_rate": 7.041203310184282e-06,
+      "loss": 0.3632,
+      "step": 15831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7798562034376875,
+      "learning_rate": 7.040843622992865e-06,
+      "loss": 0.4605,
+      "step": 15832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.448361582860072,
+      "learning_rate": 7.040483923128365e-06,
+      "loss": 0.3627,
+      "step": 15833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0897196866581984,
+      "learning_rate": 7.040124210593011e-06,
+      "loss": 0.3537,
+      "step": 15834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2618341924569498,
+      "learning_rate": 7.039764485389041e-06,
+      "loss": 0.2328,
+      "step": 15835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4544309072748753,
+      "learning_rate": 7.039404747518685e-06,
+      "loss": 0.3627,
+      "step": 15836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.050601544453869,
+      "learning_rate": 7.03904499698418e-06,
+      "loss": 0.1661,
+      "step": 15837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7618906218518537,
+      "learning_rate": 7.038685233787758e-06,
+      "loss": 0.237,
+      "step": 15838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6946349042369624,
+      "learning_rate": 7.038325457931653e-06,
+      "loss": 0.5063,
+      "step": 15839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6080456122629856,
+      "learning_rate": 7.037965669418101e-06,
+      "loss": 0.4918,
+      "step": 15840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5076882133254483,
+      "learning_rate": 7.037605868249333e-06,
+      "loss": 0.24,
+      "step": 15841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7822091799371707,
+      "learning_rate": 7.037246054427587e-06,
+      "loss": 0.3531,
+      "step": 15842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0017454979389626,
+      "learning_rate": 7.0368862279550945e-06,
+      "loss": 0.5418,
+      "step": 15843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8187420213989705,
+      "learning_rate": 7.036526388834091e-06,
+      "loss": 0.5332,
+      "step": 15844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9557885814095661,
+      "learning_rate": 7.036166537066811e-06,
+      "loss": 0.441,
+      "step": 15845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6719646964566894,
+      "learning_rate": 7.035806672655488e-06,
+      "loss": 0.5136,
+      "step": 15846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.134771572150125,
+      "learning_rate": 7.03544679560236e-06,
+      "loss": 0.363,
+      "step": 15847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6558639733281488,
+      "learning_rate": 7.0350869059096575e-06,
+      "loss": 0.1548,
+      "step": 15848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0257107730879844,
+      "learning_rate": 7.034727003579619e-06,
+      "loss": 0.4228,
+      "step": 15849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.16673362533713,
+      "learning_rate": 7.034367088614476e-06,
+      "loss": 0.5004,
+      "step": 15850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1113449507715063,
+      "learning_rate": 7.034007161016466e-06,
+      "loss": 0.295,
+      "step": 15851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8312834451830564,
+      "learning_rate": 7.0336472207878224e-06,
+      "loss": 0.2949,
+      "step": 15852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6286274559326885,
+      "learning_rate": 7.033287267930781e-06,
+      "loss": 0.4338,
+      "step": 15853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.230539607475287,
+      "learning_rate": 7.032927302447578e-06,
+      "loss": 0.257,
+      "step": 15854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0447152260147146,
+      "learning_rate": 7.032567324340448e-06,
+      "loss": 0.4841,
+      "step": 15855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5457315505477025,
+      "learning_rate": 7.032207333611626e-06,
+      "loss": 0.4342,
+      "step": 15856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6051854059922137,
+      "learning_rate": 7.031847330263347e-06,
+      "loss": 0.4267,
+      "step": 15857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4862460378168132,
+      "learning_rate": 7.031487314297847e-06,
+      "loss": 0.2139,
+      "step": 15858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2942391403841296,
+      "learning_rate": 7.031127285717363e-06,
+      "loss": 0.2078,
+      "step": 15859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4040515006173873,
+      "learning_rate": 7.030767244524129e-06,
+      "loss": 0.3938,
+      "step": 15860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0241495006498753,
+      "learning_rate": 7.030407190720381e-06,
+      "loss": 0.5177,
+      "step": 15861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7134078537851147,
+      "learning_rate": 7.030047124308355e-06,
+      "loss": 0.5652,
+      "step": 15862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3829914451377283,
+      "learning_rate": 7.029687045290285e-06,
+      "loss": 0.1646,
+      "step": 15863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3494604217598363,
+      "learning_rate": 7.029326953668412e-06,
+      "loss": 0.2028,
+      "step": 15864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4757440218513587,
+      "learning_rate": 7.028966849444967e-06,
+      "loss": 0.1709,
+      "step": 15865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6671935967116709,
+      "learning_rate": 7.02860673262219e-06,
+      "loss": 0.4255,
+      "step": 15866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.516677697328481,
+      "learning_rate": 7.0282466032023135e-06,
+      "loss": 0.3487,
+      "step": 15867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9850236121998855,
+      "learning_rate": 7.027886461187576e-06,
+      "loss": 0.1351,
+      "step": 15868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3333406221543205,
+      "learning_rate": 7.027526306580214e-06,
+      "loss": 0.1595,
+      "step": 15869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.832129912983325,
+      "learning_rate": 7.027166139382463e-06,
+      "loss": 0.4561,
+      "step": 15870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9957179594804568,
+      "learning_rate": 7.0268059595965606e-06,
+      "loss": 0.6743,
+      "step": 15871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9353897654556462,
+      "learning_rate": 7.026445767224743e-06,
+      "loss": 0.3285,
+      "step": 15872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8913244360522576,
+      "learning_rate": 7.026085562269246e-06,
+      "loss": 0.2845,
+      "step": 15873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9412557938913615,
+      "learning_rate": 7.025725344732307e-06,
+      "loss": 0.3334,
+      "step": 15874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7825302701348738,
+      "learning_rate": 7.025365114616164e-06,
+      "loss": 0.4047,
+      "step": 15875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.487217270003492,
+      "learning_rate": 7.0250048719230525e-06,
+      "loss": 0.3087,
+      "step": 15876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9919174138365081,
+      "learning_rate": 7.024644616655209e-06,
+      "loss": 0.1703,
+      "step": 15877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4383796457245626,
+      "learning_rate": 7.024284348814874e-06,
+      "loss": 0.3488,
+      "step": 15878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6161446871821572,
+      "learning_rate": 7.02392406840428e-06,
+      "loss": 0.3989,
+      "step": 15879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.651408935461316,
+      "learning_rate": 7.023563775425667e-06,
+      "loss": 0.2149,
+      "step": 15880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.67595063978454,
+      "learning_rate": 7.023203469881272e-06,
+      "loss": 0.3559,
+      "step": 15881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7690583243402482,
+      "learning_rate": 7.022843151773332e-06,
+      "loss": 0.2277,
+      "step": 15882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2879607940223554,
+      "learning_rate": 7.0224828211040855e-06,
+      "loss": 0.2122,
+      "step": 15883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6197818365340506,
+      "learning_rate": 7.022122477875769e-06,
+      "loss": 0.3643,
+      "step": 15884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8489561004804282,
+      "learning_rate": 7.021762122090619e-06,
+      "loss": 0.3992,
+      "step": 15885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6247891280094953,
+      "learning_rate": 7.021401753750876e-06,
+      "loss": 0.4898,
+      "step": 15886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.070050985985713,
+      "learning_rate": 7.0210413728587744e-06,
+      "loss": 0.2749,
+      "step": 15887
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0126796838921015,
+      "learning_rate": 7.0206809794165565e-06,
+      "loss": 0.2266,
+      "step": 15888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9328231894491246,
+      "learning_rate": 7.020320573426456e-06,
+      "loss": 0.5291,
+      "step": 15889
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0414704505060426,
+      "learning_rate": 7.019960154890713e-06,
+      "loss": 0.3118,
+      "step": 15890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3473633434043761,
+      "learning_rate": 7.019599723811566e-06,
+      "loss": 0.2605,
+      "step": 15891
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1881095049839498,
+      "learning_rate": 7.019239280191252e-06,
+      "loss": 0.3051,
+      "step": 15892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1166086426113404,
+      "learning_rate": 7.0188788240320095e-06,
+      "loss": 0.7632,
+      "step": 15893
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0099519211382737,
+      "learning_rate": 7.018518355336077e-06,
+      "loss": 0.5951,
+      "step": 15894
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3771550294282546,
+      "learning_rate": 7.018157874105692e-06,
+      "loss": 0.251,
+      "step": 15895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.845855387709492,
+      "learning_rate": 7.017797380343097e-06,
+      "loss": 0.3341,
+      "step": 15896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7817173645849143,
+      "learning_rate": 7.0174368740505235e-06,
+      "loss": 0.4794,
+      "step": 15897
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0646373204441808,
+      "learning_rate": 7.017076355230217e-06,
+      "loss": 0.1777,
+      "step": 15898
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0450230585257936,
+      "learning_rate": 7.016715823884412e-06,
+      "loss": 0.5282,
+      "step": 15899
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.575065353197594,
+      "learning_rate": 7.016355280015349e-06,
+      "loss": 0.3068,
+      "step": 15900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6726498947129724,
+      "learning_rate": 7.015994723625267e-06,
+      "loss": 0.306,
+      "step": 15901
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1943949063357433,
+      "learning_rate": 7.015634154716405e-06,
+      "loss": 0.229,
+      "step": 15902
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6982433281796503,
+      "learning_rate": 7.015273573291001e-06,
+      "loss": 0.2678,
+      "step": 15903
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6380328545384486,
+      "learning_rate": 7.0149129793512935e-06,
+      "loss": 0.2245,
+      "step": 15904
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6728096960633236,
+      "learning_rate": 7.014552372899525e-06,
+      "loss": 0.368,
+      "step": 15905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.875511226896052,
+      "learning_rate": 7.014191753937931e-06,
+      "loss": 0.5109,
+      "step": 15906
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1639963793697243,
+      "learning_rate": 7.013831122468753e-06,
+      "loss": 0.2392,
+      "step": 15907
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6813226414696483,
+      "learning_rate": 7.01347047849423e-06,
+      "loss": 0.3891,
+      "step": 15908
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3839307293897056,
+      "learning_rate": 7.013109822016602e-06,
+      "loss": 0.1794,
+      "step": 15909
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.59373059447947,
+      "learning_rate": 7.012749153038107e-06,
+      "loss": 0.4971,
+      "step": 15910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1798790656164317,
+      "learning_rate": 7.012388471560985e-06,
+      "loss": 0.3528,
+      "step": 15911
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.706468605091236,
+      "learning_rate": 7.012027777587478e-06,
+      "loss": 0.4515,
+      "step": 15912
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3683917763478664,
+      "learning_rate": 7.011667071119823e-06,
+      "loss": 0.2467,
+      "step": 15913
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1563181099935733,
+      "learning_rate": 7.011306352160262e-06,
+      "loss": 0.1527,
+      "step": 15914
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0125639001600466,
+      "learning_rate": 7.0109456207110325e-06,
+      "loss": 0.1857,
+      "step": 15915
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4002769189956739,
+      "learning_rate": 7.010584876774378e-06,
+      "loss": 0.1672,
+      "step": 15916
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6871453150891018,
+      "learning_rate": 7.0102241203525364e-06,
+      "loss": 0.3388,
+      "step": 15917
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5680570726745846,
+      "learning_rate": 7.009863351447748e-06,
+      "loss": 0.3339,
+      "step": 15918
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4115946901504957,
+      "learning_rate": 7.009502570062252e-06,
+      "loss": 0.2857,
+      "step": 15919
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4677943953428374,
+      "learning_rate": 7.0091417761982916e-06,
+      "loss": 0.2254,
+      "step": 15920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.351501193337473,
+      "learning_rate": 7.008780969858106e-06,
+      "loss": 0.2595,
+      "step": 15921
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8337092574104374,
+      "learning_rate": 7.0084201510439335e-06,
+      "loss": 0.4927,
+      "step": 15922
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4717304015961474,
+      "learning_rate": 7.008059319758016e-06,
+      "loss": 0.4475,
+      "step": 15923
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.001843194501955,
+      "learning_rate": 7.007698476002597e-06,
+      "loss": 0.1948,
+      "step": 15924
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8228776891265164,
+      "learning_rate": 7.007337619779914e-06,
+      "loss": 0.3955,
+      "step": 15925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8423925126353287,
+      "learning_rate": 7.006976751092208e-06,
+      "loss": 0.152,
+      "step": 15926
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.278101690378266,
+      "learning_rate": 7.006615869941722e-06,
+      "loss": 0.1495,
+      "step": 15927
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1693631954306056,
+      "learning_rate": 7.006254976330696e-06,
+      "loss": 0.4882,
+      "step": 15928
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7662618425367755,
+      "learning_rate": 7.00589407026137e-06,
+      "loss": 0.5382,
+      "step": 15929
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9456624516768017,
+      "learning_rate": 7.005533151735986e-06,
+      "loss": 0.3102,
+      "step": 15930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9451685089992405,
+      "learning_rate": 7.005172220756784e-06,
+      "loss": 0.4299,
+      "step": 15931
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7139558339899617,
+      "learning_rate": 7.004811277326008e-06,
+      "loss": 0.2038,
+      "step": 15932
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0504042069924053,
+      "learning_rate": 7.004450321445897e-06,
+      "loss": 0.236,
+      "step": 15933
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7506010607111189,
+      "learning_rate": 7.004089353118692e-06,
+      "loss": 0.3629,
+      "step": 15934
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1216560026032174,
+      "learning_rate": 7.003728372346638e-06,
+      "loss": 0.2859,
+      "step": 15935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2457078303634335,
+      "learning_rate": 7.003367379131974e-06,
+      "loss": 0.2035,
+      "step": 15936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5695077295283149,
+      "learning_rate": 7.0030063734769395e-06,
+      "loss": 0.5045,
+      "step": 15937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5979086809402252,
+      "learning_rate": 7.0026453553837804e-06,
+      "loss": 0.5235,
+      "step": 15938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.582410312498287,
+      "learning_rate": 7.002284324854737e-06,
+      "loss": 0.2396,
+      "step": 15939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.820891501063542,
+      "learning_rate": 7.001923281892049e-06,
+      "loss": 0.3398,
+      "step": 15940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4513766528054377,
+      "learning_rate": 7.001562226497962e-06,
+      "loss": 0.1786,
+      "step": 15941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6404757227423814,
+      "learning_rate": 7.001201158674716e-06,
+      "loss": 0.2543,
+      "step": 15942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5098965762001553,
+      "learning_rate": 7.000840078424555e-06,
+      "loss": 0.2772,
+      "step": 15943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2562621495622541,
+      "learning_rate": 7.000478985749717e-06,
+      "loss": 0.2355,
+      "step": 15944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5302362881784488,
+      "learning_rate": 7.00011788065245e-06,
+      "loss": 0.2212,
+      "step": 15945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9413356422770942,
+      "learning_rate": 6.99975676313499e-06,
+      "loss": 0.7521,
+      "step": 15946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6362441109885717,
+      "learning_rate": 6.999395633199586e-06,
+      "loss": 0.312,
+      "step": 15947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3727562666003843,
+      "learning_rate": 6.999034490848475e-06,
+      "loss": 0.2678,
+      "step": 15948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1555861697138539,
+      "learning_rate": 6.998673336083903e-06,
+      "loss": 0.2555,
+      "step": 15949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6924212980214373,
+      "learning_rate": 6.998312168908112e-06,
+      "loss": 0.5398,
+      "step": 15950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6630602984151663,
+      "learning_rate": 6.9979509893233435e-06,
+      "loss": 0.2167,
+      "step": 15951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1918995945817854,
+      "learning_rate": 6.997589797331843e-06,
+      "loss": 0.2968,
+      "step": 15952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8423415619490426,
+      "learning_rate": 6.9972285929358495e-06,
+      "loss": 0.3953,
+      "step": 15953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.650743737902586,
+      "learning_rate": 6.996867376137609e-06,
+      "loss": 0.4622,
+      "step": 15954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.551277393564887,
+      "learning_rate": 6.996506146939364e-06,
+      "loss": 0.2003,
+      "step": 15955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8522430655009743,
+      "learning_rate": 6.996144905343357e-06,
+      "loss": 0.1472,
+      "step": 15956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4571383535328015,
+      "learning_rate": 6.9957836513518305e-06,
+      "loss": 0.1486,
+      "step": 15957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8579409823573503,
+      "learning_rate": 6.995422384967031e-06,
+      "loss": 0.1817,
+      "step": 15958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0971963154450197,
+      "learning_rate": 6.995061106191198e-06,
+      "loss": 0.5424,
+      "step": 15959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.213515073323872,
+      "learning_rate": 6.994699815026577e-06,
+      "loss": 0.2942,
+      "step": 15960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6007958135505553,
+      "learning_rate": 6.994338511475412e-06,
+      "loss": 0.4653,
+      "step": 15961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.696593697875112,
+      "learning_rate": 6.993977195539945e-06,
+      "loss": 0.3095,
+      "step": 15962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0522482284545687,
+      "learning_rate": 6.993615867222421e-06,
+      "loss": 0.3416,
+      "step": 15963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.807492077174966,
+      "learning_rate": 6.993254526525082e-06,
+      "loss": 0.279,
+      "step": 15964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2114289670792386,
+      "learning_rate": 6.992893173450173e-06,
+      "loss": 0.2434,
+      "step": 15965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9592990337289813,
+      "learning_rate": 6.992531807999938e-06,
+      "loss": 0.5652,
+      "step": 15966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8023044325595778,
+      "learning_rate": 6.992170430176621e-06,
+      "loss": 0.3333,
+      "step": 15967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1868738300029267,
+      "learning_rate": 6.991809039982467e-06,
+      "loss": 0.3979,
+      "step": 15968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6156827126825657,
+      "learning_rate": 6.991447637419718e-06,
+      "loss": 0.205,
+      "step": 15969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7592465400381245,
+      "learning_rate": 6.991086222490619e-06,
+      "loss": 0.5178,
+      "step": 15970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8715538383050327,
+      "learning_rate": 6.9907247951974155e-06,
+      "loss": 0.4134,
+      "step": 15971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5754226950879273,
+      "learning_rate": 6.99036335554235e-06,
+      "loss": 0.4354,
+      "step": 15972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3119352465349396,
+      "learning_rate": 6.990001903527669e-06,
+      "loss": 0.2666,
+      "step": 15973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.802304591811126,
+      "learning_rate": 6.989640439155614e-06,
+      "loss": 0.3507,
+      "step": 15974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.484556869608578,
+      "learning_rate": 6.989278962428433e-06,
+      "loss": 0.2292,
+      "step": 15975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.797097449423757,
+      "learning_rate": 6.988917473348369e-06,
+      "loss": 0.4005,
+      "step": 15976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6502756699474217,
+      "learning_rate": 6.988555971917665e-06,
+      "loss": 0.2763,
+      "step": 15977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.047269229925774,
+      "learning_rate": 6.988194458138571e-06,
+      "loss": 0.3397,
+      "step": 15978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6566877371455813,
+      "learning_rate": 6.987832932013326e-06,
+      "loss": 0.3258,
+      "step": 15979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.664031811999827,
+      "learning_rate": 6.987471393544179e-06,
+      "loss": 0.2834,
+      "step": 15980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.948077010798809,
+      "learning_rate": 6.987109842733375e-06,
+      "loss": 0.2596,
+      "step": 15981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.761358989852079,
+      "learning_rate": 6.986748279583156e-06,
+      "loss": 0.4888,
+      "step": 15982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8782642362081237,
+      "learning_rate": 6.986386704095769e-06,
+      "loss": 0.5825,
+      "step": 15983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1502741436502637,
+      "learning_rate": 6.986025116273458e-06,
+      "loss": 0.2775,
+      "step": 15984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5973313101428037,
+      "learning_rate": 6.985663516118473e-06,
+      "loss": 0.1391,
+      "step": 15985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6583794409086527,
+      "learning_rate": 6.985301903633053e-06,
+      "loss": 0.1986,
+      "step": 15986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7110790276889,
+      "learning_rate": 6.984940278819449e-06,
+      "loss": 0.2222,
+      "step": 15987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2135401321232253,
+      "learning_rate": 6.984578641679902e-06,
+      "loss": 0.1678,
+      "step": 15988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9409580442501815,
+      "learning_rate": 6.9842169922166615e-06,
+      "loss": 0.3327,
+      "step": 15989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6930499822670708,
+      "learning_rate": 6.98385533043197e-06,
+      "loss": 0.4457,
+      "step": 15990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5384405036236757,
+      "learning_rate": 6.983493656328076e-06,
+      "loss": 0.4402,
+      "step": 15991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9686404379376743,
+      "learning_rate": 6.983131969907225e-06,
+      "loss": 0.2322,
+      "step": 15992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8264569146776033,
+      "learning_rate": 6.982770271171662e-06,
+      "loss": 0.2777,
+      "step": 15993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5978742045099368,
+      "learning_rate": 6.982408560123632e-06,
+      "loss": 0.4125,
+      "step": 15994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6190883687862356,
+      "learning_rate": 6.9820468367653835e-06,
+      "loss": 0.4385,
+      "step": 15995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3129607495535036,
+      "learning_rate": 6.981685101099163e-06,
+      "loss": 0.4429,
+      "step": 15996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.509705099395302,
+      "learning_rate": 6.981323353127213e-06,
+      "loss": 0.238,
+      "step": 15997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.480145799142287,
+      "learning_rate": 6.9809615928517824e-06,
+      "loss": 0.198,
+      "step": 15998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.03079363758183,
+      "learning_rate": 6.98059982027512e-06,
+      "loss": 0.6574,
+      "step": 15999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.465817121749173,
+      "learning_rate": 6.980238035399467e-06,
+      "loss": 0.4088,
+      "step": 16000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8913437604133805,
+      "learning_rate": 6.9798762382270746e-06,
+      "loss": 0.5014,
+      "step": 16001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8635033689017584,
+      "learning_rate": 6.979514428760188e-06,
+      "loss": 0.4344,
+      "step": 16002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3977703548996852,
+      "learning_rate": 6.979152607001053e-06,
+      "loss": 0.3472,
+      "step": 16003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7239880203925588,
+      "learning_rate": 6.978790772951917e-06,
+      "loss": 0.3156,
+      "step": 16004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5529562785573503,
+      "learning_rate": 6.978428926615027e-06,
+      "loss": 0.2722,
+      "step": 16005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4044027036600577,
+      "learning_rate": 6.97806706799263e-06,
+      "loss": 0.3058,
+      "step": 16006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9911499612778576,
+      "learning_rate": 6.977705197086974e-06,
+      "loss": 0.5047,
+      "step": 16007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.745209429437242,
+      "learning_rate": 6.977343313900304e-06,
+      "loss": 0.38,
+      "step": 16008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7698136645333218,
+      "learning_rate": 6.976981418434868e-06,
+      "loss": 0.3333,
+      "step": 16009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0408932907410158,
+      "learning_rate": 6.976619510692914e-06,
+      "loss": 0.1288,
+      "step": 16010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4260338716233538,
+      "learning_rate": 6.97625759067669e-06,
+      "loss": 0.2199,
+      "step": 16011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.101994247237177,
+      "learning_rate": 6.97589565838844e-06,
+      "loss": 0.2014,
+      "step": 16012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.134416769509248,
+      "learning_rate": 6.975533713830416e-06,
+      "loss": 0.1209,
+      "step": 16013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.804140381634423,
+      "learning_rate": 6.975171757004862e-06,
+      "loss": 0.4111,
+      "step": 16014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8194920569621507,
+      "learning_rate": 6.974809787914028e-06,
+      "loss": 0.1502,
+      "step": 16015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6762150367526143,
+      "learning_rate": 6.97444780656016e-06,
+      "loss": 0.2486,
+      "step": 16016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6677819645484189,
+      "learning_rate": 6.974085812945507e-06,
+      "loss": 0.2228,
+      "step": 16017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.644741964545555,
+      "learning_rate": 6.973723807072317e-06,
+      "loss": 0.2529,
+      "step": 16018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2541765037766788,
+      "learning_rate": 6.9733617889428365e-06,
+      "loss": 0.364,
+      "step": 16019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0975594240712034,
+      "learning_rate": 6.972999758559315e-06,
+      "loss": 0.3357,
+      "step": 16020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2842698533346844,
+      "learning_rate": 6.972637715924e-06,
+      "loss": 0.2237,
+      "step": 16021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6005113664610473,
+      "learning_rate": 6.9722756610391395e-06,
+      "loss": 0.1422,
+      "step": 16022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8926037765984869,
+      "learning_rate": 6.9719135939069815e-06,
+      "loss": 0.2096,
+      "step": 16023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6281963374860937,
+      "learning_rate": 6.9715515145297755e-06,
+      "loss": 0.3428,
+      "step": 16024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6590272767205607,
+      "learning_rate": 6.9711894229097695e-06,
+      "loss": 0.434,
+      "step": 16025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.606511083955132,
+      "learning_rate": 6.970827319049212e-06,
+      "loss": 0.301,
+      "step": 16026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6768779167302061,
+      "learning_rate": 6.970465202950352e-06,
+      "loss": 0.3439,
+      "step": 16027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7037269902022427,
+      "learning_rate": 6.970103074615435e-06,
+      "loss": 0.3388,
+      "step": 16028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.732846114586756,
+      "learning_rate": 6.969740934046715e-06,
+      "loss": 0.3222,
+      "step": 16029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5457476351376083,
+      "learning_rate": 6.969378781246436e-06,
+      "loss": 0.1402,
+      "step": 16030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5200230016034608,
+      "learning_rate": 6.96901661621685e-06,
+      "loss": 0.2631,
+      "step": 16031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8286382492662914,
+      "learning_rate": 6.968654438960205e-06,
+      "loss": 0.3624,
+      "step": 16032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.163550295618953,
+      "learning_rate": 6.96829224947875e-06,
+      "loss": 0.5778,
+      "step": 16033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3014378952267036,
+      "learning_rate": 6.967930047774733e-06,
+      "loss": 0.1885,
+      "step": 16034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6801854552325297,
+      "learning_rate": 6.967567833850404e-06,
+      "loss": 0.5074,
+      "step": 16035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.505153287609238,
+      "learning_rate": 6.967205607708014e-06,
+      "loss": 0.272,
+      "step": 16036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6618687066593494,
+      "learning_rate": 6.9668433693498085e-06,
+      "loss": 0.1928,
+      "step": 16037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9073936947801955,
+      "learning_rate": 6.966481118778041e-06,
+      "loss": 0.5109,
+      "step": 16038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7896547982039255,
+      "learning_rate": 6.9661188559949575e-06,
+      "loss": 0.1849,
+      "step": 16039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4669130078703718,
+      "learning_rate": 6.965756581002811e-06,
+      "loss": 0.2425,
+      "step": 16040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.229805512035271,
+      "learning_rate": 6.965394293803847e-06,
+      "loss": 0.4308,
+      "step": 16041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9718051018394596,
+      "learning_rate": 6.96503199440032e-06,
+      "loss": 0.382,
+      "step": 16042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.921849878658701,
+      "learning_rate": 6.964669682794475e-06,
+      "loss": 0.6264,
+      "step": 16043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9964789487331738,
+      "learning_rate": 6.964307358988565e-06,
+      "loss": 0.4765,
+      "step": 16044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2990081606415615,
+      "learning_rate": 6.963945022984838e-06,
+      "loss": 0.2413,
+      "step": 16045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.200154500109541,
+      "learning_rate": 6.963582674785547e-06,
+      "loss": 0.3878,
+      "step": 16046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7817446373250907,
+      "learning_rate": 6.963220314392939e-06,
+      "loss": 0.46,
+      "step": 16047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1596598446210085,
+      "learning_rate": 6.962857941809266e-06,
+      "loss": 0.1193,
+      "step": 16048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7160169083795105,
+      "learning_rate": 6.962495557036776e-06,
+      "loss": 0.6905,
+      "step": 16049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8302116508178807,
+      "learning_rate": 6.962133160077722e-06,
+      "loss": 0.4945,
+      "step": 16050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.707967771886354,
+      "learning_rate": 6.9617707509343536e-06,
+      "loss": 0.2997,
+      "step": 16051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5528104847422617,
+      "learning_rate": 6.9614083296089184e-06,
+      "loss": 0.2407,
+      "step": 16052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.791816798470077,
+      "learning_rate": 6.961045896103673e-06,
+      "loss": 0.5201,
+      "step": 16053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5403674881543081,
+      "learning_rate": 6.960683450420862e-06,
+      "loss": 0.2863,
+      "step": 16054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.909833627302746,
+      "learning_rate": 6.960320992562739e-06,
+      "loss": 0.4604,
+      "step": 16055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3309205925972158,
+      "learning_rate": 6.959958522531555e-06,
+      "loss": 0.22,
+      "step": 16056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6749497019987587,
+      "learning_rate": 6.95959604032956e-06,
+      "loss": 0.2269,
+      "step": 16057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7478615689223003,
+      "learning_rate": 6.959233545959006e-06,
+      "loss": 0.3205,
+      "step": 16058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.861953934362925,
+      "learning_rate": 6.95887103942214e-06,
+      "loss": 0.2726,
+      "step": 16059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6751201788469468,
+      "learning_rate": 6.958508520721219e-06,
+      "loss": 0.4133,
+      "step": 16060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3125952845000624,
+      "learning_rate": 6.958145989858489e-06,
+      "loss": 0.2544,
+      "step": 16061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.797922301882167,
+      "learning_rate": 6.957783446836206e-06,
+      "loss": 0.5218,
+      "step": 16062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5511403232613479,
+      "learning_rate": 6.957420891656617e-06,
+      "loss": 0.3172,
+      "step": 16063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6843755659233375,
+      "learning_rate": 6.957058324321976e-06,
+      "loss": 0.2792,
+      "step": 16064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4842755645147965,
+      "learning_rate": 6.956695744834532e-06,
+      "loss": 0.2091,
+      "step": 16065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3444874286991322,
+      "learning_rate": 6.95633315319654e-06,
+      "loss": 0.1352,
+      "step": 16066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8631085609688465,
+      "learning_rate": 6.955970549410248e-06,
+      "loss": 0.3275,
+      "step": 16067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5499733879319764,
+      "learning_rate": 6.955607933477911e-06,
+      "loss": 0.2265,
+      "step": 16068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8166416959037999,
+      "learning_rate": 6.9552453054017785e-06,
+      "loss": 0.5018,
+      "step": 16069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.471988820172179,
+      "learning_rate": 6.954882665184103e-06,
+      "loss": 0.3241,
+      "step": 16070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1886289588183523,
+      "learning_rate": 6.954520012827136e-06,
+      "loss": 0.152,
+      "step": 16071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5607483582177144,
+      "learning_rate": 6.9541573483331306e-06,
+      "loss": 0.4781,
+      "step": 16072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2071807810672888,
+      "learning_rate": 6.953794671704337e-06,
+      "loss": 0.3501,
+      "step": 16073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7465196469848479,
+      "learning_rate": 6.953431982943009e-06,
+      "loss": 0.4071,
+      "step": 16074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.512315755451607,
+      "learning_rate": 6.953069282051397e-06,
+      "loss": 0.238,
+      "step": 16075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1762256613613298,
+      "learning_rate": 6.952706569031755e-06,
+      "loss": 0.3233,
+      "step": 16076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4291229045213272,
+      "learning_rate": 6.9523438438863355e-06,
+      "loss": 0.2444,
+      "step": 16077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9015755953059659,
+      "learning_rate": 6.951981106617389e-06,
+      "loss": 0.5432,
+      "step": 16078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7962684602790955,
+      "learning_rate": 6.951618357227171e-06,
+      "loss": 0.5048,
+      "step": 16079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5182094046504504,
+      "learning_rate": 6.951255595717931e-06,
+      "loss": 0.2738,
+      "step": 16080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4393005031384047,
+      "learning_rate": 6.950892822091924e-06,
+      "loss": 0.1264,
+      "step": 16081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.519399002223769,
+      "learning_rate": 6.950530036351402e-06,
+      "loss": 0.1623,
+      "step": 16082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4023716214526325,
+      "learning_rate": 6.950167238498615e-06,
+      "loss": 0.293,
+      "step": 16083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5679449719202208,
+      "learning_rate": 6.9498044285358215e-06,
+      "loss": 0.3839,
+      "step": 16084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8736164591359994,
+      "learning_rate": 6.949441606465268e-06,
+      "loss": 0.4009,
+      "step": 16085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.186627675070407,
+      "learning_rate": 6.949078772289214e-06,
+      "loss": 0.344,
+      "step": 16086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3741818182896284,
+      "learning_rate": 6.9487159260099065e-06,
+      "loss": 0.4262,
+      "step": 16087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9287829267369976,
+      "learning_rate": 6.948353067629604e-06,
+      "loss": 0.332,
+      "step": 16088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6666262447899285,
+      "learning_rate": 6.947990197150556e-06,
+      "loss": 0.3381,
+      "step": 16089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2924808549754305,
+      "learning_rate": 6.9476273145750184e-06,
+      "loss": 0.2552,
+      "step": 16090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.375947217528804,
+      "learning_rate": 6.947264419905243e-06,
+      "loss": 0.433,
+      "step": 16091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0144297989195095,
+      "learning_rate": 6.9469015131434825e-06,
+      "loss": 0.4621,
+      "step": 16092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.493100650855478,
+      "learning_rate": 6.946538594291993e-06,
+      "loss": 0.3582,
+      "step": 16093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9595238236667643,
+      "learning_rate": 6.9461756633530255e-06,
+      "loss": 0.309,
+      "step": 16094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.936606912742294,
+      "learning_rate": 6.945812720328836e-06,
+      "loss": 0.3205,
+      "step": 16095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5909851180874037,
+      "learning_rate": 6.945449765221677e-06,
+      "loss": 0.3236,
+      "step": 16096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.914640642940412,
+      "learning_rate": 6.945086798033802e-06,
+      "loss": 0.4589,
+      "step": 16097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.016191184935538,
+      "learning_rate": 6.944723818767466e-06,
+      "loss": 0.2724,
+      "step": 16098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.735887950117946,
+      "learning_rate": 6.944360827424921e-06,
+      "loss": 0.2445,
+      "step": 16099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5200512739053773,
+      "learning_rate": 6.943997824008425e-06,
+      "loss": 0.2053,
+      "step": 16100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2552831508438134,
+      "learning_rate": 6.943634808520227e-06,
+      "loss": 0.2259,
+      "step": 16101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7931903419609054,
+      "learning_rate": 6.9432717809625875e-06,
+      "loss": 0.38,
+      "step": 16102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5342472870112613,
+      "learning_rate": 6.942908741337755e-06,
+      "loss": 0.4141,
+      "step": 16103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8997032827909903,
+      "learning_rate": 6.942545689647987e-06,
+      "loss": 0.2461,
+      "step": 16104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8151895139505922,
+      "learning_rate": 6.942182625895536e-06,
+      "loss": 0.4129,
+      "step": 16105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8774121464506655,
+      "learning_rate": 6.941819550082659e-06,
+      "loss": 0.3467,
+      "step": 16106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0982990520856393,
+      "learning_rate": 6.9414564622116085e-06,
+      "loss": 0.1827,
+      "step": 16107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.473079638241967,
+      "learning_rate": 6.94109336228464e-06,
+      "loss": 0.2164,
+      "step": 16108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4603482399290848,
+      "learning_rate": 6.9407302503040075e-06,
+      "loss": 0.3397,
+      "step": 16109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7630506585292234,
+      "learning_rate": 6.940367126271967e-06,
+      "loss": 0.4955,
+      "step": 16110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.5742592488099625,
+      "learning_rate": 6.940003990190774e-06,
+      "loss": 0.3141,
+      "step": 16111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3307956508481333,
+      "learning_rate": 6.9396408420626814e-06,
+      "loss": 0.416,
+      "step": 16112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4934385358028968,
+      "learning_rate": 6.939277681889946e-06,
+      "loss": 0.2162,
+      "step": 16113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.094734430187184,
+      "learning_rate": 6.93891450967482e-06,
+      "loss": 0.6454,
+      "step": 16114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6616991345925187,
+      "learning_rate": 6.938551325419563e-06,
+      "loss": 0.3851,
+      "step": 16115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3956215814086481,
+      "learning_rate": 6.938188129126426e-06,
+      "loss": 0.3294,
+      "step": 16116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6727177643811963,
+      "learning_rate": 6.937824920797669e-06,
+      "loss": 0.2893,
+      "step": 16117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5250318498490476,
+      "learning_rate": 6.937461700435542e-06,
+      "loss": 0.1563,
+      "step": 16118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5563409437853546,
+      "learning_rate": 6.937098468042306e-06,
+      "loss": 0.3293,
+      "step": 16119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.686225599751485,
+      "learning_rate": 6.9367352236202115e-06,
+      "loss": 0.2297,
+      "step": 16120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8174425839901014,
+      "learning_rate": 6.936371967171518e-06,
+      "loss": 0.3037,
+      "step": 16121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.623563484575753,
+      "learning_rate": 6.936008698698479e-06,
+      "loss": 0.1959,
+      "step": 16122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8425948112529635,
+      "learning_rate": 6.935645418203352e-06,
+      "loss": 0.5116,
+      "step": 16123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5067651822282728,
+      "learning_rate": 6.935282125688389e-06,
+      "loss": 0.4438,
+      "step": 16124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5826487007379102,
+      "learning_rate": 6.934918821155851e-06,
+      "loss": 0.2803,
+      "step": 16125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9597272085289736,
+      "learning_rate": 6.9345555046079935e-06,
+      "loss": 0.3773,
+      "step": 16126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4009867307878154,
+      "learning_rate": 6.934192176047069e-06,
+      "loss": 0.1718,
+      "step": 16127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0038230611030783,
+      "learning_rate": 6.9338288354753355e-06,
+      "loss": 0.2423,
+      "step": 16128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7940722416977222,
+      "learning_rate": 6.93346548289505e-06,
+      "loss": 0.3045,
+      "step": 16129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3145590451828613,
+      "learning_rate": 6.933102118308469e-06,
+      "loss": 0.3156,
+      "step": 16130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3430721047904,
+      "learning_rate": 6.932738741717846e-06,
+      "loss": 0.1829,
+      "step": 16131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6638350980592576,
+      "learning_rate": 6.932375353125442e-06,
+      "loss": 0.2689,
+      "step": 16132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8621305395275565,
+      "learning_rate": 6.93201195253351e-06,
+      "loss": 0.3553,
+      "step": 16133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6117512003025978,
+      "learning_rate": 6.931648539944308e-06,
+      "loss": 0.3577,
+      "step": 16134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3545767077731108,
+      "learning_rate": 6.931285115360093e-06,
+      "loss": 0.1857,
+      "step": 16135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9645140354182313,
+      "learning_rate": 6.93092167878312e-06,
+      "loss": 0.5021,
+      "step": 16136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.549387809719365,
+      "learning_rate": 6.9305582302156475e-06,
+      "loss": 0.2783,
+      "step": 16137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7614732939251954,
+      "learning_rate": 6.930194769659931e-06,
+      "loss": 0.1646,
+      "step": 16138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.016296134522614,
+      "learning_rate": 6.92983129711823e-06,
+      "loss": 0.287,
+      "step": 16139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.38174791358989,
+      "learning_rate": 6.929467812592799e-06,
+      "loss": 0.2343,
+      "step": 16140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8079594577118305,
+      "learning_rate": 6.929104316085897e-06,
+      "loss": 0.3001,
+      "step": 16141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0849188271051156,
+      "learning_rate": 6.9287408075997795e-06,
+      "loss": 0.3837,
+      "step": 16142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.698910976906092,
+      "learning_rate": 6.928377287136706e-06,
+      "loss": 0.4231,
+      "step": 16143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6765027482618635,
+      "learning_rate": 6.928013754698931e-06,
+      "loss": 0.2144,
+      "step": 16144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2979337380645366,
+      "learning_rate": 6.927650210288716e-06,
+      "loss": 0.321,
+      "step": 16145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1809432880394026,
+      "learning_rate": 6.9272866539083136e-06,
+      "loss": 0.1944,
+      "step": 16146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6529480458145764,
+      "learning_rate": 6.926923085559985e-06,
+      "loss": 0.3895,
+      "step": 16147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6767685456062946,
+      "learning_rate": 6.926559505245984e-06,
+      "loss": 0.29,
+      "step": 16148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.45097614678061,
+      "learning_rate": 6.926195912968574e-06,
+      "loss": 0.1978,
+      "step": 16149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3038576707881515,
+      "learning_rate": 6.925832308730008e-06,
+      "loss": 0.1569,
+      "step": 16150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.672198696530291,
+      "learning_rate": 6.925468692532548e-06,
+      "loss": 0.1834,
+      "step": 16151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9044279082773048,
+      "learning_rate": 6.925105064378448e-06,
+      "loss": 0.5843,
+      "step": 16152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.464281793250174,
+      "learning_rate": 6.924741424269966e-06,
+      "loss": 0.3734,
+      "step": 16153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8496273648153518,
+      "learning_rate": 6.924377772209364e-06,
+      "loss": 0.2114,
+      "step": 16154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6413442942253078,
+      "learning_rate": 6.924014108198898e-06,
+      "loss": 0.2093,
+      "step": 16155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8425363975459856,
+      "learning_rate": 6.923650432240825e-06,
+      "loss": 0.3035,
+      "step": 16156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8211367218695012,
+      "learning_rate": 6.923286744337407e-06,
+      "loss": 0.5962,
+      "step": 16157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.572986053889996,
+      "learning_rate": 6.922923044490897e-06,
+      "loss": 0.1995,
+      "step": 16158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4919127515931363,
+      "learning_rate": 6.922559332703558e-06,
+      "loss": 0.4219,
+      "step": 16159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8887704312666702,
+      "learning_rate": 6.922195608977645e-06,
+      "loss": 0.5781,
+      "step": 16160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4019026049133765,
+      "learning_rate": 6.921831873315421e-06,
+      "loss": 0.3288,
+      "step": 16161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.278422759263078,
+      "learning_rate": 6.92146812571914e-06,
+      "loss": 0.7701,
+      "step": 16162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8271073867370118,
+      "learning_rate": 6.921104366191066e-06,
+      "loss": 0.6732,
+      "step": 16163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7811597081773631,
+      "learning_rate": 6.920740594733453e-06,
+      "loss": 0.2604,
+      "step": 16164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.606185567431848,
+      "learning_rate": 6.920376811348562e-06,
+      "loss": 0.273,
+      "step": 16165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5094078622127292,
+      "learning_rate": 6.920013016038651e-06,
+      "loss": 0.3595,
+      "step": 16166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6781323974417914,
+      "learning_rate": 6.919649208805982e-06,
+      "loss": 0.3471,
+      "step": 16167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.278528300035155,
+      "learning_rate": 6.919285389652811e-06,
+      "loss": 0.316,
+      "step": 16168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6012718774371564,
+      "learning_rate": 6.918921558581397e-06,
+      "loss": 0.3354,
+      "step": 16169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1568482526189063,
+      "learning_rate": 6.918557715594003e-06,
+      "loss": 0.3383,
+      "step": 16170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.567694113171572,
+      "learning_rate": 6.918193860692884e-06,
+      "loss": 0.3519,
+      "step": 16171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8180383630398964,
+      "learning_rate": 6.917829993880303e-06,
+      "loss": 0.4784,
+      "step": 16172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9895763618658683,
+      "learning_rate": 6.917466115158516e-06,
+      "loss": 0.1849,
+      "step": 16173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0861173973660407,
+      "learning_rate": 6.917102224529785e-06,
+      "loss": 0.5365,
+      "step": 16174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4695041323425557,
+      "learning_rate": 6.9167383219963706e-06,
+      "loss": 0.3755,
+      "step": 16175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5216521536659708,
+      "learning_rate": 6.916374407560529e-06,
+      "loss": 0.2489,
+      "step": 16176
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2337012383642858,
+      "learning_rate": 6.916010481224524e-06,
+      "loss": 0.2239,
+      "step": 16177
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5146740370750844,
+      "learning_rate": 6.915646542990611e-06,
+      "loss": 0.257,
+      "step": 16178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1886958152749525,
+      "learning_rate": 6.915282592861055e-06,
+      "loss": 0.1907,
+      "step": 16179
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7840511868726727,
+      "learning_rate": 6.914918630838111e-06,
+      "loss": 0.2192,
+      "step": 16180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3746884613506358,
+      "learning_rate": 6.914554656924044e-06,
+      "loss": 0.2543,
+      "step": 16181
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6979613727308571,
+      "learning_rate": 6.914190671121111e-06,
+      "loss": 0.3466,
+      "step": 16182
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6893415185491802,
+      "learning_rate": 6.913826673431573e-06,
+      "loss": 0.1835,
+      "step": 16183
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6697592583611103,
+      "learning_rate": 6.913462663857689e-06,
+      "loss": 0.4235,
+      "step": 16184
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3670181932522338,
+      "learning_rate": 6.913098642401723e-06,
+      "loss": 0.3226,
+      "step": 16185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.208920957075987,
+      "learning_rate": 6.9127346090659334e-06,
+      "loss": 0.189,
+      "step": 16186
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6238972631495947,
+      "learning_rate": 6.9123705638525795e-06,
+      "loss": 0.238,
+      "step": 16187
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7227439598538936,
+      "learning_rate": 6.912006506763923e-06,
+      "loss": 0.4261,
+      "step": 16188
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9531512608584383,
+      "learning_rate": 6.9116424378022255e-06,
+      "loss": 0.1395,
+      "step": 16189
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7672078350694889,
+      "learning_rate": 6.911278356969746e-06,
+      "loss": 0.3512,
+      "step": 16190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2216652704791064,
+      "learning_rate": 6.910914264268747e-06,
+      "loss": 0.2824,
+      "step": 16191
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5287751054339906,
+      "learning_rate": 6.910550159701489e-06,
+      "loss": 0.271,
+      "step": 16192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.03120204795577,
+      "learning_rate": 6.91018604327023e-06,
+      "loss": 0.2501,
+      "step": 16193
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6752018137953886,
+      "learning_rate": 6.909821914977237e-06,
+      "loss": 0.251,
+      "step": 16194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7505038166287832,
+      "learning_rate": 6.909457774824766e-06,
+      "loss": 0.3797,
+      "step": 16195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4930544152427971,
+      "learning_rate": 6.909093622815082e-06,
+      "loss": 0.2986,
+      "step": 16196
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5716435593844404,
+      "learning_rate": 6.9087294589504425e-06,
+      "loss": 0.361,
+      "step": 16197
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4351423614859922,
+      "learning_rate": 6.908365283233111e-06,
+      "loss": 0.2801,
+      "step": 16198
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0940402216401393,
+      "learning_rate": 6.908001095665349e-06,
+      "loss": 0.1949,
+      "step": 16199
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4462336530252013,
+      "learning_rate": 6.907636896249418e-06,
+      "loss": 0.3155,
+      "step": 16200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6059193176128024,
+      "learning_rate": 6.907272684987578e-06,
+      "loss": 0.1861,
+      "step": 16201
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.101351787681962,
+      "learning_rate": 6.9069084618820925e-06,
+      "loss": 0.4929,
+      "step": 16202
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.762376345361691,
+      "learning_rate": 6.9065442269352235e-06,
+      "loss": 0.5421,
+      "step": 16203
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.128961520396223,
+      "learning_rate": 6.90617998014923e-06,
+      "loss": 0.3855,
+      "step": 16204
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5434614638229,
+      "learning_rate": 6.905815721526377e-06,
+      "loss": 0.2957,
+      "step": 16205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6441081334096308,
+      "learning_rate": 6.905451451068925e-06,
+      "loss": 0.3428,
+      "step": 16206
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7162082257503897,
+      "learning_rate": 6.9050871687791375e-06,
+      "loss": 0.2666,
+      "step": 16207
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.392874767053267,
+      "learning_rate": 6.9047228746592744e-06,
+      "loss": 0.3772,
+      "step": 16208
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7329952920964211,
+      "learning_rate": 6.904358568711599e-06,
+      "loss": 0.3877,
+      "step": 16209
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.838279433681176,
+      "learning_rate": 6.903994250938373e-06,
+      "loss": 0.1467,
+      "step": 16210
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3493723875173322,
+      "learning_rate": 6.903629921341859e-06,
+      "loss": 0.2486,
+      "step": 16211
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5542579805157908,
+      "learning_rate": 6.903265579924321e-06,
+      "loss": 0.3802,
+      "step": 16212
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8232207233164652,
+      "learning_rate": 6.902901226688018e-06,
+      "loss": 0.3256,
+      "step": 16213
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.828605268840272,
+      "learning_rate": 6.9025368616352164e-06,
+      "loss": 0.2448,
+      "step": 16214
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3832643371477964,
+      "learning_rate": 6.9021724847681745e-06,
+      "loss": 0.2342,
+      "step": 16215
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6407786842299248,
+      "learning_rate": 6.901808096089159e-06,
+      "loss": 0.4619,
+      "step": 16216
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2817841518500726,
+      "learning_rate": 6.90144369560043e-06,
+      "loss": 0.3773,
+      "step": 16217
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3919924712338032,
+      "learning_rate": 6.901079283304252e-06,
+      "loss": 0.2307,
+      "step": 16218
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7705242579138758,
+      "learning_rate": 6.900714859202886e-06,
+      "loss": 0.5048,
+      "step": 16219
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5390261389217934,
+      "learning_rate": 6.900350423298599e-06,
+      "loss": 0.3561,
+      "step": 16220
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.712810662381182,
+      "learning_rate": 6.8999859755936495e-06,
+      "loss": 0.2225,
+      "step": 16221
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.697574230441955,
+      "learning_rate": 6.899621516090303e-06,
+      "loss": 0.207,
+      "step": 16222
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6261372495415367,
+      "learning_rate": 6.89925704479082e-06,
+      "loss": 0.2348,
+      "step": 16223
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4839659202454394,
+      "learning_rate": 6.898892561697468e-06,
+      "loss": 0.3186,
+      "step": 16224
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.039278640107289,
+      "learning_rate": 6.898528066812506e-06,
+      "loss": 0.3766,
+      "step": 16225
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6020421496167194,
+      "learning_rate": 6.898163560138202e-06,
+      "loss": 0.3365,
+      "step": 16226
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6424538999704257,
+      "learning_rate": 6.897799041676816e-06,
+      "loss": 0.3925,
+      "step": 16227
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2446123184170255,
+      "learning_rate": 6.897434511430612e-06,
+      "loss": 0.2109,
+      "step": 16228
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9725164684966765,
+      "learning_rate": 6.897069969401856e-06,
+      "loss": 0.4163,
+      "step": 16229
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.734961114256649,
+      "learning_rate": 6.896705415592808e-06,
+      "loss": 0.4872,
+      "step": 16230
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6846076916956594,
+      "learning_rate": 6.896340850005736e-06,
+      "loss": 0.3637,
+      "step": 16231
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5518642986772082,
+      "learning_rate": 6.8959762726429014e-06,
+      "loss": 0.3661,
+      "step": 16232
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5780629627384686,
+      "learning_rate": 6.895611683506567e-06,
+      "loss": 0.3843,
+      "step": 16233
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.417675045691458,
+      "learning_rate": 6.895247082598999e-06,
+      "loss": 0.3198,
+      "step": 16234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.258010152723685,
+      "learning_rate": 6.894882469922459e-06,
+      "loss": 0.1854,
+      "step": 16235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9953153402596544,
+      "learning_rate": 6.894517845479215e-06,
+      "loss": 0.5135,
+      "step": 16236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3771898052714777,
+      "learning_rate": 6.894153209271528e-06,
+      "loss": 0.3152,
+      "step": 16237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3396187251336389,
+      "learning_rate": 6.893788561301664e-06,
+      "loss": 0.1388,
+      "step": 16238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7680840079193345,
+      "learning_rate": 6.8934239015718854e-06,
+      "loss": 0.3804,
+      "step": 16239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5804708721242808,
+      "learning_rate": 6.8930592300844575e-06,
+      "loss": 0.4641,
+      "step": 16240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4128153132541648,
+      "learning_rate": 6.892694546841647e-06,
+      "loss": 0.2156,
+      "step": 16241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.294544688622955,
+      "learning_rate": 6.892329851845716e-06,
+      "loss": 0.1539,
+      "step": 16242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1014133102612664,
+      "learning_rate": 6.89196514509893e-06,
+      "loss": 0.3391,
+      "step": 16243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8600426444938385,
+      "learning_rate": 6.891600426603553e-06,
+      "loss": 0.5256,
+      "step": 16244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3326494706408905,
+      "learning_rate": 6.891235696361851e-06,
+      "loss": 0.1925,
+      "step": 16245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3757603018130986,
+      "learning_rate": 6.8908709543760875e-06,
+      "loss": 0.6932,
+      "step": 16246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7316452917805851,
+      "learning_rate": 6.890506200648529e-06,
+      "loss": 0.3955,
+      "step": 16247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.210104760223856,
+      "learning_rate": 6.890141435181439e-06,
+      "loss": 0.3707,
+      "step": 16248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.324514305417666,
+      "learning_rate": 6.889776657977083e-06,
+      "loss": 0.173,
+      "step": 16249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2000766195731452,
+      "learning_rate": 6.889411869037728e-06,
+      "loss": 0.2229,
+      "step": 16250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.374936871813427,
+      "learning_rate": 6.889047068365635e-06,
+      "loss": 0.2043,
+      "step": 16251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.172115806633646,
+      "learning_rate": 6.888682255963075e-06,
+      "loss": 0.2612,
+      "step": 16252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6610121596271374,
+      "learning_rate": 6.888317431832307e-06,
+      "loss": 0.253,
+      "step": 16253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5594706255655721,
+      "learning_rate": 6.887952595975604e-06,
+      "loss": 0.1564,
+      "step": 16254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.22910646382762,
+      "learning_rate": 6.887587748395225e-06,
+      "loss": 0.1303,
+      "step": 16255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9237227280302862,
+      "learning_rate": 6.887222889093439e-06,
+      "loss": 0.4695,
+      "step": 16256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6315064522790708,
+      "learning_rate": 6.8868580180725095e-06,
+      "loss": 0.3157,
+      "step": 16257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.936460381432063,
+      "learning_rate": 6.8864931353347045e-06,
+      "loss": 0.7044,
+      "step": 16258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5277692051006397,
+      "learning_rate": 6.886128240882288e-06,
+      "loss": 0.4264,
+      "step": 16259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6087756151843986,
+      "learning_rate": 6.885763334717528e-06,
+      "loss": 0.3641,
+      "step": 16260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2402472097378536,
+      "learning_rate": 6.885398416842687e-06,
+      "loss": 0.325,
+      "step": 16261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1740879338860233,
+      "learning_rate": 6.885033487260036e-06,
+      "loss": 0.6112,
+      "step": 16262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9749652002383242,
+      "learning_rate": 6.884668545971836e-06,
+      "loss": 0.4821,
+      "step": 16263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4166014692676008,
+      "learning_rate": 6.884303592980356e-06,
+      "loss": 0.3701,
+      "step": 16264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9107996826193672,
+      "learning_rate": 6.883938628287862e-06,
+      "loss": 0.523,
+      "step": 16265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4674260203553728,
+      "learning_rate": 6.88357365189662e-06,
+      "loss": 0.2625,
+      "step": 16266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5247625359540429,
+      "learning_rate": 6.8832086638088955e-06,
+      "loss": 0.2135,
+      "step": 16267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5017185598485816,
+      "learning_rate": 6.882843664026957e-06,
+      "loss": 0.3132,
+      "step": 16268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3104937566802604,
+      "learning_rate": 6.882478652553071e-06,
+      "loss": 0.1471,
+      "step": 16269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4547283132791526,
+      "learning_rate": 6.8821136293895e-06,
+      "loss": 0.2058,
+      "step": 16270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.994431148516907,
+      "learning_rate": 6.881748594538517e-06,
+      "loss": 0.369,
+      "step": 16271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6111175156717004,
+      "learning_rate": 6.8813835480023835e-06,
+      "loss": 0.2668,
+      "step": 16272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.098207820357707,
+      "learning_rate": 6.881018489783369e-06,
+      "loss": 0.2345,
+      "step": 16273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.629179092353337,
+      "learning_rate": 6.8806534198837405e-06,
+      "loss": 0.3458,
+      "step": 16274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.200679950867679,
+      "learning_rate": 6.880288338305764e-06,
+      "loss": 0.3343,
+      "step": 16275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7723479684185766,
+      "learning_rate": 6.879923245051706e-06,
+      "loss": 0.4419,
+      "step": 16276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.505768099045888,
+      "learning_rate": 6.879558140123834e-06,
+      "loss": 0.1875,
+      "step": 16277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5470364298539692,
+      "learning_rate": 6.8791930235244175e-06,
+      "loss": 0.2671,
+      "step": 16278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.601411847262874,
+      "learning_rate": 6.878827895255722e-06,
+      "loss": 0.3524,
+      "step": 16279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8914459589902746,
+      "learning_rate": 6.878462755320014e-06,
+      "loss": 0.2198,
+      "step": 16280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9142753132437673,
+      "learning_rate": 6.878097603719561e-06,
+      "loss": 0.1563,
+      "step": 16281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.840436382376659,
+      "learning_rate": 6.877732440456634e-06,
+      "loss": 0.5398,
+      "step": 16282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4329580149915553,
+      "learning_rate": 6.877367265533494e-06,
+      "loss": 0.2742,
+      "step": 16283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2750770101477509,
+      "learning_rate": 6.877002078952415e-06,
+      "loss": 0.2136,
+      "step": 16284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5194601714075615,
+      "learning_rate": 6.876636880715661e-06,
+      "loss": 0.32,
+      "step": 16285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9784211628573312,
+      "learning_rate": 6.876271670825503e-06,
+      "loss": 0.5044,
+      "step": 16286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4053326488901372,
+      "learning_rate": 6.8759064492842055e-06,
+      "loss": 0.1792,
+      "step": 16287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2046777687718742,
+      "learning_rate": 6.875541216094037e-06,
+      "loss": 0.1891,
+      "step": 16288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4556414953683665,
+      "learning_rate": 6.875175971257268e-06,
+      "loss": 0.5117,
+      "step": 16289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7098991474379386,
+      "learning_rate": 6.874810714776162e-06,
+      "loss": 0.4707,
+      "step": 16290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.486594421903228,
+      "learning_rate": 6.874445446652992e-06,
+      "loss": 0.1886,
+      "step": 16291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1912381452793572,
+      "learning_rate": 6.874080166890023e-06,
+      "loss": 0.4501,
+      "step": 16292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.708106677693022,
+      "learning_rate": 6.8737148754895255e-06,
+      "loss": 0.4589,
+      "step": 16293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8821813111042773,
+      "learning_rate": 6.8733495724537654e-06,
+      "loss": 0.4656,
+      "step": 16294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9111732084890818,
+      "learning_rate": 6.872984257785013e-06,
+      "loss": 0.2047,
+      "step": 16295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8975338610164696,
+      "learning_rate": 6.872618931485537e-06,
+      "loss": 0.3771,
+      "step": 16296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4637120491054765,
+      "learning_rate": 6.872253593557606e-06,
+      "loss": 0.1838,
+      "step": 16297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7774459264540687,
+      "learning_rate": 6.871888244003488e-06,
+      "loss": 0.3771,
+      "step": 16298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7633966588156973,
+      "learning_rate": 6.87152288282545e-06,
+      "loss": 0.6285,
+      "step": 16299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.837244739205317,
+      "learning_rate": 6.871157510025762e-06,
+      "loss": 0.3573,
+      "step": 16300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7291730848063047,
+      "learning_rate": 6.8707921256066955e-06,
+      "loss": 0.4487,
+      "step": 16301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5712722198930293,
+      "learning_rate": 6.870426729570515e-06,
+      "loss": 0.1577,
+      "step": 16302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8910190707535643,
+      "learning_rate": 6.870061321919493e-06,
+      "loss": 0.5788,
+      "step": 16303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.240514962169163,
+      "learning_rate": 6.869695902655898e-06,
+      "loss": 0.2781,
+      "step": 16304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7944071366831476,
+      "learning_rate": 6.869330471781997e-06,
+      "loss": 0.4765,
+      "step": 16305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4734953253069643,
+      "learning_rate": 6.868965029300062e-06,
+      "loss": 0.2309,
+      "step": 16306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7525702912985346,
+      "learning_rate": 6.86859957521236e-06,
+      "loss": 0.2239,
+      "step": 16307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.182690134898933,
+      "learning_rate": 6.868234109521164e-06,
+      "loss": 0.3137,
+      "step": 16308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8351132145460274,
+      "learning_rate": 6.867868632228738e-06,
+      "loss": 0.3501,
+      "step": 16309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4104938578460087,
+      "learning_rate": 6.867503143337355e-06,
+      "loss": 0.2717,
+      "step": 16310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4107362949079798,
+      "learning_rate": 6.867137642849285e-06,
+      "loss": 0.1642,
+      "step": 16311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7650252648671505,
+      "learning_rate": 6.866772130766796e-06,
+      "loss": 0.3779,
+      "step": 16312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3887066858083514,
+      "learning_rate": 6.866406607092159e-06,
+      "loss": 0.2363,
+      "step": 16313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8051581367701401,
+      "learning_rate": 6.866041071827641e-06,
+      "loss": 0.4882,
+      "step": 16314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4228865410374316,
+      "learning_rate": 6.865675524975516e-06,
+      "loss": 0.2997,
+      "step": 16315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.711990289465324,
+      "learning_rate": 6.8653099665380515e-06,
+      "loss": 0.3195,
+      "step": 16316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6549586624128791,
+      "learning_rate": 6.8649443965175175e-06,
+      "loss": 0.3606,
+      "step": 16317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8891844088530072,
+      "learning_rate": 6.8645788149161855e-06,
+      "loss": 0.3501,
+      "step": 16318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.84551807868614,
+      "learning_rate": 6.864213221736325e-06,
+      "loss": 0.2483,
+      "step": 16319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3126746271538687,
+      "learning_rate": 6.8638476169802045e-06,
+      "loss": 0.209,
+      "step": 16320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8071306584743199,
+      "learning_rate": 6.8634820006500965e-06,
+      "loss": 0.4842,
+      "step": 16321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4841957184791366,
+      "learning_rate": 6.863116372748271e-06,
+      "loss": 0.3335,
+      "step": 16322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4258786733232625,
+      "learning_rate": 6.862750733276997e-06,
+      "loss": 0.5561,
+      "step": 16323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5279357546975425,
+      "learning_rate": 6.862385082238549e-06,
+      "loss": 0.2181,
+      "step": 16324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8208429275917533,
+      "learning_rate": 6.862019419635191e-06,
+      "loss": 0.5164,
+      "step": 16325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9410737518555337,
+      "learning_rate": 6.861653745469199e-06,
+      "loss": 0.453,
+      "step": 16326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.223411734932078,
+      "learning_rate": 6.8612880597428435e-06,
+      "loss": 0.1812,
+      "step": 16327
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6613535075667027,
+      "learning_rate": 6.860922362458391e-06,
+      "loss": 0.3586,
+      "step": 16328
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1125090598021445,
+      "learning_rate": 6.8605566536181166e-06,
+      "loss": 0.1087,
+      "step": 16329
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.547002366871971,
+      "learning_rate": 6.860190933224291e-06,
+      "loss": 0.1826,
+      "step": 16330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6930653295883749,
+      "learning_rate": 6.859825201279185e-06,
+      "loss": 0.3082,
+      "step": 16331
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.833533982293069,
+      "learning_rate": 6.859459457785067e-06,
+      "loss": 0.6057,
+      "step": 16332
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9430140471230428,
+      "learning_rate": 6.85909370274421e-06,
+      "loss": 0.6016,
+      "step": 16333
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2017753325768807,
+      "learning_rate": 6.858727936158886e-06,
+      "loss": 0.2686,
+      "step": 16334
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3438865748985234,
+      "learning_rate": 6.858362158031366e-06,
+      "loss": 0.256,
+      "step": 16335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.826001580605862,
+      "learning_rate": 6.85799636836392e-06,
+      "loss": 0.2985,
+      "step": 16336
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.259969156474659,
+      "learning_rate": 6.857630567158822e-06,
+      "loss": 0.5273,
+      "step": 16337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.380057099141959,
+      "learning_rate": 6.8572647544183405e-06,
+      "loss": 0.2372,
+      "step": 16338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6374963219797891,
+      "learning_rate": 6.856898930144748e-06,
+      "loss": 0.2599,
+      "step": 16339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.54615919332222,
+      "learning_rate": 6.856533094340319e-06,
+      "loss": 0.3079,
+      "step": 16340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4241153730790133,
+      "learning_rate": 6.856167247007322e-06,
+      "loss": 0.1463,
+      "step": 16341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3434944145412526,
+      "learning_rate": 6.855801388148029e-06,
+      "loss": 0.155,
+      "step": 16342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9939634373300972,
+      "learning_rate": 6.855435517764714e-06,
+      "loss": 0.2686,
+      "step": 16343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7658705920341748,
+      "learning_rate": 6.855069635859647e-06,
+      "loss": 0.3488,
+      "step": 16344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8794680249261437,
+      "learning_rate": 6.8547037424351e-06,
+      "loss": 0.1704,
+      "step": 16345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.055755501879584,
+      "learning_rate": 6.854337837493346e-06,
+      "loss": 0.1819,
+      "step": 16346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5025250198063085,
+      "learning_rate": 6.853971921036657e-06,
+      "loss": 0.3723,
+      "step": 16347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3236503534616846,
+      "learning_rate": 6.853605993067306e-06,
+      "loss": 0.2992,
+      "step": 16348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.516603036303765,
+      "learning_rate": 6.853240053587564e-06,
+      "loss": 0.1792,
+      "step": 16349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7260511426448755,
+      "learning_rate": 6.852874102599705e-06,
+      "loss": 0.2149,
+      "step": 16350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2274590778498604,
+      "learning_rate": 6.852508140105998e-06,
+      "loss": 0.4689,
+      "step": 16351
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.267142764770582,
+      "learning_rate": 6.852142166108718e-06,
+      "loss": 0.2633,
+      "step": 16352
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.154838507757033,
+      "learning_rate": 6.851776180610139e-06,
+      "loss": 0.6665,
+      "step": 16353
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6449233559461607,
+      "learning_rate": 6.851410183612531e-06,
+      "loss": 0.1763,
+      "step": 16354
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.258268205652536,
+      "learning_rate": 6.851044175118169e-06,
+      "loss": 0.1853,
+      "step": 16355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5388474386773034,
+      "learning_rate": 6.850678155129324e-06,
+      "loss": 0.2904,
+      "step": 16356
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7955448493303754,
+      "learning_rate": 6.85031212364827e-06,
+      "loss": 0.2769,
+      "step": 16357
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1169618898949536,
+      "learning_rate": 6.849946080677279e-06,
+      "loss": 0.18,
+      "step": 16358
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7216863821400414,
+      "learning_rate": 6.849580026218625e-06,
+      "loss": 0.238,
+      "step": 16359
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.721393218417467,
+      "learning_rate": 6.849213960274581e-06,
+      "loss": 0.213,
+      "step": 16360
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.820557721037798,
+      "learning_rate": 6.84884788284742e-06,
+      "loss": 0.3554,
+      "step": 16361
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7291877603584807,
+      "learning_rate": 6.848481793939413e-06,
+      "loss": 0.5447,
+      "step": 16362
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5105977152695,
+      "learning_rate": 6.848115693552838e-06,
+      "loss": 0.2307,
+      "step": 16363
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3160252854674015,
+      "learning_rate": 6.847749581689965e-06,
+      "loss": 0.2526,
+      "step": 16364
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.327388693961866,
+      "learning_rate": 6.847383458353068e-06,
+      "loss": 0.2109,
+      "step": 16365
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3473311930622838,
+      "learning_rate": 6.84701732354442e-06,
+      "loss": 0.2809,
+      "step": 16366
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6103143080455622,
+      "learning_rate": 6.8466511772662956e-06,
+      "loss": 0.3265,
+      "step": 16367
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7330133144601299,
+      "learning_rate": 6.846285019520969e-06,
+      "loss": 0.4607,
+      "step": 16368
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7603206500178452,
+      "learning_rate": 6.845918850310713e-06,
+      "loss": 0.2434,
+      "step": 16369
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5224789636118576,
+      "learning_rate": 6.845552669637802e-06,
+      "loss": 0.2524,
+      "step": 16370
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.466263824043652,
+      "learning_rate": 6.845186477504507e-06,
+      "loss": 0.1931,
+      "step": 16371
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2512951638000318,
+      "learning_rate": 6.844820273913108e-06,
+      "loss": 0.2224,
+      "step": 16372
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.386475786227751,
+      "learning_rate": 6.844454058865873e-06,
+      "loss": 0.298,
+      "step": 16373
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.699770771991385,
+      "learning_rate": 6.844087832365078e-06,
+      "loss": 0.1917,
+      "step": 16374
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8420596334781774,
+      "learning_rate": 6.8437215944129984e-06,
+      "loss": 0.2222,
+      "step": 16375
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8965659570874598,
+      "learning_rate": 6.84335534501191e-06,
+      "loss": 0.4157,
+      "step": 16376
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1613790797947259,
+      "learning_rate": 6.842989084164082e-06,
+      "loss": 0.2045,
+      "step": 16377
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7072595988072,
+      "learning_rate": 6.842622811871792e-06,
+      "loss": 0.195,
+      "step": 16378
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2912981781679755,
+      "learning_rate": 6.842256528137315e-06,
+      "loss": 0.1584,
+      "step": 16379
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4760258723804633,
+      "learning_rate": 6.841890232962923e-06,
+      "loss": 0.1918,
+      "step": 16380
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6005564763148965,
+      "learning_rate": 6.841523926350895e-06,
+      "loss": 0.1792,
+      "step": 16381
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.320800431510611,
+      "learning_rate": 6.841157608303501e-06,
+      "loss": 0.2592,
+      "step": 16382
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4646732790618568,
+      "learning_rate": 6.8407912788230176e-06,
+      "loss": 0.2804,
+      "step": 16383
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8292918597618364,
+      "learning_rate": 6.840424937911719e-06,
+      "loss": 0.6496,
+      "step": 16384
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8242732352343682,
+      "learning_rate": 6.840058585571883e-06,
+      "loss": 0.1888,
+      "step": 16385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0625087853446886,
+      "learning_rate": 6.83969222180578e-06,
+      "loss": 0.496,
+      "step": 16386
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9418972169791824,
+      "learning_rate": 6.839325846615689e-06,
+      "loss": 0.3593,
+      "step": 16387
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.802668259256545,
+      "learning_rate": 6.838959460003882e-06,
+      "loss": 0.3963,
+      "step": 16388
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6516850513545909,
+      "learning_rate": 6.838593061972636e-06,
+      "loss": 0.2709,
+      "step": 16389
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4228641061747014,
+      "learning_rate": 6.838226652524226e-06,
+      "loss": 0.2406,
+      "step": 16390
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3885175777510215,
+      "learning_rate": 6.837860231660926e-06,
+      "loss": 0.2191,
+      "step": 16391
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5566710836477107,
+      "learning_rate": 6.837493799385013e-06,
+      "loss": 0.2446,
+      "step": 16392
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.105910981964151,
+      "learning_rate": 6.837127355698762e-06,
+      "loss": 0.3593,
+      "step": 16393
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3208195429401932,
+      "learning_rate": 6.836760900604447e-06,
+      "loss": 0.2818,
+      "step": 16394
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3834264478011493,
+      "learning_rate": 6.836394434104346e-06,
+      "loss": 0.2543,
+      "step": 16395
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6289250397913486,
+      "learning_rate": 6.8360279562007346e-06,
+      "loss": 0.2126,
+      "step": 16396
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6423022143933261,
+      "learning_rate": 6.8356614668958875e-06,
+      "loss": 0.3534,
+      "step": 16397
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8967579389929907,
+      "learning_rate": 6.83529496619208e-06,
+      "loss": 0.2644,
+      "step": 16398
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9300485304119408,
+      "learning_rate": 6.8349284540915885e-06,
+      "loss": 0.2296,
+      "step": 16399
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.609090247007264,
+      "learning_rate": 6.8345619305966885e-06,
+      "loss": 0.3443,
+      "step": 16400
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1327827416819294,
+      "learning_rate": 6.834195395709655e-06,
+      "loss": 0.1949,
+      "step": 16401
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5226552837989213,
+      "learning_rate": 6.833828849432769e-06,
+      "loss": 0.3778,
+      "step": 16402
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.135533183493369,
+      "learning_rate": 6.8334622917683e-06,
+      "loss": 0.2929,
+      "step": 16403
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4805953115149348,
+      "learning_rate": 6.833095722718528e-06,
+      "loss": 0.3605,
+      "step": 16404
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.091365442031061,
+      "learning_rate": 6.83272914228573e-06,
+      "loss": 0.6474,
+      "step": 16405
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5964926747217647,
+      "learning_rate": 6.83236255047218e-06,
+      "loss": 0.2496,
+      "step": 16406
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1981374404129037,
+      "learning_rate": 6.831995947280157e-06,
+      "loss": 0.2059,
+      "step": 16407
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6757094627956624,
+      "learning_rate": 6.8316293327119355e-06,
+      "loss": 0.4132,
+      "step": 16408
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9720989195373444,
+      "learning_rate": 6.831262706769792e-06,
+      "loss": 0.2088,
+      "step": 16409
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7683057056841585,
+      "learning_rate": 6.8308960694560046e-06,
+      "loss": 0.4643,
+      "step": 16410
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6234347985018445,
+      "learning_rate": 6.830529420772848e-06,
+      "loss": 0.3635,
+      "step": 16411
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.281790125207084,
+      "learning_rate": 6.8301627607226e-06,
+      "loss": 0.3329,
+      "step": 16412
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3336011355641966,
+      "learning_rate": 6.829796089307539e-06,
+      "loss": 0.2306,
+      "step": 16413
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6142367865271399,
+      "learning_rate": 6.829429406529941e-06,
+      "loss": 0.4062,
+      "step": 16414
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.173460925169377,
+      "learning_rate": 6.82906271239208e-06,
+      "loss": 0.1646,
+      "step": 16415
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.455203431129618,
+      "learning_rate": 6.8286960068962384e-06,
+      "loss": 0.2103,
+      "step": 16416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0721770976839133,
+      "learning_rate": 6.828329290044688e-06,
+      "loss": 0.564,
+      "step": 16417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8020909774118343,
+      "learning_rate": 6.827962561839711e-06,
+      "loss": 0.2494,
+      "step": 16418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1537862170967919,
+      "learning_rate": 6.8275958222835815e-06,
+      "loss": 0.1517,
+      "step": 16419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.446498716562963,
+      "learning_rate": 6.827229071378576e-06,
+      "loss": 0.1979,
+      "step": 16420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.023305244247553,
+      "learning_rate": 6.826862309126975e-06,
+      "loss": 0.429,
+      "step": 16421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4461556063323207,
+      "learning_rate": 6.8264955355310534e-06,
+      "loss": 0.4485,
+      "step": 16422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.162707980840033,
+      "learning_rate": 6.826128750593092e-06,
+      "loss": 0.2628,
+      "step": 16423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4373278229680395,
+      "learning_rate": 6.8257619543153645e-06,
+      "loss": 0.2724,
+      "step": 16424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6099845139428053,
+      "learning_rate": 6.825395146700151e-06,
+      "loss": 0.2318,
+      "step": 16425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8066542836265782,
+      "learning_rate": 6.825028327749727e-06,
+      "loss": 0.4801,
+      "step": 16426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.570509656229729,
+      "learning_rate": 6.824661497466374e-06,
+      "loss": 0.3523,
+      "step": 16427
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3817210752917763,
+      "learning_rate": 6.824294655852368e-06,
+      "loss": 0.1752,
+      "step": 16428
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.242417649343303,
+      "learning_rate": 6.823927802909987e-06,
+      "loss": 0.3847,
+      "step": 16429
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7791473313266644,
+      "learning_rate": 6.823560938641508e-06,
+      "loss": 0.4839,
+      "step": 16430
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1126905521812813,
+      "learning_rate": 6.8231940630492096e-06,
+      "loss": 0.3889,
+      "step": 16431
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.921369222442575,
+      "learning_rate": 6.822827176135372e-06,
+      "loss": 0.3341,
+      "step": 16432
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2522999837941655,
+      "learning_rate": 6.822460277902272e-06,
+      "loss": 0.2555,
+      "step": 16433
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7152780882152563,
+      "learning_rate": 6.822093368352187e-06,
+      "loss": 0.4349,
+      "step": 16434
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3735128130351586,
+      "learning_rate": 6.821726447487398e-06,
+      "loss": 0.3125,
+      "step": 16435
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9865042646812616,
+      "learning_rate": 6.821359515310181e-06,
+      "loss": 0.357,
+      "step": 16436
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7751535460890786,
+      "learning_rate": 6.820992571822814e-06,
+      "loss": 0.3564,
+      "step": 16437
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.603229536678544,
+      "learning_rate": 6.820625617027579e-06,
+      "loss": 0.2259,
+      "step": 16438
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7654212986067017,
+      "learning_rate": 6.820258650926751e-06,
+      "loss": 0.5006,
+      "step": 16439
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5259893807664155,
+      "learning_rate": 6.819891673522612e-06,
+      "loss": 0.2371,
+      "step": 16440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6681760158340573,
+      "learning_rate": 6.819524684817439e-06,
+      "loss": 0.2356,
+      "step": 16441
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4869111680774687,
+      "learning_rate": 6.81915768481351e-06,
+      "loss": 0.386,
+      "step": 16442
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5112633659960466,
+      "learning_rate": 6.818790673513106e-06,
+      "loss": 0.2193,
+      "step": 16443
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2190243994586365,
+      "learning_rate": 6.8184236509185046e-06,
+      "loss": 0.6976,
+      "step": 16444
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7929473441431465,
+      "learning_rate": 6.818056617031987e-06,
+      "loss": 0.206,
+      "step": 16445
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.635398140780085,
+      "learning_rate": 6.817689571855829e-06,
+      "loss": 0.3151,
+      "step": 16446
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8795632108025435,
+      "learning_rate": 6.817322515392313e-06,
+      "loss": 0.5343,
+      "step": 16447
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.553139896340574,
+      "learning_rate": 6.816955447643716e-06,
+      "loss": 0.3292,
+      "step": 16448
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2766031107051696,
+      "learning_rate": 6.81658836861232e-06,
+      "loss": 0.2114,
+      "step": 16449
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1560053755723256,
+      "learning_rate": 6.8162212783004015e-06,
+      "loss": 0.6736,
+      "step": 16450
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6072673074376664,
+      "learning_rate": 6.815854176710241e-06,
+      "loss": 0.3461,
+      "step": 16451
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.57973492380656,
+      "learning_rate": 6.8154870638441196e-06,
+      "loss": 0.5432,
+      "step": 16452
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4174337507506405,
+      "learning_rate": 6.815119939704315e-06,
+      "loss": 0.339,
+      "step": 16453
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6248848649648509,
+      "learning_rate": 6.81475280429311e-06,
+      "loss": 0.2545,
+      "step": 16454
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4496347647694374,
+      "learning_rate": 6.8143856576127795e-06,
+      "loss": 0.2885,
+      "step": 16455
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.244075741296056,
+      "learning_rate": 6.814018499665609e-06,
+      "loss": 0.5094,
+      "step": 16456
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5489295647247074,
+      "learning_rate": 6.813651330453873e-06,
+      "loss": 0.2311,
+      "step": 16457
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9559943103661213,
+      "learning_rate": 6.813284149979855e-06,
+      "loss": 0.5142,
+      "step": 16458
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6356091163368367,
+      "learning_rate": 6.812916958245835e-06,
+      "loss": 0.2046,
+      "step": 16459
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9059570919974453,
+      "learning_rate": 6.812549755254092e-06,
+      "loss": 0.3159,
+      "step": 16460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.722675120105457,
+      "learning_rate": 6.8121825410069066e-06,
+      "loss": 0.1976,
+      "step": 16461
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4290389533806804,
+      "learning_rate": 6.811815315506558e-06,
+      "loss": 0.2393,
+      "step": 16462
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3908979970600126,
+      "learning_rate": 6.81144807875533e-06,
+      "loss": 0.3356,
+      "step": 16463
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4640209328481835,
+      "learning_rate": 6.811080830755498e-06,
+      "loss": 0.2078,
+      "step": 16464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8698023857315953,
+      "learning_rate": 6.810713571509347e-06,
+      "loss": 0.3868,
+      "step": 16465
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4442571808116647,
+      "learning_rate": 6.810346301019155e-06,
+      "loss": 0.2992,
+      "step": 16466
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9203611668813123,
+      "learning_rate": 6.809979019287204e-06,
+      "loss": 0.3169,
+      "step": 16467
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6208768998307979,
+      "learning_rate": 6.809611726315773e-06,
+      "loss": 0.5213,
+      "step": 16468
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.436564579108084,
+      "learning_rate": 6.809244422107147e-06,
+      "loss": 0.2293,
+      "step": 16469
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4881233171584072,
+      "learning_rate": 6.8088771066636e-06,
+      "loss": 0.3989,
+      "step": 16470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.924396145418867,
+      "learning_rate": 6.808509779987419e-06,
+      "loss": 0.1575,
+      "step": 16471
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4113671632884812,
+      "learning_rate": 6.808142442080883e-06,
+      "loss": 0.416,
+      "step": 16472
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9629443792318002,
+      "learning_rate": 6.8077750929462715e-06,
+      "loss": 0.4919,
+      "step": 16473
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6667897258468445,
+      "learning_rate": 6.807407732585868e-06,
+      "loss": 0.5267,
+      "step": 16474
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8826385849054204,
+      "learning_rate": 6.807040361001951e-06,
+      "loss": 0.3414,
+      "step": 16475
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6494705795804943,
+      "learning_rate": 6.806672978196804e-06,
+      "loss": 0.4323,
+      "step": 16476
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7123316414940402,
+      "learning_rate": 6.80630558417271e-06,
+      "loss": 0.2891,
+      "step": 16477
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8672043887376129,
+      "learning_rate": 6.805938178931946e-06,
+      "loss": 0.3136,
+      "step": 16478
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.755912062859053,
+      "learning_rate": 6.805570762476797e-06,
+      "loss": 0.2252,
+      "step": 16479
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7308327849525307,
+      "learning_rate": 6.805203334809542e-06,
+      "loss": 0.315,
+      "step": 16480
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1558499898151555,
+      "learning_rate": 6.804835895932462e-06,
+      "loss": 0.2615,
+      "step": 16481
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2530697218791351,
+      "learning_rate": 6.804468445847845e-06,
+      "loss": 0.173,
+      "step": 16482
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4042378626747634,
+      "learning_rate": 6.804100984557967e-06,
+      "loss": 0.308,
+      "step": 16483
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6502526976154188,
+      "learning_rate": 6.803733512065109e-06,
+      "loss": 0.342,
+      "step": 16484
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7468574825015895,
+      "learning_rate": 6.803366028371558e-06,
+      "loss": 0.365,
+      "step": 16485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5624961328458613,
+      "learning_rate": 6.80299853347959e-06,
+      "loss": 0.4372,
+      "step": 16486
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8136006170599484,
+      "learning_rate": 6.8026310273914934e-06,
+      "loss": 0.2317,
+      "step": 16487
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2993368658256235,
+      "learning_rate": 6.8022635101095434e-06,
+      "loss": 0.1761,
+      "step": 16488
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8091339906960535,
+      "learning_rate": 6.801895981636028e-06,
+      "loss": 0.1992,
+      "step": 16489
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7294348390054257,
+      "learning_rate": 6.8015284419732275e-06,
+      "loss": 0.6648,
+      "step": 16490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.332006403895469,
+      "learning_rate": 6.801160891123424e-06,
+      "loss": 0.2369,
+      "step": 16491
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5894003959903917,
+      "learning_rate": 6.800793329088899e-06,
+      "loss": 0.213,
+      "step": 16492
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8253879639344723,
+      "learning_rate": 6.800425755871935e-06,
+      "loss": 0.2311,
+      "step": 16493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.134480819480593,
+      "learning_rate": 6.800058171474818e-06,
+      "loss": 0.6723,
+      "step": 16494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5775867526265666,
+      "learning_rate": 6.799690575899826e-06,
+      "loss": 0.2232,
+      "step": 16495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3738797891516168,
+      "learning_rate": 6.799322969149245e-06,
+      "loss": 0.1902,
+      "step": 16496
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6526634006322107,
+      "learning_rate": 6.798955351225354e-06,
+      "loss": 0.4367,
+      "step": 16497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3785039014850147,
+      "learning_rate": 6.798587722130439e-06,
+      "loss": 0.233,
+      "step": 16498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9036193744384314,
+      "learning_rate": 6.798220081866783e-06,
+      "loss": 0.1979,
+      "step": 16499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.408337312884135,
+      "learning_rate": 6.797852430436668e-06,
+      "loss": 0.3512,
+      "step": 16500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.801142058563145,
+      "learning_rate": 6.797484767842376e-06,
+      "loss": 0.2558,
+      "step": 16501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.008990988367252,
+      "learning_rate": 6.797117094086191e-06,
+      "loss": 0.3832,
+      "step": 16502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7985975674835672,
+      "learning_rate": 6.796749409170398e-06,
+      "loss": 0.3782,
+      "step": 16503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.57188854552968,
+      "learning_rate": 6.7963817130972765e-06,
+      "loss": 0.3228,
+      "step": 16504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.94662417959104,
+      "learning_rate": 6.796014005869112e-06,
+      "loss": 0.4835,
+      "step": 16505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9728866593664667,
+      "learning_rate": 6.795646287488189e-06,
+      "loss": 0.3333,
+      "step": 16506
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6227824228644625,
+      "learning_rate": 6.795278557956788e-06,
+      "loss": 0.2277,
+      "step": 16507
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.160184256456601,
+      "learning_rate": 6.794910817277195e-06,
+      "loss": 0.2257,
+      "step": 16508
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5710311497793792,
+      "learning_rate": 6.794543065451693e-06,
+      "loss": 0.239,
+      "step": 16509
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.832141844414681,
+      "learning_rate": 6.794175302482564e-06,
+      "loss": 0.379,
+      "step": 16510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8225487328647736,
+      "learning_rate": 6.793807528372094e-06,
+      "loss": 0.4047,
+      "step": 16511
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4472589737176889,
+      "learning_rate": 6.7934397431225656e-06,
+      "loss": 0.1637,
+      "step": 16512
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1174977979569767,
+      "learning_rate": 6.793071946736263e-06,
+      "loss": 0.602,
+      "step": 16513
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8098466871070296,
+      "learning_rate": 6.792704139215469e-06,
+      "loss": 0.5352,
+      "step": 16514
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.66787318328882,
+      "learning_rate": 6.7923363205624695e-06,
+      "loss": 0.2547,
+      "step": 16515
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8167649109841213,
+      "learning_rate": 6.791968490779547e-06,
+      "loss": 0.2295,
+      "step": 16516
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.092225882938077,
+      "learning_rate": 6.791600649868986e-06,
+      "loss": 0.497,
+      "step": 16517
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.285723543618662,
+      "learning_rate": 6.791232797833072e-06,
+      "loss": 0.3081,
+      "step": 16518
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0044165959694493,
+      "learning_rate": 6.790864934674086e-06,
+      "loss": 0.3735,
+      "step": 16519
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7945346642919668,
+      "learning_rate": 6.790497060394316e-06,
+      "loss": 0.4013,
+      "step": 16520
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9954113811048004,
+      "learning_rate": 6.790129174996044e-06,
+      "loss": 0.6038,
+      "step": 16521
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.556338061863055,
+      "learning_rate": 6.789761278481556e-06,
+      "loss": 0.4168,
+      "step": 16522
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9772339536040102,
+      "learning_rate": 6.789393370853136e-06,
+      "loss": 0.3959,
+      "step": 16523
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2093507368773642,
+      "learning_rate": 6.789025452113068e-06,
+      "loss": 0.2492,
+      "step": 16524
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.728751331357795,
+      "learning_rate": 6.788657522263636e-06,
+      "loss": 0.3034,
+      "step": 16525
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4185296832862528,
+      "learning_rate": 6.788289581307128e-06,
+      "loss": 0.2597,
+      "step": 16526
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3786396114620652,
+      "learning_rate": 6.787921629245826e-06,
+      "loss": 0.2343,
+      "step": 16527
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6624675940728995,
+      "learning_rate": 6.787553666082016e-06,
+      "loss": 0.2247,
+      "step": 16528
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8135895085328317,
+      "learning_rate": 6.787185691817981e-06,
+      "loss": 0.3763,
+      "step": 16529
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.558967914026147,
+      "learning_rate": 6.786817706456008e-06,
+      "loss": 0.4003,
+      "step": 16530
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.54321826897064,
+      "learning_rate": 6.786449709998383e-06,
+      "loss": 0.208,
+      "step": 16531
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5545970655760866,
+      "learning_rate": 6.786081702447388e-06,
+      "loss": 0.318,
+      "step": 16532
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7464235739881278,
+      "learning_rate": 6.785713683805311e-06,
+      "loss": 0.1839,
+      "step": 16533
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8943045981989122,
+      "learning_rate": 6.785345654074436e-06,
+      "loss": 0.2114,
+      "step": 16534
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5651861942607068,
+      "learning_rate": 6.78497761325705e-06,
+      "loss": 0.2929,
+      "step": 16535
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.463700527402443,
+      "learning_rate": 6.784609561355435e-06,
+      "loss": 0.2124,
+      "step": 16536
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.96380423112901,
+      "learning_rate": 6.784241498371881e-06,
+      "loss": 0.3649,
+      "step": 16537
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7425068896094498,
+      "learning_rate": 6.78387342430867e-06,
+      "loss": 0.2264,
+      "step": 16538
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4638481724212182,
+      "learning_rate": 6.783505339168088e-06,
+      "loss": 0.282,
+      "step": 16539
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.099692480527339,
+      "learning_rate": 6.783137242952423e-06,
+      "loss": 0.4135,
+      "step": 16540
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6088076952686203,
+      "learning_rate": 6.782769135663958e-06,
+      "loss": 0.2534,
+      "step": 16541
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1492297147063109,
+      "learning_rate": 6.782401017304981e-06,
+      "loss": 0.1882,
+      "step": 16542
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1255124004700654,
+      "learning_rate": 6.782032887877778e-06,
+      "loss": 0.1496,
+      "step": 16543
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0974782819814901,
+      "learning_rate": 6.781664747384633e-06,
+      "loss": 0.1549,
+      "step": 16544
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0820942309159722,
+      "learning_rate": 6.781296595827832e-06,
+      "loss": 0.2501,
+      "step": 16545
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.590623460536579,
+      "learning_rate": 6.780928433209663e-06,
+      "loss": 0.319,
+      "step": 16546
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1103201771966489,
+      "learning_rate": 6.780560259532413e-06,
+      "loss": 0.168,
+      "step": 16547
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1671119061915047,
+      "learning_rate": 6.7801920747983656e-06,
+      "loss": 0.1965,
+      "step": 16548
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8073129835896236,
+      "learning_rate": 6.779823879009808e-06,
+      "loss": 0.3925,
+      "step": 16549
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.69118057203451,
+      "learning_rate": 6.7794556721690265e-06,
+      "loss": 0.2699,
+      "step": 16550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3019190042432787,
+      "learning_rate": 6.779087454278308e-06,
+      "loss": 0.2115,
+      "step": 16551
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.814131434146971,
+      "learning_rate": 6.77871922533994e-06,
+      "loss": 0.5142,
+      "step": 16552
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6960468327917995,
+      "learning_rate": 6.778350985356206e-06,
+      "loss": 0.2438,
+      "step": 16553
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2762395040424224,
+      "learning_rate": 6.777982734329397e-06,
+      "loss": 0.3109,
+      "step": 16554
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.369541984152631,
+      "learning_rate": 6.777614472261795e-06,
+      "loss": 0.2124,
+      "step": 16555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5167501859432162,
+      "learning_rate": 6.77724619915569e-06,
+      "loss": 0.2581,
+      "step": 16556
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4304842943381364,
+      "learning_rate": 6.776877915013369e-06,
+      "loss": 0.197,
+      "step": 16557
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5261871234121303,
+      "learning_rate": 6.776509619837117e-06,
+      "loss": 0.3743,
+      "step": 16558
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5913059109351784,
+      "learning_rate": 6.776141313629224e-06,
+      "loss": 0.3793,
+      "step": 16559
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8126056369054802,
+      "learning_rate": 6.775772996391974e-06,
+      "loss": 0.3766,
+      "step": 16560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0657742260284846,
+      "learning_rate": 6.7754046681276554e-06,
+      "loss": 0.5253,
+      "step": 16561
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6986860470666802,
+      "learning_rate": 6.775036328838556e-06,
+      "loss": 0.3116,
+      "step": 16562
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8416399236574255,
+      "learning_rate": 6.774667978526961e-06,
+      "loss": 0.3448,
+      "step": 16563
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8259954692546595,
+      "learning_rate": 6.774299617195159e-06,
+      "loss": 0.191,
+      "step": 16564
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5420399441610697,
+      "learning_rate": 6.773931244845439e-06,
+      "loss": 0.3324,
+      "step": 16565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9525576911992333,
+      "learning_rate": 6.773562861480086e-06,
+      "loss": 0.2644,
+      "step": 16566
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.614241805923222,
+      "learning_rate": 6.773194467101388e-06,
+      "loss": 0.4603,
+      "step": 16567
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.589906199210548,
+      "learning_rate": 6.772826061711635e-06,
+      "loss": 0.129,
+      "step": 16568
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0859181484721134,
+      "learning_rate": 6.772457645313112e-06,
+      "loss": 0.1713,
+      "step": 16569
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5437111046559149,
+      "learning_rate": 6.772089217908109e-06,
+      "loss": 0.3639,
+      "step": 16570
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5724030997314895,
+      "learning_rate": 6.771720779498912e-06,
+      "loss": 0.1834,
+      "step": 16571
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7071177157363735,
+      "learning_rate": 6.771352330087808e-06,
+      "loss": 0.3731,
+      "step": 16572
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7118737520127691,
+      "learning_rate": 6.770983869677088e-06,
+      "loss": 0.3337,
+      "step": 16573
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6506481263210013,
+      "learning_rate": 6.770615398269037e-06,
+      "loss": 0.4511,
+      "step": 16574
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8728947249501207,
+      "learning_rate": 6.7702469158659455e-06,
+      "loss": 0.2207,
+      "step": 16575
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6721100820172639,
+      "learning_rate": 6.769878422470101e-06,
+      "loss": 0.4975,
+      "step": 16576
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7873115883744948,
+      "learning_rate": 6.769509918083792e-06,
+      "loss": 0.4227,
+      "step": 16577
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.352518706736612,
+      "learning_rate": 6.769141402709305e-06,
+      "loss": 0.2331,
+      "step": 16578
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9161209754095292,
+      "learning_rate": 6.768772876348929e-06,
+      "loss": 0.3238,
+      "step": 16579
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.640239518155935,
+      "learning_rate": 6.768404339004954e-06,
+      "loss": 0.3959,
+      "step": 16580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6324667130971675,
+      "learning_rate": 6.768035790679669e-06,
+      "loss": 0.2342,
+      "step": 16581
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8074047246102392,
+      "learning_rate": 6.767667231375361e-06,
+      "loss": 0.3435,
+      "step": 16582
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.873472714272142,
+      "learning_rate": 6.767298661094317e-06,
+      "loss": 0.4866,
+      "step": 16583
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9198748096739424,
+      "learning_rate": 6.766930079838829e-06,
+      "loss": 0.2475,
+      "step": 16584
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.448099966829352,
+      "learning_rate": 6.766561487611184e-06,
+      "loss": 0.4018,
+      "step": 16585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2458876905043996,
+      "learning_rate": 6.766192884413671e-06,
+      "loss": 0.336,
+      "step": 16586
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6377107156610367,
+      "learning_rate": 6.76582427024858e-06,
+      "loss": 0.198,
+      "step": 16587
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0512477852718156,
+      "learning_rate": 6.7654556451182e-06,
+      "loss": 0.5627,
+      "step": 16588
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.399865295324401,
+      "learning_rate": 6.765087009024818e-06,
+      "loss": 0.2247,
+      "step": 16589
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8496384442148812,
+      "learning_rate": 6.764718361970725e-06,
+      "loss": 0.1875,
+      "step": 16590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7159430862124634,
+      "learning_rate": 6.764349703958209e-06,
+      "loss": 0.3142,
+      "step": 16591
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3756731834621603,
+      "learning_rate": 6.763981034989559e-06,
+      "loss": 0.1985,
+      "step": 16592
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7295714650435772,
+      "learning_rate": 6.763612355067066e-06,
+      "loss": 0.3495,
+      "step": 16593
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2901049502553499,
+      "learning_rate": 6.763243664193019e-06,
+      "loss": 0.1715,
+      "step": 16594
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4749221166964832,
+      "learning_rate": 6.7628749623697075e-06,
+      "loss": 0.3075,
+      "step": 16595
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7633127351687086,
+      "learning_rate": 6.762506249599419e-06,
+      "loss": 0.2678,
+      "step": 16596
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.585781315220835,
+      "learning_rate": 6.762137525884446e-06,
+      "loss": 0.4092,
+      "step": 16597
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9806424258773507,
+      "learning_rate": 6.761768791227076e-06,
+      "loss": 0.1303,
+      "step": 16598
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0987845625448402,
+      "learning_rate": 6.761400045629601e-06,
+      "loss": 0.1417,
+      "step": 16599
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8573123967812235,
+      "learning_rate": 6.761031289094307e-06,
+      "loss": 0.4725,
+      "step": 16600
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6870413881113675,
+      "learning_rate": 6.760662521623489e-06,
+      "loss": 0.223,
+      "step": 16601
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.508823937565467,
+      "learning_rate": 6.760293743219432e-06,
+      "loss": 0.2945,
+      "step": 16602
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7280469334406299,
+      "learning_rate": 6.759924953884429e-06,
+      "loss": 0.465,
+      "step": 16603
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.739943608483165,
+      "learning_rate": 6.75955615362077e-06,
+      "loss": 0.1801,
+      "step": 16604
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7767511490747943,
+      "learning_rate": 6.759187342430743e-06,
+      "loss": 0.3435,
+      "step": 16605
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1886617003828701,
+      "learning_rate": 6.758818520316641e-06,
+      "loss": 0.3266,
+      "step": 16606
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6984927500956641,
+      "learning_rate": 6.758449687280752e-06,
+      "loss": 0.2186,
+      "step": 16607
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7552087276318689,
+      "learning_rate": 6.75808084332537e-06,
+      "loss": 0.4419,
+      "step": 16608
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7949014585743095,
+      "learning_rate": 6.75771198845278e-06,
+      "loss": 0.2184,
+      "step": 16609
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0317625132800026,
+      "learning_rate": 6.757343122665277e-06,
+      "loss": 0.3364,
+      "step": 16610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3022562233266646,
+      "learning_rate": 6.756974245965148e-06,
+      "loss": 0.2222,
+      "step": 16611
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7627683397207925,
+      "learning_rate": 6.7566053583546866e-06,
+      "loss": 0.3071,
+      "step": 16612
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7571573011234418,
+      "learning_rate": 6.756236459836182e-06,
+      "loss": 0.5167,
+      "step": 16613
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.677156517601962,
+      "learning_rate": 6.755867550411925e-06,
+      "loss": 0.1582,
+      "step": 16614
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1514483491511944,
+      "learning_rate": 6.755498630084208e-06,
+      "loss": 0.3664,
+      "step": 16615
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6317494116626179,
+      "learning_rate": 6.755129698855319e-06,
+      "loss": 0.2762,
+      "step": 16616
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.833364436759956,
+      "learning_rate": 6.754760756727552e-06,
+      "loss": 0.3947,
+      "step": 16617
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7863087252029137,
+      "learning_rate": 6.754391803703195e-06,
+      "loss": 0.1935,
+      "step": 16618
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.180843238285745,
+      "learning_rate": 6.7540228397845435e-06,
+      "loss": 0.2663,
+      "step": 16619
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.019231194759855,
+      "learning_rate": 6.7536538649738835e-06,
+      "loss": 0.5987,
+      "step": 16620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.570379270016556,
+      "learning_rate": 6.7532848792735096e-06,
+      "loss": 0.3198,
+      "step": 16621
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.378531299601439,
+      "learning_rate": 6.752915882685711e-06,
+      "loss": 0.2767,
+      "step": 16622
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3031582064251321,
+      "learning_rate": 6.752546875212781e-06,
+      "loss": 0.2303,
+      "step": 16623
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7010699350009773,
+      "learning_rate": 6.75217785685701e-06,
+      "loss": 0.4597,
+      "step": 16624
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5233136444607123,
+      "learning_rate": 6.751808827620689e-06,
+      "loss": 0.2977,
+      "step": 16625
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.917466569873939,
+      "learning_rate": 6.751439787506112e-06,
+      "loss": 0.5068,
+      "step": 16626
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7779086944302849,
+      "learning_rate": 6.751070736515568e-06,
+      "loss": 0.2654,
+      "step": 16627
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.640660101651298,
+      "learning_rate": 6.750701674651348e-06,
+      "loss": 0.2659,
+      "step": 16628
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6988681782351214,
+      "learning_rate": 6.750332601915748e-06,
+      "loss": 0.3465,
+      "step": 16629
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.758082852236391,
+      "learning_rate": 6.749963518311056e-06,
+      "loss": 0.2329,
+      "step": 16630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7875217233018075,
+      "learning_rate": 6.749594423839565e-06,
+      "loss": 0.3863,
+      "step": 16631
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2056532125944215,
+      "learning_rate": 6.749225318503567e-06,
+      "loss": 0.3519,
+      "step": 16632
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4901296046313637,
+      "learning_rate": 6.748856202305355e-06,
+      "loss": 0.3703,
+      "step": 16633
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8626138303521778,
+      "learning_rate": 6.748487075247221e-06,
+      "loss": 0.3911,
+      "step": 16634
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9101659778433078,
+      "learning_rate": 6.748117937331454e-06,
+      "loss": 0.4632,
+      "step": 16635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3582124918761977,
+      "learning_rate": 6.7477487885603506e-06,
+      "loss": 0.2828,
+      "step": 16636
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1670533966228716,
+      "learning_rate": 6.747379628936201e-06,
+      "loss": 0.2383,
+      "step": 16637
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7015188653813953,
+      "learning_rate": 6.747010458461298e-06,
+      "loss": 0.492,
+      "step": 16638
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2609410694194696,
+      "learning_rate": 6.746641277137933e-06,
+      "loss": 0.1586,
+      "step": 16639
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1338022389003595,
+      "learning_rate": 6.746272084968399e-06,
+      "loss": 0.1385,
+      "step": 16640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7463792998381316,
+      "learning_rate": 6.745902881954991e-06,
+      "loss": 0.4861,
+      "step": 16641
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.749081924286998,
+      "learning_rate": 6.745533668099998e-06,
+      "loss": 0.4143,
+      "step": 16642
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.736158540766596,
+      "learning_rate": 6.745164443405715e-06,
+      "loss": 0.4803,
+      "step": 16643
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7735286050661545,
+      "learning_rate": 6.744795207874434e-06,
+      "loss": 0.4878,
+      "step": 16644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7910611888381935,
+      "learning_rate": 6.744425961508447e-06,
+      "loss": 0.3165,
+      "step": 16645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2150891368235577,
+      "learning_rate": 6.744056704310049e-06,
+      "loss": 0.3387,
+      "step": 16646
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.309942373266783,
+      "learning_rate": 6.743687436281532e-06,
+      "loss": 0.2208,
+      "step": 16647
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4576623900927508,
+      "learning_rate": 6.743318157425189e-06,
+      "loss": 0.2301,
+      "step": 16648
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2526383031709782,
+      "learning_rate": 6.742948867743312e-06,
+      "loss": 0.7587,
+      "step": 16649
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4339048739232667,
+      "learning_rate": 6.742579567238196e-06,
+      "loss": 0.2871,
+      "step": 16650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4769620985410787,
+      "learning_rate": 6.7422102559121326e-06,
+      "loss": 0.4535,
+      "step": 16651
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8553130797285189,
+      "learning_rate": 6.7418409337674175e-06,
+      "loss": 0.3705,
+      "step": 16652
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7764096037428456,
+      "learning_rate": 6.741471600806341e-06,
+      "loss": 0.341,
+      "step": 16653
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.676646740477726,
+      "learning_rate": 6.741102257031198e-06,
+      "loss": 0.5059,
+      "step": 16654
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0175314872436585,
+      "learning_rate": 6.740732902444285e-06,
+      "loss": 0.1698,
+      "step": 16655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6244379914880989,
+      "learning_rate": 6.74036353704789e-06,
+      "loss": 0.4811,
+      "step": 16656
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0209227919850727,
+      "learning_rate": 6.7399941608443096e-06,
+      "loss": 0.6121,
+      "step": 16657
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9748633548690946,
+      "learning_rate": 6.7396247738358375e-06,
+      "loss": 0.2662,
+      "step": 16658
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8881610339276755,
+      "learning_rate": 6.739255376024769e-06,
+      "loss": 0.2051,
+      "step": 16659
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8385903347065031,
+      "learning_rate": 6.738885967413394e-06,
+      "loss": 0.4452,
+      "step": 16660
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.867021764797718,
+      "learning_rate": 6.738516548004012e-06,
+      "loss": 0.2139,
+      "step": 16661
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9119278584905235,
+      "learning_rate": 6.73814711779891e-06,
+      "loss": 0.5415,
+      "step": 16662
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.838670662352156,
+      "learning_rate": 6.7377776768003885e-06,
+      "loss": 0.4948,
+      "step": 16663
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5246700142860288,
+      "learning_rate": 6.7374082250107376e-06,
+      "loss": 0.3862,
+      "step": 16664
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3343326510514517,
+      "learning_rate": 6.7370387624322534e-06,
+      "loss": 0.1936,
+      "step": 16665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1670395518039447,
+      "learning_rate": 6.736669289067229e-06,
+      "loss": 0.6191,
+      "step": 16666
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.672336193895596,
+      "learning_rate": 6.736299804917961e-06,
+      "loss": 0.5383,
+      "step": 16667
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.844550805992453,
+      "learning_rate": 6.735930309986739e-06,
+      "loss": 0.1808,
+      "step": 16668
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2521493632693725,
+      "learning_rate": 6.7355608042758634e-06,
+      "loss": 0.3143,
+      "step": 16669
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4522361061912734,
+      "learning_rate": 6.735191287787626e-06,
+      "loss": 0.1904,
+      "step": 16670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5163392566581078,
+      "learning_rate": 6.734821760524319e-06,
+      "loss": 0.442,
+      "step": 16671
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9566681576446407,
+      "learning_rate": 6.734452222488241e-06,
+      "loss": 0.6103,
+      "step": 16672
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0455477453395081,
+      "learning_rate": 6.734082673681683e-06,
+      "loss": 0.1854,
+      "step": 16673
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8128957973966189,
+      "learning_rate": 6.733713114106944e-06,
+      "loss": 0.524,
+      "step": 16674
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6315685626174699,
+      "learning_rate": 6.733343543766316e-06,
+      "loss": 0.3481,
+      "step": 16675
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.469332622568452,
+      "learning_rate": 6.732973962662096e-06,
+      "loss": 0.1548,
+      "step": 16676
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8538832104035368,
+      "learning_rate": 6.732604370796576e-06,
+      "loss": 0.425,
+      "step": 16677
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2150497737130939,
+      "learning_rate": 6.732234768172055e-06,
+      "loss": 0.2026,
+      "step": 16678
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9287140510630527,
+      "learning_rate": 6.731865154790823e-06,
+      "loss": 0.2717,
+      "step": 16679
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2955594686707994,
+      "learning_rate": 6.731495530655179e-06,
+      "loss": 0.1565,
+      "step": 16680
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.587012523237549,
+      "learning_rate": 6.731125895767419e-06,
+      "loss": 0.1997,
+      "step": 16681
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.696983540033524,
+      "learning_rate": 6.730756250129834e-06,
+      "loss": 0.171,
+      "step": 16682
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5499740248472513,
+      "learning_rate": 6.730386593744724e-06,
+      "loss": 0.1943,
+      "step": 16683
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6615277941503075,
+      "learning_rate": 6.7300169266143815e-06,
+      "loss": 0.4973,
+      "step": 16684
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.326105023414047,
+      "learning_rate": 6.729647248741105e-06,
+      "loss": 0.1994,
+      "step": 16685
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6604403779978465,
+      "learning_rate": 6.729277560127185e-06,
+      "loss": 0.225,
+      "step": 16686
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5696878009095645,
+      "learning_rate": 6.728907860774923e-06,
+      "loss": 0.1933,
+      "step": 16687
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7882974563070935,
+      "learning_rate": 6.7285381506866106e-06,
+      "loss": 0.4135,
+      "step": 16688
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5647524619935609,
+      "learning_rate": 6.728168429864545e-06,
+      "loss": 0.3243,
+      "step": 16689
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8552927917569697,
+      "learning_rate": 6.727798698311024e-06,
+      "loss": 0.6169,
+      "step": 16690
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8462041185007452,
+      "learning_rate": 6.72742895602834e-06,
+      "loss": 0.3083,
+      "step": 16691
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7420924721020186,
+      "learning_rate": 6.727059203018792e-06,
+      "loss": 0.3124,
+      "step": 16692
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.092894308112672,
+      "learning_rate": 6.726689439284674e-06,
+      "loss": 0.5173,
+      "step": 16693
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.544399650651793,
+      "learning_rate": 6.726319664828283e-06,
+      "loss": 0.3723,
+      "step": 16694
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.57061313009881,
+      "learning_rate": 6.725949879651914e-06,
+      "loss": 0.209,
+      "step": 16695
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5670065205306152,
+      "learning_rate": 6.725580083757865e-06,
+      "loss": 0.2601,
+      "step": 16696
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8138920108348986,
+      "learning_rate": 6.725210277148433e-06,
+      "loss": 0.2435,
+      "step": 16697
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4489642071370634,
+      "learning_rate": 6.724840459825912e-06,
+      "loss": 0.3283,
+      "step": 16698
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2443331624857696,
+      "learning_rate": 6.724470631792599e-06,
+      "loss": 0.2796,
+      "step": 16699
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9920567967088637,
+      "learning_rate": 6.724100793050792e-06,
+      "loss": 0.1906,
+      "step": 16700
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7279840201460213,
+      "learning_rate": 6.7237309436027866e-06,
+      "loss": 0.3963,
+      "step": 16701
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2924170714517782,
+      "learning_rate": 6.723361083450879e-06,
+      "loss": 0.1663,
+      "step": 16702
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.552975094927442,
+      "learning_rate": 6.722991212597366e-06,
+      "loss": 0.289,
+      "step": 16703
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.174839363798066,
+      "learning_rate": 6.722621331044546e-06,
+      "loss": 0.4143,
+      "step": 16704
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7338942159943422,
+      "learning_rate": 6.722251438794714e-06,
+      "loss": 0.1797,
+      "step": 16705
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6428796375793757,
+      "learning_rate": 6.7218815358501666e-06,
+      "loss": 0.1707,
+      "step": 16706
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8900332529716457,
+      "learning_rate": 6.721511622213204e-06,
+      "loss": 0.4336,
+      "step": 16707
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3970236023963534,
+      "learning_rate": 6.72114169788612e-06,
+      "loss": 0.3342,
+      "step": 16708
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8563528002745393,
+      "learning_rate": 6.720771762871213e-06,
+      "loss": 0.3733,
+      "step": 16709
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8552659335468071,
+      "learning_rate": 6.72040181717078e-06,
+      "loss": 0.2825,
+      "step": 16710
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8621724026230113,
+      "learning_rate": 6.720031860787118e-06,
+      "loss": 0.3395,
+      "step": 16711
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.835461067715551,
+      "learning_rate": 6.719661893722527e-06,
+      "loss": 0.2371,
+      "step": 16712
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3242525023263365,
+      "learning_rate": 6.7192919159792995e-06,
+      "loss": 0.2507,
+      "step": 16713
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1191186636455952,
+      "learning_rate": 6.718921927559736e-06,
+      "loss": 0.1951,
+      "step": 16714
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9174890908984095,
+      "learning_rate": 6.718551928466133e-06,
+      "loss": 0.4699,
+      "step": 16715
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9175507581401117,
+      "learning_rate": 6.71818191870079e-06,
+      "loss": 0.3847,
+      "step": 16716
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3780633589849027,
+      "learning_rate": 6.717811898266002e-06,
+      "loss": 0.1833,
+      "step": 16717
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8870073338665785,
+      "learning_rate": 6.717441867164068e-06,
+      "loss": 0.479,
+      "step": 16718
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.65377780107383,
+      "learning_rate": 6.717071825397287e-06,
+      "loss": 0.3537,
+      "step": 16719
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.002233264210146,
+      "learning_rate": 6.716701772967955e-06,
+      "loss": 0.4518,
+      "step": 16720
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6629998257038234,
+      "learning_rate": 6.71633170987837e-06,
+      "loss": 0.2815,
+      "step": 16721
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.507005764778878,
+      "learning_rate": 6.715961636130831e-06,
+      "loss": 0.2022,
+      "step": 16722
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8799059581570248,
+      "learning_rate": 6.715591551727636e-06,
+      "loss": 0.4554,
+      "step": 16723
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4370151977289558,
+      "learning_rate": 6.715221456671083e-06,
+      "loss": 0.2012,
+      "step": 16724
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7557683290972117,
+      "learning_rate": 6.714851350963469e-06,
+      "loss": 0.2002,
+      "step": 16725
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7761944418908795,
+      "learning_rate": 6.714481234607093e-06,
+      "loss": 0.4944,
+      "step": 16726
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.530597306811373,
+      "learning_rate": 6.714111107604255e-06,
+      "loss": 0.2103,
+      "step": 16727
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5089132461551569,
+      "learning_rate": 6.71374096995725e-06,
+      "loss": 0.2891,
+      "step": 16728
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.502693485204954,
+      "learning_rate": 6.713370821668379e-06,
+      "loss": 0.1408,
+      "step": 16729
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.238166866076035,
+      "learning_rate": 6.713000662739942e-06,
+      "loss": 0.4331,
+      "step": 16730
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.323264751734614,
+      "learning_rate": 6.7126304931742324e-06,
+      "loss": 0.2319,
+      "step": 16731
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7252150624389697,
+      "learning_rate": 6.712260312973552e-06,
+      "loss": 0.4393,
+      "step": 16732
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4344914533469206,
+      "learning_rate": 6.711890122140201e-06,
+      "loss": 0.3267,
+      "step": 16733
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9294823294228567,
+      "learning_rate": 6.711519920676477e-06,
+      "loss": 0.4133,
+      "step": 16734
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6897370844000443,
+      "learning_rate": 6.711149708584677e-06,
+      "loss": 0.2321,
+      "step": 16735
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6985221617806836,
+      "learning_rate": 6.710779485867103e-06,
+      "loss": 0.4342,
+      "step": 16736
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2600015634134674,
+      "learning_rate": 6.710409252526051e-06,
+      "loss": 0.3443,
+      "step": 16737
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.816764394256127,
+      "learning_rate": 6.7100390085638226e-06,
+      "loss": 0.4446,
+      "step": 16738
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.452113296709549,
+      "learning_rate": 6.709668753982714e-06,
+      "loss": 0.2269,
+      "step": 16739
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.189325773021664,
+      "learning_rate": 6.7092984887850275e-06,
+      "loss": 0.6533,
+      "step": 16740
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9131539610141528,
+      "learning_rate": 6.70892821297306e-06,
+      "loss": 0.1298,
+      "step": 16741
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.753720466952397,
+      "learning_rate": 6.708557926549113e-06,
+      "loss": 0.4336,
+      "step": 16742
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0639613999608093,
+      "learning_rate": 6.708187629515483e-06,
+      "loss": 0.4371,
+      "step": 16743
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6775533936021976,
+      "learning_rate": 6.7078173218744734e-06,
+      "loss": 0.2442,
+      "step": 16744
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4479977587305817,
+      "learning_rate": 6.70744700362838e-06,
+      "loss": 0.3821,
+      "step": 16745
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4518093612420115,
+      "learning_rate": 6.707076674779504e-06,
+      "loss": 0.1983,
+      "step": 16746
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3687841683307853,
+      "learning_rate": 6.706706335330145e-06,
+      "loss": 0.2555,
+      "step": 16747
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4065537322929473,
+      "learning_rate": 6.706335985282602e-06,
+      "loss": 0.5078,
+      "step": 16748
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.471798766038924,
+      "learning_rate": 6.705965624639176e-06,
+      "loss": 0.1626,
+      "step": 16749
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7015795783034435,
+      "learning_rate": 6.705595253402165e-06,
+      "loss": 0.4055,
+      "step": 16750
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2207252027422444,
+      "learning_rate": 6.705224871573871e-06,
+      "loss": 0.2731,
+      "step": 16751
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3523821467739556,
+      "learning_rate": 6.704854479156593e-06,
+      "loss": 0.1881,
+      "step": 16752
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1297118351386324,
+      "learning_rate": 6.7044840761526304e-06,
+      "loss": 0.2479,
+      "step": 16753
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7945891975875043,
+      "learning_rate": 6.7041136625642846e-06,
+      "loss": 0.3759,
+      "step": 16754
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8644521740827646,
+      "learning_rate": 6.703743238393854e-06,
+      "loss": 0.5507,
+      "step": 16755
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2826272522897542,
+      "learning_rate": 6.7033728036436415e-06,
+      "loss": 0.3093,
+      "step": 16756
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6766015114916104,
+      "learning_rate": 6.7030023583159445e-06,
+      "loss": 0.5424,
+      "step": 16757
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5484788604278155,
+      "learning_rate": 6.702631902413066e-06,
+      "loss": 0.3144,
+      "step": 16758
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9773905943355732,
+      "learning_rate": 6.702261435937304e-06,
+      "loss": 0.2048,
+      "step": 16759
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0359840649763914,
+      "learning_rate": 6.701890958890962e-06,
+      "loss": 0.4021,
+      "step": 16760
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5389019742021857,
+      "learning_rate": 6.701520471276337e-06,
+      "loss": 0.3834,
+      "step": 16761
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2974612479358114,
+      "learning_rate": 6.701149973095732e-06,
+      "loss": 0.2124,
+      "step": 16762
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8318232111148647,
+      "learning_rate": 6.700779464351445e-06,
+      "loss": 0.4776,
+      "step": 16763
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5100164661551767,
+      "learning_rate": 6.700408945045782e-06,
+      "loss": 0.297,
+      "step": 16764
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4934860540967467,
+      "learning_rate": 6.700038415181039e-06,
+      "loss": 0.4689,
+      "step": 16765
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6958533624064547,
+      "learning_rate": 6.699667874759519e-06,
+      "loss": 0.5066,
+      "step": 16766
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0787697379883525,
+      "learning_rate": 6.6992973237835225e-06,
+      "loss": 0.3745,
+      "step": 16767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7166270279435905,
+      "learning_rate": 6.69892676225535e-06,
+      "loss": 0.4001,
+      "step": 16768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3307383210998907,
+      "learning_rate": 6.698556190177304e-06,
+      "loss": 0.1254,
+      "step": 16769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8189437697445103,
+      "learning_rate": 6.698185607551684e-06,
+      "loss": 0.4959,
+      "step": 16770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8016031917399977,
+      "learning_rate": 6.697815014380792e-06,
+      "loss": 0.2938,
+      "step": 16771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5328567794019266,
+      "learning_rate": 6.697444410666929e-06,
+      "loss": 0.2036,
+      "step": 16772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.030247822536188,
+      "learning_rate": 6.697073796412397e-06,
+      "loss": 0.4924,
+      "step": 16773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.875165782990376,
+      "learning_rate": 6.6967031716194965e-06,
+      "loss": 0.2537,
+      "step": 16774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3039404524179512,
+      "learning_rate": 6.69633253629053e-06,
+      "loss": 0.1619,
+      "step": 16775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4547228206562737,
+      "learning_rate": 6.695961890427798e-06,
+      "loss": 0.209,
+      "step": 16776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8006837926049708,
+      "learning_rate": 6.695591234033603e-06,
+      "loss": 0.4468,
+      "step": 16777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4010329735201423,
+      "learning_rate": 6.695220567110245e-06,
+      "loss": 0.2774,
+      "step": 16778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8835439883493716,
+      "learning_rate": 6.694849889660029e-06,
+      "loss": 0.498,
+      "step": 16779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3886944256600005,
+      "learning_rate": 6.694479201685253e-06,
+      "loss": 0.1576,
+      "step": 16780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.630405659864698,
+      "learning_rate": 6.694108503188222e-06,
+      "loss": 0.2203,
+      "step": 16781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.758962644899837,
+      "learning_rate": 6.693737794171236e-06,
+      "loss": 0.3541,
+      "step": 16782
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6349351081817525,
+      "learning_rate": 6.693367074636598e-06,
+      "loss": 0.1221,
+      "step": 16783
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8979622516817607,
+      "learning_rate": 6.69299634458661e-06,
+      "loss": 0.3538,
+      "step": 16784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2605842353893624,
+      "learning_rate": 6.692625604023572e-06,
+      "loss": 0.2153,
+      "step": 16785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.50828582864934,
+      "learning_rate": 6.692254852949791e-06,
+      "loss": 0.4521,
+      "step": 16786
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3238152674660584,
+      "learning_rate": 6.691884091367564e-06,
+      "loss": 0.2758,
+      "step": 16787
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.36444803712575,
+      "learning_rate": 6.691513319279197e-06,
+      "loss": 0.1778,
+      "step": 16788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9297673791885357,
+      "learning_rate": 6.691142536686992e-06,
+      "loss": 0.3835,
+      "step": 16789
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2341716303066597,
+      "learning_rate": 6.690771743593248e-06,
+      "loss": 0.2649,
+      "step": 16790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6748964201158685,
+      "learning_rate": 6.690400940000273e-06,
+      "loss": 0.3704,
+      "step": 16791
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6056267226712202,
+      "learning_rate": 6.6900301259103624e-06,
+      "loss": 0.4403,
+      "step": 16792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8831317164063601,
+      "learning_rate": 6.6896593013258275e-06,
+      "loss": 0.3835,
+      "step": 16793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1418227445664957,
+      "learning_rate": 6.689288466248963e-06,
+      "loss": 0.1856,
+      "step": 16794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2952555507193497,
+      "learning_rate": 6.688917620682078e-06,
+      "loss": 0.8348,
+      "step": 16795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7896900908616424,
+      "learning_rate": 6.68854676462747e-06,
+      "loss": 0.35,
+      "step": 16796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.754031608026792,
+      "learning_rate": 6.688175898087447e-06,
+      "loss": 0.2224,
+      "step": 16797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8978251815867169,
+      "learning_rate": 6.6878050210643085e-06,
+      "loss": 0.547,
+      "step": 16798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7654744733938283,
+      "learning_rate": 6.687434133560359e-06,
+      "loss": 0.3,
+      "step": 16799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.198735466904183,
+      "learning_rate": 6.6870632355779006e-06,
+      "loss": 0.5464,
+      "step": 16800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5881796664107497,
+      "learning_rate": 6.686692327119237e-06,
+      "loss": 0.3576,
+      "step": 16801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6047312395324766,
+      "learning_rate": 6.686321408186672e-06,
+      "loss": 0.214,
+      "step": 16802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5552440634308955,
+      "learning_rate": 6.685950478782508e-06,
+      "loss": 0.2792,
+      "step": 16803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4513361543929295,
+      "learning_rate": 6.6855795389090486e-06,
+      "loss": 0.3336,
+      "step": 16804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.582662126942804,
+      "learning_rate": 6.6852085885685965e-06,
+      "loss": 0.1714,
+      "step": 16805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9280173375197984,
+      "learning_rate": 6.684837627763457e-06,
+      "loss": 0.5529,
+      "step": 16806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8465883463394948,
+      "learning_rate": 6.684466656495934e-06,
+      "loss": 0.2608,
+      "step": 16807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4113621680437645,
+      "learning_rate": 6.684095674768329e-06,
+      "loss": 0.3368,
+      "step": 16808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6728787060888746,
+      "learning_rate": 6.683724682582946e-06,
+      "loss": 0.5156,
+      "step": 16809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4384312256753489,
+      "learning_rate": 6.683353679942091e-06,
+      "loss": 0.4965,
+      "step": 16810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5939745557823155,
+      "learning_rate": 6.6829826668480644e-06,
+      "loss": 0.3198,
+      "step": 16811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6734665354356644,
+      "learning_rate": 6.682611643303172e-06,
+      "loss": 0.4598,
+      "step": 16812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1614589858659128,
+      "learning_rate": 6.682240609309719e-06,
+      "loss": 0.1402,
+      "step": 16813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5320465500012663,
+      "learning_rate": 6.6818695648700075e-06,
+      "loss": 0.3335,
+      "step": 16814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8153781237161333,
+      "learning_rate": 6.681498509986342e-06,
+      "loss": 0.5018,
+      "step": 16815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.805562723636663,
+      "learning_rate": 6.6811274446610265e-06,
+      "loss": 0.2642,
+      "step": 16816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1777723639957998,
+      "learning_rate": 6.6807563688963664e-06,
+      "loss": 0.1827,
+      "step": 16817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6111214800022329,
+      "learning_rate": 6.680385282694664e-06,
+      "loss": 0.344,
+      "step": 16818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7790527490328183,
+      "learning_rate": 6.680014186058224e-06,
+      "loss": 0.3773,
+      "step": 16819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.24399955854042,
+      "learning_rate": 6.679643078989353e-06,
+      "loss": 0.2218,
+      "step": 16820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0290777659051225,
+      "learning_rate": 6.679271961490354e-06,
+      "loss": 0.4256,
+      "step": 16821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8039711153734947,
+      "learning_rate": 6.6789008335635315e-06,
+      "loss": 0.2253,
+      "step": 16822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7886851529130294,
+      "learning_rate": 6.6785296952111885e-06,
+      "loss": 0.2168,
+      "step": 16823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5284940931586186,
+      "learning_rate": 6.678158546435633e-06,
+      "loss": 0.1804,
+      "step": 16824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.496073398895161,
+      "learning_rate": 6.677787387239168e-06,
+      "loss": 0.6744,
+      "step": 16825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2176815753908745,
+      "learning_rate": 6.677416217624099e-06,
+      "loss": 0.3254,
+      "step": 16826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5925769439124002,
+      "learning_rate": 6.677045037592729e-06,
+      "loss": 0.3338,
+      "step": 16827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0960606138029116,
+      "learning_rate": 6.676673847147365e-06,
+      "loss": 0.252,
+      "step": 16828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4646528094142883,
+      "learning_rate": 6.67630264629031e-06,
+      "loss": 0.4071,
+      "step": 16829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8192424883264744,
+      "learning_rate": 6.67593143502387e-06,
+      "loss": 0.4239,
+      "step": 16830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1326880024866393,
+      "learning_rate": 6.675560213350352e-06,
+      "loss": 0.1309,
+      "step": 16831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.627159767262539,
+      "learning_rate": 6.675188981272058e-06,
+      "loss": 0.1879,
+      "step": 16832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.518582648605556,
+      "learning_rate": 6.674817738791296e-06,
+      "loss": 0.4362,
+      "step": 16833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3111962927991654,
+      "learning_rate": 6.674446485910369e-06,
+      "loss": 0.2041,
+      "step": 16834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7199802070826182,
+      "learning_rate": 6.6740752226315845e-06,
+      "loss": 0.2983,
+      "step": 16835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9917660619765315,
+      "learning_rate": 6.673703948957245e-06,
+      "loss": 0.0927,
+      "step": 16836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6893433728152474,
+      "learning_rate": 6.67333266488966e-06,
+      "loss": 0.2208,
+      "step": 16837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7329729659455972,
+      "learning_rate": 6.672961370431131e-06,
+      "loss": 0.4316,
+      "step": 16838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.581412340300771,
+      "learning_rate": 6.672590065583968e-06,
+      "loss": 0.3551,
+      "step": 16839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8340656274706144,
+      "learning_rate": 6.6722187503504725e-06,
+      "loss": 0.3194,
+      "step": 16840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5453431800442723,
+      "learning_rate": 6.671847424732953e-06,
+      "loss": 0.3625,
+      "step": 16841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7809113883363212,
+      "learning_rate": 6.6714760887337136e-06,
+      "loss": 0.3258,
+      "step": 16842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6619087595736788,
+      "learning_rate": 6.671104742355061e-06,
+      "loss": 0.4743,
+      "step": 16843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6967689217557398,
+      "learning_rate": 6.670733385599303e-06,
+      "loss": 0.392,
+      "step": 16844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6036961554915847,
+      "learning_rate": 6.670362018468743e-06,
+      "loss": 0.4264,
+      "step": 16845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5155899225678764,
+      "learning_rate": 6.669990640965687e-06,
+      "loss": 0.2509,
+      "step": 16846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2227407615143366,
+      "learning_rate": 6.669619253092442e-06,
+      "loss": 0.2593,
+      "step": 16847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.973300537799255,
+      "learning_rate": 6.669247854851315e-06,
+      "loss": 0.4828,
+      "step": 16848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.20057989112712,
+      "learning_rate": 6.66887644624461e-06,
+      "loss": 0.4675,
+      "step": 16849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6876693216783039,
+      "learning_rate": 6.668505027274636e-06,
+      "loss": 0.2993,
+      "step": 16850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2051984890717375,
+      "learning_rate": 6.668133597943699e-06,
+      "loss": 0.2015,
+      "step": 16851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.172867808763439,
+      "learning_rate": 6.667762158254104e-06,
+      "loss": 0.5834,
+      "step": 16852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4498895496197803,
+      "learning_rate": 6.667390708208158e-06,
+      "loss": 0.3208,
+      "step": 16853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.877550091716531,
+      "learning_rate": 6.6670192478081684e-06,
+      "loss": 0.5595,
+      "step": 16854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.203709701405611,
+      "learning_rate": 6.666647777056441e-06,
+      "loss": 0.1334,
+      "step": 16855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5821627456270098,
+      "learning_rate": 6.666276295955284e-06,
+      "loss": 0.3497,
+      "step": 16856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1132016789292218,
+      "learning_rate": 6.665904804507002e-06,
+      "loss": 0.1326,
+      "step": 16857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.442725784549679,
+      "learning_rate": 6.665533302713902e-06,
+      "loss": 0.3231,
+      "step": 16858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4472919569824698,
+      "learning_rate": 6.665161790578294e-06,
+      "loss": 0.2441,
+      "step": 16859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4768226384289496,
+      "learning_rate": 6.664790268102481e-06,
+      "loss": 0.2968,
+      "step": 16860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5253182881258553,
+      "learning_rate": 6.664418735288774e-06,
+      "loss": 0.2189,
+      "step": 16861
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7639414516845746,
+      "learning_rate": 6.664047192139477e-06,
+      "loss": 0.4372,
+      "step": 16862
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0425973558277497,
+      "learning_rate": 6.663675638656899e-06,
+      "loss": 0.652,
+      "step": 16863
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4454497581465566,
+      "learning_rate": 6.6633040748433445e-06,
+      "loss": 0.271,
+      "step": 16864
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.57852348876973,
+      "learning_rate": 6.662932500701125e-06,
+      "loss": 0.3354,
+      "step": 16865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5458128707252006,
+      "learning_rate": 6.662560916232546e-06,
+      "loss": 0.3842,
+      "step": 16866
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2855859838035226,
+      "learning_rate": 6.662189321439912e-06,
+      "loss": 0.3783,
+      "step": 16867
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4850482041509823,
+      "learning_rate": 6.661817716325535e-06,
+      "loss": 0.2554,
+      "step": 16868
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6252785155145277,
+      "learning_rate": 6.66144610089172e-06,
+      "loss": 0.1793,
+      "step": 16869
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.697632287984689,
+      "learning_rate": 6.6610744751407755e-06,
+      "loss": 0.355,
+      "step": 16870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8608506321750709,
+      "learning_rate": 6.660702839075008e-06,
+      "loss": 0.226,
+      "step": 16871
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8461288990415057,
+      "learning_rate": 6.660331192696729e-06,
+      "loss": 0.283,
+      "step": 16872
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9910273296572725,
+      "learning_rate": 6.65995953600824e-06,
+      "loss": 0.4705,
+      "step": 16873
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0603242329003457,
+      "learning_rate": 6.659587869011856e-06,
+      "loss": 0.3754,
+      "step": 16874
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.029974821441337,
+      "learning_rate": 6.6592161917098785e-06,
+      "loss": 0.3418,
+      "step": 16875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7961360199655467,
+      "learning_rate": 6.65884450410462e-06,
+      "loss": 0.5191,
+      "step": 16876
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5350271272592095,
+      "learning_rate": 6.6584728061983875e-06,
+      "loss": 0.2824,
+      "step": 16877
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8220013900153786,
+      "learning_rate": 6.658101097993488e-06,
+      "loss": 0.3755,
+      "step": 16878
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.802744707303179,
+      "learning_rate": 6.657729379492229e-06,
+      "loss": 0.4623,
+      "step": 16879
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6235643496075685,
+      "learning_rate": 6.657357650696922e-06,
+      "loss": 0.2035,
+      "step": 16880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6032391773336243,
+      "learning_rate": 6.656985911609872e-06,
+      "loss": 0.3066,
+      "step": 16881
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0987050494711228,
+      "learning_rate": 6.65661416223339e-06,
+      "loss": 0.1511,
+      "step": 16882
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7977386643261288,
+      "learning_rate": 6.656242402569784e-06,
+      "loss": 0.2355,
+      "step": 16883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6940613741300086,
+      "learning_rate": 6.65587063262136e-06,
+      "loss": 0.2314,
+      "step": 16884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.813948874319847,
+      "learning_rate": 6.655498852390429e-06,
+      "loss": 0.1949,
+      "step": 16885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3815383361514388,
+      "learning_rate": 6.6551270618793e-06,
+      "loss": 0.2321,
+      "step": 16886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6872893423810473,
+      "learning_rate": 6.65475526109028e-06,
+      "loss": 0.2098,
+      "step": 16887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0678804760849114,
+      "learning_rate": 6.65438345002568e-06,
+      "loss": 0.2164,
+      "step": 16888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6338072899476,
+      "learning_rate": 6.6540116286878056e-06,
+      "loss": 0.4288,
+      "step": 16889
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3889083714575716,
+      "learning_rate": 6.653639797078969e-06,
+      "loss": 0.3563,
+      "step": 16890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6570788504792715,
+      "learning_rate": 6.653267955201476e-06,
+      "loss": 0.249,
+      "step": 16891
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8245628244827488,
+      "learning_rate": 6.65289610305764e-06,
+      "loss": 0.4064,
+      "step": 16892
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2374402990293634,
+      "learning_rate": 6.652524240649765e-06,
+      "loss": 0.185,
+      "step": 16893
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6014631007913303,
+      "learning_rate": 6.652152367980164e-06,
+      "loss": 0.2265,
+      "step": 16894
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7805241184906937,
+      "learning_rate": 6.651780485051144e-06,
+      "loss": 0.3225,
+      "step": 16895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.857531539294974,
+      "learning_rate": 6.651408591865016e-06,
+      "loss": 0.4546,
+      "step": 16896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5853727655331107,
+      "learning_rate": 6.651036688424089e-06,
+      "loss": 0.1791,
+      "step": 16897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4265587972737086,
+      "learning_rate": 6.650664774730671e-06,
+      "loss": 0.3373,
+      "step": 16898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4605635149023934,
+      "learning_rate": 6.650292850787072e-06,
+      "loss": 0.3056,
+      "step": 16899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8524696485900476,
+      "learning_rate": 6.649920916595603e-06,
+      "loss": 0.3468,
+      "step": 16900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0963056894387755,
+      "learning_rate": 6.649548972158573e-06,
+      "loss": 0.7569,
+      "step": 16901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.296655136452864,
+      "learning_rate": 6.649177017478289e-06,
+      "loss": 0.1863,
+      "step": 16902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.622430521398769,
+      "learning_rate": 6.648805052557065e-06,
+      "loss": 0.3671,
+      "step": 16903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.870826260076149,
+      "learning_rate": 6.648433077397207e-06,
+      "loss": 0.2742,
+      "step": 16904
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3086356227495348,
+      "learning_rate": 6.648061092001028e-06,
+      "loss": 0.271,
+      "step": 16905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8654940619118572,
+      "learning_rate": 6.647689096370836e-06,
+      "loss": 0.2871,
+      "step": 16906
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7055044678963913,
+      "learning_rate": 6.647317090508942e-06,
+      "loss": 0.3409,
+      "step": 16907
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.815714382361436,
+      "learning_rate": 6.646945074417656e-06,
+      "loss": 0.2224,
+      "step": 16908
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.846027822039712,
+      "learning_rate": 6.646573048099287e-06,
+      "loss": 0.5297,
+      "step": 16909
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0049004465899234,
+      "learning_rate": 6.646201011556147e-06,
+      "loss": 0.4646,
+      "step": 16910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8832659155635083,
+      "learning_rate": 6.645828964790544e-06,
+      "loss": 0.3523,
+      "step": 16911
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.567195729056268,
+      "learning_rate": 6.64545690780479e-06,
+      "loss": 0.282,
+      "step": 16912
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4850172510573982,
+      "learning_rate": 6.6450848406011945e-06,
+      "loss": 0.2835,
+      "step": 16913
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4918741501349084,
+      "learning_rate": 6.644712763182069e-06,
+      "loss": 0.2864,
+      "step": 16914
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1367923612310875,
+      "learning_rate": 6.644340675549723e-06,
+      "loss": 0.3992,
+      "step": 16915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4462798464093154,
+      "learning_rate": 6.643968577706468e-06,
+      "loss": 0.1758,
+      "step": 16916
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8661176816914478,
+      "learning_rate": 6.643596469654612e-06,
+      "loss": 0.2061,
+      "step": 16917
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3301373509176067,
+      "learning_rate": 6.643224351396469e-06,
+      "loss": 0.2189,
+      "step": 16918
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.368213810584658,
+      "learning_rate": 6.6428522229343486e-06,
+      "loss": 0.3317,
+      "step": 16919
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.684173973009213,
+      "learning_rate": 6.6424800842705615e-06,
+      "loss": 0.3948,
+      "step": 16920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1740613474013981,
+      "learning_rate": 6.6421079354074195e-06,
+      "loss": 0.1709,
+      "step": 16921
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.736911932994241,
+      "learning_rate": 6.64173577634723e-06,
+      "loss": 0.234,
+      "step": 16922
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5150113127275784,
+      "learning_rate": 6.641363607092308e-06,
+      "loss": 0.566,
+      "step": 16923
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7086992778031003,
+      "learning_rate": 6.640991427644963e-06,
+      "loss": 0.3562,
+      "step": 16924
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.506332572972784,
+      "learning_rate": 6.6406192380075065e-06,
+      "loss": 0.3317,
+      "step": 16925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5380453461634445,
+      "learning_rate": 6.6402470381822494e-06,
+      "loss": 0.4271,
+      "step": 16926
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.547089821897517,
+      "learning_rate": 6.639874828171503e-06,
+      "loss": 0.3213,
+      "step": 16927
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3062903985479544,
+      "learning_rate": 6.639502607977578e-06,
+      "loss": 0.1511,
+      "step": 16928
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2798168683480267,
+      "learning_rate": 6.6391303776027885e-06,
+      "loss": 0.173,
+      "step": 16929
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3651981567250147,
+      "learning_rate": 6.638758137049441e-06,
+      "loss": 0.1683,
+      "step": 16930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4750945903464625,
+      "learning_rate": 6.638385886319853e-06,
+      "loss": 0.1482,
+      "step": 16931
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3615396901694083,
+      "learning_rate": 6.638013625416331e-06,
+      "loss": 0.2068,
+      "step": 16932
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4661139004235386,
+      "learning_rate": 6.637641354341188e-06,
+      "loss": 0.1951,
+      "step": 16933
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4022695925558322,
+      "learning_rate": 6.637269073096738e-06,
+      "loss": 0.3224,
+      "step": 16934
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.841204899043246,
+      "learning_rate": 6.6368967816852895e-06,
+      "loss": 0.3429,
+      "step": 16935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5115854933998756,
+      "learning_rate": 6.636524480109158e-06,
+      "loss": 0.3252,
+      "step": 16936
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1457847201987852,
+      "learning_rate": 6.636152168370651e-06,
+      "loss": 0.1605,
+      "step": 16937
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7017451698816939,
+      "learning_rate": 6.6357798464720845e-06,
+      "loss": 0.3899,
+      "step": 16938
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4521303566972368,
+      "learning_rate": 6.635407514415768e-06,
+      "loss": 0.3416,
+      "step": 16939
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3464426325787109,
+      "learning_rate": 6.635035172204015e-06,
+      "loss": 0.2654,
+      "step": 16940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1767614269596798,
+      "learning_rate": 6.634662819839137e-06,
+      "loss": 0.161,
+      "step": 16941
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7959021920377018,
+      "learning_rate": 6.634290457323446e-06,
+      "loss": 0.4055,
+      "step": 16942
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3228281168575693,
+      "learning_rate": 6.633918084659254e-06,
+      "loss": 0.2566,
+      "step": 16943
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5529829294171673,
+      "learning_rate": 6.633545701848875e-06,
+      "loss": 0.3898,
+      "step": 16944
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0224621106918004,
+      "learning_rate": 6.63317330889462e-06,
+      "loss": 0.1313,
+      "step": 16945
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9849528500100257,
+      "learning_rate": 6.6328009057988e-06,
+      "loss": 0.4558,
+      "step": 16946
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.832282693684438,
+      "learning_rate": 6.632428492563731e-06,
+      "loss": 0.6136,
+      "step": 16947
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6952996958552866,
+      "learning_rate": 6.632056069191723e-06,
+      "loss": 0.2226,
+      "step": 16948
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5601490072159612,
+      "learning_rate": 6.631683635685089e-06,
+      "loss": 0.1754,
+      "step": 16949
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.426921292024848,
+      "learning_rate": 6.631311192046142e-06,
+      "loss": 0.3523,
+      "step": 16950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.538282680946719,
+      "learning_rate": 6.6309387382771975e-06,
+      "loss": 0.1604,
+      "step": 16951
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.108226092363478,
+      "learning_rate": 6.630566274380563e-06,
+      "loss": 0.4573,
+      "step": 16952
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5178527289013837,
+      "learning_rate": 6.630193800358555e-06,
+      "loss": 0.3297,
+      "step": 16953
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4736470928979066,
+      "learning_rate": 6.6298213162134865e-06,
+      "loss": 0.328,
+      "step": 16954
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9506265682054036,
+      "learning_rate": 6.6294488219476685e-06,
+      "loss": 0.6359,
+      "step": 16955
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7205493546536321,
+      "learning_rate": 6.629076317563414e-06,
+      "loss": 0.3492,
+      "step": 16956
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.981064957130749,
+      "learning_rate": 6.6287038030630405e-06,
+      "loss": 0.1514,
+      "step": 16957
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.541870356379963,
+      "learning_rate": 6.628331278448855e-06,
+      "loss": 0.2637,
+      "step": 16958
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5398099459146068,
+      "learning_rate": 6.627958743723175e-06,
+      "loss": 0.1716,
+      "step": 16959
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.847451140166749,
+      "learning_rate": 6.627586198888314e-06,
+      "loss": 0.2604,
+      "step": 16960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.949604185408634,
+      "learning_rate": 6.6272136439465825e-06,
+      "loss": 0.4262,
+      "step": 16961
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5593002792604869,
+      "learning_rate": 6.626841078900297e-06,
+      "loss": 0.4214,
+      "step": 16962
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7996505861757421,
+      "learning_rate": 6.6264685037517695e-06,
+      "loss": 0.5439,
+      "step": 16963
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3299533880731356,
+      "learning_rate": 6.626095918503313e-06,
+      "loss": 0.2403,
+      "step": 16964
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.88261940270634,
+      "learning_rate": 6.625723323157242e-06,
+      "loss": 0.4558,
+      "step": 16965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.45394789432655,
+      "learning_rate": 6.625350717715871e-06,
+      "loss": 0.2043,
+      "step": 16966
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9026391506938238,
+      "learning_rate": 6.624978102181512e-06,
+      "loss": 0.3882,
+      "step": 16967
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6889926238696487,
+      "learning_rate": 6.624605476556479e-06,
+      "loss": 0.2099,
+      "step": 16968
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3727404989257916,
+      "learning_rate": 6.624232840843086e-06,
+      "loss": 0.5005,
+      "step": 16969
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5591692375932937,
+      "learning_rate": 6.62386019504365e-06,
+      "loss": 0.1784,
+      "step": 16970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0719909455118044,
+      "learning_rate": 6.623487539160481e-06,
+      "loss": 0.1648,
+      "step": 16971
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6720979956282596,
+      "learning_rate": 6.623114873195895e-06,
+      "loss": 0.1932,
+      "step": 16972
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5412257840504153,
+      "learning_rate": 6.6227421971522065e-06,
+      "loss": 0.3308,
+      "step": 16973
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.017189823093155,
+      "learning_rate": 6.6223695110317285e-06,
+      "loss": 0.399,
+      "step": 16974
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4450250507302689,
+      "learning_rate": 6.621996814836775e-06,
+      "loss": 0.195,
+      "step": 16975
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3356051408030303,
+      "learning_rate": 6.621624108569663e-06,
+      "loss": 0.3139,
+      "step": 16976
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3224656560878718,
+      "learning_rate": 6.621251392232704e-06,
+      "loss": 0.321,
+      "step": 16977
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.277568738995392,
+      "learning_rate": 6.620878665828214e-06,
+      "loss": 0.1779,
+      "step": 16978
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.707743973853573,
+      "learning_rate": 6.620505929358506e-06,
+      "loss": 0.176,
+      "step": 16979
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8597496661002466,
+      "learning_rate": 6.620133182825897e-06,
+      "loss": 0.4584,
+      "step": 16980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.598850680485039,
+      "learning_rate": 6.6197604262326985e-06,
+      "loss": 0.4017,
+      "step": 16981
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6614303456961663,
+      "learning_rate": 6.619387659581228e-06,
+      "loss": 0.2925,
+      "step": 16982
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6865733481102991,
+      "learning_rate": 6.6190148828738e-06,
+      "loss": 0.2059,
+      "step": 16983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5504779723309385,
+      "learning_rate": 6.618642096112728e-06,
+      "loss": 0.3385,
+      "step": 16984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.909889763899829,
+      "learning_rate": 6.618269299300329e-06,
+      "loss": 0.148,
+      "step": 16985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.151794003378522,
+      "learning_rate": 6.617896492438913e-06,
+      "loss": 0.4054,
+      "step": 16986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5566946496527077,
+      "learning_rate": 6.617523675530803e-06,
+      "loss": 0.2824,
+      "step": 16987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5388454669088123,
+      "learning_rate": 6.617150848578306e-06,
+      "loss": 0.3441,
+      "step": 16988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5977050953325826,
+      "learning_rate": 6.616778011583744e-06,
+      "loss": 0.2091,
+      "step": 16989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4911023445785108,
+      "learning_rate": 6.616405164549427e-06,
+      "loss": 0.1928,
+      "step": 16990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.792623873185262,
+      "learning_rate": 6.616032307477673e-06,
+      "loss": 0.41,
+      "step": 16991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2078414510136772,
+      "learning_rate": 6.615659440370795e-06,
+      "loss": 0.3328,
+      "step": 16992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.682502937533973,
+      "learning_rate": 6.615286563231113e-06,
+      "loss": 0.2007,
+      "step": 16993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.427768451163996,
+      "learning_rate": 6.6149136760609364e-06,
+      "loss": 0.2209,
+      "step": 16994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9491645790171552,
+      "learning_rate": 6.614540778862586e-06,
+      "loss": 0.5564,
+      "step": 16995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7998688490261996,
+      "learning_rate": 6.614167871638375e-06,
+      "loss": 0.4302,
+      "step": 16996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3525406861553706,
+      "learning_rate": 6.613794954390619e-06,
+      "loss": 0.3449,
+      "step": 16997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2955048344181486,
+      "learning_rate": 6.613422027121634e-06,
+      "loss": 0.348,
+      "step": 16998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1805780425181813,
+      "learning_rate": 6.613049089833736e-06,
+      "loss": 0.2618,
+      "step": 16999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5959603068528079,
+      "learning_rate": 6.612676142529241e-06,
+      "loss": 0.3768,
+      "step": 17000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3048114161395357,
+      "learning_rate": 6.6123031852104625e-06,
+      "loss": 0.3841,
+      "step": 17001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3597025476656264,
+      "learning_rate": 6.611930217879721e-06,
+      "loss": 0.1918,
+      "step": 17002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.743369420027123,
+      "learning_rate": 6.611557240539328e-06,
+      "loss": 0.3349,
+      "step": 17003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1682729588036211,
+      "learning_rate": 6.611184253191603e-06,
+      "loss": 0.1214,
+      "step": 17004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3633856070781238,
+      "learning_rate": 6.61081125583886e-06,
+      "loss": 0.2031,
+      "step": 17005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5464488174501667,
+      "learning_rate": 6.610438248483417e-06,
+      "loss": 0.3944,
+      "step": 17006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5989212307265004,
+      "learning_rate": 6.610065231127587e-06,
+      "loss": 0.3641,
+      "step": 17007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8442191379756465,
+      "learning_rate": 6.60969220377369e-06,
+      "loss": 0.307,
+      "step": 17008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.567647827180561,
+      "learning_rate": 6.609319166424042e-06,
+      "loss": 0.3288,
+      "step": 17009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8093228034493969,
+      "learning_rate": 6.6089461190809565e-06,
+      "loss": 0.3023,
+      "step": 17010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9238593723451067,
+      "learning_rate": 6.608573061746754e-06,
+      "loss": 0.3322,
+      "step": 17011
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.725370366653581,
+      "learning_rate": 6.608199994423747e-06,
+      "loss": 0.361,
+      "step": 17012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3951124267838686,
+      "learning_rate": 6.607826917114255e-06,
+      "loss": 0.3655,
+      "step": 17013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0665294677183175,
+      "learning_rate": 6.607453829820594e-06,
+      "loss": 0.2692,
+      "step": 17014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.593453730365682,
+      "learning_rate": 6.607080732545082e-06,
+      "loss": 0.1901,
+      "step": 17015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.484964344081461,
+      "learning_rate": 6.606707625290032e-06,
+      "loss": 0.3731,
+      "step": 17016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.085361754865999,
+      "learning_rate": 6.606334508057766e-06,
+      "loss": 0.5365,
+      "step": 17017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6808005437677132,
+      "learning_rate": 6.6059613808505964e-06,
+      "loss": 0.2216,
+      "step": 17018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4994420886603754,
+      "learning_rate": 6.605588243670843e-06,
+      "loss": 0.3988,
+      "step": 17019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7565565772432632,
+      "learning_rate": 6.6052150965208215e-06,
+      "loss": 0.228,
+      "step": 17020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9402195471605408,
+      "learning_rate": 6.604841939402849e-06,
+      "loss": 0.443,
+      "step": 17021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.288638341362736,
+      "learning_rate": 6.604468772319246e-06,
+      "loss": 0.2243,
+      "step": 17022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6518702583879996,
+      "learning_rate": 6.604095595272324e-06,
+      "loss": 0.3833,
+      "step": 17023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1439953553089186,
+      "learning_rate": 6.6037224082644055e-06,
+      "loss": 0.3588,
+      "step": 17024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.166148536807629,
+      "learning_rate": 6.603349211297805e-06,
+      "loss": 0.2929,
+      "step": 17025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1093092731549596,
+      "learning_rate": 6.602976004374842e-06,
+      "loss": 0.4768,
+      "step": 17026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8904450937851913,
+      "learning_rate": 6.602602787497829e-06,
+      "loss": 0.3545,
+      "step": 17027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6056812312836155,
+      "learning_rate": 6.602229560669091e-06,
+      "loss": 0.317,
+      "step": 17028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7748387856627705,
+      "learning_rate": 6.601856323890942e-06,
+      "loss": 0.1867,
+      "step": 17029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.096593633839335,
+      "learning_rate": 6.601483077165697e-06,
+      "loss": 0.367,
+      "step": 17030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3625496394570742,
+      "learning_rate": 6.601109820495677e-06,
+      "loss": 0.1541,
+      "step": 17031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2353469166272522,
+      "learning_rate": 6.600736553883201e-06,
+      "loss": 0.1589,
+      "step": 17032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5789644940301353,
+      "learning_rate": 6.6003632773305834e-06,
+      "loss": 0.2695,
+      "step": 17033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6792168933840042,
+      "learning_rate": 6.599989990840145e-06,
+      "loss": 0.2058,
+      "step": 17034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5689534574012562,
+      "learning_rate": 6.599616694414203e-06,
+      "loss": 0.1917,
+      "step": 17035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0811053394757058,
+      "learning_rate": 6.599243388055073e-06,
+      "loss": 0.598,
+      "step": 17036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2181989542783858,
+      "learning_rate": 6.5988700717650774e-06,
+      "loss": 0.2493,
+      "step": 17037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.294144724767905,
+      "learning_rate": 6.598496745546532e-06,
+      "loss": 0.2435,
+      "step": 17038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.556857083149911,
+      "learning_rate": 6.5981234094017554e-06,
+      "loss": 0.6464,
+      "step": 17039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3000110058364456,
+      "learning_rate": 6.597750063333066e-06,
+      "loss": 0.1247,
+      "step": 17040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7883143693149388,
+      "learning_rate": 6.597376707342781e-06,
+      "loss": 0.2515,
+      "step": 17041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3664327227392883,
+      "learning_rate": 6.597003341433222e-06,
+      "loss": 0.285,
+      "step": 17042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6883327628083251,
+      "learning_rate": 6.5966299656067024e-06,
+      "loss": 0.3362,
+      "step": 17043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.948299338284096,
+      "learning_rate": 6.596256579865546e-06,
+      "loss": 0.3802,
+      "step": 17044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.65233330576777,
+      "learning_rate": 6.595883184212067e-06,
+      "loss": 0.1968,
+      "step": 17045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8806570465182044,
+      "learning_rate": 6.595509778648588e-06,
+      "loss": 0.5005,
+      "step": 17046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.292399370382292,
+      "learning_rate": 6.595136363177426e-06,
+      "loss": 0.2195,
+      "step": 17047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5548467075356958,
+      "learning_rate": 6.594762937800899e-06,
+      "loss": 0.3097,
+      "step": 17048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.461022166657626,
+      "learning_rate": 6.594389502521327e-06,
+      "loss": 0.2562,
+      "step": 17049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4466474244939118,
+      "learning_rate": 6.594016057341028e-06,
+      "loss": 0.3327,
+      "step": 17050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8973937874307492,
+      "learning_rate": 6.593642602262324e-06,
+      "loss": 0.3621,
+      "step": 17051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3463246414838117,
+      "learning_rate": 6.593269137287529e-06,
+      "loss": 0.3687,
+      "step": 17052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5582052143319123,
+      "learning_rate": 6.592895662418965e-06,
+      "loss": 0.2233,
+      "step": 17053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3649411766639936,
+      "learning_rate": 6.592522177658952e-06,
+      "loss": 0.1378,
+      "step": 17054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6862144341758312,
+      "learning_rate": 6.592148683009808e-06,
+      "loss": 0.4806,
+      "step": 17055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.358610911313034,
+      "learning_rate": 6.591775178473852e-06,
+      "loss": 0.1811,
+      "step": 17056
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8199205918071126,
+      "learning_rate": 6.591401664053403e-06,
+      "loss": 0.504,
+      "step": 17057
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.772707213219388,
+      "learning_rate": 6.5910281397507835e-06,
+      "loss": 0.4895,
+      "step": 17058
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6559110240615855,
+      "learning_rate": 6.590654605568308e-06,
+      "loss": 0.4871,
+      "step": 17059
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3555512921900736,
+      "learning_rate": 6.5902810615083e-06,
+      "loss": 0.1828,
+      "step": 17060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3079234684019099,
+      "learning_rate": 6.589907507573079e-06,
+      "loss": 0.208,
+      "step": 17061
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3173170023797456,
+      "learning_rate": 6.589533943764964e-06,
+      "loss": 0.3145,
+      "step": 17062
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2998067402984665,
+      "learning_rate": 6.589160370086272e-06,
+      "loss": 0.2284,
+      "step": 17063
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7780828448328427,
+      "learning_rate": 6.5887867865393275e-06,
+      "loss": 0.2916,
+      "step": 17064
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8005782766000227,
+      "learning_rate": 6.588413193126446e-06,
+      "loss": 0.267,
+      "step": 17065
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3576903044204087,
+      "learning_rate": 6.588039589849951e-06,
+      "loss": 0.2398,
+      "step": 17066
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5064314313496547,
+      "learning_rate": 6.587665976712159e-06,
+      "loss": 0.4385,
+      "step": 17067
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3367062024631775,
+      "learning_rate": 6.587292353715393e-06,
+      "loss": 0.3407,
+      "step": 17068
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7174426828978997,
+      "learning_rate": 6.586918720861971e-06,
+      "loss": 0.4518,
+      "step": 17069
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8048842516900492,
+      "learning_rate": 6.5865450781542165e-06,
+      "loss": 0.1853,
+      "step": 17070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.35385891247742,
+      "learning_rate": 6.586171425594444e-06,
+      "loss": 0.2397,
+      "step": 17071
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.520100981983634,
+      "learning_rate": 6.58579776318498e-06,
+      "loss": 0.3132,
+      "step": 17072
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2772401210311197,
+      "learning_rate": 6.585424090928141e-06,
+      "loss": 0.2332,
+      "step": 17073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0801883584347265,
+      "learning_rate": 6.585050408826248e-06,
+      "loss": 0.1768,
+      "step": 17074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6237202015448517,
+      "learning_rate": 6.584676716881622e-06,
+      "loss": 0.3521,
+      "step": 17075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7752664436580117,
+      "learning_rate": 6.584303015096582e-06,
+      "loss": 0.5792,
+      "step": 17076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8225758974273543,
+      "learning_rate": 6.583929303473453e-06,
+      "loss": 0.2805,
+      "step": 17077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6325898725969872,
+      "learning_rate": 6.58355558201455e-06,
+      "loss": 0.3501,
+      "step": 17078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.760399267197196,
+      "learning_rate": 6.583181850722198e-06,
+      "loss": 0.2731,
+      "step": 17079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5224269645379942,
+      "learning_rate": 6.5828081095987155e-06,
+      "loss": 0.5566,
+      "step": 17080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.503777181363057,
+      "learning_rate": 6.5824343586464236e-06,
+      "loss": 0.3881,
+      "step": 17081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8576007199352145,
+      "learning_rate": 6.582060597867643e-06,
+      "loss": 0.377,
+      "step": 17082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8410050479059774,
+      "learning_rate": 6.581686827264695e-06,
+      "loss": 0.6887,
+      "step": 17083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6770255042307953,
+      "learning_rate": 6.5813130468399036e-06,
+      "loss": 0.461,
+      "step": 17084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3999225932992851,
+      "learning_rate": 6.5809392565955845e-06,
+      "loss": 0.3596,
+      "step": 17085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4469363395261476,
+      "learning_rate": 6.580565456534061e-06,
+      "loss": 0.2127,
+      "step": 17086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4263744766193271,
+      "learning_rate": 6.5801916466576555e-06,
+      "loss": 0.2488,
+      "step": 17087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.440688836013567,
+      "learning_rate": 6.579817826968688e-06,
+      "loss": 0.1639,
+      "step": 17088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.134297655541739,
+      "learning_rate": 6.579443997469481e-06,
+      "loss": 0.2775,
+      "step": 17089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9351056170471388,
+      "learning_rate": 6.5790701581623565e-06,
+      "loss": 0.3361,
+      "step": 17090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4304342820969764,
+      "learning_rate": 6.578696309049632e-06,
+      "loss": 0.2581,
+      "step": 17091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8128861722951077,
+      "learning_rate": 6.578322450133633e-06,
+      "loss": 0.3095,
+      "step": 17092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.091288878004795,
+      "learning_rate": 6.5779485814166785e-06,
+      "loss": 0.604,
+      "step": 17093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6150382036969626,
+      "learning_rate": 6.5775747029010926e-06,
+      "loss": 0.2173,
+      "step": 17094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1932437258496407,
+      "learning_rate": 6.577200814589195e-06,
+      "loss": 0.1488,
+      "step": 17095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5973217574378686,
+      "learning_rate": 6.576826916483308e-06,
+      "loss": 0.1938,
+      "step": 17096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4686811689724701,
+      "learning_rate": 6.576453008585755e-06,
+      "loss": 0.3663,
+      "step": 17097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8146650808521811,
+      "learning_rate": 6.576079090898853e-06,
+      "loss": 0.3558,
+      "step": 17098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2347839770451208,
+      "learning_rate": 6.57570516342493e-06,
+      "loss": 0.1815,
+      "step": 17099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.177102285436489,
+      "learning_rate": 6.575331226166304e-06,
+      "loss": 0.342,
+      "step": 17100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0592732909444855,
+      "learning_rate": 6.5749572791253e-06,
+      "loss": 0.4296,
+      "step": 17101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.917897374598382,
+      "learning_rate": 6.574583322304236e-06,
+      "loss": 0.1994,
+      "step": 17102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.500429504216552,
+      "learning_rate": 6.5742093557054385e-06,
+      "loss": 0.3651,
+      "step": 17103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.429646058862975,
+      "learning_rate": 6.573835379331227e-06,
+      "loss": 0.3149,
+      "step": 17104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.938036370443654,
+      "learning_rate": 6.5734613931839254e-06,
+      "loss": 0.2297,
+      "step": 17105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4487012166566193,
+      "learning_rate": 6.5730873972658546e-06,
+      "loss": 0.1603,
+      "step": 17106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8759549570105414,
+      "learning_rate": 6.572713391579338e-06,
+      "loss": 0.3067,
+      "step": 17107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.382322658155977,
+      "learning_rate": 6.572339376126698e-06,
+      "loss": 0.2892,
+      "step": 17108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.815616895555802,
+      "learning_rate": 6.571965350910255e-06,
+      "loss": 0.4837,
+      "step": 17109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.024774742843352,
+      "learning_rate": 6.571591315932337e-06,
+      "loss": 0.2315,
+      "step": 17110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8915971545205972,
+      "learning_rate": 6.571217271195259e-06,
+      "loss": 0.1877,
+      "step": 17111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7509356364686721,
+      "learning_rate": 6.570843216701352e-06,
+      "loss": 0.3671,
+      "step": 17112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.043586947414248,
+      "learning_rate": 6.570469152452932e-06,
+      "loss": 0.2973,
+      "step": 17113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.673389963913834,
+      "learning_rate": 6.570095078452325e-06,
+      "loss": 0.2106,
+      "step": 17114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4156588940320687,
+      "learning_rate": 6.569720994701854e-06,
+      "loss": 0.2436,
+      "step": 17115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5413941763295287,
+      "learning_rate": 6.5693469012038416e-06,
+      "loss": 0.287,
+      "step": 17116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9543832006913724,
+      "learning_rate": 6.568972797960609e-06,
+      "loss": 0.3309,
+      "step": 17117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6587042124436069,
+      "learning_rate": 6.568598684974482e-06,
+      "loss": 0.3094,
+      "step": 17118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9815434009598794,
+      "learning_rate": 6.5682245622477826e-06,
+      "loss": 0.3093,
+      "step": 17119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6083775966570257,
+      "learning_rate": 6.567850429782834e-06,
+      "loss": 0.4543,
+      "step": 17120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.724525112629551,
+      "learning_rate": 6.5674762875819585e-06,
+      "loss": 0.4986,
+      "step": 17121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.331982575363452,
+      "learning_rate": 6.5671021356474805e-06,
+      "loss": 0.1737,
+      "step": 17122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6600446237173414,
+      "learning_rate": 6.566727973981725e-06,
+      "loss": 0.1914,
+      "step": 17123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9132362742840732,
+      "learning_rate": 6.566353802587011e-06,
+      "loss": 0.1859,
+      "step": 17124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2520733506296127,
+      "learning_rate": 6.565979621465668e-06,
+      "loss": 0.2229,
+      "step": 17125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0201630849086065,
+      "learning_rate": 6.565605430620014e-06,
+      "loss": 0.4496,
+      "step": 17126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4303472799648593,
+      "learning_rate": 6.565231230052375e-06,
+      "loss": 0.299,
+      "step": 17127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2100844203351344,
+      "learning_rate": 6.564857019765075e-06,
+      "loss": 0.2297,
+      "step": 17128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.700985058447785,
+      "learning_rate": 6.564482799760437e-06,
+      "loss": 0.2724,
+      "step": 17129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9037542538398111,
+      "learning_rate": 6.564108570040785e-06,
+      "loss": 0.5516,
+      "step": 17130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7652318314164457,
+      "learning_rate": 6.563734330608443e-06,
+      "loss": 0.2927,
+      "step": 17131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.952373714911621,
+      "learning_rate": 6.563360081465735e-06,
+      "loss": 0.4451,
+      "step": 17132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6166553819727891,
+      "learning_rate": 6.562985822614986e-06,
+      "loss": 0.3138,
+      "step": 17133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.694496331881799,
+      "learning_rate": 6.562611554058518e-06,
+      "loss": 0.3734,
+      "step": 17134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6611308345076476,
+      "learning_rate": 6.562237275798655e-06,
+      "loss": 0.35,
+      "step": 17135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.946704519654462,
+      "learning_rate": 6.5618629878377235e-06,
+      "loss": 0.5414,
+      "step": 17136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.558804892098853,
+      "learning_rate": 6.561488690178046e-06,
+      "loss": 0.2465,
+      "step": 17137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.142060261598853,
+      "learning_rate": 6.5611143828219485e-06,
+      "loss": 0.5018,
+      "step": 17138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.174450752044272,
+      "learning_rate": 6.5607400657717526e-06,
+      "loss": 0.206,
+      "step": 17139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4489579390199163,
+      "learning_rate": 6.560365739029784e-06,
+      "loss": 0.3077,
+      "step": 17140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9698351677385155,
+      "learning_rate": 6.559991402598369e-06,
+      "loss": 0.494,
+      "step": 17141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9129882908669913,
+      "learning_rate": 6.559617056479828e-06,
+      "loss": 0.2125,
+      "step": 17142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.682276921064947,
+      "learning_rate": 6.559242700676491e-06,
+      "loss": 0.3768,
+      "step": 17143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.688693573719878,
+      "learning_rate": 6.5588683351906764e-06,
+      "loss": 0.4647,
+      "step": 17144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6912075447993902,
+      "learning_rate": 6.558493960024716e-06,
+      "loss": 0.1922,
+      "step": 17145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8446615034982856,
+      "learning_rate": 6.558119575180928e-06,
+      "loss": 0.5429,
+      "step": 17146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7520827157223218,
+      "learning_rate": 6.557745180661641e-06,
+      "loss": 0.2603,
+      "step": 17147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5954235583737912,
+      "learning_rate": 6.5573707764691774e-06,
+      "loss": 0.3066,
+      "step": 17148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0550895703650873,
+      "learning_rate": 6.556996362605865e-06,
+      "loss": 0.2612,
+      "step": 17149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2735565062655776,
+      "learning_rate": 6.556621939074028e-06,
+      "loss": 0.2225,
+      "step": 17150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9098051524699606,
+      "learning_rate": 6.55624750587599e-06,
+      "loss": 0.2216,
+      "step": 17151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6575097025936831,
+      "learning_rate": 6.555873063014079e-06,
+      "loss": 0.2289,
+      "step": 17152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4311558942383085,
+      "learning_rate": 6.555498610490616e-06,
+      "loss": 0.2761,
+      "step": 17153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0006861040713417,
+      "learning_rate": 6.555124148307929e-06,
+      "loss": 0.2212,
+      "step": 17154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.325875649494218,
+      "learning_rate": 6.554749676468343e-06,
+      "loss": 0.2386,
+      "step": 17155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.334453960892122,
+      "learning_rate": 6.554375194974183e-06,
+      "loss": 0.2025,
+      "step": 17156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7415684896120043,
+      "learning_rate": 6.554000703827774e-06,
+      "loss": 0.3058,
+      "step": 17157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3323618682706653,
+      "learning_rate": 6.553626203031442e-06,
+      "loss": 0.3688,
+      "step": 17158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.649676422339199,
+      "learning_rate": 6.553251692587515e-06,
+      "loss": 0.3363,
+      "step": 17159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9233907095813663,
+      "learning_rate": 6.552877172498313e-06,
+      "loss": 0.2759,
+      "step": 17160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0369944059724916,
+      "learning_rate": 6.552502642766167e-06,
+      "loss": 0.1797,
+      "step": 17161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6997780153552584,
+      "learning_rate": 6.5521281033934005e-06,
+      "loss": 0.2077,
+      "step": 17162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6536446043195188,
+      "learning_rate": 6.551753554382339e-06,
+      "loss": 0.3111,
+      "step": 17163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7023128948767934,
+      "learning_rate": 6.551378995735309e-06,
+      "loss": 0.3865,
+      "step": 17164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7510186662095668,
+      "learning_rate": 6.551004427454636e-06,
+      "loss": 0.4536,
+      "step": 17165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8409914923692268,
+      "learning_rate": 6.5506298495426455e-06,
+      "loss": 0.2628,
+      "step": 17166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1959043452509297,
+      "learning_rate": 6.550255262001666e-06,
+      "loss": 0.1864,
+      "step": 17167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.232833458053888,
+      "learning_rate": 6.54988066483402e-06,
+      "loss": 0.264,
+      "step": 17168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4757899427990937,
+      "learning_rate": 6.549506058042037e-06,
+      "loss": 0.2245,
+      "step": 17169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6384335914112642,
+      "learning_rate": 6.549131441628041e-06,
+      "loss": 0.4422,
+      "step": 17170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8098729596419274,
+      "learning_rate": 6.548756815594358e-06,
+      "loss": 0.273,
+      "step": 17171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.682918430844382,
+      "learning_rate": 6.548382179943316e-06,
+      "loss": 0.2305,
+      "step": 17172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.131555728434617,
+      "learning_rate": 6.5480075346772395e-06,
+      "loss": 0.5542,
+      "step": 17173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1425345354679703,
+      "learning_rate": 6.5476328797984575e-06,
+      "loss": 0.5848,
+      "step": 17174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7262326093677505,
+      "learning_rate": 6.547258215309295e-06,
+      "loss": 0.5123,
+      "step": 17175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3643972889429077,
+      "learning_rate": 6.546883541212078e-06,
+      "loss": 0.1822,
+      "step": 17176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0155154834241067,
+      "learning_rate": 6.546508857509133e-06,
+      "loss": 0.246,
+      "step": 17177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9876141731289279,
+      "learning_rate": 6.546134164202789e-06,
+      "loss": 0.7384,
+      "step": 17178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4710413656566237,
+      "learning_rate": 6.54575946129537e-06,
+      "loss": 0.2708,
+      "step": 17179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3674069378362887,
+      "learning_rate": 6.545384748789205e-06,
+      "loss": 0.169,
+      "step": 17180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9210037978206737,
+      "learning_rate": 6.545010026686619e-06,
+      "loss": 0.0837,
+      "step": 17181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.631543005421247,
+      "learning_rate": 6.5446352949899404e-06,
+      "loss": 0.1771,
+      "step": 17182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.133856985726394,
+      "learning_rate": 6.544260553701495e-06,
+      "loss": 0.6237,
+      "step": 17183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8835788844885448,
+      "learning_rate": 6.543885802823611e-06,
+      "loss": 0.2974,
+      "step": 17184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4167602843551994,
+      "learning_rate": 6.5435110423586136e-06,
+      "loss": 0.1732,
+      "step": 17185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9407608410074129,
+      "learning_rate": 6.543136272308831e-06,
+      "loss": 0.3066,
+      "step": 17186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.602097512644593,
+      "learning_rate": 6.542761492676593e-06,
+      "loss": 0.1914,
+      "step": 17187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8214740306748951,
+      "learning_rate": 6.542386703464222e-06,
+      "loss": 0.1306,
+      "step": 17188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9219987857326517,
+      "learning_rate": 6.5420119046740495e-06,
+      "loss": 0.3419,
+      "step": 17189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6867553648656277,
+      "learning_rate": 6.5416370963084005e-06,
+      "loss": 0.3028,
+      "step": 17190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8605422034916765,
+      "learning_rate": 6.541262278369605e-06,
+      "loss": 0.4486,
+      "step": 17191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4426161590859246,
+      "learning_rate": 6.540887450859987e-06,
+      "loss": 0.2377,
+      "step": 17192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8969848671695648,
+      "learning_rate": 6.540512613781875e-06,
+      "loss": 0.2085,
+      "step": 17193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6956989181947932,
+      "learning_rate": 6.5401377671376e-06,
+      "loss": 0.543,
+      "step": 17194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5475945388683787,
+      "learning_rate": 6.539762910929486e-06,
+      "loss": 0.4146,
+      "step": 17195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9113540326453362,
+      "learning_rate": 6.5393880451598615e-06,
+      "loss": 0.4651,
+      "step": 17196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.298090594850038,
+      "learning_rate": 6.539013169831055e-06,
+      "loss": 0.257,
+      "step": 17197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0532610949752184,
+      "learning_rate": 6.538638284945394e-06,
+      "loss": 0.361,
+      "step": 17198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2076966047509878,
+      "learning_rate": 6.538263390505206e-06,
+      "loss": 0.192,
+      "step": 17199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2552330389269166,
+      "learning_rate": 6.53788848651282e-06,
+      "loss": 0.308,
+      "step": 17200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8945184176295535,
+      "learning_rate": 6.5375135729705644e-06,
+      "loss": 0.3048,
+      "step": 17201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6043420351042645,
+      "learning_rate": 6.537138649880766e-06,
+      "loss": 0.2555,
+      "step": 17202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.963421101793841,
+      "learning_rate": 6.536763717245753e-06,
+      "loss": 0.2788,
+      "step": 17203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.864498688376649,
+      "learning_rate": 6.536388775067854e-06,
+      "loss": 0.3437,
+      "step": 17204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4761243099353492,
+      "learning_rate": 6.536013823349399e-06,
+      "loss": 0.221,
+      "step": 17205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7840905169841093,
+      "learning_rate": 6.535638862092712e-06,
+      "loss": 0.1813,
+      "step": 17206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8642328939076367,
+      "learning_rate": 6.535263891300126e-06,
+      "loss": 0.3147,
+      "step": 17207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2889357706833475,
+      "learning_rate": 6.534888910973967e-06,
+      "loss": 0.2299,
+      "step": 17208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.595322812259085,
+      "learning_rate": 6.534513921116565e-06,
+      "loss": 0.1799,
+      "step": 17209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2781824572363296,
+      "learning_rate": 6.534138921730247e-06,
+      "loss": 0.1576,
+      "step": 17210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.325835447940103,
+      "learning_rate": 6.533763912817342e-06,
+      "loss": 0.2208,
+      "step": 17211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2386679677261037,
+      "learning_rate": 6.5333888943801795e-06,
+      "loss": 0.1237,
+      "step": 17212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5789190126683743,
+      "learning_rate": 6.533013866421088e-06,
+      "loss": 0.3427,
+      "step": 17213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6530858418310421,
+      "learning_rate": 6.532638828942397e-06,
+      "loss": 0.3545,
+      "step": 17214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5973988957640557,
+      "learning_rate": 6.532263781946433e-06,
+      "loss": 0.3357,
+      "step": 17215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0367420776108225,
+      "learning_rate": 6.531888725435528e-06,
+      "loss": 0.5999,
+      "step": 17216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3218487092147426,
+      "learning_rate": 6.531513659412008e-06,
+      "loss": 0.292,
+      "step": 17217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2006776863067619,
+      "learning_rate": 6.5311385838782045e-06,
+      "loss": 0.18,
+      "step": 17218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0971105634393186,
+      "learning_rate": 6.530763498836446e-06,
+      "loss": 0.1273,
+      "step": 17219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.55044451115568,
+      "learning_rate": 6.530388404289059e-06,
+      "loss": 0.1785,
+      "step": 17220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3378322406002146,
+      "learning_rate": 6.530013300238377e-06,
+      "loss": 0.4091,
+      "step": 17221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8580407235863725,
+      "learning_rate": 6.529638186686727e-06,
+      "loss": 0.2677,
+      "step": 17222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8222222992549646,
+      "learning_rate": 6.529263063636438e-06,
+      "loss": 0.3283,
+      "step": 17223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6829374808864215,
+      "learning_rate": 6.528887931089842e-06,
+      "loss": 0.2268,
+      "step": 17224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.963105418661105,
+      "learning_rate": 6.528512789049266e-06,
+      "loss": 0.5728,
+      "step": 17225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4278796757395504,
+      "learning_rate": 6.528137637517039e-06,
+      "loss": 0.3167,
+      "step": 17226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4511348185679023,
+      "learning_rate": 6.5277624764954925e-06,
+      "loss": 0.3231,
+      "step": 17227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7407351243266775,
+      "learning_rate": 6.5273873059869556e-06,
+      "loss": 0.2571,
+      "step": 17228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5980198453427648,
+      "learning_rate": 6.527012125993757e-06,
+      "loss": 0.2717,
+      "step": 17229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3872558641549522,
+      "learning_rate": 6.526636936518227e-06,
+      "loss": 0.2917,
+      "step": 17230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3586123110954897,
+      "learning_rate": 6.526261737562697e-06,
+      "loss": 0.1891,
+      "step": 17231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3453681763832983,
+      "learning_rate": 6.525886529129495e-06,
+      "loss": 0.2703,
+      "step": 17232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4259786142633084,
+      "learning_rate": 6.525511311220951e-06,
+      "loss": 0.2054,
+      "step": 17233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3004952107636112,
+      "learning_rate": 6.525136083839395e-06,
+      "loss": 0.2562,
+      "step": 17234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3273212300725847,
+      "learning_rate": 6.524760846987158e-06,
+      "loss": 0.3359,
+      "step": 17235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8965497698052571,
+      "learning_rate": 6.524385600666571e-06,
+      "loss": 0.4334,
+      "step": 17236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.147186278890959,
+      "learning_rate": 6.5240103448799605e-06,
+      "loss": 0.2407,
+      "step": 17237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.433572568918859,
+      "learning_rate": 6.523635079629661e-06,
+      "loss": 0.2737,
+      "step": 17238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5425417925027953,
+      "learning_rate": 6.523259804918001e-06,
+      "loss": 0.2024,
+      "step": 17239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9114720312003648,
+      "learning_rate": 6.5228845207473105e-06,
+      "loss": 0.3996,
+      "step": 17240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.482285977091518,
+      "learning_rate": 6.522509227119919e-06,
+      "loss": 0.1339,
+      "step": 17241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.014926242212036,
+      "learning_rate": 6.5221339240381584e-06,
+      "loss": 0.3817,
+      "step": 17242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3496267452861133,
+      "learning_rate": 6.521758611504359e-06,
+      "loss": 0.2755,
+      "step": 17243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.035186554102875,
+      "learning_rate": 6.521383289520851e-06,
+      "loss": 0.2421,
+      "step": 17244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2672533874221863,
+      "learning_rate": 6.521007958089967e-06,
+      "loss": 0.2347,
+      "step": 17245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5169032350119895,
+      "learning_rate": 6.520632617214036e-06,
+      "loss": 0.461,
+      "step": 17246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5647341776979011,
+      "learning_rate": 6.520257266895388e-06,
+      "loss": 0.2433,
+      "step": 17247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9204435949375027,
+      "learning_rate": 6.519881907136354e-06,
+      "loss": 0.6809,
+      "step": 17248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0776665307066629,
+      "learning_rate": 6.519506537939267e-06,
+      "loss": 0.0997,
+      "step": 17249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.458707342694768,
+      "learning_rate": 6.519131159306454e-06,
+      "loss": 0.264,
+      "step": 17250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0128428568429024,
+      "learning_rate": 6.5187557712402515e-06,
+      "loss": 0.5768,
+      "step": 17251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5601735581616063,
+      "learning_rate": 6.518380373742985e-06,
+      "loss": 0.2272,
+      "step": 17252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5537974311351408,
+      "learning_rate": 6.518004966816991e-06,
+      "loss": 0.2506,
+      "step": 17253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1189852634994495,
+      "learning_rate": 6.517629550464596e-06,
+      "loss": 0.2017,
+      "step": 17254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3839516796542821,
+      "learning_rate": 6.517254124688134e-06,
+      "loss": 0.1855,
+      "step": 17255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9379586130775333,
+      "learning_rate": 6.5168786894899335e-06,
+      "loss": 0.227,
+      "step": 17256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.931503279444216,
+      "learning_rate": 6.51650324487233e-06,
+      "loss": 0.4878,
+      "step": 17257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.525965197755334,
+      "learning_rate": 6.516127790837651e-06,
+      "loss": 0.1859,
+      "step": 17258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4859025574034372,
+      "learning_rate": 6.515752327388231e-06,
+      "loss": 0.2239,
+      "step": 17259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.553730099582054,
+      "learning_rate": 6.515376854526398e-06,
+      "loss": 0.2302,
+      "step": 17260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2214228774089793,
+      "learning_rate": 6.515001372254487e-06,
+      "loss": 0.129,
+      "step": 17261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9242904107661762,
+      "learning_rate": 6.514625880574829e-06,
+      "loss": 0.1969,
+      "step": 17262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6158683114039172,
+      "learning_rate": 6.514250379489754e-06,
+      "loss": 0.3591,
+      "step": 17263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.943774672563239,
+      "learning_rate": 6.513874869001595e-06,
+      "loss": 0.3385,
+      "step": 17264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7734842840401224,
+      "learning_rate": 6.513499349112684e-06,
+      "loss": 0.3434,
+      "step": 17265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5397796775265273,
+      "learning_rate": 6.513123819825354e-06,
+      "loss": 0.2078,
+      "step": 17266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7849447638449578,
+      "learning_rate": 6.512748281141933e-06,
+      "loss": 0.2606,
+      "step": 17267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5082645825617373,
+      "learning_rate": 6.512372733064757e-06,
+      "loss": 0.3277,
+      "step": 17268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7381833766858037,
+      "learning_rate": 6.511997175596156e-06,
+      "loss": 0.2206,
+      "step": 17269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.704341648535775,
+      "learning_rate": 6.511621608738462e-06,
+      "loss": 0.3149,
+      "step": 17270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1889930891758493,
+      "learning_rate": 6.511246032494009e-06,
+      "loss": 0.1951,
+      "step": 17271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3072049243868247,
+      "learning_rate": 6.510870446865126e-06,
+      "loss": 0.2459,
+      "step": 17272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.898532904280269,
+      "learning_rate": 6.51049485185415e-06,
+      "loss": 0.4912,
+      "step": 17273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8026495941494207,
+      "learning_rate": 6.510119247463408e-06,
+      "loss": 0.481,
+      "step": 17274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.373813290444059,
+      "learning_rate": 6.509743633695237e-06,
+      "loss": 0.1945,
+      "step": 17275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9796956902916383,
+      "learning_rate": 6.509368010551966e-06,
+      "loss": 0.4815,
+      "step": 17276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7584128308366527,
+      "learning_rate": 6.50899237803593e-06,
+      "loss": 0.5171,
+      "step": 17277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7649828937453957,
+      "learning_rate": 6.50861673614946e-06,
+      "loss": 0.3699,
+      "step": 17278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7473747367741348,
+      "learning_rate": 6.5082410848948905e-06,
+      "loss": 0.3762,
+      "step": 17279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5640059235579087,
+      "learning_rate": 6.507865424274552e-06,
+      "loss": 0.2521,
+      "step": 17280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8300462809573828,
+      "learning_rate": 6.507489754290778e-06,
+      "loss": 0.2199,
+      "step": 17281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.643398160728371,
+      "learning_rate": 6.507114074945902e-06,
+      "loss": 0.2122,
+      "step": 17282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7380029176911422,
+      "learning_rate": 6.506738386242255e-06,
+      "loss": 0.5561,
+      "step": 17283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.676888022590826,
+      "learning_rate": 6.506362688182172e-06,
+      "loss": 0.3061,
+      "step": 17284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.456185079834293,
+      "learning_rate": 6.505986980767985e-06,
+      "loss": 0.1685,
+      "step": 17285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4113258045049701,
+      "learning_rate": 6.505611264002027e-06,
+      "loss": 0.324,
+      "step": 17286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8008925058085024,
+      "learning_rate": 6.505235537886631e-06,
+      "loss": 0.1736,
+      "step": 17287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9701737126591432,
+      "learning_rate": 6.504859802424133e-06,
+      "loss": 0.3902,
+      "step": 17288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7075628155705613,
+      "learning_rate": 6.504484057616862e-06,
+      "loss": 0.6024,
+      "step": 17289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.448836060145111,
+      "learning_rate": 6.504108303467153e-06,
+      "loss": 0.2775,
+      "step": 17290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7058958559391677,
+      "learning_rate": 6.503732539977341e-06,
+      "loss": 0.2887,
+      "step": 17291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9619956903864788,
+      "learning_rate": 6.503356767149757e-06,
+      "loss": 0.5018,
+      "step": 17292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3521964234793147,
+      "learning_rate": 6.502980984986734e-06,
+      "loss": 0.2405,
+      "step": 17293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7603921230192807,
+      "learning_rate": 6.502605193490607e-06,
+      "loss": 0.4527,
+      "step": 17294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9936979475499053,
+      "learning_rate": 6.5022293926637105e-06,
+      "loss": 0.1603,
+      "step": 17295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.078265271436221,
+      "learning_rate": 6.501853582508375e-06,
+      "loss": 0.3579,
+      "step": 17296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7568821627953641,
+      "learning_rate": 6.501477763026938e-06,
+      "loss": 0.5677,
+      "step": 17297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8459951906374052,
+      "learning_rate": 6.5011019342217294e-06,
+      "loss": 0.17,
+      "step": 17298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1074559330066376,
+      "learning_rate": 6.500726096095087e-06,
+      "loss": 0.2116,
+      "step": 17299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5067081632106711,
+      "learning_rate": 6.500350248649341e-06,
+      "loss": 0.2206,
+      "step": 17300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.269495896476058,
+      "learning_rate": 6.499974391886829e-06,
+      "loss": 0.5343,
+      "step": 17301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.297207921739731,
+      "learning_rate": 6.4995985258098804e-06,
+      "loss": 0.2375,
+      "step": 17302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5638145448398344,
+      "learning_rate": 6.499222650420832e-06,
+      "loss": 0.3004,
+      "step": 17303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.133193866069935,
+      "learning_rate": 6.498846765722019e-06,
+      "loss": 0.1656,
+      "step": 17304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9036717176090943,
+      "learning_rate": 6.498470871715772e-06,
+      "loss": 0.3946,
+      "step": 17305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5961679881079696,
+      "learning_rate": 6.4980949684044295e-06,
+      "loss": 0.3447,
+      "step": 17306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6258444884217518,
+      "learning_rate": 6.497719055790321e-06,
+      "loss": 0.3786,
+      "step": 17307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7494872163615973,
+      "learning_rate": 6.4973431338757864e-06,
+      "loss": 0.1974,
+      "step": 17308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.947541419177641,
+      "learning_rate": 6.496967202663155e-06,
+      "loss": 0.4707,
+      "step": 17309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4002724626974254,
+      "learning_rate": 6.496591262154762e-06,
+      "loss": 0.2951,
+      "step": 17310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3180933945399689,
+      "learning_rate": 6.496215312352946e-06,
+      "loss": 0.2295,
+      "step": 17311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5510375678258281,
+      "learning_rate": 6.495839353260037e-06,
+      "loss": 0.1788,
+      "step": 17312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5593059318092537,
+      "learning_rate": 6.495463384878372e-06,
+      "loss": 0.2663,
+      "step": 17313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4436207148848128,
+      "learning_rate": 6.495087407210284e-06,
+      "loss": 0.3115,
+      "step": 17314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4838796931129024,
+      "learning_rate": 6.494711420258111e-06,
+      "loss": 0.3426,
+      "step": 17315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7428224779219263,
+      "learning_rate": 6.494335424024183e-06,
+      "loss": 0.2881,
+      "step": 17316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0722567457168455,
+      "learning_rate": 6.493959418510838e-06,
+      "loss": 0.2058,
+      "step": 17317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7108106348390377,
+      "learning_rate": 6.49358340372041e-06,
+      "loss": 0.4086,
+      "step": 17318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9669758621375644,
+      "learning_rate": 6.493207379655236e-06,
+      "loss": 0.1627,
+      "step": 17319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6187618074759733,
+      "learning_rate": 6.4928313463176466e-06,
+      "loss": 0.3248,
+      "step": 17320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8090930850866327,
+      "learning_rate": 6.492455303709982e-06,
+      "loss": 0.3682,
+      "step": 17321
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9387473121491623,
+      "learning_rate": 6.492079251834571e-06,
+      "loss": 0.6456,
+      "step": 17322
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.841648370898609,
+      "learning_rate": 6.4917031906937555e-06,
+      "loss": 0.3958,
+      "step": 17323
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0004522587123925,
+      "learning_rate": 6.4913271202898675e-06,
+      "loss": 0.1607,
+      "step": 17324
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.584135483426154,
+      "learning_rate": 6.490951040625242e-06,
+      "loss": 0.3472,
+      "step": 17325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.219159525028827,
+      "learning_rate": 6.490574951702215e-06,
+      "loss": 0.2386,
+      "step": 17326
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5513930471186785,
+      "learning_rate": 6.490198853523122e-06,
+      "loss": 0.3411,
+      "step": 17327
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4992178203784632,
+      "learning_rate": 6.489822746090297e-06,
+      "loss": 0.2288,
+      "step": 17328
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.924885446682249,
+      "learning_rate": 6.4894466294060775e-06,
+      "loss": 0.4109,
+      "step": 17329
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.522019514990829,
+      "learning_rate": 6.4890705034728e-06,
+      "loss": 0.2121,
+      "step": 17330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2087753805476258,
+      "learning_rate": 6.488694368292796e-06,
+      "loss": 0.2168,
+      "step": 17331
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6667564149375222,
+      "learning_rate": 6.488318223868406e-06,
+      "loss": 0.4493,
+      "step": 17332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5448182284651806,
+      "learning_rate": 6.487942070201961e-06,
+      "loss": 0.4554,
+      "step": 17333
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6912050942495025,
+      "learning_rate": 6.487565907295802e-06,
+      "loss": 0.2111,
+      "step": 17334
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1122759379086777,
+      "learning_rate": 6.487189735152259e-06,
+      "loss": 0.1897,
+      "step": 17335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5119826212217276,
+      "learning_rate": 6.486813553773673e-06,
+      "loss": 0.3832,
+      "step": 17336
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0039072110954463,
+      "learning_rate": 6.486437363162379e-06,
+      "loss": 0.2114,
+      "step": 17337
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7993645753455811,
+      "learning_rate": 6.486061163320711e-06,
+      "loss": 0.2515,
+      "step": 17338
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5815380740583147,
+      "learning_rate": 6.485684954251006e-06,
+      "loss": 0.25,
+      "step": 17339
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8128871504247501,
+      "learning_rate": 6.4853087359556e-06,
+      "loss": 0.3602,
+      "step": 17340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7174382752983475,
+      "learning_rate": 6.484932508436831e-06,
+      "loss": 0.3331,
+      "step": 17341
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6454353203355705,
+      "learning_rate": 6.484556271697034e-06,
+      "loss": 0.3536,
+      "step": 17342
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4930853788652605,
+      "learning_rate": 6.484180025738544e-06,
+      "loss": 0.3108,
+      "step": 17343
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.701318294386212,
+      "learning_rate": 6.483803770563698e-06,
+      "loss": 0.4126,
+      "step": 17344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.317293246075624,
+      "learning_rate": 6.4834275061748345e-06,
+      "loss": 0.2317,
+      "step": 17345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1819056284701552,
+      "learning_rate": 6.4830512325742874e-06,
+      "loss": 0.2497,
+      "step": 17346
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6747004262429246,
+      "learning_rate": 6.482674949764395e-06,
+      "loss": 0.5145,
+      "step": 17347
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1620091257398983,
+      "learning_rate": 6.482298657747493e-06,
+      "loss": 0.3731,
+      "step": 17348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6202961165833123,
+      "learning_rate": 6.4819223565259184e-06,
+      "loss": 0.3515,
+      "step": 17349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6773989412051964,
+      "learning_rate": 6.481546046102008e-06,
+      "loss": 0.2326,
+      "step": 17350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3178859397501426,
+      "learning_rate": 6.481169726478098e-06,
+      "loss": 0.2653,
+      "step": 17351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6093016024398386,
+      "learning_rate": 6.480793397656526e-06,
+      "loss": 0.5308,
+      "step": 17352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2388400913663034,
+      "learning_rate": 6.4804170596396284e-06,
+      "loss": 0.1929,
+      "step": 17353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2962906509120695,
+      "learning_rate": 6.480040712429744e-06,
+      "loss": 0.3947,
+      "step": 17354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8979471067785354,
+      "learning_rate": 6.479664356029206e-06,
+      "loss": 0.5012,
+      "step": 17355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6176491622047342,
+      "learning_rate": 6.479287990440355e-06,
+      "loss": 0.4596,
+      "step": 17356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.157758665276494,
+      "learning_rate": 6.478911615665526e-06,
+      "loss": 0.5642,
+      "step": 17357
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2990895302235925,
+      "learning_rate": 6.478535231707057e-06,
+      "loss": 0.2366,
+      "step": 17358
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0577491460851003,
+      "learning_rate": 6.478158838567286e-06,
+      "loss": 0.5309,
+      "step": 17359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7414924492429984,
+      "learning_rate": 6.4777824362485486e-06,
+      "loss": 0.3167,
+      "step": 17360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.578852783120807,
+      "learning_rate": 6.477406024753183e-06,
+      "loss": 0.1803,
+      "step": 17361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.084724759521002,
+      "learning_rate": 6.477029604083527e-06,
+      "loss": 0.6644,
+      "step": 17362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0338756354188376,
+      "learning_rate": 6.4766531742419185e-06,
+      "loss": 0.401,
+      "step": 17363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.121295785853763,
+      "learning_rate": 6.476276735230692e-06,
+      "loss": 0.2257,
+      "step": 17364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6706787357329485,
+      "learning_rate": 6.475900287052189e-06,
+      "loss": 0.2997,
+      "step": 17365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.767587528941393,
+      "learning_rate": 6.4755238297087455e-06,
+      "loss": 0.2759,
+      "step": 17366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6528673762679575,
+      "learning_rate": 6.475147363202699e-06,
+      "loss": 0.314,
+      "step": 17367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.148676008566627,
+      "learning_rate": 6.474770887536388e-06,
+      "loss": 0.5328,
+      "step": 17368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.943604981894216,
+      "learning_rate": 6.47439440271215e-06,
+      "loss": 0.1717,
+      "step": 17369
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.560119448625566,
+      "learning_rate": 6.474017908732322e-06,
+      "loss": 0.4887,
+      "step": 17370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0086105148642535,
+      "learning_rate": 6.473641405599241e-06,
+      "loss": 0.2894,
+      "step": 17371
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7032863645491674,
+      "learning_rate": 6.473264893315249e-06,
+      "loss": 0.4459,
+      "step": 17372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9282303687043074,
+      "learning_rate": 6.472888371882679e-06,
+      "loss": 0.3546,
+      "step": 17373
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2754977942125785,
+      "learning_rate": 6.472511841303874e-06,
+      "loss": 0.2404,
+      "step": 17374
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3575674514100657,
+      "learning_rate": 6.472135301581169e-06,
+      "loss": 0.1902,
+      "step": 17375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.265840257087147,
+      "learning_rate": 6.471758752716903e-06,
+      "loss": 0.3097,
+      "step": 17376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5893992053223984,
+      "learning_rate": 6.471382194713415e-06,
+      "loss": 0.5391,
+      "step": 17377
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5426616411357355,
+      "learning_rate": 6.471005627573041e-06,
+      "loss": 0.4035,
+      "step": 17378
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.26718558463207,
+      "learning_rate": 6.470629051298124e-06,
+      "loss": 0.5436,
+      "step": 17379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.384784472381642,
+      "learning_rate": 6.470252465890996e-06,
+      "loss": 0.2144,
+      "step": 17380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0208299614898455,
+      "learning_rate": 6.469875871354001e-06,
+      "loss": 0.4374,
+      "step": 17381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4623675563540315,
+      "learning_rate": 6.469499267689474e-06,
+      "loss": 0.3155,
+      "step": 17382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7531368887595087,
+      "learning_rate": 6.469122654899758e-06,
+      "loss": 0.3185,
+      "step": 17383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.573751536437505,
+      "learning_rate": 6.468746032987185e-06,
+      "loss": 0.3234,
+      "step": 17384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.224289122143053,
+      "learning_rate": 6.468369401954099e-06,
+      "loss": 0.1937,
+      "step": 17385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.943728027648868,
+      "learning_rate": 6.467992761802838e-06,
+      "loss": 0.6929,
+      "step": 17386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6973532356679113,
+      "learning_rate": 6.46761611253574e-06,
+      "loss": 0.3268,
+      "step": 17387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6269095358899746,
+      "learning_rate": 6.4672394541551416e-06,
+      "loss": 0.2156,
+      "step": 17388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9529097705507967,
+      "learning_rate": 6.466862786663387e-06,
+      "loss": 0.3014,
+      "step": 17389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.01032417995476,
+      "learning_rate": 6.466486110062812e-06,
+      "loss": 0.3315,
+      "step": 17390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6106085124076734,
+      "learning_rate": 6.466109424355754e-06,
+      "loss": 0.3634,
+      "step": 17391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4381952366759456,
+      "learning_rate": 6.465732729544556e-06,
+      "loss": 0.3103,
+      "step": 17392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4470421403595153,
+      "learning_rate": 6.465356025631555e-06,
+      "loss": 0.1863,
+      "step": 17393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2591624198263294,
+      "learning_rate": 6.464979312619091e-06,
+      "loss": 0.2947,
+      "step": 17394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3038775862576804,
+      "learning_rate": 6.4646025905095015e-06,
+      "loss": 0.2189,
+      "step": 17395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3232231171423776,
+      "learning_rate": 6.4642258593051264e-06,
+      "loss": 0.2116,
+      "step": 17396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.452945152043499,
+      "learning_rate": 6.4638491190083075e-06,
+      "loss": 0.381,
+      "step": 17397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5115411910431467,
+      "learning_rate": 6.463472369621382e-06,
+      "loss": 0.1948,
+      "step": 17398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.677847177889955,
+      "learning_rate": 6.46309561114669e-06,
+      "loss": 0.4424,
+      "step": 17399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7639433777478768,
+      "learning_rate": 6.462718843586572e-06,
+      "loss": 0.3658,
+      "step": 17400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2627366214848568,
+      "learning_rate": 6.462342066943366e-06,
+      "loss": 0.2766,
+      "step": 17401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5765596784825404,
+      "learning_rate": 6.461965281219411e-06,
+      "loss": 0.3939,
+      "step": 17402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7440164229200483,
+      "learning_rate": 6.4615884864170494e-06,
+      "loss": 0.3047,
+      "step": 17403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0221149806267085,
+      "learning_rate": 6.46121168253862e-06,
+      "loss": 0.504,
+      "step": 17404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4774835138663842,
+      "learning_rate": 6.460834869586462e-06,
+      "loss": 0.2291,
+      "step": 17405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.795125300156963,
+      "learning_rate": 6.460458047562915e-06,
+      "loss": 0.3283,
+      "step": 17406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6136281515683617,
+      "learning_rate": 6.460081216470321e-06,
+      "loss": 0.5352,
+      "step": 17407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7080238531285314,
+      "learning_rate": 6.459704376311017e-06,
+      "loss": 0.3265,
+      "step": 17408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2822145961194378,
+      "learning_rate": 6.459327527087347e-06,
+      "loss": 0.2119,
+      "step": 17409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.669246882942795,
+      "learning_rate": 6.458950668801646e-06,
+      "loss": 0.3276,
+      "step": 17410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.858671143288617,
+      "learning_rate": 6.458573801456258e-06,
+      "loss": 0.3166,
+      "step": 17411
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6462463238493326,
+      "learning_rate": 6.458196925053522e-06,
+      "loss": 0.3771,
+      "step": 17412
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.541273301003353,
+      "learning_rate": 6.457820039595779e-06,
+      "loss": 0.4069,
+      "step": 17413
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.715208992263122,
+      "learning_rate": 6.45744314508537e-06,
+      "loss": 0.5176,
+      "step": 17414
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5494552824172416,
+      "learning_rate": 6.457066241524633e-06,
+      "loss": 0.2815,
+      "step": 17415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9038415139407785,
+      "learning_rate": 6.456689328915911e-06,
+      "loss": 0.3312,
+      "step": 17416
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2806970868689405,
+      "learning_rate": 6.456312407261543e-06,
+      "loss": 0.2731,
+      "step": 17417
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7538465041538673,
+      "learning_rate": 6.45593547656387e-06,
+      "loss": 0.2774,
+      "step": 17418
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1853601506604552,
+      "learning_rate": 6.45555853682523e-06,
+      "loss": 0.2554,
+      "step": 17419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.682081550186232,
+      "learning_rate": 6.455181588047969e-06,
+      "loss": 0.4244,
+      "step": 17420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4552211614365107,
+      "learning_rate": 6.454804630234426e-06,
+      "loss": 0.315,
+      "step": 17421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.728510769336152,
+      "learning_rate": 6.454427663386937e-06,
+      "loss": 0.4452,
+      "step": 17422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7589323016207357,
+      "learning_rate": 6.45405068750785e-06,
+      "loss": 0.2669,
+      "step": 17423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8631070444565246,
+      "learning_rate": 6.453673702599501e-06,
+      "loss": 0.1592,
+      "step": 17424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.583386667089497,
+      "learning_rate": 6.453296708664234e-06,
+      "loss": 0.2329,
+      "step": 17425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5730561371941012,
+      "learning_rate": 6.452919705704387e-06,
+      "loss": 0.134,
+      "step": 17426
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1849909005873278,
+      "learning_rate": 6.452542693722303e-06,
+      "loss": 0.144,
+      "step": 17427
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8498558355555763,
+      "learning_rate": 6.452165672720323e-06,
+      "loss": 0.6061,
+      "step": 17428
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5637648255364296,
+      "learning_rate": 6.451788642700788e-06,
+      "loss": 0.2824,
+      "step": 17429
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0547172542189676,
+      "learning_rate": 6.451411603666039e-06,
+      "loss": 0.1025,
+      "step": 17430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7533694169990464,
+      "learning_rate": 6.451034555618419e-06,
+      "loss": 0.2663,
+      "step": 17431
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7861626290428803,
+      "learning_rate": 6.450657498560266e-06,
+      "loss": 0.2543,
+      "step": 17432
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9120689832027387,
+      "learning_rate": 6.450280432493925e-06,
+      "loss": 0.2233,
+      "step": 17433
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.591137008569681,
+      "learning_rate": 6.449903357421734e-06,
+      "loss": 0.3194,
+      "step": 17434
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6680071600750606,
+      "learning_rate": 6.449526273346038e-06,
+      "loss": 0.3829,
+      "step": 17435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6057909076118526,
+      "learning_rate": 6.449149180269175e-06,
+      "loss": 0.2572,
+      "step": 17436
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8677850868302999,
+      "learning_rate": 6.44877207819349e-06,
+      "loss": 0.2715,
+      "step": 17437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5244504739737033,
+      "learning_rate": 6.448394967121324e-06,
+      "loss": 0.3357,
+      "step": 17438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7703877748678793,
+      "learning_rate": 6.448017847055017e-06,
+      "loss": 0.3204,
+      "step": 17439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.579914975790954,
+      "learning_rate": 6.447640717996912e-06,
+      "loss": 0.3835,
+      "step": 17440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.997170320352516,
+      "learning_rate": 6.447263579949351e-06,
+      "loss": 0.4553,
+      "step": 17441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5657550809123937,
+      "learning_rate": 6.446886432914676e-06,
+      "loss": 0.3395,
+      "step": 17442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.061190160341945,
+      "learning_rate": 6.4465092768952285e-06,
+      "loss": 0.3779,
+      "step": 17443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6461729431655792,
+      "learning_rate": 6.4461321118933505e-06,
+      "loss": 0.4692,
+      "step": 17444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.97507267619063,
+      "learning_rate": 6.445754937911385e-06,
+      "loss": 0.2502,
+      "step": 17445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4899861140212156,
+      "learning_rate": 6.445377754951672e-06,
+      "loss": 0.2861,
+      "step": 17446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7365665679713749,
+      "learning_rate": 6.445000563016557e-06,
+      "loss": 0.4578,
+      "step": 17447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.815466475665494,
+      "learning_rate": 6.444623362108378e-06,
+      "loss": 0.427,
+      "step": 17448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5024827040884363,
+      "learning_rate": 6.444246152229482e-06,
+      "loss": 0.2089,
+      "step": 17449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9488020625448625,
+      "learning_rate": 6.443868933382207e-06,
+      "loss": 0.4626,
+      "step": 17450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5971146950485953,
+      "learning_rate": 6.4434917055689e-06,
+      "loss": 0.2517,
+      "step": 17451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4639960875455633,
+      "learning_rate": 6.443114468791898e-06,
+      "loss": 0.3531,
+      "step": 17452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2600350631483415,
+      "learning_rate": 6.442737223053549e-06,
+      "loss": 0.2182,
+      "step": 17453
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3700456009187996,
+      "learning_rate": 6.4423599683561934e-06,
+      "loss": 0.1288,
+      "step": 17454
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2608753358603755,
+      "learning_rate": 6.441982704702171e-06,
+      "loss": 0.1707,
+      "step": 17455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0182038029229634,
+      "learning_rate": 6.441605432093829e-06,
+      "loss": 0.1675,
+      "step": 17456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2013952125154426,
+      "learning_rate": 6.441228150533507e-06,
+      "loss": 0.1863,
+      "step": 17457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7065433846010676,
+      "learning_rate": 6.4408508600235506e-06,
+      "loss": 0.192,
+      "step": 17458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8386189338254126,
+      "learning_rate": 6.4404735605663e-06,
+      "loss": 0.1672,
+      "step": 17459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6804029271825032,
+      "learning_rate": 6.440096252164099e-06,
+      "loss": 0.3531,
+      "step": 17460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.996946850530234,
+      "learning_rate": 6.439718934819292e-06,
+      "loss": 0.7483,
+      "step": 17461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.255543129292119,
+      "learning_rate": 6.43934160853422e-06,
+      "loss": 0.5681,
+      "step": 17462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.777328501625465,
+      "learning_rate": 6.438964273311228e-06,
+      "loss": 0.3549,
+      "step": 17463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3208871287891617,
+      "learning_rate": 6.438586929152657e-06,
+      "loss": 0.203,
+      "step": 17464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7720285524038804,
+      "learning_rate": 6.438209576060852e-06,
+      "loss": 0.1919,
+      "step": 17465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7584176950269867,
+      "learning_rate": 6.437832214038154e-06,
+      "loss": 0.1872,
+      "step": 17466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1660264361484076,
+      "learning_rate": 6.43745484308691e-06,
+      "loss": 0.5619,
+      "step": 17467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4804046863594917,
+      "learning_rate": 6.43707746320946e-06,
+      "loss": 0.1798,
+      "step": 17468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.213758508869449,
+      "learning_rate": 6.43670007440815e-06,
+      "loss": 0.411,
+      "step": 17469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7472382065523322,
+      "learning_rate": 6.43632267668532e-06,
+      "loss": 0.3699,
+      "step": 17470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5165969445722616,
+      "learning_rate": 6.435945270043319e-06,
+      "loss": 0.3487,
+      "step": 17471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7825207360360276,
+      "learning_rate": 6.4355678544844845e-06,
+      "loss": 0.5434,
+      "step": 17472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6490034458440024,
+      "learning_rate": 6.435190430011165e-06,
+      "loss": 0.2923,
+      "step": 17473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5003085688265574,
+      "learning_rate": 6.4348129966257e-06,
+      "loss": 0.2184,
+      "step": 17474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4248657082765814,
+      "learning_rate": 6.4344355543304375e-06,
+      "loss": 0.1536,
+      "step": 17475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9391795846169748,
+      "learning_rate": 6.434058103127718e-06,
+      "loss": 0.5381,
+      "step": 17476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6262839132800155,
+      "learning_rate": 6.433680643019886e-06,
+      "loss": 0.3669,
+      "step": 17477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5401397991433825,
+      "learning_rate": 6.4333031740092875e-06,
+      "loss": 0.3959,
+      "step": 17478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.101265573581645,
+      "learning_rate": 6.432925696098264e-06,
+      "loss": 0.2054,
+      "step": 17479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8216290522217302,
+      "learning_rate": 6.432548209289162e-06,
+      "loss": 0.4502,
+      "step": 17480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.91009556882617,
+      "learning_rate": 6.432170713584322e-06,
+      "loss": 0.5928,
+      "step": 17481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0139706313235006,
+      "learning_rate": 6.431793208986093e-06,
+      "loss": 0.4215,
+      "step": 17482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8350593272533475,
+      "learning_rate": 6.4314156954968134e-06,
+      "loss": 0.2439,
+      "step": 17483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.361323215689215,
+      "learning_rate": 6.431038173118832e-06,
+      "loss": 0.2637,
+      "step": 17484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.644756736609954,
+      "learning_rate": 6.430660641854492e-06,
+      "loss": 0.3525,
+      "step": 17485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6917325006928348,
+      "learning_rate": 6.430283101706136e-06,
+      "loss": 0.3985,
+      "step": 17486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6728047210139072,
+      "learning_rate": 6.429905552676111e-06,
+      "loss": 0.1813,
+      "step": 17487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3167841351961351,
+      "learning_rate": 6.42952799476676e-06,
+      "loss": 0.1893,
+      "step": 17488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.438017399595095,
+      "learning_rate": 6.429150427980429e-06,
+      "loss": 0.1973,
+      "step": 17489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4673934985385721,
+      "learning_rate": 6.428772852319459e-06,
+      "loss": 0.1882,
+      "step": 17490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4669570085585732,
+      "learning_rate": 6.428395267786199e-06,
+      "loss": 0.2309,
+      "step": 17491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9842703896830638,
+      "learning_rate": 6.4280176743829915e-06,
+      "loss": 0.3644,
+      "step": 17492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8197543021021658,
+      "learning_rate": 6.4276400721121825e-06,
+      "loss": 0.466,
+      "step": 17493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3014169219462604,
+      "learning_rate": 6.427262460976113e-06,
+      "loss": 0.1479,
+      "step": 17494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.4643069234071735,
+      "learning_rate": 6.426884840977133e-06,
+      "loss": 0.5346,
+      "step": 17495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.039429931128802,
+      "learning_rate": 6.426507212117586e-06,
+      "loss": 0.3532,
+      "step": 17496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6179067606266173,
+      "learning_rate": 6.426129574399814e-06,
+      "loss": 0.2,
+      "step": 17497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.648448431060128,
+      "learning_rate": 6.425751927826165e-06,
+      "loss": 0.2422,
+      "step": 17498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4854351373819739,
+      "learning_rate": 6.425374272398983e-06,
+      "loss": 0.2875,
+      "step": 17499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6895533228802668,
+      "learning_rate": 6.424996608120614e-06,
+      "loss": 0.2559,
+      "step": 17500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0171544367786463,
+      "learning_rate": 6.424618934993401e-06,
+      "loss": 0.4146,
+      "step": 17501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1690296709921413,
+      "learning_rate": 6.424241253019692e-06,
+      "loss": 0.1771,
+      "step": 17502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5661934210961708,
+      "learning_rate": 6.423863562201831e-06,
+      "loss": 0.2203,
+      "step": 17503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0405357750739455,
+      "learning_rate": 6.423485862542162e-06,
+      "loss": 0.2079,
+      "step": 17504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7066676234831524,
+      "learning_rate": 6.423108154043034e-06,
+      "loss": 0.3929,
+      "step": 17505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7849971212001903,
+      "learning_rate": 6.42273043670679e-06,
+      "loss": 0.2973,
+      "step": 17506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2238191293799194,
+      "learning_rate": 6.422352710535774e-06,
+      "loss": 0.3172,
+      "step": 17507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.820330992183159,
+      "learning_rate": 6.421974975532335e-06,
+      "loss": 0.4785,
+      "step": 17508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5162797550818208,
+      "learning_rate": 6.421597231698816e-06,
+      "loss": 0.1788,
+      "step": 17509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4308424346966604,
+      "learning_rate": 6.421219479037565e-06,
+      "loss": 0.2138,
+      "step": 17510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8519423815390066,
+      "learning_rate": 6.4208417175509255e-06,
+      "loss": 0.4283,
+      "step": 17511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8680003124842801,
+      "learning_rate": 6.4204639472412455e-06,
+      "loss": 0.2094,
+      "step": 17512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9935203282439475,
+      "learning_rate": 6.4200861681108685e-06,
+      "loss": 0.3509,
+      "step": 17513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5422734492660815,
+      "learning_rate": 6.419708380162141e-06,
+      "loss": 0.3745,
+      "step": 17514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.896856711439301,
+      "learning_rate": 6.41933058339741e-06,
+      "loss": 0.2361,
+      "step": 17515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5875003542218717,
+      "learning_rate": 6.418952777819021e-06,
+      "loss": 0.4584,
+      "step": 17516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.852019688138835,
+      "learning_rate": 6.418574963429322e-06,
+      "loss": 0.3834,
+      "step": 17517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8789948345405671,
+      "learning_rate": 6.418197140230654e-06,
+      "loss": 0.2429,
+      "step": 17518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2312560665881582,
+      "learning_rate": 6.417819308225369e-06,
+      "loss": 0.3628,
+      "step": 17519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9556237449288416,
+      "learning_rate": 6.4174414674158105e-06,
+      "loss": 0.456,
+      "step": 17520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6044059261228745,
+      "learning_rate": 6.417063617804323e-06,
+      "loss": 0.3088,
+      "step": 17521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6893469099051484,
+      "learning_rate": 6.416685759393257e-06,
+      "loss": 0.3766,
+      "step": 17522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.194507836610448,
+      "learning_rate": 6.416307892184954e-06,
+      "loss": 0.3192,
+      "step": 17523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3008176294457594,
+      "learning_rate": 6.415930016181766e-06,
+      "loss": 0.2788,
+      "step": 17524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2273620441416146,
+      "learning_rate": 6.415552131386034e-06,
+      "loss": 0.1916,
+      "step": 17525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3948040462991014,
+      "learning_rate": 6.4151742378001095e-06,
+      "loss": 0.3071,
+      "step": 17526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.614131864883346,
+      "learning_rate": 6.4147963354263345e-06,
+      "loss": 0.3904,
+      "step": 17527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5176845676449324,
+      "learning_rate": 6.414418424267059e-06,
+      "loss": 0.1697,
+      "step": 17528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.067476559018398,
+      "learning_rate": 6.414040504324629e-06,
+      "loss": 0.2121,
+      "step": 17529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.951712106350532,
+      "learning_rate": 6.413662575601391e-06,
+      "loss": 0.162,
+      "step": 17530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4347037350054466,
+      "learning_rate": 6.413284638099693e-06,
+      "loss": 0.2606,
+      "step": 17531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.151284768970169,
+      "learning_rate": 6.4129066918218785e-06,
+      "loss": 0.2689,
+      "step": 17532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3403065180845313,
+      "learning_rate": 6.412528736770297e-06,
+      "loss": 0.4743,
+      "step": 17533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4794753271974015,
+      "learning_rate": 6.412150772947295e-06,
+      "loss": 0.3676,
+      "step": 17534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2565911328047352,
+      "learning_rate": 6.411772800355221e-06,
+      "loss": 0.2581,
+      "step": 17535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8177466051468452,
+      "learning_rate": 6.411394818996419e-06,
+      "loss": 0.3331,
+      "step": 17536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5864132369635358,
+      "learning_rate": 6.411016828873239e-06,
+      "loss": 0.2569,
+      "step": 17537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6509043238301366,
+      "learning_rate": 6.410638829988028e-06,
+      "loss": 0.3392,
+      "step": 17538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9561928268462956,
+      "learning_rate": 6.41026082234313e-06,
+      "loss": 0.2423,
+      "step": 17539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.137609376395528,
+      "learning_rate": 6.4098828059408965e-06,
+      "loss": 0.2017,
+      "step": 17540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6063310715888728,
+      "learning_rate": 6.409504780783673e-06,
+      "loss": 0.3072,
+      "step": 17541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.456613117591516,
+      "learning_rate": 6.409126746873808e-06,
+      "loss": 0.2784,
+      "step": 17542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.750286091682074,
+      "learning_rate": 6.408748704213646e-06,
+      "loss": 0.5137,
+      "step": 17543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5562147733041238,
+      "learning_rate": 6.408370652805539e-06,
+      "loss": 0.1782,
+      "step": 17544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7544216426768593,
+      "learning_rate": 6.4079925926518305e-06,
+      "loss": 0.3118,
+      "step": 17545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3033089921809238,
+      "learning_rate": 6.40761452375487e-06,
+      "loss": 0.2656,
+      "step": 17546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5509759194632518,
+      "learning_rate": 6.4072364461170054e-06,
+      "loss": 0.206,
+      "step": 17547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.490462089529291,
+      "learning_rate": 6.4068583597405845e-06,
+      "loss": 0.308,
+      "step": 17548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6439206808918183,
+      "learning_rate": 6.406480264627955e-06,
+      "loss": 0.2394,
+      "step": 17549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.157754873960619,
+      "learning_rate": 6.406102160781463e-06,
+      "loss": 0.584,
+      "step": 17550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8387876220928279,
+      "learning_rate": 6.4057240482034585e-06,
+      "loss": 0.2032,
+      "step": 17551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6139464727170854,
+      "learning_rate": 6.405345926896291e-06,
+      "loss": 0.3589,
+      "step": 17552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4971806124112943,
+      "learning_rate": 6.404967796862305e-06,
+      "loss": 0.2512,
+      "step": 17553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1825856251322677,
+      "learning_rate": 6.404589658103848e-06,
+      "loss": 0.2543,
+      "step": 17554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5701852149049371,
+      "learning_rate": 6.404211510623272e-06,
+      "loss": 0.3189,
+      "step": 17555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3628113655535223,
+      "learning_rate": 6.403833354422924e-06,
+      "loss": 0.1509,
+      "step": 17556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8265053347737588,
+      "learning_rate": 6.40345518950515e-06,
+      "loss": 0.3593,
+      "step": 17557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6075492891820564,
+      "learning_rate": 6.403077015872301e-06,
+      "loss": 0.3176,
+      "step": 17558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3285262287168738,
+      "learning_rate": 6.402698833526724e-06,
+      "loss": 0.1542,
+      "step": 17559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9882231668258614,
+      "learning_rate": 6.402320642470767e-06,
+      "loss": 0.4798,
+      "step": 17560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.195529981273359,
+      "learning_rate": 6.40194244270678e-06,
+      "loss": 0.6729,
+      "step": 17561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2781198036885666,
+      "learning_rate": 6.401564234237109e-06,
+      "loss": 0.1211,
+      "step": 17562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6943841493106966,
+      "learning_rate": 6.401186017064106e-06,
+      "loss": 0.3726,
+      "step": 17563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0100213562630214,
+      "learning_rate": 6.400807791190118e-06,
+      "loss": 0.1811,
+      "step": 17564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5110432325487673,
+      "learning_rate": 6.400429556617492e-06,
+      "loss": 0.3683,
+      "step": 17565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7911902064907248,
+      "learning_rate": 6.400051313348581e-06,
+      "loss": 0.3387,
+      "step": 17566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5945665614448281,
+      "learning_rate": 6.399673061385728e-06,
+      "loss": 0.3192,
+      "step": 17567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.808447376696885,
+      "learning_rate": 6.399294800731286e-06,
+      "loss": 0.3773,
+      "step": 17568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7127869508310123,
+      "learning_rate": 6.398916531387604e-06,
+      "loss": 0.1388,
+      "step": 17569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8312559052850597,
+      "learning_rate": 6.398538253357028e-06,
+      "loss": 0.2416,
+      "step": 17570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4997422374347038,
+      "learning_rate": 6.398159966641909e-06,
+      "loss": 0.2034,
+      "step": 17571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0347198015422077,
+      "learning_rate": 6.397781671244597e-06,
+      "loss": 0.3331,
+      "step": 17572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7143480414281553,
+      "learning_rate": 6.397403367167438e-06,
+      "loss": 0.2022,
+      "step": 17573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8740569703569327,
+      "learning_rate": 6.397025054412786e-06,
+      "loss": 0.2713,
+      "step": 17574
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2499787740236012,
+      "learning_rate": 6.396646732982985e-06,
+      "loss": 0.4953,
+      "step": 17575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8231003863988182,
+      "learning_rate": 6.396268402880387e-06,
+      "loss": 0.2057,
+      "step": 17576
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7109368816357196,
+      "learning_rate": 6.3958900641073415e-06,
+      "loss": 0.4696,
+      "step": 17577
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6874392109977423,
+      "learning_rate": 6.395511716666196e-06,
+      "loss": 0.497,
+      "step": 17578
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4441843630353455,
+      "learning_rate": 6.3951333605593025e-06,
+      "loss": 0.2003,
+      "step": 17579
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1512069914257244,
+      "learning_rate": 6.3947549957890074e-06,
+      "loss": 0.1834,
+      "step": 17580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5995061745615362,
+      "learning_rate": 6.394376622357664e-06,
+      "loss": 0.3104,
+      "step": 17581
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0666607458756205,
+      "learning_rate": 6.393998240267619e-06,
+      "loss": 0.213,
+      "step": 17582
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5308245157749036,
+      "learning_rate": 6.393619849521225e-06,
+      "loss": 0.1991,
+      "step": 17583
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7518056767058026,
+      "learning_rate": 6.393241450120826e-06,
+      "loss": 0.4669,
+      "step": 17584
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1561924172574267,
+      "learning_rate": 6.392863042068779e-06,
+      "loss": 0.1929,
+      "step": 17585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7026806697965409,
+      "learning_rate": 6.3924846253674275e-06,
+      "loss": 0.2922,
+      "step": 17586
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8023761608283537,
+      "learning_rate": 6.392106200019126e-06,
+      "loss": 0.2233,
+      "step": 17587
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.453532787243206,
+      "learning_rate": 6.391727766026221e-06,
+      "loss": 0.2304,
+      "step": 17588
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3692729018086796,
+      "learning_rate": 6.391349323391065e-06,
+      "loss": 0.2565,
+      "step": 17589
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.727792965962597,
+      "learning_rate": 6.390970872116007e-06,
+      "loss": 0.4091,
+      "step": 17590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8039562015217607,
+      "learning_rate": 6.390592412203397e-06,
+      "loss": 0.3562,
+      "step": 17591
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5544684989074231,
+      "learning_rate": 6.390213943655585e-06,
+      "loss": 0.2607,
+      "step": 17592
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.335697662518615,
+      "learning_rate": 6.389835466474922e-06,
+      "loss": 0.3125,
+      "step": 17593
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.697166444363012,
+      "learning_rate": 6.389456980663758e-06,
+      "loss": 0.4208,
+      "step": 17594
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2293823896107448,
+      "learning_rate": 6.389078486224441e-06,
+      "loss": 0.125,
+      "step": 17595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5231153343138184,
+      "learning_rate": 6.388699983159325e-06,
+      "loss": 0.3336,
+      "step": 17596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.265552690601715,
+      "learning_rate": 6.38832147147076e-06,
+      "loss": 0.2265,
+      "step": 17597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4056968300834627,
+      "learning_rate": 6.387942951161092e-06,
+      "loss": 0.3138,
+      "step": 17598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6827224038780964,
+      "learning_rate": 6.387564422232677e-06,
+      "loss": 0.4127,
+      "step": 17599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5735513488735628,
+      "learning_rate": 6.387185884687863e-06,
+      "loss": 0.3644,
+      "step": 17600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9593165906082028,
+      "learning_rate": 6.386807338529001e-06,
+      "loss": 0.2572,
+      "step": 17601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.749037307522388,
+      "learning_rate": 6.38642878375844e-06,
+      "loss": 0.5505,
+      "step": 17602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.021566700756304,
+      "learning_rate": 6.386050220378535e-06,
+      "loss": 0.2346,
+      "step": 17603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8159350491587456,
+      "learning_rate": 6.385671648391632e-06,
+      "loss": 0.3135,
+      "step": 17604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8367807724255671,
+      "learning_rate": 6.385293067800084e-06,
+      "loss": 0.2895,
+      "step": 17605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9382207206798594,
+      "learning_rate": 6.384914478606241e-06,
+      "loss": 0.4944,
+      "step": 17606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5418054855411647,
+      "learning_rate": 6.384535880812457e-06,
+      "loss": 0.2835,
+      "step": 17607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4106025781991784,
+      "learning_rate": 6.384157274421079e-06,
+      "loss": 0.1807,
+      "step": 17608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7276772788279877,
+      "learning_rate": 6.38377865943446e-06,
+      "loss": 0.5523,
+      "step": 17609
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.589618498195326,
+      "learning_rate": 6.383400035854951e-06,
+      "loss": 0.4517,
+      "step": 17610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9024989712971778,
+      "learning_rate": 6.383021403684903e-06,
+      "loss": 0.376,
+      "step": 17611
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.600131269933627,
+      "learning_rate": 6.382642762926666e-06,
+      "loss": 0.1507,
+      "step": 17612
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5120346986093285,
+      "learning_rate": 6.382264113582594e-06,
+      "loss": 0.261,
+      "step": 17613
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8839415690703956,
+      "learning_rate": 6.381885455655036e-06,
+      "loss": 0.406,
+      "step": 17614
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4939723674305794,
+      "learning_rate": 6.381506789146343e-06,
+      "loss": 0.1654,
+      "step": 17615
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.262322957306404,
+      "learning_rate": 6.3811281140588695e-06,
+      "loss": 0.3381,
+      "step": 17616
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.540836024001774,
+      "learning_rate": 6.3807494303949635e-06,
+      "loss": 0.287,
+      "step": 17617
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5251979217721918,
+      "learning_rate": 6.38037073815698e-06,
+      "loss": 0.4178,
+      "step": 17618
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9773592285521053,
+      "learning_rate": 6.379992037347268e-06,
+      "loss": 0.3741,
+      "step": 17619
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5578970630769637,
+      "learning_rate": 6.379613327968179e-06,
+      "loss": 0.2066,
+      "step": 17620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0385724059685053,
+      "learning_rate": 6.379234610022065e-06,
+      "loss": 0.2347,
+      "step": 17621
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.403345010921039,
+      "learning_rate": 6.378855883511278e-06,
+      "loss": 0.2231,
+      "step": 17622
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5585486876395838,
+      "learning_rate": 6.3784771484381715e-06,
+      "loss": 0.3049,
+      "step": 17623
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7505446250711414,
+      "learning_rate": 6.378098404805094e-06,
+      "loss": 0.474,
+      "step": 17624
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9306098224158148,
+      "learning_rate": 6.3777196526144e-06,
+      "loss": 0.5237,
+      "step": 17625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8544761894062327,
+      "learning_rate": 6.37734089186844e-06,
+      "loss": 0.6094,
+      "step": 17626
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4011538911446482,
+      "learning_rate": 6.3769621225695675e-06,
+      "loss": 0.2874,
+      "step": 17627
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4620263713336579,
+      "learning_rate": 6.376583344720133e-06,
+      "loss": 0.2111,
+      "step": 17628
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4418422950446694,
+      "learning_rate": 6.376204558322491e-06,
+      "loss": 0.5914,
+      "step": 17629
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6920409652511363,
+      "learning_rate": 6.375825763378991e-06,
+      "loss": 0.328,
+      "step": 17630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2802108663284897,
+      "learning_rate": 6.375446959891985e-06,
+      "loss": 0.3522,
+      "step": 17631
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7873623278092807,
+      "learning_rate": 6.375068147863828e-06,
+      "loss": 0.5503,
+      "step": 17632
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9483635771371217,
+      "learning_rate": 6.374689327296868e-06,
+      "loss": 0.3431,
+      "step": 17633
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.828746429199988,
+      "learning_rate": 6.374310498193463e-06,
+      "loss": 0.5267,
+      "step": 17634
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.635503185518685,
+      "learning_rate": 6.373931660555961e-06,
+      "loss": 0.2014,
+      "step": 17635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8918276736839907,
+      "learning_rate": 6.373552814386717e-06,
+      "loss": 0.526,
+      "step": 17636
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2007593640608234,
+      "learning_rate": 6.373173959688082e-06,
+      "loss": 0.1585,
+      "step": 17637
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1721017618128489,
+      "learning_rate": 6.372795096462407e-06,
+      "loss": 0.1545,
+      "step": 17638
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8924846724515618,
+      "learning_rate": 6.37241622471205e-06,
+      "loss": 0.5072,
+      "step": 17639
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7045697025539521,
+      "learning_rate": 6.3720373444393595e-06,
+      "loss": 0.3292,
+      "step": 17640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5813185487559662,
+      "learning_rate": 6.371658455646689e-06,
+      "loss": 0.2252,
+      "step": 17641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.940119245756464,
+      "learning_rate": 6.3712795583363916e-06,
+      "loss": 0.3261,
+      "step": 17642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.831458895001242,
+      "learning_rate": 6.370900652510821e-06,
+      "loss": 0.4941,
+      "step": 17643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4871190406719672,
+      "learning_rate": 6.370521738172327e-06,
+      "loss": 0.3368,
+      "step": 17644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8237056643677199,
+      "learning_rate": 6.370142815323266e-06,
+      "loss": 0.1819,
+      "step": 17645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.264233527737778,
+      "learning_rate": 6.369763883965989e-06,
+      "loss": 0.2857,
+      "step": 17646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0825906703494885,
+      "learning_rate": 6.369384944102851e-06,
+      "loss": 0.2019,
+      "step": 17647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7469062408287857,
+      "learning_rate": 6.369005995736203e-06,
+      "loss": 0.3002,
+      "step": 17648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5691099805358477,
+      "learning_rate": 6.3686270388684006e-06,
+      "loss": 0.2119,
+      "step": 17649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3041473329792372,
+      "learning_rate": 6.368248073501794e-06,
+      "loss": 0.1939,
+      "step": 17650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9420822969762752,
+      "learning_rate": 6.367869099638737e-06,
+      "loss": 0.2683,
+      "step": 17651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7412235518430217,
+      "learning_rate": 6.367490117281585e-06,
+      "loss": 0.2581,
+      "step": 17652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.088038475860969,
+      "learning_rate": 6.36711112643269e-06,
+      "loss": 0.2169,
+      "step": 17653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.002532156967063,
+      "learning_rate": 6.3667321270944075e-06,
+      "loss": 0.3561,
+      "step": 17654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1229194560764193,
+      "learning_rate": 6.366353119269087e-06,
+      "loss": 0.5943,
+      "step": 17655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3438982659980798,
+      "learning_rate": 6.365974102959085e-06,
+      "loss": 0.2389,
+      "step": 17656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5774745190711656,
+      "learning_rate": 6.365595078166753e-06,
+      "loss": 0.335,
+      "step": 17657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2748629109505627,
+      "learning_rate": 6.365216044894448e-06,
+      "loss": 0.4284,
+      "step": 17658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1983436381427723,
+      "learning_rate": 6.364837003144519e-06,
+      "loss": 0.3555,
+      "step": 17659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8692635363243901,
+      "learning_rate": 6.364457952919325e-06,
+      "loss": 0.2013,
+      "step": 17660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3886790054073845,
+      "learning_rate": 6.364078894221215e-06,
+      "loss": 0.2752,
+      "step": 17661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5562186345229607,
+      "learning_rate": 6.3636998270525474e-06,
+      "loss": 0.1702,
+      "step": 17662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.608113424020045,
+      "learning_rate": 6.363320751415671e-06,
+      "loss": 0.2702,
+      "step": 17663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7406175750217927,
+      "learning_rate": 6.362941667312944e-06,
+      "loss": 0.2283,
+      "step": 17664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.691531720694492,
+      "learning_rate": 6.362562574746719e-06,
+      "loss": 0.2764,
+      "step": 17665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8319050596979745,
+      "learning_rate": 6.36218347371935e-06,
+      "loss": 0.3457,
+      "step": 17666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9305399014327997,
+      "learning_rate": 6.361804364233191e-06,
+      "loss": 0.1485,
+      "step": 17667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3587527796770387,
+      "learning_rate": 6.361425246290596e-06,
+      "loss": 0.1662,
+      "step": 17668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6907028471181158,
+      "learning_rate": 6.361046119893919e-06,
+      "loss": 0.372,
+      "step": 17669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7876183332275308,
+      "learning_rate": 6.360666985045515e-06,
+      "loss": 0.4748,
+      "step": 17670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3713392760582368,
+      "learning_rate": 6.360287841747739e-06,
+      "loss": 0.1871,
+      "step": 17671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4549062896239515,
+      "learning_rate": 6.359908690002944e-06,
+      "loss": 0.1311,
+      "step": 17672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6197687824100555,
+      "learning_rate": 6.359529529813484e-06,
+      "loss": 0.1474,
+      "step": 17673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1359639813554283,
+      "learning_rate": 6.3591503611817155e-06,
+      "loss": 0.2091,
+      "step": 17674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2863693019880493,
+      "learning_rate": 6.358771184109991e-06,
+      "loss": 0.2988,
+      "step": 17675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8584110742681568,
+      "learning_rate": 6.358391998600666e-06,
+      "loss": 0.2021,
+      "step": 17676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0717944329478732,
+      "learning_rate": 6.358012804656095e-06,
+      "loss": 0.3128,
+      "step": 17677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3079374119252778,
+      "learning_rate": 6.357633602278634e-06,
+      "loss": 0.1868,
+      "step": 17678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2089371371554394,
+      "learning_rate": 6.357254391470636e-06,
+      "loss": 0.2855,
+      "step": 17679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7013306855660975,
+      "learning_rate": 6.3568751722344555e-06,
+      "loss": 0.2791,
+      "step": 17680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7282561759416242,
+      "learning_rate": 6.3564959445724475e-06,
+      "loss": 0.3294,
+      "step": 17681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2397247634140478,
+      "learning_rate": 6.35611670848697e-06,
+      "loss": 0.2484,
+      "step": 17682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6989346584617249,
+      "learning_rate": 6.355737463980374e-06,
+      "loss": 0.1778,
+      "step": 17683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4435804040388724,
+      "learning_rate": 6.355358211055016e-06,
+      "loss": 0.3969,
+      "step": 17684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5546890107129965,
+      "learning_rate": 6.354978949713252e-06,
+      "loss": 0.2912,
+      "step": 17685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2830332348912696,
+      "learning_rate": 6.354599679957435e-06,
+      "loss": 0.1698,
+      "step": 17686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7272826155908234,
+      "learning_rate": 6.354220401789921e-06,
+      "loss": 0.4498,
+      "step": 17687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8388084770955515,
+      "learning_rate": 6.353841115213067e-06,
+      "loss": 0.3541,
+      "step": 17688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5427341053400756,
+      "learning_rate": 6.353461820229227e-06,
+      "loss": 0.2012,
+      "step": 17689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.081947772301071,
+      "learning_rate": 6.353082516840754e-06,
+      "loss": 0.7043,
+      "step": 17690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.426363201780887,
+      "learning_rate": 6.352703205050008e-06,
+      "loss": 0.2197,
+      "step": 17691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1763383497642637,
+      "learning_rate": 6.35232388485934e-06,
+      "loss": 0.1856,
+      "step": 17692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5857523354361733,
+      "learning_rate": 6.351944556271109e-06,
+      "loss": 0.3835,
+      "step": 17693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4342112030053489,
+      "learning_rate": 6.351565219287669e-06,
+      "loss": 0.2854,
+      "step": 17694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4741135199327924,
+      "learning_rate": 6.351185873911374e-06,
+      "loss": 0.1512,
+      "step": 17695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7557473916060604,
+      "learning_rate": 6.350806520144584e-06,
+      "loss": 0.4671,
+      "step": 17696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4139294588087719,
+      "learning_rate": 6.350427157989649e-06,
+      "loss": 0.2608,
+      "step": 17697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1645827236180795,
+      "learning_rate": 6.350047787448928e-06,
+      "loss": 0.3295,
+      "step": 17698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9967059038469754,
+      "learning_rate": 6.349668408524778e-06,
+      "loss": 0.3918,
+      "step": 17699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7119293487194938,
+      "learning_rate": 6.349289021219551e-06,
+      "loss": 0.4737,
+      "step": 17700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4961462385168254,
+      "learning_rate": 6.348909625535606e-06,
+      "loss": 0.3128,
+      "step": 17701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9830264968765812,
+      "learning_rate": 6.3485302214752995e-06,
+      "loss": 0.5612,
+      "step": 17702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3961082418256643,
+      "learning_rate": 6.348150809040985e-06,
+      "loss": 0.4494,
+      "step": 17703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.696001165684123,
+      "learning_rate": 6.34777138823502e-06,
+      "loss": 0.3344,
+      "step": 17704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7877308405943868,
+      "learning_rate": 6.34739195905976e-06,
+      "loss": 0.2353,
+      "step": 17705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3927649978512517,
+      "learning_rate": 6.34701252151756e-06,
+      "loss": 0.3088,
+      "step": 17706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.587369579961541,
+      "learning_rate": 6.3466330756107795e-06,
+      "loss": 0.1883,
+      "step": 17707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4143394045979467,
+      "learning_rate": 6.346253621341772e-06,
+      "loss": 0.2895,
+      "step": 17708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3662475951855044,
+      "learning_rate": 6.3458741587128944e-06,
+      "loss": 0.2197,
+      "step": 17709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2984947474940167,
+      "learning_rate": 6.345494687726503e-06,
+      "loss": 0.4012,
+      "step": 17710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.009448590866567,
+      "learning_rate": 6.345115208384955e-06,
+      "loss": 0.4714,
+      "step": 17711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6982899681595447,
+      "learning_rate": 6.344735720690607e-06,
+      "loss": 0.1622,
+      "step": 17712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8536577019539715,
+      "learning_rate": 6.344356224645814e-06,
+      "loss": 0.33,
+      "step": 17713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8403861143404499,
+      "learning_rate": 6.343976720252933e-06,
+      "loss": 0.5081,
+      "step": 17714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.413273530646942,
+      "learning_rate": 6.343597207514321e-06,
+      "loss": 0.3208,
+      "step": 17715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7485474663551177,
+      "learning_rate": 6.343217686432337e-06,
+      "loss": 0.4002,
+      "step": 17716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5201979219067403,
+      "learning_rate": 6.342838157009333e-06,
+      "loss": 0.2198,
+      "step": 17717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6775587831564527,
+      "learning_rate": 6.34245861924767e-06,
+      "loss": 0.3812,
+      "step": 17718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5425407443778456,
+      "learning_rate": 6.342079073149702e-06,
+      "loss": 0.3315,
+      "step": 17719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5031741664106055,
+      "learning_rate": 6.3416995187177875e-06,
+      "loss": 0.1741,
+      "step": 17720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8835442968875544,
+      "learning_rate": 6.341319955954281e-06,
+      "loss": 0.459,
+      "step": 17721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3038286105995114,
+      "learning_rate": 6.340940384861545e-06,
+      "loss": 0.1574,
+      "step": 17722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5842058753613444,
+      "learning_rate": 6.34056080544193e-06,
+      "loss": 0.2809,
+      "step": 17723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.985440763992411,
+      "learning_rate": 6.340181217697797e-06,
+      "loss": 0.3974,
+      "step": 17724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6323554292479918,
+      "learning_rate": 6.339801621631501e-06,
+      "loss": 0.144,
+      "step": 17725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4525949577932338,
+      "learning_rate": 6.339422017245403e-06,
+      "loss": 0.2506,
+      "step": 17726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4242840966923702,
+      "learning_rate": 6.339042404541855e-06,
+      "loss": 0.2089,
+      "step": 17727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7772936426162989,
+      "learning_rate": 6.3386627835232175e-06,
+      "loss": 0.4008,
+      "step": 17728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5471065810279023,
+      "learning_rate": 6.3382831541918475e-06,
+      "loss": 0.1897,
+      "step": 17729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.592047639998864,
+      "learning_rate": 6.337903516550101e-06,
+      "loss": 0.3974,
+      "step": 17730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.165197302535155,
+      "learning_rate": 6.3375238706003374e-06,
+      "loss": 0.0939,
+      "step": 17731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7047547626949093,
+      "learning_rate": 6.337144216344913e-06,
+      "loss": 0.455,
+      "step": 17732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7644019997152787,
+      "learning_rate": 6.336764553786187e-06,
+      "loss": 0.526,
+      "step": 17733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5671497540097696,
+      "learning_rate": 6.336384882926514e-06,
+      "loss": 0.3537,
+      "step": 17734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7367794471564024,
+      "learning_rate": 6.336005203768254e-06,
+      "loss": 0.3725,
+      "step": 17735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1218965567222754,
+      "learning_rate": 6.335625516313764e-06,
+      "loss": 0.4385,
+      "step": 17736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7352520898614854,
+      "learning_rate": 6.335245820565401e-06,
+      "loss": 0.3352,
+      "step": 17737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3077148574807451,
+      "learning_rate": 6.334866116525524e-06,
+      "loss": 0.2227,
+      "step": 17738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6880272553755375,
+      "learning_rate": 6.334486404196491e-06,
+      "loss": 0.4405,
+      "step": 17739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.294181840767673,
+      "learning_rate": 6.334106683580659e-06,
+      "loss": 0.3269,
+      "step": 17740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8140118640335046,
+      "learning_rate": 6.333726954680386e-06,
+      "loss": 0.2951,
+      "step": 17741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.483704977903084,
+      "learning_rate": 6.3333472174980305e-06,
+      "loss": 0.1264,
+      "step": 17742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1350354253076087,
+      "learning_rate": 6.3329674720359505e-06,
+      "loss": 0.1601,
+      "step": 17743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7598274846694046,
+      "learning_rate": 6.332587718296505e-06,
+      "loss": 0.4902,
+      "step": 17744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7007948614232045,
+      "learning_rate": 6.332207956282049e-06,
+      "loss": 0.2797,
+      "step": 17745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8913755267049677,
+      "learning_rate": 6.3318281859949445e-06,
+      "loss": 0.5092,
+      "step": 17746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.667063807070259,
+      "learning_rate": 6.331448407437547e-06,
+      "loss": 0.3354,
+      "step": 17747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5459482374140336,
+      "learning_rate": 6.331068620612217e-06,
+      "loss": 0.3793,
+      "step": 17748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7746528124756913,
+      "learning_rate": 6.330688825521313e-06,
+      "loss": 0.3432,
+      "step": 17749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2345232391676888,
+      "learning_rate": 6.33030902216719e-06,
+      "loss": 0.1677,
+      "step": 17750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4580309764455375,
+      "learning_rate": 6.3299292105522105e-06,
+      "loss": 0.2259,
+      "step": 17751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.802302745919823,
+      "learning_rate": 6.32954939067873e-06,
+      "loss": 0.3853,
+      "step": 17752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.365827276634838,
+      "learning_rate": 6.329169562549109e-06,
+      "loss": 0.6973,
+      "step": 17753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7014678452351488,
+      "learning_rate": 6.328789726165705e-06,
+      "loss": 0.3603,
+      "step": 17754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2600826781164196,
+      "learning_rate": 6.328409881530878e-06,
+      "loss": 0.1504,
+      "step": 17755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4530065806138759,
+      "learning_rate": 6.328030028646985e-06,
+      "loss": 0.1803,
+      "step": 17756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6549912383749419,
+      "learning_rate": 6.3276501675163875e-06,
+      "loss": 0.4425,
+      "step": 17757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3909869794459004,
+      "learning_rate": 6.32727029814144e-06,
+      "loss": 0.3179,
+      "step": 17758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9491073331925783,
+      "learning_rate": 6.326890420524506e-06,
+      "loss": 0.7886,
+      "step": 17759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6017940377133675,
+      "learning_rate": 6.326510534667941e-06,
+      "loss": 0.3526,
+      "step": 17760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8833962323085776,
+      "learning_rate": 6.326130640574107e-06,
+      "loss": 0.318,
+      "step": 17761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3668797337310972,
+      "learning_rate": 6.325750738245359e-06,
+      "loss": 0.2522,
+      "step": 17762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5570158942843983,
+      "learning_rate": 6.325370827684059e-06,
+      "loss": 0.3463,
+      "step": 17763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9965325337966073,
+      "learning_rate": 6.324990908892566e-06,
+      "loss": 0.2974,
+      "step": 17764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2972015650484547,
+      "learning_rate": 6.324610981873238e-06,
+      "loss": 0.1272,
+      "step": 17765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3745451955292747,
+      "learning_rate": 6.324231046628437e-06,
+      "loss": 0.1609,
+      "step": 17766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8788025761986837,
+      "learning_rate": 6.3238511031605186e-06,
+      "loss": 0.6247,
+      "step": 17767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0128430382174356,
+      "learning_rate": 6.323471151471846e-06,
+      "loss": 0.4985,
+      "step": 17768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1374866097704885,
+      "learning_rate": 6.323091191564774e-06,
+      "loss": 0.1633,
+      "step": 17769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7842101630587106,
+      "learning_rate": 6.322711223441667e-06,
+      "loss": 0.2144,
+      "step": 17770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6569024186543935,
+      "learning_rate": 6.322331247104881e-06,
+      "loss": 0.4981,
+      "step": 17771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0034005529031145,
+      "learning_rate": 6.321951262556776e-06,
+      "loss": 0.424,
+      "step": 17772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.848221093712188,
+      "learning_rate": 6.321571269799713e-06,
+      "loss": 0.5106,
+      "step": 17773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.076377679021282,
+      "learning_rate": 6.32119126883605e-06,
+      "loss": 0.7158,
+      "step": 17774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2646872421302398,
+      "learning_rate": 6.320811259668148e-06,
+      "loss": 0.3462,
+      "step": 17775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4697440466363605,
+      "learning_rate": 6.320431242298366e-06,
+      "loss": 0.1678,
+      "step": 17776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7532550520435881,
+      "learning_rate": 6.320051216729065e-06,
+      "loss": 0.3206,
+      "step": 17777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8828560282000806,
+      "learning_rate": 6.3196711829626035e-06,
+      "loss": 0.4393,
+      "step": 17778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.629377552427482,
+      "learning_rate": 6.3192911410013414e-06,
+      "loss": 0.3637,
+      "step": 17779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3896752634361111,
+      "learning_rate": 6.318911090847639e-06,
+      "loss": 0.2836,
+      "step": 17780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8551556674432979,
+      "learning_rate": 6.318531032503857e-06,
+      "loss": 0.2322,
+      "step": 17781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8099529687292966,
+      "learning_rate": 6.318150965972355e-06,
+      "loss": 0.6134,
+      "step": 17782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.927658759801393,
+      "learning_rate": 6.3177708912554925e-06,
+      "loss": 0.38,
+      "step": 17783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.755626711849656,
+      "learning_rate": 6.317390808355631e-06,
+      "loss": 0.3034,
+      "step": 17784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.331882356617094,
+      "learning_rate": 6.317010717275129e-06,
+      "loss": 0.2054,
+      "step": 17785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.389221182020764,
+      "learning_rate": 6.316630618016348e-06,
+      "loss": 0.3131,
+      "step": 17786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4422426743249428,
+      "learning_rate": 6.316250510581648e-06,
+      "loss": 0.2254,
+      "step": 17787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3792988593757851,
+      "learning_rate": 6.315870394973388e-06,
+      "loss": 0.2485,
+      "step": 17788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5926884874696263,
+      "learning_rate": 6.3154902711939315e-06,
+      "loss": 0.3514,
+      "step": 17789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7622732359852922,
+      "learning_rate": 6.315110139245635e-06,
+      "loss": 0.3671,
+      "step": 17790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6370559586275688,
+      "learning_rate": 6.314729999130864e-06,
+      "loss": 0.4109,
+      "step": 17791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9027109359647671,
+      "learning_rate": 6.314349850851974e-06,
+      "loss": 0.3301,
+      "step": 17792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.683427767083722,
+      "learning_rate": 6.313969694411329e-06,
+      "loss": 0.4015,
+      "step": 17793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4200624530156418,
+      "learning_rate": 6.313589529811288e-06,
+      "loss": 0.2746,
+      "step": 17794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2261448744480419,
+      "learning_rate": 6.313209357054213e-06,
+      "loss": 0.2457,
+      "step": 17795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.107105167790899,
+      "learning_rate": 6.312829176142463e-06,
+      "loss": 0.2181,
+      "step": 17796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6313446213877543,
+      "learning_rate": 6.312448987078401e-06,
+      "loss": 0.3492,
+      "step": 17797
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4501627240870318,
+      "learning_rate": 6.312068789864386e-06,
+      "loss": 0.1842,
+      "step": 17798
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5836277853155156,
+      "learning_rate": 6.311688584502781e-06,
+      "loss": 0.2252,
+      "step": 17799
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5934328717383175,
+      "learning_rate": 6.3113083709959436e-06,
+      "loss": 0.3101,
+      "step": 17800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4655745984624564,
+      "learning_rate": 6.310928149346238e-06,
+      "loss": 0.5026,
+      "step": 17801
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8707331526953652,
+      "learning_rate": 6.3105479195560245e-06,
+      "loss": 0.4497,
+      "step": 17802
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5988840384624141,
+      "learning_rate": 6.310167681627663e-06,
+      "loss": 0.1999,
+      "step": 17803
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9707585337385254,
+      "learning_rate": 6.309787435563517e-06,
+      "loss": 0.4471,
+      "step": 17804
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7134080668566412,
+      "learning_rate": 6.309407181365945e-06,
+      "loss": 0.4749,
+      "step": 17805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4824970261494108,
+      "learning_rate": 6.30902691903731e-06,
+      "loss": 0.351,
+      "step": 17806
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7224325523542101,
+      "learning_rate": 6.308646648579971e-06,
+      "loss": 0.5642,
+      "step": 17807
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9154074672943866,
+      "learning_rate": 6.308266369996294e-06,
+      "loss": 0.615,
+      "step": 17808
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7816072992387493,
+      "learning_rate": 6.307886083288636e-06,
+      "loss": 0.2239,
+      "step": 17809
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.579922785154853,
+      "learning_rate": 6.307505788459361e-06,
+      "loss": 0.1837,
+      "step": 17810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.934923161963559,
+      "learning_rate": 6.307125485510829e-06,
+      "loss": 0.3898,
+      "step": 17811
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7727992972250992,
+      "learning_rate": 6.306745174445402e-06,
+      "loss": 0.5095,
+      "step": 17812
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.370351034038546,
+      "learning_rate": 6.306364855265442e-06,
+      "loss": 0.2894,
+      "step": 17813
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7270095470834967,
+      "learning_rate": 6.30598452797331e-06,
+      "loss": 0.307,
+      "step": 17814
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6319865939658202,
+      "learning_rate": 6.3056041925713705e-06,
+      "loss": 0.4136,
+      "step": 17815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0049374216142433,
+      "learning_rate": 6.3052238490619814e-06,
+      "loss": 0.1563,
+      "step": 17816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2306915479242115,
+      "learning_rate": 6.304843497447506e-06,
+      "loss": 0.6585,
+      "step": 17817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7343460286787953,
+      "learning_rate": 6.304463137730307e-06,
+      "loss": 0.4427,
+      "step": 17818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7867274809149893,
+      "learning_rate": 6.304082769912745e-06,
+      "loss": 0.4846,
+      "step": 17819
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3088141355402751,
+      "learning_rate": 6.303702393997183e-06,
+      "loss": 0.2065,
+      "step": 17820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7841281392752655,
+      "learning_rate": 6.303322009985984e-06,
+      "loss": 0.5181,
+      "step": 17821
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0936213349548973,
+      "learning_rate": 6.302941617881506e-06,
+      "loss": 0.2042,
+      "step": 17822
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.666557256992309,
+      "learning_rate": 6.302561217686118e-06,
+      "loss": 0.5282,
+      "step": 17823
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6061389066954628,
+      "learning_rate": 6.302180809402174e-06,
+      "loss": 0.4209,
+      "step": 17824
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7919115635667753,
+      "learning_rate": 6.301800393032043e-06,
+      "loss": 0.3805,
+      "step": 17825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9965774375618257,
+      "learning_rate": 6.301419968578084e-06,
+      "loss": 0.2231,
+      "step": 17826
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7962383718120856,
+      "learning_rate": 6.301039536042659e-06,
+      "loss": 0.3605,
+      "step": 17827
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6043357563903382,
+      "learning_rate": 6.300659095428132e-06,
+      "loss": 0.1698,
+      "step": 17828
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8383706182569133,
+      "learning_rate": 6.300278646736865e-06,
+      "loss": 0.3928,
+      "step": 17829
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.643822723491477,
+      "learning_rate": 6.2998981899712195e-06,
+      "loss": 0.1857,
+      "step": 17830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8573305313793925,
+      "learning_rate": 6.2995177251335595e-06,
+      "loss": 0.0987,
+      "step": 17831
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0048258903884655,
+      "learning_rate": 6.299137252226246e-06,
+      "loss": 0.3569,
+      "step": 17832
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1599047602384784,
+      "learning_rate": 6.298756771251644e-06,
+      "loss": 0.2406,
+      "step": 17833
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.897241883093018,
+      "learning_rate": 6.298376282212114e-06,
+      "loss": 0.5895,
+      "step": 17834
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.964686393124654,
+      "learning_rate": 6.2979957851100185e-06,
+      "loss": 0.4506,
+      "step": 17835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7946692902313546,
+      "learning_rate": 6.297615279947722e-06,
+      "loss": 0.439,
+      "step": 17836
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7639072383675436,
+      "learning_rate": 6.297234766727587e-06,
+      "loss": 0.4542,
+      "step": 17837
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1301685265798915,
+      "learning_rate": 6.296854245451976e-06,
+      "loss": 0.2361,
+      "step": 17838
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5298594418097784,
+      "learning_rate": 6.296473716123251e-06,
+      "loss": 0.3084,
+      "step": 17839
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0950477976228647,
+      "learning_rate": 6.296093178743775e-06,
+      "loss": 0.3567,
+      "step": 17840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4237780908864488,
+      "learning_rate": 6.295712633315914e-06,
+      "loss": 0.1285,
+      "step": 17841
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5916736086994425,
+      "learning_rate": 6.295332079842026e-06,
+      "loss": 0.3565,
+      "step": 17842
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2932226675051803,
+      "learning_rate": 6.294951518324479e-06,
+      "loss": 0.1989,
+      "step": 17843
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.425728052270351,
+      "learning_rate": 6.294570948765635e-06,
+      "loss": 0.3089,
+      "step": 17844
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5792167859807364,
+      "learning_rate": 6.294190371167855e-06,
+      "loss": 0.2395,
+      "step": 17845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2362530099429063,
+      "learning_rate": 6.2938097855335046e-06,
+      "loss": 0.2084,
+      "step": 17846
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0065297693750517,
+      "learning_rate": 6.293429191864947e-06,
+      "loss": 0.2144,
+      "step": 17847
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.019901610296953,
+      "learning_rate": 6.293048590164545e-06,
+      "loss": 0.1563,
+      "step": 17848
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6940716276220233,
+      "learning_rate": 6.29266798043466e-06,
+      "loss": 0.4982,
+      "step": 17849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6079429347418348,
+      "learning_rate": 6.292287362677659e-06,
+      "loss": 0.4228,
+      "step": 17850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9173766997652673,
+      "learning_rate": 6.291906736895903e-06,
+      "loss": 0.2165,
+      "step": 17851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6182473527635786,
+      "learning_rate": 6.291526103091758e-06,
+      "loss": 0.4854,
+      "step": 17852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.24469282959619,
+      "learning_rate": 6.291145461267585e-06,
+      "loss": 0.1991,
+      "step": 17853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.833136593131374,
+      "learning_rate": 6.290764811425749e-06,
+      "loss": 0.5459,
+      "step": 17854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5904898584211002,
+      "learning_rate": 6.290384153568614e-06,
+      "loss": 0.2685,
+      "step": 17855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1745958721373013,
+      "learning_rate": 6.2900034876985425e-06,
+      "loss": 0.2065,
+      "step": 17856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.786615376386437,
+      "learning_rate": 6.2896228138179e-06,
+      "loss": 0.2486,
+      "step": 17857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9446482002888485,
+      "learning_rate": 6.28924213192905e-06,
+      "loss": 0.3536,
+      "step": 17858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4668415847798,
+      "learning_rate": 6.288861442034355e-06,
+      "loss": 0.3031,
+      "step": 17859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8327008184893447,
+      "learning_rate": 6.2884807441361805e-06,
+      "loss": 0.1836,
+      "step": 17860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0144277203022956,
+      "learning_rate": 6.288100038236889e-06,
+      "loss": 0.6275,
+      "step": 17861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2777163460840653,
+      "learning_rate": 6.287719324338846e-06,
+      "loss": 0.188,
+      "step": 17862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2878089416073368,
+      "learning_rate": 6.287338602444416e-06,
+      "loss": 0.2769,
+      "step": 17863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3911044815765277,
+      "learning_rate": 6.286957872555963e-06,
+      "loss": 0.3755,
+      "step": 17864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6573275263911047,
+      "learning_rate": 6.286577134675849e-06,
+      "loss": 0.3321,
+      "step": 17865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.73935871377764,
+      "learning_rate": 6.286196388806441e-06,
+      "loss": 0.4539,
+      "step": 17866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9539970320424778,
+      "learning_rate": 6.2858156349501005e-06,
+      "loss": 0.1948,
+      "step": 17867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2602980930011904,
+      "learning_rate": 6.285434873109197e-06,
+      "loss": 0.1812,
+      "step": 17868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0618605164201025,
+      "learning_rate": 6.285054103286088e-06,
+      "loss": 0.4104,
+      "step": 17869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1224822492890656,
+      "learning_rate": 6.284673325483144e-06,
+      "loss": 0.3278,
+      "step": 17870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.105604376166805,
+      "learning_rate": 6.284292539702725e-06,
+      "loss": 0.4174,
+      "step": 17871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8077468701898542,
+      "learning_rate": 6.2839117459471995e-06,
+      "loss": 0.411,
+      "step": 17872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.72424874255873,
+      "learning_rate": 6.28353094421893e-06,
+      "loss": 0.5192,
+      "step": 17873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.937317251231431,
+      "learning_rate": 6.2831501345202815e-06,
+      "loss": 0.6229,
+      "step": 17874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7585717320599648,
+      "learning_rate": 6.282769316853617e-06,
+      "loss": 0.5137,
+      "step": 17875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4692657204446287,
+      "learning_rate": 6.282388491221305e-06,
+      "loss": 0.28,
+      "step": 17876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9491958731491859,
+      "learning_rate": 6.282007657625707e-06,
+      "loss": 0.7274,
+      "step": 17877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1608668119242638,
+      "learning_rate": 6.281626816069191e-06,
+      "loss": 0.5784,
+      "step": 17878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.016450853235046,
+      "learning_rate": 6.281245966554118e-06,
+      "loss": 0.4257,
+      "step": 17879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0432375918825727,
+      "learning_rate": 6.280865109082856e-06,
+      "loss": 0.7795,
+      "step": 17880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.809654011568857,
+      "learning_rate": 6.280484243657769e-06,
+      "loss": 0.3385,
+      "step": 17881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.055623729609152,
+      "learning_rate": 6.280103370281221e-06,
+      "loss": 0.5318,
+      "step": 17882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4161702341320739,
+      "learning_rate": 6.27972248895558e-06,
+      "loss": 0.2027,
+      "step": 17883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7289112220275513,
+      "learning_rate": 6.2793415996832074e-06,
+      "loss": 0.345,
+      "step": 17884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1736197427436736,
+      "learning_rate": 6.278960702466472e-06,
+      "loss": 0.3949,
+      "step": 17885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4359662513843168,
+      "learning_rate": 6.278579797307737e-06,
+      "loss": 0.1472,
+      "step": 17886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4444991787475359,
+      "learning_rate": 6.278198884209369e-06,
+      "loss": 0.307,
+      "step": 17887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0334179349760917,
+      "learning_rate": 6.277817963173731e-06,
+      "loss": 0.3219,
+      "step": 17888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7687624088431695,
+      "learning_rate": 6.277437034203191e-06,
+      "loss": 0.4505,
+      "step": 17889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.849891822300506,
+      "learning_rate": 6.277056097300114e-06,
+      "loss": 0.492,
+      "step": 17890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7128220808696968,
+      "learning_rate": 6.276675152466863e-06,
+      "loss": 0.3406,
+      "step": 17891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9465235651653916,
+      "learning_rate": 6.276294199705808e-06,
+      "loss": 0.4624,
+      "step": 17892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.508964474058246,
+      "learning_rate": 6.27591323901931e-06,
+      "loss": 0.1416,
+      "step": 17893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6573836927041408,
+      "learning_rate": 6.2755322704097385e-06,
+      "loss": 0.2419,
+      "step": 17894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0064590388760588,
+      "learning_rate": 6.2751512938794556e-06,
+      "loss": 0.1538,
+      "step": 17895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5115994577629528,
+      "learning_rate": 6.274770309430831e-06,
+      "loss": 0.1453,
+      "step": 17896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6361549619569105,
+      "learning_rate": 6.2743893170662275e-06,
+      "loss": 0.3346,
+      "step": 17897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0853009091933634,
+      "learning_rate": 6.274008316788013e-06,
+      "loss": 0.3507,
+      "step": 17898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7353617421518726,
+      "learning_rate": 6.273627308598551e-06,
+      "loss": 0.4637,
+      "step": 17899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5303753865746357,
+      "learning_rate": 6.27324629250021e-06,
+      "loss": 0.3197,
+      "step": 17900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5592726182677874,
+      "learning_rate": 6.272865268495354e-06,
+      "loss": 0.3407,
+      "step": 17901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8304448935777222,
+      "learning_rate": 6.272484236586351e-06,
+      "loss": 0.5745,
+      "step": 17902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3909225528016984,
+      "learning_rate": 6.272103196775565e-06,
+      "loss": 0.1949,
+      "step": 17903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6183967906430399,
+      "learning_rate": 6.271722149065362e-06,
+      "loss": 0.1936,
+      "step": 17904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.431893689322084,
+      "learning_rate": 6.271341093458112e-06,
+      "loss": 0.4742,
+      "step": 17905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9385612181043632,
+      "learning_rate": 6.270960029956177e-06,
+      "loss": 0.378,
+      "step": 17906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7982494902252741,
+      "learning_rate": 6.270578958561926e-06,
+      "loss": 0.3363,
+      "step": 17907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2769263486718347,
+      "learning_rate": 6.270197879277722e-06,
+      "loss": 0.324,
+      "step": 17908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.666095217450392,
+      "learning_rate": 6.269816792105936e-06,
+      "loss": 0.3331,
+      "step": 17909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7932741075509617,
+      "learning_rate": 6.269435697048931e-06,
+      "loss": 0.4681,
+      "step": 17910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7707319543840756,
+      "learning_rate": 6.269054594109075e-06,
+      "loss": 0.3399,
+      "step": 17911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9384489273676602,
+      "learning_rate": 6.268673483288734e-06,
+      "loss": 0.4446,
+      "step": 17912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7724778944313058,
+      "learning_rate": 6.268292364590275e-06,
+      "loss": 0.4365,
+      "step": 17913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9068371775204145,
+      "learning_rate": 6.267911238016063e-06,
+      "loss": 0.3833,
+      "step": 17914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6601174832473966,
+      "learning_rate": 6.2675301035684685e-06,
+      "loss": 0.2322,
+      "step": 17915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.406151508484974,
+      "learning_rate": 6.267148961249854e-06,
+      "loss": 0.1917,
+      "step": 17916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5897631050669905,
+      "learning_rate": 6.266767811062587e-06,
+      "loss": 0.4638,
+      "step": 17917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7367210286923869,
+      "learning_rate": 6.2663866530090374e-06,
+      "loss": 0.3981,
+      "step": 17918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4293806809936707,
+      "learning_rate": 6.26600548709157e-06,
+      "loss": 0.2799,
+      "step": 17919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7409515743956898,
+      "learning_rate": 6.265624313312553e-06,
+      "loss": 0.458,
+      "step": 17920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1122373002884152,
+      "learning_rate": 6.26524313167435e-06,
+      "loss": 0.1769,
+      "step": 17921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5607040144732636,
+      "learning_rate": 6.264861942179332e-06,
+      "loss": 0.2644,
+      "step": 17922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7350013027371782,
+      "learning_rate": 6.264480744829864e-06,
+      "loss": 0.3909,
+      "step": 17923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3972390606519656,
+      "learning_rate": 6.264099539628312e-06,
+      "loss": 0.2129,
+      "step": 17924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3151519739947806,
+      "learning_rate": 6.263718326577046e-06,
+      "loss": 0.2113,
+      "step": 17925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4685222418704464,
+      "learning_rate": 6.263337105678432e-06,
+      "loss": 0.1613,
+      "step": 17926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4821474552596166,
+      "learning_rate": 6.262955876934838e-06,
+      "loss": 0.1522,
+      "step": 17927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.041366298794108,
+      "learning_rate": 6.262574640348629e-06,
+      "loss": 0.3364,
+      "step": 17928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8969395147347283,
+      "learning_rate": 6.262193395922175e-06,
+      "loss": 0.5735,
+      "step": 17929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6778941206103701,
+      "learning_rate": 6.26181214365784e-06,
+      "loss": 0.336,
+      "step": 17930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8844248804604398,
+      "learning_rate": 6.2614308835579976e-06,
+      "loss": 0.5409,
+      "step": 17931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.511086412999132,
+      "learning_rate": 6.261049615625008e-06,
+      "loss": 0.3132,
+      "step": 17932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.621722405114646,
+      "learning_rate": 6.260668339861244e-06,
+      "loss": 0.2201,
+      "step": 17933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.562423286939008,
+      "learning_rate": 6.260287056269072e-06,
+      "loss": 0.3072,
+      "step": 17934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6687575625844746,
+      "learning_rate": 6.2599057648508574e-06,
+      "loss": 0.5636,
+      "step": 17935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6225469406862034,
+      "learning_rate": 6.259524465608972e-06,
+      "loss": 0.2794,
+      "step": 17936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7828030424677936,
+      "learning_rate": 6.25914315854578e-06,
+      "loss": 0.4256,
+      "step": 17937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1856261671469759,
+      "learning_rate": 6.25876184366365e-06,
+      "loss": 0.1792,
+      "step": 17938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7013356450782602,
+      "learning_rate": 6.258380520964951e-06,
+      "loss": 0.3684,
+      "step": 17939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4934155044763826,
+      "learning_rate": 6.257999190452049e-06,
+      "loss": 0.2635,
+      "step": 17940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6781851503755598,
+      "learning_rate": 6.257617852127316e-06,
+      "loss": 0.2886,
+      "step": 17941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1761922181446973,
+      "learning_rate": 6.257236505993114e-06,
+      "loss": 0.1383,
+      "step": 17942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7523324273024672,
+      "learning_rate": 6.256855152051815e-06,
+      "loss": 0.5831,
+      "step": 17943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0032376927174234,
+      "learning_rate": 6.256473790305788e-06,
+      "loss": 0.2778,
+      "step": 17944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9431975714259444,
+      "learning_rate": 6.256092420757399e-06,
+      "loss": 0.3689,
+      "step": 17945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8038329375954278,
+      "learning_rate": 6.2557110434090166e-06,
+      "loss": 0.3726,
+      "step": 17946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5535184164312317,
+      "learning_rate": 6.255329658263009e-06,
+      "loss": 0.3686,
+      "step": 17947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7479258937633826,
+      "learning_rate": 6.254948265321744e-06,
+      "loss": 0.2356,
+      "step": 17948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6340364381765833,
+      "learning_rate": 6.254566864587592e-06,
+      "loss": 0.304,
+      "step": 17949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6828848481454344,
+      "learning_rate": 6.2541854560629176e-06,
+      "loss": 0.1808,
+      "step": 17950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4251804366190792,
+      "learning_rate": 6.253804039750095e-06,
+      "loss": 0.3103,
+      "step": 17951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2363813102428027,
+      "learning_rate": 6.253422615651487e-06,
+      "loss": 0.5547,
+      "step": 17952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.566678725062627,
+      "learning_rate": 6.253041183769466e-06,
+      "loss": 0.3614,
+      "step": 17953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9819624420119044,
+      "learning_rate": 6.2526597441063975e-06,
+      "loss": 0.214,
+      "step": 17954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5134168890022786,
+      "learning_rate": 6.2522782966646535e-06,
+      "loss": 0.275,
+      "step": 17955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9022357604533737,
+      "learning_rate": 6.2518968414466e-06,
+      "loss": 0.3952,
+      "step": 17956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.770295005671497,
+      "learning_rate": 6.251515378454606e-06,
+      "loss": 0.4867,
+      "step": 17957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.184661776792435,
+      "learning_rate": 6.251133907691043e-06,
+      "loss": 0.1824,
+      "step": 17958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0749984181064876,
+      "learning_rate": 6.2507524291582745e-06,
+      "loss": 0.1969,
+      "step": 17959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4763160555408203,
+      "learning_rate": 6.250370942858675e-06,
+      "loss": 0.1774,
+      "step": 17960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1966503173949636,
+      "learning_rate": 6.24998944879461e-06,
+      "loss": 0.1615,
+      "step": 17961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.998691652435195,
+      "learning_rate": 6.24960794696845e-06,
+      "loss": 0.4299,
+      "step": 17962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5686190332023673,
+      "learning_rate": 6.249226437382563e-06,
+      "loss": 0.3559,
+      "step": 17963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9340165417761634,
+      "learning_rate": 6.24884492003932e-06,
+      "loss": 0.4926,
+      "step": 17964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.433318955021043,
+      "learning_rate": 6.248463394941088e-06,
+      "loss": 0.2461,
+      "step": 17965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4522867758965934,
+      "learning_rate": 6.248081862090236e-06,
+      "loss": 0.2367,
+      "step": 17966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.44012102861058,
+      "learning_rate": 6.247700321489135e-06,
+      "loss": 0.39,
+      "step": 17967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4170226994915935,
+      "learning_rate": 6.247318773140153e-06,
+      "loss": 0.2831,
+      "step": 17968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.404910407185797,
+      "learning_rate": 6.24693721704566e-06,
+      "loss": 0.2174,
+      "step": 17969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.108675996321914,
+      "learning_rate": 6.246555653208025e-06,
+      "loss": 0.1885,
+      "step": 17970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.308387545095985,
+      "learning_rate": 6.246174081629618e-06,
+      "loss": 0.1885,
+      "step": 17971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.715731457817369,
+      "learning_rate": 6.245792502312806e-06,
+      "loss": 0.5884,
+      "step": 17972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3232617563299842,
+      "learning_rate": 6.245410915259963e-06,
+      "loss": 0.1516,
+      "step": 17973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9950846697345277,
+      "learning_rate": 6.245029320473455e-06,
+      "loss": 0.2205,
+      "step": 17974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5730022245047008,
+      "learning_rate": 6.244647717955653e-06,
+      "loss": 0.2547,
+      "step": 17975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5800937149636767,
+      "learning_rate": 6.244266107708925e-06,
+      "loss": 0.155,
+      "step": 17976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6940263184942883,
+      "learning_rate": 6.2438844897356435e-06,
+      "loss": 0.3224,
+      "step": 17977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0118931963199334,
+      "learning_rate": 6.243502864038176e-06,
+      "loss": 0.563,
+      "step": 17978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8995075899059366,
+      "learning_rate": 6.243121230618891e-06,
+      "loss": 0.3579,
+      "step": 17979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6514094779872468,
+      "learning_rate": 6.242739589480163e-06,
+      "loss": 0.1645,
+      "step": 17980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8162741869427168,
+      "learning_rate": 6.242357940624358e-06,
+      "loss": 0.23,
+      "step": 17981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.13356431698411,
+      "learning_rate": 6.241976284053848e-06,
+      "loss": 0.3518,
+      "step": 17982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3281851754861316,
+      "learning_rate": 6.241594619771001e-06,
+      "loss": 0.2764,
+      "step": 17983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6939778326029864,
+      "learning_rate": 6.241212947778189e-06,
+      "loss": 0.1527,
+      "step": 17984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.156960011977321,
+      "learning_rate": 6.2408312680777804e-06,
+      "loss": 0.2485,
+      "step": 17985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.467788625144935,
+      "learning_rate": 6.240449580672148e-06,
+      "loss": 0.3269,
+      "step": 17986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8197366597713407,
+      "learning_rate": 6.240067885563657e-06,
+      "loss": 0.6405,
+      "step": 17987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4485289430610409,
+      "learning_rate": 6.2396861827546824e-06,
+      "loss": 0.3373,
+      "step": 17988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7159216193577758,
+      "learning_rate": 6.2393044722475935e-06,
+      "loss": 0.5393,
+      "step": 17989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2846571814193974,
+      "learning_rate": 6.23892275404476e-06,
+      "loss": 0.2645,
+      "step": 17990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0958344986294897,
+      "learning_rate": 6.23854102814855e-06,
+      "loss": 0.4582,
+      "step": 17991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3727962284113748,
+      "learning_rate": 6.238159294561338e-06,
+      "loss": 0.1736,
+      "step": 17992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.093341397868315,
+      "learning_rate": 6.2377775532854925e-06,
+      "loss": 0.3964,
+      "step": 17993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4266211870537433,
+      "learning_rate": 6.237395804323383e-06,
+      "loss": 0.2014,
+      "step": 17994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7653907835740394,
+      "learning_rate": 6.237014047677384e-06,
+      "loss": 0.1741,
+      "step": 17995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.713467498645749,
+      "learning_rate": 6.236632283349859e-06,
+      "loss": 0.2719,
+      "step": 17996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9946349686678737,
+      "learning_rate": 6.236250511343186e-06,
+      "loss": 0.4411,
+      "step": 17997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0180119715487685,
+      "learning_rate": 6.235868731659731e-06,
+      "loss": 0.3826,
+      "step": 17998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5078827700769308,
+      "learning_rate": 6.235486944301868e-06,
+      "loss": 0.2464,
+      "step": 17999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6948433503763891,
+      "learning_rate": 6.235105149271966e-06,
+      "loss": 0.2803,
+      "step": 18000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5580831135790145,
+      "learning_rate": 6.234723346572395e-06,
+      "loss": 0.3355,
+      "step": 18001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9154849391719568,
+      "learning_rate": 6.234341536205527e-06,
+      "loss": 0.4096,
+      "step": 18002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.036507375830713,
+      "learning_rate": 6.233959718173732e-06,
+      "loss": 0.4844,
+      "step": 18003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0294187360490965,
+      "learning_rate": 6.233577892479383e-06,
+      "loss": 0.2972,
+      "step": 18004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9880108508924696,
+      "learning_rate": 6.233196059124849e-06,
+      "loss": 0.2669,
+      "step": 18005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7556264211473047,
+      "learning_rate": 6.232814218112504e-06,
+      "loss": 0.3595,
+      "step": 18006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6411900296661885,
+      "learning_rate": 6.232432369444715e-06,
+      "loss": 0.3508,
+      "step": 18007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1382197350659653,
+      "learning_rate": 6.232050513123856e-06,
+      "loss": 0.2794,
+      "step": 18008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5257233343544434,
+      "learning_rate": 6.2316686491522974e-06,
+      "loss": 0.294,
+      "step": 18009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9444867108686188,
+      "learning_rate": 6.231286777532411e-06,
+      "loss": 0.4607,
+      "step": 18010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6335839999622637,
+      "learning_rate": 6.230904898266567e-06,
+      "loss": 0.1989,
+      "step": 18011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9292748006054854,
+      "learning_rate": 6.230523011357137e-06,
+      "loss": 0.4929,
+      "step": 18012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.242538643964366,
+      "learning_rate": 6.230141116806494e-06,
+      "loss": 0.4331,
+      "step": 18013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9085293493909916,
+      "learning_rate": 6.229759214617007e-06,
+      "loss": 0.2346,
+      "step": 18014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.314270376080715,
+      "learning_rate": 6.229377304791049e-06,
+      "loss": 0.5709,
+      "step": 18015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.199169136164223,
+      "learning_rate": 6.228995387330992e-06,
+      "loss": 0.1514,
+      "step": 18016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6024459867628735,
+      "learning_rate": 6.228613462239206e-06,
+      "loss": 0.3707,
+      "step": 18017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3400535085655867,
+      "learning_rate": 6.228231529518064e-06,
+      "loss": 0.2166,
+      "step": 18018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5898199548276215,
+      "learning_rate": 6.227849589169938e-06,
+      "loss": 0.3366,
+      "step": 18019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6694666905788966,
+      "learning_rate": 6.227467641197198e-06,
+      "loss": 0.433,
+      "step": 18020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5749964248525823,
+      "learning_rate": 6.227085685602218e-06,
+      "loss": 0.2917,
+      "step": 18021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6030416264334901,
+      "learning_rate": 6.226703722387368e-06,
+      "loss": 0.2545,
+      "step": 18022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.494988042833839,
+      "learning_rate": 6.2263217515550215e-06,
+      "loss": 0.2459,
+      "step": 18023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2494020730702056,
+      "learning_rate": 6.225939773107549e-06,
+      "loss": 0.2152,
+      "step": 18024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3038617393343028,
+      "learning_rate": 6.225557787047322e-06,
+      "loss": 0.2662,
+      "step": 18025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7818842478137178,
+      "learning_rate": 6.225175793376714e-06,
+      "loss": 0.4754,
+      "step": 18026
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5797994860715883,
+      "learning_rate": 6.224793792098097e-06,
+      "loss": 0.243,
+      "step": 18027
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5010210724326098,
+      "learning_rate": 6.224411783213843e-06,
+      "loss": 0.3489,
+      "step": 18028
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5283431371379947,
+      "learning_rate": 6.224029766726323e-06,
+      "loss": 0.3485,
+      "step": 18029
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6168683469670115,
+      "learning_rate": 6.223647742637911e-06,
+      "loss": 0.2946,
+      "step": 18030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.379460733901172,
+      "learning_rate": 6.223265710950977e-06,
+      "loss": 0.2241,
+      "step": 18031
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4469355105029298,
+      "learning_rate": 6.222883671667897e-06,
+      "loss": 0.2884,
+      "step": 18032
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5720500348053843,
+      "learning_rate": 6.222501624791041e-06,
+      "loss": 0.3677,
+      "step": 18033
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4928906441762948,
+      "learning_rate": 6.222119570322779e-06,
+      "loss": 0.232,
+      "step": 18034
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2806737842688234,
+      "learning_rate": 6.221737508265487e-06,
+      "loss": 0.2765,
+      "step": 18035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9064886147038826,
+      "learning_rate": 6.221355438621537e-06,
+      "loss": 0.4836,
+      "step": 18036
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5969533469955464,
+      "learning_rate": 6.220973361393301e-06,
+      "loss": 0.3985,
+      "step": 18037
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9878977344297849,
+      "learning_rate": 6.22059127658315e-06,
+      "loss": 0.1261,
+      "step": 18038
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3749186778945814,
+      "learning_rate": 6.220209184193459e-06,
+      "loss": 0.2799,
+      "step": 18039
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.355490681583509,
+      "learning_rate": 6.2198270842266005e-06,
+      "loss": 0.1003,
+      "step": 18040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7401248092496437,
+      "learning_rate": 6.219444976684946e-06,
+      "loss": 0.461,
+      "step": 18041
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7415728703670426,
+      "learning_rate": 6.219062861570871e-06,
+      "loss": 0.3784,
+      "step": 18042
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.668647777537507,
+      "learning_rate": 6.218680738886743e-06,
+      "loss": 0.3907,
+      "step": 18043
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4987641945994747,
+      "learning_rate": 6.21829860863494e-06,
+      "loss": 0.4055,
+      "step": 18044
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4661654905240273,
+      "learning_rate": 6.2179164708178315e-06,
+      "loss": 0.3645,
+      "step": 18045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.872101999889235,
+      "learning_rate": 6.217534325437794e-06,
+      "loss": 0.3982,
+      "step": 18046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3669395112877554,
+      "learning_rate": 6.2171521724971974e-06,
+      "loss": 0.1704,
+      "step": 18047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9823792589239169,
+      "learning_rate": 6.2167700119984165e-06,
+      "loss": 0.378,
+      "step": 18048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9658968550992366,
+      "learning_rate": 6.216387843943823e-06,
+      "loss": 0.2786,
+      "step": 18049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.603601160683709,
+      "learning_rate": 6.216005668335793e-06,
+      "loss": 0.2178,
+      "step": 18050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6010410878952492,
+      "learning_rate": 6.215623485176695e-06,
+      "loss": 0.3558,
+      "step": 18051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3432973831306605,
+      "learning_rate": 6.215241294468907e-06,
+      "loss": 0.349,
+      "step": 18052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.186995121049083,
+      "learning_rate": 6.2148590962147985e-06,
+      "loss": 0.3963,
+      "step": 18053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5004592530223004,
+      "learning_rate": 6.2144768904167455e-06,
+      "loss": 0.3725,
+      "step": 18054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9008938161410747,
+      "learning_rate": 6.21409467707712e-06,
+      "loss": 0.5184,
+      "step": 18055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4649363374059758,
+      "learning_rate": 6.213712456198295e-06,
+      "loss": 0.3337,
+      "step": 18056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5641677801005114,
+      "learning_rate": 6.2133302277826465e-06,
+      "loss": 0.1744,
+      "step": 18057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5029736215615788,
+      "learning_rate": 6.212947991832546e-06,
+      "loss": 0.2123,
+      "step": 18058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7651826549161391,
+      "learning_rate": 6.212565748350366e-06,
+      "loss": 0.2286,
+      "step": 18059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7471359799252864,
+      "learning_rate": 6.212183497338482e-06,
+      "loss": 0.2503,
+      "step": 18060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6247632697872896,
+      "learning_rate": 6.211801238799269e-06,
+      "loss": 0.3116,
+      "step": 18061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2476200991882958,
+      "learning_rate": 6.211418972735096e-06,
+      "loss": 0.2011,
+      "step": 18062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6313022286359304,
+      "learning_rate": 6.2110366991483415e-06,
+      "loss": 0.1718,
+      "step": 18063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5995924253578804,
+      "learning_rate": 6.210654418041378e-06,
+      "loss": 0.4567,
+      "step": 18064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3841153919675755,
+      "learning_rate": 6.210272129416578e-06,
+      "loss": 0.1787,
+      "step": 18065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.55953286734519,
+      "learning_rate": 6.209889833276317e-06,
+      "loss": 0.3421,
+      "step": 18066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.274888627501611,
+      "learning_rate": 6.209507529622967e-06,
+      "loss": 0.3909,
+      "step": 18067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5492754647381437,
+      "learning_rate": 6.209125218458905e-06,
+      "loss": 0.3023,
+      "step": 18068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.786560849917472,
+      "learning_rate": 6.208742899786502e-06,
+      "loss": 0.2319,
+      "step": 18069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.651869157852251,
+      "learning_rate": 6.208360573608135e-06,
+      "loss": 0.5261,
+      "step": 18070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7129189065927706,
+      "learning_rate": 6.207978239926174e-06,
+      "loss": 0.3527,
+      "step": 18071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.309858217904342,
+      "learning_rate": 6.207595898742998e-06,
+      "loss": 0.5175,
+      "step": 18072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.316729753458924,
+      "learning_rate": 6.207213550060976e-06,
+      "loss": 0.189,
+      "step": 18073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.664338939611306,
+      "learning_rate": 6.2068311938824885e-06,
+      "loss": 0.2716,
+      "step": 18074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6202071022941813,
+      "learning_rate": 6.206448830209906e-06,
+      "loss": 0.1861,
+      "step": 18075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1621478242684364,
+      "learning_rate": 6.2060664590456036e-06,
+      "loss": 0.1984,
+      "step": 18076
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.545838239778545,
+      "learning_rate": 6.205684080391955e-06,
+      "loss": 0.2866,
+      "step": 18077
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5627708343385776,
+      "learning_rate": 6.205301694251334e-06,
+      "loss": 0.4664,
+      "step": 18078
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.740033276173961,
+      "learning_rate": 6.204919300626119e-06,
+      "loss": 0.3382,
+      "step": 18079
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.672667251026907,
+      "learning_rate": 6.2045368995186795e-06,
+      "loss": 0.3089,
+      "step": 18080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.728474434399529,
+      "learning_rate": 6.204154490931394e-06,
+      "loss": 0.2482,
+      "step": 18081
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.063197444266751,
+      "learning_rate": 6.203772074866634e-06,
+      "loss": 0.5303,
+      "step": 18082
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4915860746587832,
+      "learning_rate": 6.203389651326779e-06,
+      "loss": 0.1894,
+      "step": 18083
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6306499049725467,
+      "learning_rate": 6.203007220314198e-06,
+      "loss": 0.4157,
+      "step": 18084
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2011962019775195,
+      "learning_rate": 6.202624781831269e-06,
+      "loss": 0.1886,
+      "step": 18085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8561983541246247,
+      "learning_rate": 6.202242335880367e-06,
+      "loss": 0.2182,
+      "step": 18086
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.871112827371012,
+      "learning_rate": 6.2018598824638655e-06,
+      "loss": 0.4901,
+      "step": 18087
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7704065444356465,
+      "learning_rate": 6.20147742158414e-06,
+      "loss": 0.2151,
+      "step": 18088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3309751390292746,
+      "learning_rate": 6.201094953243565e-06,
+      "loss": 0.1597,
+      "step": 18089
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7594495525453417,
+      "learning_rate": 6.200712477444517e-06,
+      "loss": 0.5335,
+      "step": 18090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5380113371665596,
+      "learning_rate": 6.20032999418937e-06,
+      "loss": 0.352,
+      "step": 18091
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8264402222963732,
+      "learning_rate": 6.199947503480499e-06,
+      "loss": 0.3361,
+      "step": 18092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.14552218918982,
+      "learning_rate": 6.199565005320279e-06,
+      "loss": 0.4343,
+      "step": 18093
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6896728108188936,
+      "learning_rate": 6.1991824997110876e-06,
+      "loss": 0.3736,
+      "step": 18094
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7775934463160579,
+      "learning_rate": 6.1987999866552965e-06,
+      "loss": 0.3118,
+      "step": 18095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5320504684837948,
+      "learning_rate": 6.1984174661552834e-06,
+      "loss": 0.2881,
+      "step": 18096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6998643227325503,
+      "learning_rate": 6.198034938213421e-06,
+      "loss": 0.2481,
+      "step": 18097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4679918259963483,
+      "learning_rate": 6.197652402832089e-06,
+      "loss": 0.2191,
+      "step": 18098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0663385506340646,
+      "learning_rate": 6.1972698600136596e-06,
+      "loss": 0.5645,
+      "step": 18099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.925227439591409,
+      "learning_rate": 6.196887309760509e-06,
+      "loss": 0.5063,
+      "step": 18100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7602825400553912,
+      "learning_rate": 6.196504752075012e-06,
+      "loss": 0.3271,
+      "step": 18101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.459194128694942,
+      "learning_rate": 6.196122186959546e-06,
+      "loss": 0.35,
+      "step": 18102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5609841384640848,
+      "learning_rate": 6.195739614416485e-06,
+      "loss": 0.3368,
+      "step": 18103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1602724415300587,
+      "learning_rate": 6.195357034448205e-06,
+      "loss": 0.1918,
+      "step": 18104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1334571178635637,
+      "learning_rate": 6.194974447057084e-06,
+      "loss": 0.4103,
+      "step": 18105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6144149436100035,
+      "learning_rate": 6.194591852245493e-06,
+      "loss": 0.2357,
+      "step": 18106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8156195423812387,
+      "learning_rate": 6.194209250015813e-06,
+      "loss": 0.2063,
+      "step": 18107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2613381204359682,
+      "learning_rate": 6.193826640370416e-06,
+      "loss": 0.3088,
+      "step": 18108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.572179737469976,
+      "learning_rate": 6.19344402331168e-06,
+      "loss": 0.3592,
+      "step": 18109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7839645272711364,
+      "learning_rate": 6.1930613988419795e-06,
+      "loss": 0.1804,
+      "step": 18110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.238856213716645,
+      "learning_rate": 6.192678766963691e-06,
+      "loss": 0.2192,
+      "step": 18111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.250868855946085,
+      "learning_rate": 6.1922961276791925e-06,
+      "loss": 0.3043,
+      "step": 18112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8849739394422822,
+      "learning_rate": 6.191913480990858e-06,
+      "loss": 0.3096,
+      "step": 18113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3452482188216015,
+      "learning_rate": 6.191530826901063e-06,
+      "loss": 0.1772,
+      "step": 18114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6570831893197926,
+      "learning_rate": 6.191148165412185e-06,
+      "loss": 0.361,
+      "step": 18115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7431537061538582,
+      "learning_rate": 6.1907654965266e-06,
+      "loss": 0.4674,
+      "step": 18116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8556136907886236,
+      "learning_rate": 6.190382820246684e-06,
+      "loss": 0.4749,
+      "step": 18117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.050128124321203,
+      "learning_rate": 6.190000136574814e-06,
+      "loss": 0.3862,
+      "step": 18118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.741770315898162,
+      "learning_rate": 6.189617445513366e-06,
+      "loss": 0.3899,
+      "step": 18119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5798362221405022,
+      "learning_rate": 6.189234747064717e-06,
+      "loss": 0.2358,
+      "step": 18120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3623207474175523,
+      "learning_rate": 6.188852041231241e-06,
+      "loss": 0.2571,
+      "step": 18121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5169666609833308,
+      "learning_rate": 6.188469328015317e-06,
+      "loss": 0.3709,
+      "step": 18122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.780612406891986,
+      "learning_rate": 6.1880866074193214e-06,
+      "loss": 0.3337,
+      "step": 18123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.723795109961496,
+      "learning_rate": 6.187703879445629e-06,
+      "loss": 0.4453,
+      "step": 18124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3663522362325218,
+      "learning_rate": 6.187321144096618e-06,
+      "loss": 0.2045,
+      "step": 18125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6754835348945292,
+      "learning_rate": 6.186938401374666e-06,
+      "loss": 0.369,
+      "step": 18126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4055821853003265,
+      "learning_rate": 6.186555651282148e-06,
+      "loss": 0.2055,
+      "step": 18127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1282065475345044,
+      "learning_rate": 6.186172893821439e-06,
+      "loss": 0.5263,
+      "step": 18128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3893510509658562,
+      "learning_rate": 6.1857901289949206e-06,
+      "loss": 0.2422,
+      "step": 18129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6681347480737791,
+      "learning_rate": 6.185407356804967e-06,
+      "loss": 0.3212,
+      "step": 18130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.501432819800289,
+      "learning_rate": 6.185024577253954e-06,
+      "loss": 0.3749,
+      "step": 18131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0966474998825517,
+      "learning_rate": 6.18464179034426e-06,
+      "loss": 0.2437,
+      "step": 18132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0045240985285764,
+      "learning_rate": 6.184258996078262e-06,
+      "loss": 0.2685,
+      "step": 18133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.632613799676576,
+      "learning_rate": 6.183876194458337e-06,
+      "loss": 0.2237,
+      "step": 18134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.315298842823474,
+      "learning_rate": 6.18349338548686e-06,
+      "loss": 0.3321,
+      "step": 18135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8094449511942465,
+      "learning_rate": 6.183110569166213e-06,
+      "loss": 0.2208,
+      "step": 18136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4001995964586043,
+      "learning_rate": 6.182727745498769e-06,
+      "loss": 0.1993,
+      "step": 18137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2159834633558817,
+      "learning_rate": 6.182344914486906e-06,
+      "loss": 0.1773,
+      "step": 18138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5895408440227903,
+      "learning_rate": 6.181962076133002e-06,
+      "loss": 0.1909,
+      "step": 18139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4761610608135054,
+      "learning_rate": 6.181579230439435e-06,
+      "loss": 0.2852,
+      "step": 18140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1637094395591645,
+      "learning_rate": 6.18119637740858e-06,
+      "loss": 0.4024,
+      "step": 18141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8676422156016694,
+      "learning_rate": 6.180813517042817e-06,
+      "loss": 0.2648,
+      "step": 18142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6143576515556366,
+      "learning_rate": 6.180430649344522e-06,
+      "loss": 0.3255,
+      "step": 18143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.418745997521689,
+      "learning_rate": 6.180047774316071e-06,
+      "loss": 0.5529,
+      "step": 18144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5670612587671868,
+      "learning_rate": 6.179664891959846e-06,
+      "loss": 0.2457,
+      "step": 18145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.477194461612425,
+      "learning_rate": 6.179282002278221e-06,
+      "loss": 0.2379,
+      "step": 18146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5183980143395017,
+      "learning_rate": 6.1788991052735746e-06,
+      "loss": 0.1267,
+      "step": 18147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.459308835863904,
+      "learning_rate": 6.1785162009482836e-06,
+      "loss": 0.261,
+      "step": 18148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4998355417695615,
+      "learning_rate": 6.178133289304729e-06,
+      "loss": 0.1308,
+      "step": 18149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6876608144993135,
+      "learning_rate": 6.177750370345283e-06,
+      "loss": 0.3281,
+      "step": 18150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0435466424416648,
+      "learning_rate": 6.1773674440723286e-06,
+      "loss": 0.1501,
+      "step": 18151
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.152203027379241,
+      "learning_rate": 6.176984510488242e-06,
+      "loss": 0.2186,
+      "step": 18152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.627998888344731,
+      "learning_rate": 6.176601569595399e-06,
+      "loss": 0.1726,
+      "step": 18153
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9156321103833711,
+      "learning_rate": 6.176218621396182e-06,
+      "loss": 0.286,
+      "step": 18154
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9295190127558721,
+      "learning_rate": 6.1758356658929645e-06,
+      "loss": 0.3438,
+      "step": 18155
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8346039981647102,
+      "learning_rate": 6.1754527030881275e-06,
+      "loss": 0.354,
+      "step": 18156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9575266198300011,
+      "learning_rate": 6.175069732984046e-06,
+      "loss": 0.1739,
+      "step": 18157
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2692301466469027,
+      "learning_rate": 6.1746867555831016e-06,
+      "loss": 0.2869,
+      "step": 18158
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4983987805213894,
+      "learning_rate": 6.1743037708876704e-06,
+      "loss": 0.3118,
+      "step": 18159
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.844260217901269,
+      "learning_rate": 6.173920778900132e-06,
+      "loss": 0.1924,
+      "step": 18160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.363450310397946,
+      "learning_rate": 6.173537779622863e-06,
+      "loss": 0.2728,
+      "step": 18161
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4973307692172,
+      "learning_rate": 6.173154773058244e-06,
+      "loss": 0.3268,
+      "step": 18162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0783551741168518,
+      "learning_rate": 6.172771759208652e-06,
+      "loss": 0.7676,
+      "step": 18163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.531762729409537,
+      "learning_rate": 6.172388738076465e-06,
+      "loss": 0.2107,
+      "step": 18164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9097588404603023,
+      "learning_rate": 6.172005709664061e-06,
+      "loss": 0.3579,
+      "step": 18165
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2962884883698294,
+      "learning_rate": 6.17162267397382e-06,
+      "loss": 0.3556,
+      "step": 18166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7721680664560253,
+      "learning_rate": 6.171239631008119e-06,
+      "loss": 0.2744,
+      "step": 18167
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4686952489428278,
+      "learning_rate": 6.17085658076934e-06,
+      "loss": 0.2632,
+      "step": 18168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0309881188229655,
+      "learning_rate": 6.170473523259856e-06,
+      "loss": 0.6373,
+      "step": 18169
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.537708955614851,
+      "learning_rate": 6.170090458482051e-06,
+      "loss": 0.3926,
+      "step": 18170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6061166351703808,
+      "learning_rate": 6.169707386438301e-06,
+      "loss": 0.3759,
+      "step": 18171
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1824073920419975,
+      "learning_rate": 6.169324307130986e-06,
+      "loss": 0.1763,
+      "step": 18172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.71936241423321,
+      "learning_rate": 6.1689412205624845e-06,
+      "loss": 0.3448,
+      "step": 18173
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6993005667095624,
+      "learning_rate": 6.168558126735174e-06,
+      "loss": 0.2896,
+      "step": 18174
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9882560121324724,
+      "learning_rate": 6.168175025651435e-06,
+      "loss": 0.2653,
+      "step": 18175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5687466175396234,
+      "learning_rate": 6.167791917313646e-06,
+      "loss": 0.1958,
+      "step": 18176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5389952221315857,
+      "learning_rate": 6.167408801724185e-06,
+      "loss": 0.3084,
+      "step": 18177
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5249185625399233,
+      "learning_rate": 6.167025678885433e-06,
+      "loss": 0.2178,
+      "step": 18178
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.960991854883209,
+      "learning_rate": 6.166642548799767e-06,
+      "loss": 0.5299,
+      "step": 18179
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5487710802750316,
+      "learning_rate": 6.166259411469569e-06,
+      "loss": 0.2363,
+      "step": 18180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3529341578530194,
+      "learning_rate": 6.1658762668972135e-06,
+      "loss": 0.3474,
+      "step": 18181
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6597394206778915,
+      "learning_rate": 6.165493115085085e-06,
+      "loss": 0.3006,
+      "step": 18182
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3413211945510182,
+      "learning_rate": 6.165109956035559e-06,
+      "loss": 0.2084,
+      "step": 18183
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9004104903340382,
+      "learning_rate": 6.164726789751017e-06,
+      "loss": 0.342,
+      "step": 18184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.790232413297307,
+      "learning_rate": 6.164343616233838e-06,
+      "loss": 0.2059,
+      "step": 18185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9199067403326808,
+      "learning_rate": 6.163960435486399e-06,
+      "loss": 0.2588,
+      "step": 18186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7289287956963844,
+      "learning_rate": 6.1635772475110835e-06,
+      "loss": 0.3801,
+      "step": 18187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.106411526426282,
+      "learning_rate": 6.163194052310267e-06,
+      "loss": 0.6556,
+      "step": 18188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6389331972996513,
+      "learning_rate": 6.162810849886332e-06,
+      "loss": 0.394,
+      "step": 18189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4403933478475175,
+      "learning_rate": 6.162427640241656e-06,
+      "loss": 0.2193,
+      "step": 18190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.44147619734178,
+      "learning_rate": 6.162044423378621e-06,
+      "loss": 0.2812,
+      "step": 18191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5944774173768428,
+      "learning_rate": 6.161661199299604e-06,
+      "loss": 0.3416,
+      "step": 18192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.125149030929094,
+      "learning_rate": 6.161277968006987e-06,
+      "loss": 0.555,
+      "step": 18193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3598982472607117,
+      "learning_rate": 6.160894729503148e-06,
+      "loss": 0.3417,
+      "step": 18194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0376022817822923,
+      "learning_rate": 6.160511483790468e-06,
+      "loss": 0.3453,
+      "step": 18195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4000408049580282,
+      "learning_rate": 6.160128230871327e-06,
+      "loss": 0.2718,
+      "step": 18196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8574643957523793,
+      "learning_rate": 6.159744970748103e-06,
+      "loss": 0.4987,
+      "step": 18197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5045700800368844,
+      "learning_rate": 6.159361703423179e-06,
+      "loss": 0.2466,
+      "step": 18198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4624190289750756,
+      "learning_rate": 6.158978428898931e-06,
+      "loss": 0.1988,
+      "step": 18199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8091649930930278,
+      "learning_rate": 6.158595147177744e-06,
+      "loss": 0.6217,
+      "step": 18200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6200178413674342,
+      "learning_rate": 6.1582118582619934e-06,
+      "loss": 0.2926,
+      "step": 18201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9551642319869886,
+      "learning_rate": 6.1578285621540615e-06,
+      "loss": 0.4768,
+      "step": 18202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9426773714831742,
+      "learning_rate": 6.1574452588563285e-06,
+      "loss": 0.3956,
+      "step": 18203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.567849177204681,
+      "learning_rate": 6.157061948371174e-06,
+      "loss": 0.4324,
+      "step": 18204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4439215429943233,
+      "learning_rate": 6.156678630700979e-06,
+      "loss": 0.1218,
+      "step": 18205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.755067487625431,
+      "learning_rate": 6.156295305848123e-06,
+      "loss": 0.3105,
+      "step": 18206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.213553055790297,
+      "learning_rate": 6.155911973814989e-06,
+      "loss": 0.1669,
+      "step": 18207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.365653414397915,
+      "learning_rate": 6.155528634603952e-06,
+      "loss": 0.2778,
+      "step": 18208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.63377880405596,
+      "learning_rate": 6.1551452882173975e-06,
+      "loss": 0.2372,
+      "step": 18209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.309472714057925,
+      "learning_rate": 6.1547619346577026e-06,
+      "loss": 0.1955,
+      "step": 18210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8585259438421116,
+      "learning_rate": 6.154378573927251e-06,
+      "loss": 0.6589,
+      "step": 18211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8126176423912677,
+      "learning_rate": 6.15399520602842e-06,
+      "loss": 0.3019,
+      "step": 18212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.406004553918481,
+      "learning_rate": 6.153611830963593e-06,
+      "loss": 0.2076,
+      "step": 18213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8086188048899203,
+      "learning_rate": 6.153228448735148e-06,
+      "loss": 0.4629,
+      "step": 18214
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7656193750004923,
+      "learning_rate": 6.152845059345468e-06,
+      "loss": 0.4847,
+      "step": 18215
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8469113406851527,
+      "learning_rate": 6.152461662796933e-06,
+      "loss": 0.3868,
+      "step": 18216
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1082832001029868,
+      "learning_rate": 6.1520782590919225e-06,
+      "loss": 0.3376,
+      "step": 18217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3831046248459458,
+      "learning_rate": 6.151694848232819e-06,
+      "loss": 0.1928,
+      "step": 18218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4974700906557636,
+      "learning_rate": 6.151311430222002e-06,
+      "loss": 0.2284,
+      "step": 18219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6354176851338253,
+      "learning_rate": 6.150928005061856e-06,
+      "loss": 0.3459,
+      "step": 18220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1182456240465286,
+      "learning_rate": 6.1505445727547565e-06,
+      "loss": 0.5708,
+      "step": 18221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1932984370111155,
+      "learning_rate": 6.150161133303088e-06,
+      "loss": 0.1635,
+      "step": 18222
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0057648745138246,
+      "learning_rate": 6.149777686709231e-06,
+      "loss": 0.6163,
+      "step": 18223
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3341170017447124,
+      "learning_rate": 6.149394232975567e-06,
+      "loss": 0.2096,
+      "step": 18224
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5399182985279833,
+      "learning_rate": 6.149010772104476e-06,
+      "loss": 0.3381,
+      "step": 18225
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3945917776206118,
+      "learning_rate": 6.14862730409834e-06,
+      "loss": 0.1883,
+      "step": 18226
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4442936424887343,
+      "learning_rate": 6.148243828959539e-06,
+      "loss": 0.2834,
+      "step": 18227
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2498389721505003,
+      "learning_rate": 6.147860346690457e-06,
+      "loss": 0.1761,
+      "step": 18228
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8354513295682362,
+      "learning_rate": 6.147476857293473e-06,
+      "loss": 0.3785,
+      "step": 18229
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6755067916232722,
+      "learning_rate": 6.147093360770968e-06,
+      "loss": 0.2926,
+      "step": 18230
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.953874329829008,
+      "learning_rate": 6.146709857125326e-06,
+      "loss": 0.4818,
+      "step": 18231
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2288462234964788,
+      "learning_rate": 6.146326346358924e-06,
+      "loss": 0.1607,
+      "step": 18232
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.832704204927459,
+      "learning_rate": 6.1459428284741485e-06,
+      "loss": 0.4574,
+      "step": 18233
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7724500545451338,
+      "learning_rate": 6.145559303473378e-06,
+      "loss": 0.2056,
+      "step": 18234
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2286256367925006,
+      "learning_rate": 6.145175771358996e-06,
+      "loss": 0.19,
+      "step": 18235
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3869995289619887,
+      "learning_rate": 6.144792232133381e-06,
+      "loss": 0.1693,
+      "step": 18236
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.012307416100787,
+      "learning_rate": 6.144408685798918e-06,
+      "loss": 0.2879,
+      "step": 18237
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6723503078876358,
+      "learning_rate": 6.144025132357987e-06,
+      "loss": 0.1789,
+      "step": 18238
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7781910103007759,
+      "learning_rate": 6.143641571812971e-06,
+      "loss": 0.1797,
+      "step": 18239
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3003764166933802,
+      "learning_rate": 6.14325800416625e-06,
+      "loss": 0.2208,
+      "step": 18240
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7417489834224418,
+      "learning_rate": 6.142874429420208e-06,
+      "loss": 0.2431,
+      "step": 18241
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4553873514417806,
+      "learning_rate": 6.1424908475772246e-06,
+      "loss": 0.3321,
+      "step": 18242
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4538432840449407,
+      "learning_rate": 6.142107258639684e-06,
+      "loss": 0.3168,
+      "step": 18243
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7078608868995877,
+      "learning_rate": 6.141723662609967e-06,
+      "loss": 0.3166,
+      "step": 18244
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9590703045909696,
+      "learning_rate": 6.141340059490454e-06,
+      "loss": 0.4176,
+      "step": 18245
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.227835997073503,
+      "learning_rate": 6.140956449283531e-06,
+      "loss": 0.3456,
+      "step": 18246
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4053634206501282,
+      "learning_rate": 6.140572831991574e-06,
+      "loss": 0.3635,
+      "step": 18247
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.236746280006418,
+      "learning_rate": 6.140189207616973e-06,
+      "loss": 0.4507,
+      "step": 18248
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2367026137011978,
+      "learning_rate": 6.139805576162106e-06,
+      "loss": 0.166,
+      "step": 18249
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7493235183151294,
+      "learning_rate": 6.139421937629354e-06,
+      "loss": 0.3281,
+      "step": 18250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8808223526619607,
+      "learning_rate": 6.1390382920211026e-06,
+      "loss": 0.3621,
+      "step": 18251
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5134631892908277,
+      "learning_rate": 6.138654639339731e-06,
+      "loss": 0.2886,
+      "step": 18252
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8776209949968514,
+      "learning_rate": 6.138270979587622e-06,
+      "loss": 0.2337,
+      "step": 18253
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5739996038162074,
+      "learning_rate": 6.13788731276716e-06,
+      "loss": 0.2712,
+      "step": 18254
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.660260464258336,
+      "learning_rate": 6.137503638880727e-06,
+      "loss": 0.3173,
+      "step": 18255
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8501356195988221,
+      "learning_rate": 6.137119957930704e-06,
+      "loss": 0.4858,
+      "step": 18256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.426610151792363,
+      "learning_rate": 6.136736269919474e-06,
+      "loss": 0.2601,
+      "step": 18257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.775292938365041,
+      "learning_rate": 6.136352574849421e-06,
+      "loss": 0.4199,
+      "step": 18258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4774679228605392,
+      "learning_rate": 6.135968872722927e-06,
+      "loss": 0.2814,
+      "step": 18259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6071363962702083,
+      "learning_rate": 6.135585163542374e-06,
+      "loss": 0.3146,
+      "step": 18260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3432630943144312,
+      "learning_rate": 6.135201447310145e-06,
+      "loss": 0.1286,
+      "step": 18261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7850768324133464,
+      "learning_rate": 6.134817724028623e-06,
+      "loss": 0.3332,
+      "step": 18262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4442287222148475,
+      "learning_rate": 6.1344339937001905e-06,
+      "loss": 0.2283,
+      "step": 18263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7982801996636195,
+      "learning_rate": 6.134050256327231e-06,
+      "loss": 0.2943,
+      "step": 18264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4896303155021071,
+      "learning_rate": 6.133666511912126e-06,
+      "loss": 0.2392,
+      "step": 18265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.874780145312231,
+      "learning_rate": 6.13328276045726e-06,
+      "loss": 0.5155,
+      "step": 18266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9441134231132922,
+      "learning_rate": 6.132899001965016e-06,
+      "loss": 0.5986,
+      "step": 18267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8974275217927612,
+      "learning_rate": 6.1325152364377745e-06,
+      "loss": 0.2587,
+      "step": 18268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3943193418918116,
+      "learning_rate": 6.132131463877923e-06,
+      "loss": 0.3168,
+      "step": 18269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2756166827366646,
+      "learning_rate": 6.1317476842878394e-06,
+      "loss": 0.1621,
+      "step": 18270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.651443795491912,
+      "learning_rate": 6.131363897669911e-06,
+      "loss": 0.2394,
+      "step": 18271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.230028414274487,
+      "learning_rate": 6.130980104026521e-06,
+      "loss": 0.1494,
+      "step": 18272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7708033269322088,
+      "learning_rate": 6.13059630336005e-06,
+      "loss": 0.4036,
+      "step": 18273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5883580841754188,
+      "learning_rate": 6.1302124956728824e-06,
+      "loss": 0.2405,
+      "step": 18274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6149396662100945,
+      "learning_rate": 6.129828680967402e-06,
+      "loss": 0.3562,
+      "step": 18275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1257207643875575,
+      "learning_rate": 6.129444859245992e-06,
+      "loss": 0.5891,
+      "step": 18276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5377493910968911,
+      "learning_rate": 6.129061030511037e-06,
+      "loss": 0.2894,
+      "step": 18277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5958985520669655,
+      "learning_rate": 6.128677194764917e-06,
+      "loss": 0.226,
+      "step": 18278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.425788802296742,
+      "learning_rate": 6.128293352010019e-06,
+      "loss": 0.241,
+      "step": 18279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.015475496765385,
+      "learning_rate": 6.127909502248725e-06,
+      "loss": 0.4309,
+      "step": 18280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.077368804835064,
+      "learning_rate": 6.127525645483419e-06,
+      "loss": 0.3276,
+      "step": 18281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3132142978601185,
+      "learning_rate": 6.127141781716483e-06,
+      "loss": 0.148,
+      "step": 18282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8704460352916632,
+      "learning_rate": 6.126757910950305e-06,
+      "loss": 0.4355,
+      "step": 18283
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6656182721988009,
+      "learning_rate": 6.126374033187266e-06,
+      "loss": 0.2778,
+      "step": 18284
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9803942303303237,
+      "learning_rate": 6.125990148429747e-06,
+      "loss": 0.2492,
+      "step": 18285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.563575064840142,
+      "learning_rate": 6.125606256680137e-06,
+      "loss": 0.2086,
+      "step": 18286
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6385839936580795,
+      "learning_rate": 6.1252223579408155e-06,
+      "loss": 0.2152,
+      "step": 18287
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3553253064947945,
+      "learning_rate": 6.124838452214169e-06,
+      "loss": 0.2487,
+      "step": 18288
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6315249335598863,
+      "learning_rate": 6.12445453950258e-06,
+      "loss": 0.5109,
+      "step": 18289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.631611030644479,
+      "learning_rate": 6.124070619808435e-06,
+      "loss": 0.2911,
+      "step": 18290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4883883930797748,
+      "learning_rate": 6.123686693134115e-06,
+      "loss": 0.1689,
+      "step": 18291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0973534217222487,
+      "learning_rate": 6.123302759482006e-06,
+      "loss": 0.2447,
+      "step": 18292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2411647250788986,
+      "learning_rate": 6.122918818854492e-06,
+      "loss": 0.1937,
+      "step": 18293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0314695092949657,
+      "learning_rate": 6.122534871253954e-06,
+      "loss": 0.5518,
+      "step": 18294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.077339016556413,
+      "learning_rate": 6.122150916682782e-06,
+      "loss": 0.5615,
+      "step": 18295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6702215450062992,
+      "learning_rate": 6.121766955143355e-06,
+      "loss": 0.493,
+      "step": 18296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8856177622334092,
+      "learning_rate": 6.121382986638063e-06,
+      "loss": 0.311,
+      "step": 18297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9852939790630084,
+      "learning_rate": 6.120999011169283e-06,
+      "loss": 0.36,
+      "step": 18298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4089044789927887,
+      "learning_rate": 6.120615028739405e-06,
+      "loss": 0.2118,
+      "step": 18299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6854614974375377,
+      "learning_rate": 6.1202310393508105e-06,
+      "loss": 0.2367,
+      "step": 18300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.741276199050615,
+      "learning_rate": 6.119847043005887e-06,
+      "loss": 0.2398,
+      "step": 18301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5642920219442507,
+      "learning_rate": 6.119463039707014e-06,
+      "loss": 0.1976,
+      "step": 18302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.654170314086557,
+      "learning_rate": 6.119079029456582e-06,
+      "loss": 0.2622,
+      "step": 18303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6726635806703887,
+      "learning_rate": 6.11869501225697e-06,
+      "loss": 0.3511,
+      "step": 18304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.013978840382954,
+      "learning_rate": 6.118310988110568e-06,
+      "loss": 0.3286,
+      "step": 18305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2358524437921055,
+      "learning_rate": 6.117926957019758e-06,
+      "loss": 0.1158,
+      "step": 18306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4263032845875203,
+      "learning_rate": 6.1175429189869225e-06,
+      "loss": 0.175,
+      "step": 18307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4266399175563726,
+      "learning_rate": 6.1171588740144506e-06,
+      "loss": 0.3339,
+      "step": 18308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.493707561688125,
+      "learning_rate": 6.116774822104724e-06,
+      "loss": 0.2693,
+      "step": 18309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5336941777021793,
+      "learning_rate": 6.116390763260129e-06,
+      "loss": 0.2665,
+      "step": 18310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6653880494716051,
+      "learning_rate": 6.11600669748305e-06,
+      "loss": 0.3863,
+      "step": 18311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7085704258537715,
+      "learning_rate": 6.115622624775872e-06,
+      "loss": 0.28,
+      "step": 18312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7939720670961221,
+      "learning_rate": 6.11523854514098e-06,
+      "loss": 0.3093,
+      "step": 18313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8475680219299917,
+      "learning_rate": 6.11485445858076e-06,
+      "loss": 0.3543,
+      "step": 18314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8240357936216374,
+      "learning_rate": 6.114470365097596e-06,
+      "loss": 0.1624,
+      "step": 18315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3663891504961554,
+      "learning_rate": 6.114086264693872e-06,
+      "loss": 0.2359,
+      "step": 18316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7530830422478691,
+      "learning_rate": 6.113702157371974e-06,
+      "loss": 0.2491,
+      "step": 18317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2380917250664512,
+      "learning_rate": 6.113318043134291e-06,
+      "loss": 0.2232,
+      "step": 18318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1284054298320325,
+      "learning_rate": 6.112933921983202e-06,
+      "loss": 0.6218,
+      "step": 18319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8625872937169552,
+      "learning_rate": 6.112549793921095e-06,
+      "loss": 0.4556,
+      "step": 18320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.089839974515838,
+      "learning_rate": 6.112165658950358e-06,
+      "loss": 0.562,
+      "step": 18321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7945529280446353,
+      "learning_rate": 6.111781517073371e-06,
+      "loss": 0.3234,
+      "step": 18322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1857387821033623,
+      "learning_rate": 6.111397368292524e-06,
+      "loss": 0.1477,
+      "step": 18323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2292928374993588,
+      "learning_rate": 6.1110132126102e-06,
+      "loss": 0.2774,
+      "step": 18324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5483760799425135,
+      "learning_rate": 6.110629050028787e-06,
+      "loss": 0.3328,
+      "step": 18325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6555178279001745,
+      "learning_rate": 6.110244880550667e-06,
+      "loss": 0.2816,
+      "step": 18326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9738190877780395,
+      "learning_rate": 6.109860704178228e-06,
+      "loss": 0.3976,
+      "step": 18327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4893432767765598,
+      "learning_rate": 6.1094765209138554e-06,
+      "loss": 0.2141,
+      "step": 18328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9323912339550504,
+      "learning_rate": 6.109092330759934e-06,
+      "loss": 0.2151,
+      "step": 18329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.456415524407667,
+      "learning_rate": 6.10870813371885e-06,
+      "loss": 0.3805,
+      "step": 18330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.307412711958857,
+      "learning_rate": 6.108323929792988e-06,
+      "loss": 0.2181,
+      "step": 18331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9975197925014198,
+      "learning_rate": 6.1079397189847375e-06,
+      "loss": 0.1405,
+      "step": 18332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9633786918452178,
+      "learning_rate": 6.10755550129648e-06,
+      "loss": 0.2329,
+      "step": 18333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5673783917231026,
+      "learning_rate": 6.107171276730605e-06,
+      "loss": 0.2689,
+      "step": 18334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2553359464089608,
+      "learning_rate": 6.106787045289494e-06,
+      "loss": 0.1578,
+      "step": 18335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.604070332863086,
+      "learning_rate": 6.106402806975538e-06,
+      "loss": 0.3352,
+      "step": 18336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6575267836764316,
+      "learning_rate": 6.106018561791119e-06,
+      "loss": 0.4784,
+      "step": 18337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.364121983947328,
+      "learning_rate": 6.105634309738626e-06,
+      "loss": 0.1816,
+      "step": 18338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4504412612685456,
+      "learning_rate": 6.105250050820443e-06,
+      "loss": 0.1713,
+      "step": 18339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2062703128992665,
+      "learning_rate": 6.104865785038957e-06,
+      "loss": 0.613,
+      "step": 18340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3397686689821506,
+      "learning_rate": 6.104481512396555e-06,
+      "loss": 0.2441,
+      "step": 18341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0996336860136413,
+      "learning_rate": 6.10409723289562e-06,
+      "loss": 0.4511,
+      "step": 18342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.873750214490821,
+      "learning_rate": 6.1037129465385425e-06,
+      "loss": 0.6892,
+      "step": 18343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6327212070879786,
+      "learning_rate": 6.103328653327707e-06,
+      "loss": 0.1801,
+      "step": 18344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.961511436759188,
+      "learning_rate": 6.102944353265499e-06,
+      "loss": 0.5873,
+      "step": 18345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.792885605955772,
+      "learning_rate": 6.102560046354305e-06,
+      "loss": 0.3633,
+      "step": 18346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2570640339200723,
+      "learning_rate": 6.102175732596514e-06,
+      "loss": 0.1414,
+      "step": 18347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.940902913353067,
+      "learning_rate": 6.101791411994511e-06,
+      "loss": 0.4523,
+      "step": 18348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3339963776986097,
+      "learning_rate": 6.10140708455068e-06,
+      "loss": 0.1433,
+      "step": 18349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9655768996571283,
+      "learning_rate": 6.101022750267411e-06,
+      "loss": 0.5765,
+      "step": 18350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5088191797898614,
+      "learning_rate": 6.1006384091470885e-06,
+      "loss": 0.3154,
+      "step": 18351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4621240188595015,
+      "learning_rate": 6.100254061192101e-06,
+      "loss": 0.1882,
+      "step": 18352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9925124969174282,
+      "learning_rate": 6.099869706404834e-06,
+      "loss": 0.5397,
+      "step": 18353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2271252475583134,
+      "learning_rate": 6.0994853447876746e-06,
+      "loss": 0.1791,
+      "step": 18354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.096281897837299,
+      "learning_rate": 6.0991009763430084e-06,
+      "loss": 0.3797,
+      "step": 18355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2552012473680658,
+      "learning_rate": 6.0987166010732244e-06,
+      "loss": 0.187,
+      "step": 18356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1238663544581653,
+      "learning_rate": 6.098332218980707e-06,
+      "loss": 0.2437,
+      "step": 18357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.704013881231124,
+      "learning_rate": 6.097947830067845e-06,
+      "loss": 0.2739,
+      "step": 18358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3524549951413842,
+      "learning_rate": 6.097563434337026e-06,
+      "loss": 0.5184,
+      "step": 18359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.671116019372994,
+      "learning_rate": 6.097179031790633e-06,
+      "loss": 0.3383,
+      "step": 18360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3997220367987855,
+      "learning_rate": 6.0967946224310584e-06,
+      "loss": 0.2369,
+      "step": 18361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1666954872311384,
+      "learning_rate": 6.096410206260686e-06,
+      "loss": 0.4534,
+      "step": 18362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8778215556720692,
+      "learning_rate": 6.0960257832819025e-06,
+      "loss": 0.6026,
+      "step": 18363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.460051165698294,
+      "learning_rate": 6.0956413534970975e-06,
+      "loss": 0.4365,
+      "step": 18364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.266650605729179,
+      "learning_rate": 6.095256916908656e-06,
+      "loss": 0.6986,
+      "step": 18365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3744123189160768,
+      "learning_rate": 6.0948724735189655e-06,
+      "loss": 0.3089,
+      "step": 18366
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4495578019659738,
+      "learning_rate": 6.094488023330415e-06,
+      "loss": 0.3599,
+      "step": 18367
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4920112545724151,
+      "learning_rate": 6.09410356634539e-06,
+      "loss": 0.1347,
+      "step": 18368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1973161038852225,
+      "learning_rate": 6.093719102566278e-06,
+      "loss": 0.5353,
+      "step": 18369
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.98022802195292,
+      "learning_rate": 6.093334631995469e-06,
+      "loss": 0.508,
+      "step": 18370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7054782323709101,
+      "learning_rate": 6.092950154635346e-06,
+      "loss": 0.2979,
+      "step": 18371
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5975029738940287,
+      "learning_rate": 6.092565670488301e-06,
+      "loss": 0.4242,
+      "step": 18372
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.261851525185181,
+      "learning_rate": 6.092181179556719e-06,
+      "loss": 0.1692,
+      "step": 18373
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5755414020887601,
+      "learning_rate": 6.0917966818429875e-06,
+      "loss": 0.4607,
+      "step": 18374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5150059561525795,
+      "learning_rate": 6.091412177349495e-06,
+      "loss": 0.2558,
+      "step": 18375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8333366355179899,
+      "learning_rate": 6.091027666078629e-06,
+      "loss": 0.414,
+      "step": 18376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3576418656274365,
+      "learning_rate": 6.090643148032777e-06,
+      "loss": 0.564,
+      "step": 18377
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8143141275471055,
+      "learning_rate": 6.090258623214327e-06,
+      "loss": 0.3238,
+      "step": 18378
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4090692948782229,
+      "learning_rate": 6.089874091625666e-06,
+      "loss": 0.2844,
+      "step": 18379
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2606254718630223,
+      "learning_rate": 6.089489553269183e-06,
+      "loss": 0.2487,
+      "step": 18380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.011381798830206,
+      "learning_rate": 6.089105008147267e-06,
+      "loss": 0.3454,
+      "step": 18381
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4914042754082775,
+      "learning_rate": 6.088720456262302e-06,
+      "loss": 0.1698,
+      "step": 18382
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2487055512022396,
+      "learning_rate": 6.08833589761668e-06,
+      "loss": 0.3171,
+      "step": 18383
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.232973098885333,
+      "learning_rate": 6.087951332212785e-06,
+      "loss": 0.2213,
+      "step": 18384
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1474005776620522,
+      "learning_rate": 6.087566760053009e-06,
+      "loss": 0.1811,
+      "step": 18385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5313086498472575,
+      "learning_rate": 6.087182181139737e-06,
+      "loss": 0.3231,
+      "step": 18386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7052821073398001,
+      "learning_rate": 6.08679759547536e-06,
+      "loss": 0.2358,
+      "step": 18387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5189450587100386,
+      "learning_rate": 6.086413003062263e-06,
+      "loss": 0.1905,
+      "step": 18388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.969792232229858,
+      "learning_rate": 6.086028403902837e-06,
+      "loss": 0.1625,
+      "step": 18389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7597951484345942,
+      "learning_rate": 6.0856437979994686e-06,
+      "loss": 0.2162,
+      "step": 18390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9366575648231404,
+      "learning_rate": 6.085259185354548e-06,
+      "loss": 0.1658,
+      "step": 18391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8334328956734747,
+      "learning_rate": 6.084874565970461e-06,
+      "loss": 0.4632,
+      "step": 18392
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6536532414388048,
+      "learning_rate": 6.084489939849599e-06,
+      "loss": 0.2154,
+      "step": 18393
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3536146837786873,
+      "learning_rate": 6.084105306994346e-06,
+      "loss": 0.2424,
+      "step": 18394
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.67246558551706,
+      "learning_rate": 6.083720667407093e-06,
+      "loss": 0.47,
+      "step": 18395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9780000481912683,
+      "learning_rate": 6.08333602109023e-06,
+      "loss": 0.1319,
+      "step": 18396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1264581235654614,
+      "learning_rate": 6.082951368046143e-06,
+      "loss": 0.5003,
+      "step": 18397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.196436029865426,
+      "learning_rate": 6.082566708277224e-06,
+      "loss": 0.2713,
+      "step": 18398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1633752547915828,
+      "learning_rate": 6.082182041785857e-06,
+      "loss": 0.1413,
+      "step": 18399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6182982228571259,
+      "learning_rate": 6.081797368574434e-06,
+      "loss": 0.4741,
+      "step": 18400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.69470350692311,
+      "learning_rate": 6.081412688645342e-06,
+      "loss": 0.2326,
+      "step": 18401
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7003284086732198,
+      "learning_rate": 6.081028002000971e-06,
+      "loss": 0.2216,
+      "step": 18402
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5323141278445247,
+      "learning_rate": 6.08064330864371e-06,
+      "loss": 0.2018,
+      "step": 18403
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7719255821132938,
+      "learning_rate": 6.080258608575947e-06,
+      "loss": 0.2867,
+      "step": 18404
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0275430379243753,
+      "learning_rate": 6.079873901800071e-06,
+      "loss": 0.4457,
+      "step": 18405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.054757370223891,
+      "learning_rate": 6.079489188318469e-06,
+      "loss": 0.1313,
+      "step": 18406
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7753714548886825,
+      "learning_rate": 6.079104468133534e-06,
+      "loss": 0.3584,
+      "step": 18407
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7530223343787013,
+      "learning_rate": 6.0787197412476515e-06,
+      "loss": 0.4066,
+      "step": 18408
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9119028207812991,
+      "learning_rate": 6.0783350076632144e-06,
+      "loss": 0.1986,
+      "step": 18409
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1773382194347515,
+      "learning_rate": 6.077950267382607e-06,
+      "loss": 0.3286,
+      "step": 18410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8151125577042169,
+      "learning_rate": 6.077565520408222e-06,
+      "loss": 0.2487,
+      "step": 18411
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7131359485775985,
+      "learning_rate": 6.0771807667424455e-06,
+      "loss": 0.3929,
+      "step": 18412
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6309522408182708,
+      "learning_rate": 6.076796006387671e-06,
+      "loss": 0.187,
+      "step": 18413
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6808320692407825,
+      "learning_rate": 6.076411239346285e-06,
+      "loss": 0.2639,
+      "step": 18414
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2362423033855598,
+      "learning_rate": 6.076026465620676e-06,
+      "loss": 0.1762,
+      "step": 18415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8439008925810791,
+      "learning_rate": 6.075641685213235e-06,
+      "loss": 0.3221,
+      "step": 18416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1243372963715683,
+      "learning_rate": 6.0752568981263504e-06,
+      "loss": 0.2597,
+      "step": 18417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0247027405881943,
+      "learning_rate": 6.074872104362411e-06,
+      "loss": 0.1762,
+      "step": 18418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5695049240008327,
+      "learning_rate": 6.074487303923809e-06,
+      "loss": 0.2748,
+      "step": 18419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9643124531344125,
+      "learning_rate": 6.074102496812933e-06,
+      "loss": 0.2941,
+      "step": 18420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.006266715078154,
+      "learning_rate": 6.07371768303217e-06,
+      "loss": 0.2185,
+      "step": 18421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3275683414518744,
+      "learning_rate": 6.073332862583913e-06,
+      "loss": 0.2847,
+      "step": 18422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4757129644123577,
+      "learning_rate": 6.072948035470549e-06,
+      "loss": 0.3343,
+      "step": 18423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6649223935596016,
+      "learning_rate": 6.072563201694468e-06,
+      "loss": 0.3995,
+      "step": 18424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.766432893797593,
+      "learning_rate": 6.072178361258062e-06,
+      "loss": 0.3476,
+      "step": 18425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0983020775698935,
+      "learning_rate": 6.071793514163719e-06,
+      "loss": 0.5845,
+      "step": 18426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6899705744895932,
+      "learning_rate": 6.071408660413828e-06,
+      "loss": 0.3164,
+      "step": 18427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6861606112640684,
+      "learning_rate": 6.07102380001078e-06,
+      "loss": 0.4155,
+      "step": 18428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7776935358957937,
+      "learning_rate": 6.070638932956964e-06,
+      "loss": 0.488,
+      "step": 18429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2700765480345675,
+      "learning_rate": 6.070254059254772e-06,
+      "loss": 0.2972,
+      "step": 18430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0222793935301142,
+      "learning_rate": 6.069869178906594e-06,
+      "loss": 0.2027,
+      "step": 18431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1989828710284984,
+      "learning_rate": 6.069484291914816e-06,
+      "loss": 0.2477,
+      "step": 18432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4115497171255151,
+      "learning_rate": 6.0690993982818305e-06,
+      "loss": 0.1486,
+      "step": 18433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7228776439166822,
+      "learning_rate": 6.068714498010028e-06,
+      "loss": 0.242,
+      "step": 18434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.426799750054412,
+      "learning_rate": 6.068329591101799e-06,
+      "loss": 0.2866,
+      "step": 18435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5222018145548148,
+      "learning_rate": 6.067944677559534e-06,
+      "loss": 0.2318,
+      "step": 18436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.502294915725262,
+      "learning_rate": 6.067559757385621e-06,
+      "loss": 0.1602,
+      "step": 18437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2472701424153396,
+      "learning_rate": 6.067174830582451e-06,
+      "loss": 0.2459,
+      "step": 18438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6551825861346734,
+      "learning_rate": 6.0667898971524144e-06,
+      "loss": 0.2289,
+      "step": 18439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5605911372738517,
+      "learning_rate": 6.066404957097904e-06,
+      "loss": 0.3415,
+      "step": 18440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7855542863964486,
+      "learning_rate": 6.066020010421305e-06,
+      "loss": 0.3581,
+      "step": 18441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.383827588521616,
+      "learning_rate": 6.065635057125013e-06,
+      "loss": 0.2889,
+      "step": 18442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5896781945317107,
+      "learning_rate": 6.065250097211415e-06,
+      "loss": 0.1989,
+      "step": 18443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4652178629934067,
+      "learning_rate": 6.064865130682903e-06,
+      "loss": 0.2598,
+      "step": 18444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9219031370039351,
+      "learning_rate": 6.064480157541869e-06,
+      "loss": 0.1436,
+      "step": 18445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3291480359677093,
+      "learning_rate": 6.0640951777907e-06,
+      "loss": 0.1557,
+      "step": 18446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1609935138142555,
+      "learning_rate": 6.06371019143179e-06,
+      "loss": 0.4579,
+      "step": 18447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.466498797145158,
+      "learning_rate": 6.063325198467527e-06,
+      "loss": 0.4662,
+      "step": 18448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8341489358508871,
+      "learning_rate": 6.062940198900303e-06,
+      "loss": 0.4537,
+      "step": 18449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6461581792836952,
+      "learning_rate": 6.0625551927325084e-06,
+      "loss": 0.4043,
+      "step": 18450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9099644523702952,
+      "learning_rate": 6.062170179966535e-06,
+      "loss": 0.5047,
+      "step": 18451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4894455763222618,
+      "learning_rate": 6.061785160604773e-06,
+      "loss": 0.3068,
+      "step": 18452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.350884366713485,
+      "learning_rate": 6.061400134649613e-06,
+      "loss": 0.2593,
+      "step": 18453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.416460890644449,
+      "learning_rate": 6.061015102103446e-06,
+      "loss": 0.3737,
+      "step": 18454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.898592585441729,
+      "learning_rate": 6.0606300629686635e-06,
+      "loss": 0.4778,
+      "step": 18455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9560956565006027,
+      "learning_rate": 6.060245017247655e-06,
+      "loss": 0.1688,
+      "step": 18456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5638035557973144,
+      "learning_rate": 6.059859964942812e-06,
+      "loss": 0.1953,
+      "step": 18457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.008179271165286,
+      "learning_rate": 6.059474906056528e-06,
+      "loss": 0.3327,
+      "step": 18458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8235288945218588,
+      "learning_rate": 6.059089840591191e-06,
+      "loss": 0.6022,
+      "step": 18459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7764942864010227,
+      "learning_rate": 6.058704768549193e-06,
+      "loss": 0.3408,
+      "step": 18460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5036511224275935,
+      "learning_rate": 6.058319689932925e-06,
+      "loss": 0.259,
+      "step": 18461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6571970056383118,
+      "learning_rate": 6.057934604744781e-06,
+      "loss": 0.1635,
+      "step": 18462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1318083900476257,
+      "learning_rate": 6.0575495129871475e-06,
+      "loss": 0.1616,
+      "step": 18463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5775201932140617,
+      "learning_rate": 6.05716441466242e-06,
+      "loss": 0.523,
+      "step": 18464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8810672942182631,
+      "learning_rate": 6.0567793097729874e-06,
+      "loss": 0.5209,
+      "step": 18465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8182444264662052,
+      "learning_rate": 6.056394198321243e-06,
+      "loss": 0.2937,
+      "step": 18466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.595180297421444,
+      "learning_rate": 6.056009080309575e-06,
+      "loss": 0.1543,
+      "step": 18467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5138765061403823,
+      "learning_rate": 6.0556239557403795e-06,
+      "loss": 0.3159,
+      "step": 18468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0413459656519137,
+      "learning_rate": 6.055238824616043e-06,
+      "loss": 0.4356,
+      "step": 18469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5761862994057518,
+      "learning_rate": 6.054853686938961e-06,
+      "loss": 0.1949,
+      "step": 18470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1603085462472484,
+      "learning_rate": 6.054468542711524e-06,
+      "loss": 0.6581,
+      "step": 18471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0733068687934542,
+      "learning_rate": 6.054083391936122e-06,
+      "loss": 0.186,
+      "step": 18472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6105545082623818,
+      "learning_rate": 6.0536982346151495e-06,
+      "loss": 0.1924,
+      "step": 18473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5987287526362928,
+      "learning_rate": 6.053313070750996e-06,
+      "loss": 0.2333,
+      "step": 18474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.539678476967996,
+      "learning_rate": 6.052927900346055e-06,
+      "loss": 0.291,
+      "step": 18475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4956661514689669,
+      "learning_rate": 6.052542723402716e-06,
+      "loss": 0.2057,
+      "step": 18476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5448481401095782,
+      "learning_rate": 6.052157539923373e-06,
+      "loss": 0.3721,
+      "step": 18477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3234307833909849,
+      "learning_rate": 6.051772349910416e-06,
+      "loss": 0.1855,
+      "step": 18478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4503314895142896,
+      "learning_rate": 6.05138715336624e-06,
+      "loss": 0.2268,
+      "step": 18479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2589546403506113,
+      "learning_rate": 6.0510019502932325e-06,
+      "loss": 0.4429,
+      "step": 18480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5243471877527506,
+      "learning_rate": 6.050616740693789e-06,
+      "loss": 0.4004,
+      "step": 18481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.414801483534045,
+      "learning_rate": 6.050231524570301e-06,
+      "loss": 0.2738,
+      "step": 18482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7190981772360703,
+      "learning_rate": 6.049846301925158e-06,
+      "loss": 0.4214,
+      "step": 18483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.025537513107739,
+      "learning_rate": 6.049461072760757e-06,
+      "loss": 0.4948,
+      "step": 18484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3363972299499534,
+      "learning_rate": 6.049075837079485e-06,
+      "loss": 0.1268,
+      "step": 18485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.57240634075275,
+      "learning_rate": 6.04869059488374e-06,
+      "loss": 0.4759,
+      "step": 18486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.493089947255118,
+      "learning_rate": 6.048305346175907e-06,
+      "loss": 0.4191,
+      "step": 18487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1028603861825368,
+      "learning_rate": 6.047920090958384e-06,
+      "loss": 0.1572,
+      "step": 18488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4234884979458668,
+      "learning_rate": 6.04753482923356e-06,
+      "loss": 0.2314,
+      "step": 18489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5069855215868075,
+      "learning_rate": 6.047149561003831e-06,
+      "loss": 0.1587,
+      "step": 18490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2638772446785342,
+      "learning_rate": 6.0467642862715856e-06,
+      "loss": 0.1842,
+      "step": 18491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.639949268873533,
+      "learning_rate": 6.046379005039218e-06,
+      "loss": 0.3697,
+      "step": 18492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2082586775797057,
+      "learning_rate": 6.045993717309121e-06,
+      "loss": 0.155,
+      "step": 18493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6654587314668354,
+      "learning_rate": 6.045608423083687e-06,
+      "loss": 0.2317,
+      "step": 18494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4888438240348945,
+      "learning_rate": 6.045223122365307e-06,
+      "loss": 0.2833,
+      "step": 18495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0904615398191115,
+      "learning_rate": 6.044837815156377e-06,
+      "loss": 0.2029,
+      "step": 18496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8260481889422917,
+      "learning_rate": 6.044452501459285e-06,
+      "loss": 0.3928,
+      "step": 18497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3268870165698978,
+      "learning_rate": 6.044067181276427e-06,
+      "loss": 0.1699,
+      "step": 18498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4234229323670757,
+      "learning_rate": 6.0436818546101965e-06,
+      "loss": 0.2555,
+      "step": 18499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2783220289669068,
+      "learning_rate": 6.043296521462982e-06,
+      "loss": 0.2662,
+      "step": 18500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2739080469513462,
+      "learning_rate": 6.042911181837182e-06,
+      "loss": 0.2574,
+      "step": 18501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.372529436993945,
+      "learning_rate": 6.042525835735186e-06,
+      "loss": 0.2342,
+      "step": 18502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.621536063553849,
+      "learning_rate": 6.042140483159386e-06,
+      "loss": 0.234,
+      "step": 18503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.246185008973176,
+      "learning_rate": 6.0417551241121766e-06,
+      "loss": 0.4354,
+      "step": 18504
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 41101,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2056,
+  "total_flos": 1529942178488320.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}