diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,29533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 4216,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004743833017077799,
+      "grad_norm": 12.117147445678711,
+      "learning_rate": 1.5748031496062994e-07,
+      "loss": 0.6555,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009487666034155598,
+      "grad_norm": 12.216944694519043,
+      "learning_rate": 3.149606299212599e-07,
+      "loss": 0.7042,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014231499051233396,
+      "grad_norm": 11.745461463928223,
+      "learning_rate": 4.724409448818898e-07,
+      "loss": 0.7974,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018975332068311196,
+      "grad_norm": 14.210430145263672,
+      "learning_rate": 6.299212598425198e-07,
+      "loss": 0.6868,
+      "step": 4
+    },
+    {
+      "epoch": 0.0023719165085388993,
+      "grad_norm": 13.679463386535645,
+      "learning_rate": 7.874015748031496e-07,
+      "loss": 0.7468,
+      "step": 5
+    },
+    {
+      "epoch": 0.0028462998102466793,
+      "grad_norm": 12.984026908874512,
+      "learning_rate": 9.448818897637796e-07,
+      "loss": 0.7443,
+      "step": 6
+    },
+    {
+      "epoch": 0.003320683111954459,
+      "grad_norm": 15.170774459838867,
+      "learning_rate": 1.1023622047244096e-06,
+      "loss": 0.7255,
+      "step": 7
+    },
+    {
+      "epoch": 0.003795066413662239,
+      "grad_norm": 12.011886596679688,
+      "learning_rate": 1.2598425196850396e-06,
+      "loss": 0.6684,
+      "step": 8
+    },
+    {
+      "epoch": 0.004269449715370019,
+      "grad_norm": 29.63150405883789,
+      "learning_rate": 1.4173228346456693e-06,
+      "loss": 0.6935,
+      "step": 9
+    },
+    {
+      "epoch": 0.004743833017077799,
+      "grad_norm": 12.307829856872559,
+      "learning_rate": 1.5748031496062992e-06,
+      "loss": 0.5554,
+      "step": 10
+    },
+    {
+      "epoch": 0.005218216318785579,
+      "grad_norm": 15.281328201293945,
+      "learning_rate": 1.7322834645669292e-06,
+      "loss": 0.5863,
+      "step": 11
+    },
+    {
+      "epoch": 0.0056925996204933585,
+      "grad_norm": 10.656610488891602,
+      "learning_rate": 1.8897637795275591e-06,
+      "loss": 0.4246,
+      "step": 12
+    },
+    {
+      "epoch": 0.006166982922201139,
+      "grad_norm": 9.216360092163086,
+      "learning_rate": 2.0472440944881893e-06,
+      "loss": 0.3654,
+      "step": 13
+    },
+    {
+      "epoch": 0.006641366223908918,
+      "grad_norm": 7.334237575531006,
+      "learning_rate": 2.2047244094488192e-06,
+      "loss": 0.3389,
+      "step": 14
+    },
+    {
+      "epoch": 0.007115749525616698,
+      "grad_norm": 9.264925003051758,
+      "learning_rate": 2.362204724409449e-06,
+      "loss": 0.3582,
+      "step": 15
+    },
+    {
+      "epoch": 0.007590132827324478,
+      "grad_norm": 7.431726455688477,
+      "learning_rate": 2.519685039370079e-06,
+      "loss": 0.3778,
+      "step": 16
+    },
+    {
+      "epoch": 0.008064516129032258,
+      "grad_norm": 14.502622604370117,
+      "learning_rate": 2.677165354330709e-06,
+      "loss": 0.4209,
+      "step": 17
+    },
+    {
+      "epoch": 0.008538899430740038,
+      "grad_norm": 8.857325553894043,
+      "learning_rate": 2.8346456692913386e-06,
+      "loss": 0.3823,
+      "step": 18
+    },
+    {
+      "epoch": 0.009013282732447819,
+      "grad_norm": 10.866437911987305,
+      "learning_rate": 2.992125984251969e-06,
+      "loss": 0.4257,
+      "step": 19
+    },
+    {
+      "epoch": 0.009487666034155597,
+      "grad_norm": 7.098900318145752,
+      "learning_rate": 3.1496062992125985e-06,
+      "loss": 0.324,
+      "step": 20
+    },
+    {
+      "epoch": 0.009962049335863378,
+      "grad_norm": 7.2487616539001465,
+      "learning_rate": 3.307086614173229e-06,
+      "loss": 0.3815,
+      "step": 21
+    },
+    {
+      "epoch": 0.010436432637571158,
+      "grad_norm": 6.616808891296387,
+      "learning_rate": 3.4645669291338583e-06,
+      "loss": 0.2939,
+      "step": 22
+    },
+    {
+      "epoch": 0.010910815939278937,
+      "grad_norm": 6.747620582580566,
+      "learning_rate": 3.6220472440944887e-06,
+      "loss": 0.2914,
+      "step": 23
+    },
+    {
+      "epoch": 0.011385199240986717,
+      "grad_norm": 5.690284729003906,
+      "learning_rate": 3.7795275590551182e-06,
+      "loss": 0.2879,
+      "step": 24
+    },
+    {
+      "epoch": 0.011859582542694497,
+      "grad_norm": 6.056181907653809,
+      "learning_rate": 3.937007874015748e-06,
+      "loss": 0.2785,
+      "step": 25
+    },
+    {
+      "epoch": 0.012333965844402278,
+      "grad_norm": 7.560486316680908,
+      "learning_rate": 4.0944881889763785e-06,
+      "loss": 0.3008,
+      "step": 26
+    },
+    {
+      "epoch": 0.012808349146110056,
+      "grad_norm": 7.066635608673096,
+      "learning_rate": 4.251968503937008e-06,
+      "loss": 0.3298,
+      "step": 27
+    },
+    {
+      "epoch": 0.013282732447817837,
+      "grad_norm": 5.725413799285889,
+      "learning_rate": 4.4094488188976384e-06,
+      "loss": 0.2901,
+      "step": 28
+    },
+    {
+      "epoch": 0.013757115749525617,
+      "grad_norm": 7.520142555236816,
+      "learning_rate": 4.566929133858268e-06,
+      "loss": 0.3461,
+      "step": 29
+    },
+    {
+      "epoch": 0.014231499051233396,
+      "grad_norm": 6.214694976806641,
+      "learning_rate": 4.724409448818898e-06,
+      "loss": 0.3216,
+      "step": 30
+    },
+    {
+      "epoch": 0.014705882352941176,
+      "grad_norm": 6.9384870529174805,
+      "learning_rate": 4.881889763779528e-06,
+      "loss": 0.3779,
+      "step": 31
+    },
+    {
+      "epoch": 0.015180265654648957,
+      "grad_norm": 6.863812446594238,
+      "learning_rate": 5.039370078740158e-06,
+      "loss": 0.3194,
+      "step": 32
+    },
+    {
+      "epoch": 0.015654648956356737,
+      "grad_norm": 7.183291912078857,
+      "learning_rate": 5.196850393700788e-06,
+      "loss": 0.3293,
+      "step": 33
+    },
+    {
+      "epoch": 0.016129032258064516,
+      "grad_norm": 6.995275497436523,
+      "learning_rate": 5.354330708661418e-06,
+      "loss": 0.4026,
+      "step": 34
+    },
+    {
+      "epoch": 0.016603415559772294,
+      "grad_norm": 6.775660514831543,
+      "learning_rate": 5.511811023622048e-06,
+      "loss": 0.3289,
+      "step": 35
+    },
+    {
+      "epoch": 0.017077798861480076,
+      "grad_norm": 6.413484573364258,
+      "learning_rate": 5.669291338582677e-06,
+      "loss": 0.3315,
+      "step": 36
+    },
+    {
+      "epoch": 0.017552182163187855,
+      "grad_norm": 7.173760414123535,
+      "learning_rate": 5.8267716535433075e-06,
+      "loss": 0.3633,
+      "step": 37
+    },
+    {
+      "epoch": 0.018026565464895637,
+      "grad_norm": 4.803355693817139,
+      "learning_rate": 5.984251968503938e-06,
+      "loss": 0.3607,
+      "step": 38
+    },
+    {
+      "epoch": 0.018500948766603416,
+      "grad_norm": 5.451982498168945,
+      "learning_rate": 6.141732283464567e-06,
+      "loss": 0.3503,
+      "step": 39
+    },
+    {
+      "epoch": 0.018975332068311195,
+      "grad_norm": 5.767213821411133,
+      "learning_rate": 6.299212598425197e-06,
+      "loss": 0.2794,
+      "step": 40
+    },
+    {
+      "epoch": 0.019449715370018977,
+      "grad_norm": 6.687100410461426,
+      "learning_rate": 6.456692913385827e-06,
+      "loss": 0.3668,
+      "step": 41
+    },
+    {
+      "epoch": 0.019924098671726755,
+      "grad_norm": 6.374785900115967,
+      "learning_rate": 6.614173228346458e-06,
+      "loss": 0.3366,
+      "step": 42
+    },
+    {
+      "epoch": 0.020398481973434534,
+      "grad_norm": 6.196499347686768,
+      "learning_rate": 6.771653543307087e-06,
+      "loss": 0.3421,
+      "step": 43
+    },
+    {
+      "epoch": 0.020872865275142316,
+      "grad_norm": 6.357893943786621,
+      "learning_rate": 6.929133858267717e-06,
+      "loss": 0.3663,
+      "step": 44
+    },
+    {
+      "epoch": 0.021347248576850095,
+      "grad_norm": 5.325692653656006,
+      "learning_rate": 7.086614173228347e-06,
+      "loss": 0.3341,
+      "step": 45
+    },
+    {
+      "epoch": 0.021821631878557873,
+      "grad_norm": 6.0895538330078125,
+      "learning_rate": 7.2440944881889774e-06,
+      "loss": 0.3348,
+      "step": 46
+    },
+    {
+      "epoch": 0.022296015180265655,
+      "grad_norm": 5.129356384277344,
+      "learning_rate": 7.401574803149607e-06,
+      "loss": 0.294,
+      "step": 47
+    },
+    {
+      "epoch": 0.022770398481973434,
+      "grad_norm": 5.093695163726807,
+      "learning_rate": 7.5590551181102365e-06,
+      "loss": 0.3223,
+      "step": 48
+    },
+    {
+      "epoch": 0.023244781783681213,
+      "grad_norm": 5.516068935394287,
+      "learning_rate": 7.716535433070867e-06,
+      "loss": 0.3455,
+      "step": 49
+    },
+    {
+      "epoch": 0.023719165085388995,
+      "grad_norm": 5.486409664154053,
+      "learning_rate": 7.874015748031496e-06,
+      "loss": 0.3349,
+      "step": 50
+    },
+    {
+      "epoch": 0.024193548387096774,
+      "grad_norm": 5.390769004821777,
+      "learning_rate": 8.031496062992128e-06,
+      "loss": 0.3474,
+      "step": 51
+    },
+    {
+      "epoch": 0.024667931688804556,
+      "grad_norm": 5.160117149353027,
+      "learning_rate": 8.188976377952757e-06,
+      "loss": 0.3048,
+      "step": 52
+    },
+    {
+      "epoch": 0.025142314990512334,
+      "grad_norm": 7.194066047668457,
+      "learning_rate": 8.346456692913387e-06,
+      "loss": 0.4043,
+      "step": 53
+    },
+    {
+      "epoch": 0.025616698292220113,
+      "grad_norm": 5.130771636962891,
+      "learning_rate": 8.503937007874016e-06,
+      "loss": 0.3362,
+      "step": 54
+    },
+    {
+      "epoch": 0.026091081593927895,
+      "grad_norm": 5.493372440338135,
+      "learning_rate": 8.661417322834647e-06,
+      "loss": 0.3842,
+      "step": 55
+    },
+    {
+      "epoch": 0.026565464895635674,
+      "grad_norm": 6.465633392333984,
+      "learning_rate": 8.818897637795277e-06,
+      "loss": 0.3974,
+      "step": 56
+    },
+    {
+      "epoch": 0.027039848197343452,
+      "grad_norm": 7.080130100250244,
+      "learning_rate": 8.976377952755906e-06,
+      "loss": 0.3579,
+      "step": 57
+    },
+    {
+      "epoch": 0.027514231499051234,
+      "grad_norm": 5.774561882019043,
+      "learning_rate": 9.133858267716536e-06,
+      "loss": 0.374,
+      "step": 58
+    },
+    {
+      "epoch": 0.027988614800759013,
+      "grad_norm": 5.477121353149414,
+      "learning_rate": 9.291338582677165e-06,
+      "loss": 0.3664,
+      "step": 59
+    },
+    {
+      "epoch": 0.028462998102466792,
+      "grad_norm": 5.975531101226807,
+      "learning_rate": 9.448818897637797e-06,
+      "loss": 0.4218,
+      "step": 60
+    },
+    {
+      "epoch": 0.028937381404174574,
+      "grad_norm": 5.1018967628479,
+      "learning_rate": 9.606299212598426e-06,
+      "loss": 0.3879,
+      "step": 61
+    },
+    {
+      "epoch": 0.029411764705882353,
+      "grad_norm": 5.107121467590332,
+      "learning_rate": 9.763779527559056e-06,
+      "loss": 0.3992,
+      "step": 62
+    },
+    {
+      "epoch": 0.02988614800759013,
+      "grad_norm": 6.47650671005249,
+      "learning_rate": 9.921259842519685e-06,
+      "loss": 0.4705,
+      "step": 63
+    },
+    {
+      "epoch": 0.030360531309297913,
+      "grad_norm": 6.037809371948242,
+      "learning_rate": 1.0078740157480316e-05,
+      "loss": 0.4264,
+      "step": 64
+    },
+    {
+      "epoch": 0.030834914611005692,
+      "grad_norm": 5.349067211151123,
+      "learning_rate": 1.0236220472440946e-05,
+      "loss": 0.384,
+      "step": 65
+    },
+    {
+      "epoch": 0.031309297912713474,
+      "grad_norm": 5.131612777709961,
+      "learning_rate": 1.0393700787401575e-05,
+      "loss": 0.3905,
+      "step": 66
+    },
+    {
+      "epoch": 0.03178368121442125,
+      "grad_norm": 5.476340293884277,
+      "learning_rate": 1.0551181102362205e-05,
+      "loss": 0.427,
+      "step": 67
+    },
+    {
+      "epoch": 0.03225806451612903,
+      "grad_norm": 6.4385480880737305,
+      "learning_rate": 1.0708661417322836e-05,
+      "loss": 0.5252,
+      "step": 68
+    },
+    {
+      "epoch": 0.03273244781783681,
+      "grad_norm": 5.29050874710083,
+      "learning_rate": 1.0866141732283466e-05,
+      "loss": 0.453,
+      "step": 69
+    },
+    {
+      "epoch": 0.03320683111954459,
+      "grad_norm": 4.675810813903809,
+      "learning_rate": 1.1023622047244095e-05,
+      "loss": 0.412,
+      "step": 70
+    },
+    {
+      "epoch": 0.033681214421252374,
+      "grad_norm": 5.531217575073242,
+      "learning_rate": 1.1181102362204725e-05,
+      "loss": 0.4553,
+      "step": 71
+    },
+    {
+      "epoch": 0.03415559772296015,
+      "grad_norm": 6.203500747680664,
+      "learning_rate": 1.1338582677165354e-05,
+      "loss": 0.4826,
+      "step": 72
+    },
+    {
+      "epoch": 0.03462998102466793,
+      "grad_norm": 5.102776050567627,
+      "learning_rate": 1.1496062992125985e-05,
+      "loss": 0.3427,
+      "step": 73
+    },
+    {
+      "epoch": 0.03510436432637571,
+      "grad_norm": 4.565568923950195,
+      "learning_rate": 1.1653543307086615e-05,
+      "loss": 0.3734,
+      "step": 74
+    },
+    {
+      "epoch": 0.03557874762808349,
+      "grad_norm": 4.548391819000244,
+      "learning_rate": 1.1811023622047245e-05,
+      "loss": 0.3647,
+      "step": 75
+    },
+    {
+      "epoch": 0.036053130929791274,
+      "grad_norm": 5.1189398765563965,
+      "learning_rate": 1.1968503937007876e-05,
+      "loss": 0.4211,
+      "step": 76
+    },
+    {
+      "epoch": 0.03652751423149905,
+      "grad_norm": 6.630745887756348,
+      "learning_rate": 1.2125984251968505e-05,
+      "loss": 0.5191,
+      "step": 77
+    },
+    {
+      "epoch": 0.03700189753320683,
+      "grad_norm": 5.194817066192627,
+      "learning_rate": 1.2283464566929135e-05,
+      "loss": 0.3779,
+      "step": 78
+    },
+    {
+      "epoch": 0.03747628083491461,
+      "grad_norm": 5.23333740234375,
+      "learning_rate": 1.2440944881889764e-05,
+      "loss": 0.4235,
+      "step": 79
+    },
+    {
+      "epoch": 0.03795066413662239,
+      "grad_norm": 4.812337398529053,
+      "learning_rate": 1.2598425196850394e-05,
+      "loss": 0.3958,
+      "step": 80
+    },
+    {
+      "epoch": 0.03842504743833017,
+      "grad_norm": 4.991153240203857,
+      "learning_rate": 1.2755905511811025e-05,
+      "loss": 0.4543,
+      "step": 81
+    },
+    {
+      "epoch": 0.03889943074003795,
+      "grad_norm": 5.6834025382995605,
+      "learning_rate": 1.2913385826771655e-05,
+      "loss": 0.4253,
+      "step": 82
+    },
+    {
+      "epoch": 0.03937381404174573,
+      "grad_norm": 4.659487247467041,
+      "learning_rate": 1.3070866141732284e-05,
+      "loss": 0.3582,
+      "step": 83
+    },
+    {
+      "epoch": 0.03984819734345351,
+      "grad_norm": 6.373027324676514,
+      "learning_rate": 1.3228346456692915e-05,
+      "loss": 0.4211,
+      "step": 84
+    },
+    {
+      "epoch": 0.04032258064516129,
+      "grad_norm": 4.993396759033203,
+      "learning_rate": 1.3385826771653545e-05,
+      "loss": 0.417,
+      "step": 85
+    },
+    {
+      "epoch": 0.04079696394686907,
+      "grad_norm": 5.25504207611084,
+      "learning_rate": 1.3543307086614174e-05,
+      "loss": 0.4558,
+      "step": 86
+    },
+    {
+      "epoch": 0.04127134724857685,
+      "grad_norm": 5.829912185668945,
+      "learning_rate": 1.3700787401574804e-05,
+      "loss": 0.473,
+      "step": 87
+    },
+    {
+      "epoch": 0.04174573055028463,
+      "grad_norm": 5.047908782958984,
+      "learning_rate": 1.3858267716535433e-05,
+      "loss": 0.4056,
+      "step": 88
+    },
+    {
+      "epoch": 0.04222011385199241,
+      "grad_norm": 5.062896251678467,
+      "learning_rate": 1.4015748031496063e-05,
+      "loss": 0.4264,
+      "step": 89
+    },
+    {
+      "epoch": 0.04269449715370019,
+      "grad_norm": 5.646873950958252,
+      "learning_rate": 1.4173228346456694e-05,
+      "loss": 0.4772,
+      "step": 90
+    },
+    {
+      "epoch": 0.04316888045540797,
+      "grad_norm": 4.573570728302002,
+      "learning_rate": 1.4330708661417324e-05,
+      "loss": 0.4483,
+      "step": 91
+    },
+    {
+      "epoch": 0.04364326375711575,
+      "grad_norm": 5.0274128913879395,
+      "learning_rate": 1.4488188976377955e-05,
+      "loss": 0.4248,
+      "step": 92
+    },
+    {
+      "epoch": 0.04411764705882353,
+      "grad_norm": 5.646190643310547,
+      "learning_rate": 1.4645669291338584e-05,
+      "loss": 0.5235,
+      "step": 93
+    },
+    {
+      "epoch": 0.04459203036053131,
+      "grad_norm": 5.01361083984375,
+      "learning_rate": 1.4803149606299214e-05,
+      "loss": 0.4591,
+      "step": 94
+    },
+    {
+      "epoch": 0.04506641366223909,
+      "grad_norm": 5.54431676864624,
+      "learning_rate": 1.4960629921259843e-05,
+      "loss": 0.4361,
+      "step": 95
+    },
+    {
+      "epoch": 0.04554079696394687,
+      "grad_norm": 4.712576389312744,
+      "learning_rate": 1.5118110236220473e-05,
+      "loss": 0.3923,
+      "step": 96
+    },
+    {
+      "epoch": 0.04601518026565465,
+      "grad_norm": 4.848332405090332,
+      "learning_rate": 1.5275590551181102e-05,
+      "loss": 0.4677,
+      "step": 97
+    },
+    {
+      "epoch": 0.046489563567362426,
+      "grad_norm": 5.39124870300293,
+      "learning_rate": 1.5433070866141734e-05,
+      "loss": 0.4506,
+      "step": 98
+    },
+    {
+      "epoch": 0.04696394686907021,
+      "grad_norm": 5.186285972595215,
+      "learning_rate": 1.559055118110236e-05,
+      "loss": 0.4037,
+      "step": 99
+    },
+    {
+      "epoch": 0.04743833017077799,
+      "grad_norm": 4.8791913986206055,
+      "learning_rate": 1.5748031496062993e-05,
+      "loss": 0.4467,
+      "step": 100
+    },
+    {
+      "epoch": 0.04791271347248577,
+      "grad_norm": 5.618642330169678,
+      "learning_rate": 1.5905511811023624e-05,
+      "loss": 0.4959,
+      "step": 101
+    },
+    {
+      "epoch": 0.04838709677419355,
+      "grad_norm": 4.833317279815674,
+      "learning_rate": 1.6062992125984255e-05,
+      "loss": 0.4752,
+      "step": 102
+    },
+    {
+      "epoch": 0.048861480075901326,
+      "grad_norm": 4.536538600921631,
+      "learning_rate": 1.6220472440944883e-05,
+      "loss": 0.4186,
+      "step": 103
+    },
+    {
+      "epoch": 0.04933586337760911,
+      "grad_norm": 5.135328769683838,
+      "learning_rate": 1.6377952755905514e-05,
+      "loss": 0.4471,
+      "step": 104
+    },
+    {
+      "epoch": 0.04981024667931689,
+      "grad_norm": 5.110005855560303,
+      "learning_rate": 1.6535433070866142e-05,
+      "loss": 0.4585,
+      "step": 105
+    },
+    {
+      "epoch": 0.05028462998102467,
+      "grad_norm": 5.040478229522705,
+      "learning_rate": 1.6692913385826773e-05,
+      "loss": 0.4656,
+      "step": 106
+    },
+    {
+      "epoch": 0.05075901328273245,
+      "grad_norm": 4.914176940917969,
+      "learning_rate": 1.68503937007874e-05,
+      "loss": 0.4783,
+      "step": 107
+    },
+    {
+      "epoch": 0.051233396584440226,
+      "grad_norm": 4.72114372253418,
+      "learning_rate": 1.7007874015748032e-05,
+      "loss": 0.4365,
+      "step": 108
+    },
+    {
+      "epoch": 0.051707779886148005,
+      "grad_norm": 4.7739362716674805,
+      "learning_rate": 1.7165354330708663e-05,
+      "loss": 0.4766,
+      "step": 109
+    },
+    {
+      "epoch": 0.05218216318785579,
+      "grad_norm": 4.787454128265381,
+      "learning_rate": 1.7322834645669295e-05,
+      "loss": 0.4542,
+      "step": 110
+    },
+    {
+      "epoch": 0.05265654648956357,
+      "grad_norm": 5.132379055023193,
+      "learning_rate": 1.7480314960629923e-05,
+      "loss": 0.4978,
+      "step": 111
+    },
+    {
+      "epoch": 0.05313092979127135,
+      "grad_norm": 6.812975883483887,
+      "learning_rate": 1.7637795275590554e-05,
+      "loss": 0.4559,
+      "step": 112
+    },
+    {
+      "epoch": 0.053605313092979126,
+      "grad_norm": 5.09624719619751,
+      "learning_rate": 1.779527559055118e-05,
+      "loss": 0.5481,
+      "step": 113
+    },
+    {
+      "epoch": 0.054079696394686905,
+      "grad_norm": 5.495835781097412,
+      "learning_rate": 1.7952755905511813e-05,
+      "loss": 0.5282,
+      "step": 114
+    },
+    {
+      "epoch": 0.05455407969639469,
+      "grad_norm": 6.671237468719482,
+      "learning_rate": 1.811023622047244e-05,
+      "loss": 0.5136,
+      "step": 115
+    },
+    {
+      "epoch": 0.05502846299810247,
+      "grad_norm": 7.128397464752197,
+      "learning_rate": 1.8267716535433072e-05,
+      "loss": 0.5009,
+      "step": 116
+    },
+    {
+      "epoch": 0.05550284629981025,
+      "grad_norm": 4.9932026863098145,
+      "learning_rate": 1.8425196850393703e-05,
+      "loss": 0.4473,
+      "step": 117
+    },
+    {
+      "epoch": 0.055977229601518026,
+      "grad_norm": 5.382380485534668,
+      "learning_rate": 1.858267716535433e-05,
+      "loss": 0.5041,
+      "step": 118
+    },
+    {
+      "epoch": 0.056451612903225805,
+      "grad_norm": 4.284328937530518,
+      "learning_rate": 1.8740157480314962e-05,
+      "loss": 0.4731,
+      "step": 119
+    },
+    {
+      "epoch": 0.056925996204933584,
+      "grad_norm": 6.466545104980469,
+      "learning_rate": 1.8897637795275593e-05,
+      "loss": 0.5566,
+      "step": 120
+    },
+    {
+      "epoch": 0.05740037950664137,
+      "grad_norm": 4.7638959884643555,
+      "learning_rate": 1.905511811023622e-05,
+      "loss": 0.5096,
+      "step": 121
+    },
+    {
+      "epoch": 0.05787476280834915,
+      "grad_norm": 5.770228862762451,
+      "learning_rate": 1.9212598425196852e-05,
+      "loss": 0.5628,
+      "step": 122
+    },
+    {
+      "epoch": 0.058349146110056926,
+      "grad_norm": 5.023980140686035,
+      "learning_rate": 1.937007874015748e-05,
+      "loss": 0.4875,
+      "step": 123
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 5.017863750457764,
+      "learning_rate": 1.952755905511811e-05,
+      "loss": 0.5002,
+      "step": 124
+    },
+    {
+      "epoch": 0.059297912713472484,
+      "grad_norm": 5.194025039672852,
+      "learning_rate": 1.9685039370078743e-05,
+      "loss": 0.4668,
+      "step": 125
+    },
+    {
+      "epoch": 0.05977229601518026,
+      "grad_norm": 5.176628589630127,
+      "learning_rate": 1.984251968503937e-05,
+      "loss": 0.4949,
+      "step": 126
+    },
+    {
+      "epoch": 0.06024667931688805,
+      "grad_norm": 5.499045372009277,
+      "learning_rate": 2e-05,
+      "loss": 0.502,
+      "step": 127
+    },
+    {
+      "epoch": 0.06072106261859583,
+      "grad_norm": 4.964819431304932,
+      "learning_rate": 1.999999704854948e-05,
+      "loss": 0.5428,
+      "step": 128
+    },
+    {
+      "epoch": 0.061195445920303605,
+      "grad_norm": 4.875455379486084,
+      "learning_rate": 1.9999988194199653e-05,
+      "loss": 0.473,
+      "step": 129
+    },
+    {
+      "epoch": 0.061669829222011384,
+      "grad_norm": 4.419773578643799,
+      "learning_rate": 1.999997343695575e-05,
+      "loss": 0.4555,
+      "step": 130
+    },
+    {
+      "epoch": 0.06214421252371916,
+      "grad_norm": 5.253462314605713,
+      "learning_rate": 1.999995277682648e-05,
+      "loss": 0.563,
+      "step": 131
+    },
+    {
+      "epoch": 0.06261859582542695,
+      "grad_norm": 4.706827640533447,
+      "learning_rate": 1.999992621382404e-05,
+      "loss": 0.5296,
+      "step": 132
+    },
+    {
+      "epoch": 0.06309297912713473,
+      "grad_norm": 4.565430641174316,
+      "learning_rate": 1.9999893747964108e-05,
+      "loss": 0.4412,
+      "step": 133
+    },
+    {
+      "epoch": 0.0635673624288425,
+      "grad_norm": 4.702053070068359,
+      "learning_rate": 1.9999855379265855e-05,
+      "loss": 0.5513,
+      "step": 134
+    },
+    {
+      "epoch": 0.06404174573055028,
+      "grad_norm": 4.838067531585693,
+      "learning_rate": 1.999981110775192e-05,
+      "loss": 0.5562,
+      "step": 135
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 4.499827861785889,
+      "learning_rate": 1.9999760933448443e-05,
+      "loss": 0.549,
+      "step": 136
+    },
+    {
+      "epoch": 0.06499051233396584,
+      "grad_norm": 4.42749547958374,
+      "learning_rate": 1.9999704856385037e-05,
+      "loss": 0.5549,
+      "step": 137
+    },
+    {
+      "epoch": 0.06546489563567362,
+      "grad_norm": 4.693006992340088,
+      "learning_rate": 1.9999642876594806e-05,
+      "loss": 0.4435,
+      "step": 138
+    },
+    {
+      "epoch": 0.0659392789373814,
+      "grad_norm": 4.892571449279785,
+      "learning_rate": 1.9999574994114336e-05,
+      "loss": 0.5113,
+      "step": 139
+    },
+    {
+      "epoch": 0.06641366223908918,
+      "grad_norm": 4.446290493011475,
+      "learning_rate": 1.9999501208983692e-05,
+      "loss": 0.5456,
+      "step": 140
+    },
+    {
+      "epoch": 0.06688804554079697,
+      "grad_norm": 5.799952507019043,
+      "learning_rate": 1.999942152124644e-05,
+      "loss": 0.517,
+      "step": 141
+    },
+    {
+      "epoch": 0.06736242884250475,
+      "grad_norm": 4.448901653289795,
+      "learning_rate": 1.9999335930949612e-05,
+      "loss": 0.4831,
+      "step": 142
+    },
+    {
+      "epoch": 0.06783681214421253,
+      "grad_norm": 5.490376949310303,
+      "learning_rate": 1.999924443814373e-05,
+      "loss": 0.5247,
+      "step": 143
+    },
+    {
+      "epoch": 0.0683111954459203,
+      "grad_norm": 4.813867568969727,
+      "learning_rate": 1.9999147042882803e-05,
+      "loss": 0.4864,
+      "step": 144
+    },
+    {
+      "epoch": 0.06878557874762808,
+      "grad_norm": 4.6280293464660645,
+      "learning_rate": 1.9999043745224324e-05,
+      "loss": 0.5794,
+      "step": 145
+    },
+    {
+      "epoch": 0.06925996204933586,
+      "grad_norm": 4.530418872833252,
+      "learning_rate": 1.9998934545229266e-05,
+      "loss": 0.5069,
+      "step": 146
+    },
+    {
+      "epoch": 0.06973434535104364,
+      "grad_norm": 5.701735019683838,
+      "learning_rate": 1.9998819442962088e-05,
+      "loss": 0.657,
+      "step": 147
+    },
+    {
+      "epoch": 0.07020872865275142,
+      "grad_norm": 5.44976282119751,
+      "learning_rate": 1.999869843849074e-05,
+      "loss": 0.5559,
+      "step": 148
+    },
+    {
+      "epoch": 0.0706831119544592,
+      "grad_norm": 4.931896686553955,
+      "learning_rate": 1.999857153188664e-05,
+      "loss": 0.527,
+      "step": 149
+    },
+    {
+      "epoch": 0.07115749525616698,
+      "grad_norm": 4.865047931671143,
+      "learning_rate": 1.999843872322471e-05,
+      "loss": 0.5177,
+      "step": 150
+    },
+    {
+      "epoch": 0.07163187855787476,
+      "grad_norm": 4.696810245513916,
+      "learning_rate": 1.9998300012583333e-05,
+      "loss": 0.5715,
+      "step": 151
+    },
+    {
+      "epoch": 0.07210626185958255,
+      "grad_norm": 4.350542068481445,
+      "learning_rate": 1.99981554000444e-05,
+      "loss": 0.4627,
+      "step": 152
+    },
+    {
+      "epoch": 0.07258064516129033,
+      "grad_norm": 5.0633111000061035,
+      "learning_rate": 1.999800488569327e-05,
+      "loss": 0.5614,
+      "step": 153
+    },
+    {
+      "epoch": 0.0730550284629981,
+      "grad_norm": 4.741815090179443,
+      "learning_rate": 1.999784846961879e-05,
+      "loss": 0.5093,
+      "step": 154
+    },
+    {
+      "epoch": 0.07352941176470588,
+      "grad_norm": 6.608360290527344,
+      "learning_rate": 1.9997686151913297e-05,
+      "loss": 0.497,
+      "step": 155
+    },
+    {
+      "epoch": 0.07400379506641366,
+      "grad_norm": 5.2583441734313965,
+      "learning_rate": 1.9997517932672592e-05,
+      "loss": 0.5925,
+      "step": 156
+    },
+    {
+      "epoch": 0.07447817836812144,
+      "grad_norm": 4.322223663330078,
+      "learning_rate": 1.9997343811995985e-05,
+      "loss": 0.5369,
+      "step": 157
+    },
+    {
+      "epoch": 0.07495256166982922,
+      "grad_norm": 4.527962684631348,
+      "learning_rate": 1.9997163789986255e-05,
+      "loss": 0.5313,
+      "step": 158
+    },
+    {
+      "epoch": 0.075426944971537,
+      "grad_norm": 4.953525066375732,
+      "learning_rate": 1.999697786674966e-05,
+      "loss": 0.5229,
+      "step": 159
+    },
+    {
+      "epoch": 0.07590132827324478,
+      "grad_norm": 5.049784183502197,
+      "learning_rate": 1.999678604239596e-05,
+      "loss": 0.4546,
+      "step": 160
+    },
+    {
+      "epoch": 0.07637571157495256,
+      "grad_norm": 5.497480392456055,
+      "learning_rate": 1.9996588317038382e-05,
+      "loss": 0.5097,
+      "step": 161
+    },
+    {
+      "epoch": 0.07685009487666034,
+      "grad_norm": 4.220514297485352,
+      "learning_rate": 1.9996384690793634e-05,
+      "loss": 0.471,
+      "step": 162
+    },
+    {
+      "epoch": 0.07732447817836813,
+      "grad_norm": 4.465171813964844,
+      "learning_rate": 1.999617516378193e-05,
+      "loss": 0.4938,
+      "step": 163
+    },
+    {
+      "epoch": 0.0777988614800759,
+      "grad_norm": 5.333194255828857,
+      "learning_rate": 1.999595973612694e-05,
+      "loss": 0.5676,
+      "step": 164
+    },
+    {
+      "epoch": 0.07827324478178369,
+      "grad_norm": 4.7594780921936035,
+      "learning_rate": 1.999573840795583e-05,
+      "loss": 0.5586,
+      "step": 165
+    },
+    {
+      "epoch": 0.07874762808349146,
+      "grad_norm": 4.380119800567627,
+      "learning_rate": 1.9995511179399253e-05,
+      "loss": 0.5206,
+      "step": 166
+    },
+    {
+      "epoch": 0.07922201138519924,
+      "grad_norm": 5.4240922927856445,
+      "learning_rate": 1.9995278050591334e-05,
+      "loss": 0.6165,
+      "step": 167
+    },
+    {
+      "epoch": 0.07969639468690702,
+      "grad_norm": 5.2482075691223145,
+      "learning_rate": 1.9995039021669692e-05,
+      "loss": 0.5472,
+      "step": 168
+    },
+    {
+      "epoch": 0.0801707779886148,
+      "grad_norm": 4.465966701507568,
+      "learning_rate": 1.9994794092775418e-05,
+      "loss": 0.5151,
+      "step": 169
+    },
+    {
+      "epoch": 0.08064516129032258,
+      "grad_norm": 4.838151454925537,
+      "learning_rate": 1.9994543264053093e-05,
+      "loss": 0.5031,
+      "step": 170
+    },
+    {
+      "epoch": 0.08111954459203036,
+      "grad_norm": 4.365642070770264,
+      "learning_rate": 1.9994286535650782e-05,
+      "loss": 0.4621,
+      "step": 171
+    },
+    {
+      "epoch": 0.08159392789373814,
+      "grad_norm": 5.1148152351379395,
+      "learning_rate": 1.9994023907720027e-05,
+      "loss": 0.5432,
+      "step": 172
+    },
+    {
+      "epoch": 0.08206831119544591,
+      "grad_norm": 4.51958703994751,
+      "learning_rate": 1.9993755380415854e-05,
+      "loss": 0.526,
+      "step": 173
+    },
+    {
+      "epoch": 0.0825426944971537,
+      "grad_norm": 3.9636707305908203,
+      "learning_rate": 1.999348095389677e-05,
+      "loss": 0.4614,
+      "step": 174
+    },
+    {
+      "epoch": 0.08301707779886149,
+      "grad_norm": 5.010396480560303,
+      "learning_rate": 1.999320062832477e-05,
+      "loss": 0.5862,
+      "step": 175
+    },
+    {
+      "epoch": 0.08349146110056926,
+      "grad_norm": 4.869819164276123,
+      "learning_rate": 1.9992914403865327e-05,
+      "loss": 0.5009,
+      "step": 176
+    },
+    {
+      "epoch": 0.08396584440227704,
+      "grad_norm": 5.5586838722229,
+      "learning_rate": 1.9992622280687395e-05,
+      "loss": 0.5472,
+      "step": 177
+    },
+    {
+      "epoch": 0.08444022770398482,
+      "grad_norm": 4.700314998626709,
+      "learning_rate": 1.9992324258963414e-05,
+      "loss": 0.5833,
+      "step": 178
+    },
+    {
+      "epoch": 0.0849146110056926,
+      "grad_norm": 5.759016990661621,
+      "learning_rate": 1.99920203388693e-05,
+      "loss": 0.6045,
+      "step": 179
+    },
+    {
+      "epoch": 0.08538899430740038,
+      "grad_norm": 5.080048084259033,
+      "learning_rate": 1.999171052058445e-05,
+      "loss": 0.5256,
+      "step": 180
+    },
+    {
+      "epoch": 0.08586337760910816,
+      "grad_norm": 4.277463912963867,
+      "learning_rate": 1.999139480429176e-05,
+      "loss": 0.5763,
+      "step": 181
+    },
+    {
+      "epoch": 0.08633776091081594,
+      "grad_norm": 5.613553524017334,
+      "learning_rate": 1.999107319017758e-05,
+      "loss": 0.5423,
+      "step": 182
+    },
+    {
+      "epoch": 0.08681214421252371,
+      "grad_norm": 4.811156272888184,
+      "learning_rate": 1.9990745678431765e-05,
+      "loss": 0.4214,
+      "step": 183
+    },
+    {
+      "epoch": 0.0872865275142315,
+      "grad_norm": 4.72564697265625,
+      "learning_rate": 1.9990412269247637e-05,
+      "loss": 0.466,
+      "step": 184
+    },
+    {
+      "epoch": 0.08776091081593927,
+      "grad_norm": 7.244020938873291,
+      "learning_rate": 1.999007296282201e-05,
+      "loss": 0.5073,
+      "step": 185
+    },
+    {
+      "epoch": 0.08823529411764706,
+      "grad_norm": 4.737941741943359,
+      "learning_rate": 1.9989727759355164e-05,
+      "loss": 0.5523,
+      "step": 186
+    },
+    {
+      "epoch": 0.08870967741935484,
+      "grad_norm": 4.7859086990356445,
+      "learning_rate": 1.9989376659050878e-05,
+      "loss": 0.5216,
+      "step": 187
+    },
+    {
+      "epoch": 0.08918406072106262,
+      "grad_norm": 4.781891345977783,
+      "learning_rate": 1.9989019662116395e-05,
+      "loss": 0.5328,
+      "step": 188
+    },
+    {
+      "epoch": 0.0896584440227704,
+      "grad_norm": 4.595061779022217,
+      "learning_rate": 1.9988656768762455e-05,
+      "loss": 0.5584,
+      "step": 189
+    },
+    {
+      "epoch": 0.09013282732447818,
+      "grad_norm": 4.338557720184326,
+      "learning_rate": 1.9988287979203264e-05,
+      "loss": 0.4839,
+      "step": 190
+    },
+    {
+      "epoch": 0.09060721062618596,
+      "grad_norm": 4.124628067016602,
+      "learning_rate": 1.9987913293656515e-05,
+      "loss": 0.5612,
+      "step": 191
+    },
+    {
+      "epoch": 0.09108159392789374,
+      "grad_norm": 3.821711778640747,
+      "learning_rate": 1.9987532712343388e-05,
+      "loss": 0.4442,
+      "step": 192
+    },
+    {
+      "epoch": 0.09155597722960152,
+      "grad_norm": 3.7175636291503906,
+      "learning_rate": 1.9987146235488532e-05,
+      "loss": 0.5021,
+      "step": 193
+    },
+    {
+      "epoch": 0.0920303605313093,
+      "grad_norm": 4.476022720336914,
+      "learning_rate": 1.9986753863320077e-05,
+      "loss": 0.524,
+      "step": 194
+    },
+    {
+      "epoch": 0.09250474383301707,
+      "grad_norm": 4.444040298461914,
+      "learning_rate": 1.998635559606964e-05,
+      "loss": 0.4565,
+      "step": 195
+    },
+    {
+      "epoch": 0.09297912713472485,
+      "grad_norm": 4.800974369049072,
+      "learning_rate": 1.9985951433972313e-05,
+      "loss": 0.5845,
+      "step": 196
+    },
+    {
+      "epoch": 0.09345351043643264,
+      "grad_norm": 4.884645462036133,
+      "learning_rate": 1.9985541377266675e-05,
+      "loss": 0.5319,
+      "step": 197
+    },
+    {
+      "epoch": 0.09392789373814042,
+      "grad_norm": 4.790307521820068,
+      "learning_rate": 1.998512542619477e-05,
+      "loss": 0.505,
+      "step": 198
+    },
+    {
+      "epoch": 0.0944022770398482,
+      "grad_norm": 4.581803321838379,
+      "learning_rate": 1.998470358100213e-05,
+      "loss": 0.5228,
+      "step": 199
+    },
+    {
+      "epoch": 0.09487666034155598,
+      "grad_norm": 4.337541580200195,
+      "learning_rate": 1.9984275841937776e-05,
+      "loss": 0.5135,
+      "step": 200
+    },
+    {
+      "epoch": 0.09535104364326376,
+      "grad_norm": 4.175999641418457,
+      "learning_rate": 1.9983842209254182e-05,
+      "loss": 0.5219,
+      "step": 201
+    },
+    {
+      "epoch": 0.09582542694497154,
+      "grad_norm": 3.8409905433654785,
+      "learning_rate": 1.9983402683207334e-05,
+      "loss": 0.4495,
+      "step": 202
+    },
+    {
+      "epoch": 0.09629981024667932,
+      "grad_norm": 3.7136104106903076,
+      "learning_rate": 1.9982957264056667e-05,
+      "loss": 0.5357,
+      "step": 203
+    },
+    {
+      "epoch": 0.0967741935483871,
+      "grad_norm": 4.796589374542236,
+      "learning_rate": 1.9982505952065115e-05,
+      "loss": 0.5567,
+      "step": 204
+    },
+    {
+      "epoch": 0.09724857685009487,
+      "grad_norm": 5.0165886878967285,
+      "learning_rate": 1.9982048747499082e-05,
+      "loss": 0.5025,
+      "step": 205
+    },
+    {
+      "epoch": 0.09772296015180265,
+      "grad_norm": 4.150423049926758,
+      "learning_rate": 1.9981585650628447e-05,
+      "loss": 0.5405,
+      "step": 206
+    },
+    {
+      "epoch": 0.09819734345351043,
+      "grad_norm": 3.635175943374634,
+      "learning_rate": 1.9981116661726575e-05,
+      "loss": 0.4784,
+      "step": 207
+    },
+    {
+      "epoch": 0.09867172675521822,
+      "grad_norm": 4.339625358581543,
+      "learning_rate": 1.998064178107031e-05,
+      "loss": 0.5831,
+      "step": 208
+    },
+    {
+      "epoch": 0.099146110056926,
+      "grad_norm": 4.854864120483398,
+      "learning_rate": 1.9980161008939957e-05,
+      "loss": 0.6303,
+      "step": 209
+    },
+    {
+      "epoch": 0.09962049335863378,
+      "grad_norm": 4.287778377532959,
+      "learning_rate": 1.9979674345619322e-05,
+      "loss": 0.467,
+      "step": 210
+    },
+    {
+      "epoch": 0.10009487666034156,
+      "grad_norm": 3.7619311809539795,
+      "learning_rate": 1.997918179139567e-05,
+      "loss": 0.5403,
+      "step": 211
+    },
+    {
+      "epoch": 0.10056925996204934,
+      "grad_norm": 3.486117124557495,
+      "learning_rate": 1.9978683346559762e-05,
+      "loss": 0.4819,
+      "step": 212
+    },
+    {
+      "epoch": 0.10104364326375712,
+      "grad_norm": 4.301611423492432,
+      "learning_rate": 1.9978179011405814e-05,
+      "loss": 0.5293,
+      "step": 213
+    },
+    {
+      "epoch": 0.1015180265654649,
+      "grad_norm": 4.88173770904541,
+      "learning_rate": 1.9977668786231536e-05,
+      "loss": 0.5506,
+      "step": 214
+    },
+    {
+      "epoch": 0.10199240986717267,
+      "grad_norm": 4.954184532165527,
+      "learning_rate": 1.99771526713381e-05,
+      "loss": 0.5209,
+      "step": 215
+    },
+    {
+      "epoch": 0.10246679316888045,
+      "grad_norm": 4.852926254272461,
+      "learning_rate": 1.9976630667030175e-05,
+      "loss": 0.5922,
+      "step": 216
+    },
+    {
+      "epoch": 0.10294117647058823,
+      "grad_norm": 5.19386100769043,
+      "learning_rate": 1.9976102773615894e-05,
+      "loss": 0.5261,
+      "step": 217
+    },
+    {
+      "epoch": 0.10341555977229601,
+      "grad_norm": 4.068961143493652,
+      "learning_rate": 1.997556899140686e-05,
+      "loss": 0.5104,
+      "step": 218
+    },
+    {
+      "epoch": 0.1038899430740038,
+      "grad_norm": 3.790403366088867,
+      "learning_rate": 1.997502932071816e-05,
+      "loss": 0.4915,
+      "step": 219
+    },
+    {
+      "epoch": 0.10436432637571158,
+      "grad_norm": 5.95817756652832,
+      "learning_rate": 1.997448376186836e-05,
+      "loss": 0.4801,
+      "step": 220
+    },
+    {
+      "epoch": 0.10483870967741936,
+      "grad_norm": 4.333256244659424,
+      "learning_rate": 1.9973932315179502e-05,
+      "loss": 0.5602,
+      "step": 221
+    },
+    {
+      "epoch": 0.10531309297912714,
+      "grad_norm": 3.932823419570923,
+      "learning_rate": 1.997337498097709e-05,
+      "loss": 0.5351,
+      "step": 222
+    },
+    {
+      "epoch": 0.10578747628083492,
+      "grad_norm": 4.296048164367676,
+      "learning_rate": 1.9972811759590117e-05,
+      "loss": 0.5502,
+      "step": 223
+    },
+    {
+      "epoch": 0.1062618595825427,
+      "grad_norm": 4.166548252105713,
+      "learning_rate": 1.997224265135105e-05,
+      "loss": 0.5277,
+      "step": 224
+    },
+    {
+      "epoch": 0.10673624288425047,
+      "grad_norm": 5.517498016357422,
+      "learning_rate": 1.9971667656595824e-05,
+      "loss": 0.5458,
+      "step": 225
+    },
+    {
+      "epoch": 0.10721062618595825,
+      "grad_norm": 4.525831699371338,
+      "learning_rate": 1.9971086775663856e-05,
+      "loss": 0.5935,
+      "step": 226
+    },
+    {
+      "epoch": 0.10768500948766603,
+      "grad_norm": 4.5020012855529785,
+      "learning_rate": 1.997050000889803e-05,
+      "loss": 0.4852,
+      "step": 227
+    },
+    {
+      "epoch": 0.10815939278937381,
+      "grad_norm": 4.161679744720459,
+      "learning_rate": 1.9969907356644716e-05,
+      "loss": 0.5776,
+      "step": 228
+    },
+    {
+      "epoch": 0.10863377609108159,
+      "grad_norm": 5.031361103057861,
+      "learning_rate": 1.996930881925374e-05,
+      "loss": 0.5912,
+      "step": 229
+    },
+    {
+      "epoch": 0.10910815939278938,
+      "grad_norm": 4.38935661315918,
+      "learning_rate": 1.9968704397078422e-05,
+      "loss": 0.5005,
+      "step": 230
+    },
+    {
+      "epoch": 0.10958254269449716,
+      "grad_norm": 4.4222412109375,
+      "learning_rate": 1.996809409047554e-05,
+      "loss": 0.5669,
+      "step": 231
+    },
+    {
+      "epoch": 0.11005692599620494,
+      "grad_norm": 4.336076736450195,
+      "learning_rate": 1.996747789980536e-05,
+      "loss": 0.5975,
+      "step": 232
+    },
+    {
+      "epoch": 0.11053130929791272,
+      "grad_norm": 4.671880722045898,
+      "learning_rate": 1.9966855825431605e-05,
+      "loss": 0.6065,
+      "step": 233
+    },
+    {
+      "epoch": 0.1110056925996205,
+      "grad_norm": 3.9240012168884277,
+      "learning_rate": 1.996622786772148e-05,
+      "loss": 0.474,
+      "step": 234
+    },
+    {
+      "epoch": 0.11148007590132827,
+      "grad_norm": 4.048487663269043,
+      "learning_rate": 1.9965594027045668e-05,
+      "loss": 0.5361,
+      "step": 235
+    },
+    {
+      "epoch": 0.11195445920303605,
+      "grad_norm": 3.9906532764434814,
+      "learning_rate": 1.996495430377831e-05,
+      "loss": 0.6107,
+      "step": 236
+    },
+    {
+      "epoch": 0.11242884250474383,
+      "grad_norm": 4.5680084228515625,
+      "learning_rate": 1.996430869829704e-05,
+      "loss": 0.4812,
+      "step": 237
+    },
+    {
+      "epoch": 0.11290322580645161,
+      "grad_norm": 4.888277530670166,
+      "learning_rate": 1.9963657210982947e-05,
+      "loss": 0.479,
+      "step": 238
+    },
+    {
+      "epoch": 0.11337760910815939,
+      "grad_norm": 3.874725103378296,
+      "learning_rate": 1.9962999842220596e-05,
+      "loss": 0.5418,
+      "step": 239
+    },
+    {
+      "epoch": 0.11385199240986717,
+      "grad_norm": 4.379036903381348,
+      "learning_rate": 1.9962336592398027e-05,
+      "loss": 0.4977,
+      "step": 240
+    },
+    {
+      "epoch": 0.11432637571157495,
+      "grad_norm": 4.229637145996094,
+      "learning_rate": 1.9961667461906743e-05,
+      "loss": 0.5353,
+      "step": 241
+    },
+    {
+      "epoch": 0.11480075901328274,
+      "grad_norm": 4.034006118774414,
+      "learning_rate": 1.9960992451141737e-05,
+      "loss": 0.5121,
+      "step": 242
+    },
+    {
+      "epoch": 0.11527514231499052,
+      "grad_norm": 4.3000288009643555,
+      "learning_rate": 1.9960311560501457e-05,
+      "loss": 0.6312,
+      "step": 243
+    },
+    {
+      "epoch": 0.1157495256166983,
+      "grad_norm": 4.055355072021484,
+      "learning_rate": 1.995962479038782e-05,
+      "loss": 0.5626,
+      "step": 244
+    },
+    {
+      "epoch": 0.11622390891840607,
+      "grad_norm": 4.0171990394592285,
+      "learning_rate": 1.9958932141206224e-05,
+      "loss": 0.4723,
+      "step": 245
+    },
+    {
+      "epoch": 0.11669829222011385,
+      "grad_norm": 4.504859924316406,
+      "learning_rate": 1.9958233613365534e-05,
+      "loss": 0.5482,
+      "step": 246
+    },
+    {
+      "epoch": 0.11717267552182163,
+      "grad_norm": 4.141243934631348,
+      "learning_rate": 1.9957529207278082e-05,
+      "loss": 0.4875,
+      "step": 247
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 4.3180389404296875,
+      "learning_rate": 1.9956818923359677e-05,
+      "loss": 0.5515,
+      "step": 248
+    },
+    {
+      "epoch": 0.11812144212523719,
+      "grad_norm": 3.9524307250976562,
+      "learning_rate": 1.995610276202958e-05,
+      "loss": 0.5466,
+      "step": 249
+    },
+    {
+      "epoch": 0.11859582542694497,
+      "grad_norm": 3.8721537590026855,
+      "learning_rate": 1.995538072371055e-05,
+      "loss": 0.6034,
+      "step": 250
+    },
+    {
+      "epoch": 0.11907020872865275,
+      "grad_norm": 4.435497283935547,
+      "learning_rate": 1.9954652808828793e-05,
+      "loss": 0.5693,
+      "step": 251
+    },
+    {
+      "epoch": 0.11954459203036052,
+      "grad_norm": 3.3445560932159424,
+      "learning_rate": 1.9953919017813985e-05,
+      "loss": 0.3989,
+      "step": 252
+    },
+    {
+      "epoch": 0.12001897533206832,
+      "grad_norm": 3.7608346939086914,
+      "learning_rate": 1.9953179351099276e-05,
+      "loss": 0.5034,
+      "step": 253
+    },
+    {
+      "epoch": 0.1204933586337761,
+      "grad_norm": 4.122411727905273,
+      "learning_rate": 1.995243380912129e-05,
+      "loss": 0.5092,
+      "step": 254
+    },
+    {
+      "epoch": 0.12096774193548387,
+      "grad_norm": 4.1877007484436035,
+      "learning_rate": 1.995168239232011e-05,
+      "loss": 0.5847,
+      "step": 255
+    },
+    {
+      "epoch": 0.12144212523719165,
+      "grad_norm": 4.213406562805176,
+      "learning_rate": 1.9950925101139292e-05,
+      "loss": 0.4321,
+      "step": 256
+    },
+    {
+      "epoch": 0.12191650853889943,
+      "grad_norm": 3.9307830333709717,
+      "learning_rate": 1.995016193602585e-05,
+      "loss": 0.5166,
+      "step": 257
+    },
+    {
+      "epoch": 0.12239089184060721,
+      "grad_norm": 4.194809913635254,
+      "learning_rate": 1.9949392897430283e-05,
+      "loss": 0.5193,
+      "step": 258
+    },
+    {
+      "epoch": 0.12286527514231499,
+      "grad_norm": 5.011254787445068,
+      "learning_rate": 1.994861798580654e-05,
+      "loss": 0.6044,
+      "step": 259
+    },
+    {
+      "epoch": 0.12333965844402277,
+      "grad_norm": 3.800550937652588,
+      "learning_rate": 1.9947837201612046e-05,
+      "loss": 0.5104,
+      "step": 260
+    },
+    {
+      "epoch": 0.12381404174573055,
+      "grad_norm": 4.209667682647705,
+      "learning_rate": 1.9947050545307693e-05,
+      "loss": 0.5558,
+      "step": 261
+    },
+    {
+      "epoch": 0.12428842504743833,
+      "grad_norm": 4.35175085067749,
+      "learning_rate": 1.994625801735783e-05,
+      "loss": 0.4245,
+      "step": 262
+    },
+    {
+      "epoch": 0.1247628083491461,
+      "grad_norm": 4.295873165130615,
+      "learning_rate": 1.9945459618230282e-05,
+      "loss": 0.5044,
+      "step": 263
+    },
+    {
+      "epoch": 0.1252371916508539,
+      "grad_norm": 4.467810153961182,
+      "learning_rate": 1.9944655348396336e-05,
+      "loss": 0.5308,
+      "step": 264
+    },
+    {
+      "epoch": 0.12571157495256166,
+      "grad_norm": 3.797988176345825,
+      "learning_rate": 1.9943845208330742e-05,
+      "loss": 0.4543,
+      "step": 265
+    },
+    {
+      "epoch": 0.12618595825426945,
+      "grad_norm": 3.5803213119506836,
+      "learning_rate": 1.9943029198511724e-05,
+      "loss": 0.4598,
+      "step": 266
+    },
+    {
+      "epoch": 0.12666034155597722,
+      "grad_norm": 3.6839261054992676,
+      "learning_rate": 1.9942207319420962e-05,
+      "loss": 0.5384,
+      "step": 267
+    },
+    {
+      "epoch": 0.127134724857685,
+      "grad_norm": 3.6129837036132812,
+      "learning_rate": 1.9941379571543597e-05,
+      "loss": 0.5078,
+      "step": 268
+    },
+    {
+      "epoch": 0.12760910815939278,
+      "grad_norm": 4.039849281311035,
+      "learning_rate": 1.9940545955368247e-05,
+      "loss": 0.6595,
+      "step": 269
+    },
+    {
+      "epoch": 0.12808349146110057,
+      "grad_norm": 4.292903423309326,
+      "learning_rate": 1.993970647138699e-05,
+      "loss": 0.6002,
+      "step": 270
+    },
+    {
+      "epoch": 0.12855787476280836,
+      "grad_norm": 3.4465274810791016,
+      "learning_rate": 1.9938861120095353e-05,
+      "loss": 0.4465,
+      "step": 271
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 5.027604579925537,
+      "learning_rate": 1.993800990199235e-05,
+      "loss": 0.5447,
+      "step": 272
+    },
+    {
+      "epoch": 0.12950664136622392,
+      "grad_norm": 4.433981418609619,
+      "learning_rate": 1.993715281758044e-05,
+      "loss": 0.5327,
+      "step": 273
+    },
+    {
+      "epoch": 0.12998102466793168,
+      "grad_norm": 4.061509609222412,
+      "learning_rate": 1.9936289867365557e-05,
+      "loss": 0.4561,
+      "step": 274
+    },
+    {
+      "epoch": 0.13045540796963948,
+      "grad_norm": 4.023436546325684,
+      "learning_rate": 1.9935421051857088e-05,
+      "loss": 0.5113,
+      "step": 275
+    },
+    {
+      "epoch": 0.13092979127134724,
+      "grad_norm": 4.111091613769531,
+      "learning_rate": 1.9934546371567888e-05,
+      "loss": 0.5168,
+      "step": 276
+    },
+    {
+      "epoch": 0.13140417457305503,
+      "grad_norm": 4.130946636199951,
+      "learning_rate": 1.9933665827014272e-05,
+      "loss": 0.5131,
+      "step": 277
+    },
+    {
+      "epoch": 0.1318785578747628,
+      "grad_norm": 4.0225348472595215,
+      "learning_rate": 1.9932779418716012e-05,
+      "loss": 0.4993,
+      "step": 278
+    },
+    {
+      "epoch": 0.1323529411764706,
+      "grad_norm": 4.16855525970459,
+      "learning_rate": 1.9931887147196355e-05,
+      "loss": 0.4827,
+      "step": 279
+    },
+    {
+      "epoch": 0.13282732447817835,
+      "grad_norm": 4.637106418609619,
+      "learning_rate": 1.9930989012981992e-05,
+      "loss": 0.4109,
+      "step": 280
+    },
+    {
+      "epoch": 0.13330170777988615,
+      "grad_norm": 3.4839680194854736,
+      "learning_rate": 1.993008501660309e-05,
+      "loss": 0.4779,
+      "step": 281
+    },
+    {
+      "epoch": 0.13377609108159394,
+      "grad_norm": 4.075927734375,
+      "learning_rate": 1.9929175158593262e-05,
+      "loss": 0.531,
+      "step": 282
+    },
+    {
+      "epoch": 0.1342504743833017,
+      "grad_norm": 4.005972385406494,
+      "learning_rate": 1.992825943948959e-05,
+      "loss": 0.5608,
+      "step": 283
+    },
+    {
+      "epoch": 0.1347248576850095,
+      "grad_norm": 3.9188387393951416,
+      "learning_rate": 1.9927337859832617e-05,
+      "loss": 0.5646,
+      "step": 284
+    },
+    {
+      "epoch": 0.13519924098671726,
+      "grad_norm": 3.8422210216522217,
+      "learning_rate": 1.9926410420166343e-05,
+      "loss": 0.4807,
+      "step": 285
+    },
+    {
+      "epoch": 0.13567362428842505,
+      "grad_norm": 4.120503902435303,
+      "learning_rate": 1.9925477121038218e-05,
+      "loss": 0.6104,
+      "step": 286
+    },
+    {
+      "epoch": 0.13614800759013282,
+      "grad_norm": 4.473390579223633,
+      "learning_rate": 1.992453796299917e-05,
+      "loss": 0.5511,
+      "step": 287
+    },
+    {
+      "epoch": 0.1366223908918406,
+      "grad_norm": 4.302347183227539,
+      "learning_rate": 1.9923592946603573e-05,
+      "loss": 0.4618,
+      "step": 288
+    },
+    {
+      "epoch": 0.13709677419354838,
+      "grad_norm": 7.132138729095459,
+      "learning_rate": 1.992264207240925e-05,
+      "loss": 0.5165,
+      "step": 289
+    },
+    {
+      "epoch": 0.13757115749525617,
+      "grad_norm": 4.486799716949463,
+      "learning_rate": 1.9921685340977506e-05,
+      "loss": 0.459,
+      "step": 290
+    },
+    {
+      "epoch": 0.13804554079696393,
+      "grad_norm": 4.0745038986206055,
+      "learning_rate": 1.992072275287308e-05,
+      "loss": 0.4965,
+      "step": 291
+    },
+    {
+      "epoch": 0.13851992409867173,
+      "grad_norm": 4.047938346862793,
+      "learning_rate": 1.991975430866419e-05,
+      "loss": 0.5792,
+      "step": 292
+    },
+    {
+      "epoch": 0.13899430740037952,
+      "grad_norm": 4.198129177093506,
+      "learning_rate": 1.9918780008922484e-05,
+      "loss": 0.5546,
+      "step": 293
+    },
+    {
+      "epoch": 0.13946869070208728,
+      "grad_norm": 4.39850378036499,
+      "learning_rate": 1.9917799854223093e-05,
+      "loss": 0.4974,
+      "step": 294
+    },
+    {
+      "epoch": 0.13994307400379508,
+      "grad_norm": 5.012327194213867,
+      "learning_rate": 1.9916813845144587e-05,
+      "loss": 0.5658,
+      "step": 295
+    },
+    {
+      "epoch": 0.14041745730550284,
+      "grad_norm": 4.509042263031006,
+      "learning_rate": 1.9915821982269002e-05,
+      "loss": 0.5774,
+      "step": 296
+    },
+    {
+      "epoch": 0.14089184060721063,
+      "grad_norm": 4.060871124267578,
+      "learning_rate": 1.9914824266181818e-05,
+      "loss": 0.4573,
+      "step": 297
+    },
+    {
+      "epoch": 0.1413662239089184,
+      "grad_norm": 3.875030040740967,
+      "learning_rate": 1.9913820697471988e-05,
+      "loss": 0.4239,
+      "step": 298
+    },
+    {
+      "epoch": 0.1418406072106262,
+      "grad_norm": 4.114848613739014,
+      "learning_rate": 1.9912811276731895e-05,
+      "loss": 0.5666,
+      "step": 299
+    },
+    {
+      "epoch": 0.14231499051233396,
+      "grad_norm": 3.5044398307800293,
+      "learning_rate": 1.9911796004557397e-05,
+      "loss": 0.5189,
+      "step": 300
+    },
+    {
+      "epoch": 0.14278937381404175,
+      "grad_norm": 3.8690919876098633,
+      "learning_rate": 1.9910774881547803e-05,
+      "loss": 0.5543,
+      "step": 301
+    },
+    {
+      "epoch": 0.1432637571157495,
+      "grad_norm": 3.6126976013183594,
+      "learning_rate": 1.9909747908305866e-05,
+      "loss": 0.4635,
+      "step": 302
+    },
+    {
+      "epoch": 0.1437381404174573,
+      "grad_norm": 3.76629900932312,
+      "learning_rate": 1.99087150854378e-05,
+      "loss": 0.5394,
+      "step": 303
+    },
+    {
+      "epoch": 0.1442125237191651,
+      "grad_norm": 4.235199451446533,
+      "learning_rate": 1.990767641355327e-05,
+      "loss": 0.6441,
+      "step": 304
+    },
+    {
+      "epoch": 0.14468690702087286,
+      "grad_norm": 3.634746551513672,
+      "learning_rate": 1.9906631893265393e-05,
+      "loss": 0.5005,
+      "step": 305
+    },
+    {
+      "epoch": 0.14516129032258066,
+      "grad_norm": 3.812934160232544,
+      "learning_rate": 1.990558152519074e-05,
+      "loss": 0.4656,
+      "step": 306
+    },
+    {
+      "epoch": 0.14563567362428842,
+      "grad_norm": 3.4849369525909424,
+      "learning_rate": 1.9904525309949332e-05,
+      "loss": 0.4462,
+      "step": 307
+    },
+    {
+      "epoch": 0.1461100569259962,
+      "grad_norm": 4.69333553314209,
+      "learning_rate": 1.9903463248164643e-05,
+      "loss": 0.5539,
+      "step": 308
+    },
+    {
+      "epoch": 0.14658444022770398,
+      "grad_norm": 4.567856788635254,
+      "learning_rate": 1.99023953404636e-05,
+      "loss": 0.5745,
+      "step": 309
+    },
+    {
+      "epoch": 0.14705882352941177,
+      "grad_norm": 3.9516940116882324,
+      "learning_rate": 1.9901321587476573e-05,
+      "loss": 0.5849,
+      "step": 310
+    },
+    {
+      "epoch": 0.14753320683111953,
+      "grad_norm": 4.662928581237793,
+      "learning_rate": 1.9900241989837395e-05,
+      "loss": 0.57,
+      "step": 311
+    },
+    {
+      "epoch": 0.14800759013282733,
+      "grad_norm": 3.7832348346710205,
+      "learning_rate": 1.9899156548183332e-05,
+      "loss": 0.5359,
+      "step": 312
+    },
+    {
+      "epoch": 0.1484819734345351,
+      "grad_norm": 3.7341370582580566,
+      "learning_rate": 1.9898065263155118e-05,
+      "loss": 0.5341,
+      "step": 313
+    },
+    {
+      "epoch": 0.14895635673624288,
+      "grad_norm": 4.346981048583984,
+      "learning_rate": 1.9896968135396924e-05,
+      "loss": 0.5166,
+      "step": 314
+    },
+    {
+      "epoch": 0.14943074003795068,
+      "grad_norm": 4.063964366912842,
+      "learning_rate": 1.9895865165556375e-05,
+      "loss": 0.5726,
+      "step": 315
+    },
+    {
+      "epoch": 0.14990512333965844,
+      "grad_norm": 3.6817450523376465,
+      "learning_rate": 1.989475635428454e-05,
+      "loss": 0.4844,
+      "step": 316
+    },
+    {
+      "epoch": 0.15037950664136623,
+      "grad_norm": 3.254460096359253,
+      "learning_rate": 1.9893641702235946e-05,
+      "loss": 0.4731,
+      "step": 317
+    },
+    {
+      "epoch": 0.150853889943074,
+      "grad_norm": 4.255951404571533,
+      "learning_rate": 1.9892521210068552e-05,
+      "loss": 0.5561,
+      "step": 318
+    },
+    {
+      "epoch": 0.1513282732447818,
+      "grad_norm": 4.263412952423096,
+      "learning_rate": 1.9891394878443783e-05,
+      "loss": 0.5309,
+      "step": 319
+    },
+    {
+      "epoch": 0.15180265654648956,
+      "grad_norm": 3.499624252319336,
+      "learning_rate": 1.9890262708026497e-05,
+      "loss": 0.5453,
+      "step": 320
+    },
+    {
+      "epoch": 0.15227703984819735,
+      "grad_norm": 4.114099502563477,
+      "learning_rate": 1.9889124699485e-05,
+      "loss": 0.55,
+      "step": 321
+    },
+    {
+      "epoch": 0.1527514231499051,
+      "grad_norm": 3.8856418132781982,
+      "learning_rate": 1.988798085349105e-05,
+      "loss": 0.4584,
+      "step": 322
+    },
+    {
+      "epoch": 0.1532258064516129,
+      "grad_norm": 3.403284788131714,
+      "learning_rate": 1.9886831170719852e-05,
+      "loss": 0.4941,
+      "step": 323
+    },
+    {
+      "epoch": 0.15370018975332067,
+      "grad_norm": 3.322474479675293,
+      "learning_rate": 1.9885675651850044e-05,
+      "loss": 0.4564,
+      "step": 324
+    },
+    {
+      "epoch": 0.15417457305502846,
+      "grad_norm": 4.788414001464844,
+      "learning_rate": 1.9884514297563722e-05,
+      "loss": 0.6303,
+      "step": 325
+    },
+    {
+      "epoch": 0.15464895635673626,
+      "grad_norm": 3.2986533641815186,
+      "learning_rate": 1.9883347108546424e-05,
+      "loss": 0.4885,
+      "step": 326
+    },
+    {
+      "epoch": 0.15512333965844402,
+      "grad_norm": 3.582542896270752,
+      "learning_rate": 1.9882174085487125e-05,
+      "loss": 0.4793,
+      "step": 327
+    },
+    {
+      "epoch": 0.1555977229601518,
+      "grad_norm": 5.127254486083984,
+      "learning_rate": 1.9880995229078253e-05,
+      "loss": 0.5257,
+      "step": 328
+    },
+    {
+      "epoch": 0.15607210626185958,
+      "grad_norm": 3.8393657207489014,
+      "learning_rate": 1.9879810540015674e-05,
+      "loss": 0.4924,
+      "step": 329
+    },
+    {
+      "epoch": 0.15654648956356737,
+      "grad_norm": 4.303122043609619,
+      "learning_rate": 1.9878620018998696e-05,
+      "loss": 0.5073,
+      "step": 330
+    },
+    {
+      "epoch": 0.15702087286527514,
+      "grad_norm": 3.3921568393707275,
+      "learning_rate": 1.9877423666730075e-05,
+      "loss": 0.4634,
+      "step": 331
+    },
+    {
+      "epoch": 0.15749525616698293,
+      "grad_norm": 4.370551109313965,
+      "learning_rate": 1.9876221483916006e-05,
+      "loss": 0.641,
+      "step": 332
+    },
+    {
+      "epoch": 0.1579696394686907,
+      "grad_norm": 7.480087757110596,
+      "learning_rate": 1.9875013471266124e-05,
+      "loss": 0.5516,
+      "step": 333
+    },
+    {
+      "epoch": 0.15844402277039848,
+      "grad_norm": 3.2443172931671143,
+      "learning_rate": 1.9873799629493507e-05,
+      "loss": 0.4536,
+      "step": 334
+    },
+    {
+      "epoch": 0.15891840607210625,
+      "grad_norm": 3.3742051124572754,
+      "learning_rate": 1.9872579959314675e-05,
+      "loss": 0.4506,
+      "step": 335
+    },
+    {
+      "epoch": 0.15939278937381404,
+      "grad_norm": 4.108852386474609,
+      "learning_rate": 1.987135446144959e-05,
+      "loss": 0.5022,
+      "step": 336
+    },
+    {
+      "epoch": 0.15986717267552183,
+      "grad_norm": 3.776676893234253,
+      "learning_rate": 1.9870123136621638e-05,
+      "loss": 0.4441,
+      "step": 337
+    },
+    {
+      "epoch": 0.1603415559772296,
+      "grad_norm": 4.488317966461182,
+      "learning_rate": 1.9868885985557675e-05,
+      "loss": 0.5622,
+      "step": 338
+    },
+    {
+      "epoch": 0.1608159392789374,
+      "grad_norm": 4.163279056549072,
+      "learning_rate": 1.986764300898797e-05,
+      "loss": 0.5567,
+      "step": 339
+    },
+    {
+      "epoch": 0.16129032258064516,
+      "grad_norm": 3.326791524887085,
+      "learning_rate": 1.986639420764624e-05,
+      "loss": 0.4917,
+      "step": 340
+    },
+    {
+      "epoch": 0.16176470588235295,
+      "grad_norm": 4.889130115509033,
+      "learning_rate": 1.9865139582269642e-05,
+      "loss": 0.5238,
+      "step": 341
+    },
+    {
+      "epoch": 0.16223908918406071,
+      "grad_norm": 3.6789140701293945,
+      "learning_rate": 1.986387913359877e-05,
+      "loss": 0.6143,
+      "step": 342
+    },
+    {
+      "epoch": 0.1627134724857685,
+      "grad_norm": 3.560218572616577,
+      "learning_rate": 1.9862612862377652e-05,
+      "loss": 0.4937,
+      "step": 343
+    },
+    {
+      "epoch": 0.16318785578747627,
+      "grad_norm": 3.0971603393554688,
+      "learning_rate": 1.9861340769353753e-05,
+      "loss": 0.4614,
+      "step": 344
+    },
+    {
+      "epoch": 0.16366223908918406,
+      "grad_norm": 3.41616153717041,
+      "learning_rate": 1.9860062855277982e-05,
+      "loss": 0.4853,
+      "step": 345
+    },
+    {
+      "epoch": 0.16413662239089183,
+      "grad_norm": 3.4327688217163086,
+      "learning_rate": 1.985877912090468e-05,
+      "loss": 0.4745,
+      "step": 346
+    },
+    {
+      "epoch": 0.16461100569259962,
+      "grad_norm": 3.024824380874634,
+      "learning_rate": 1.9857489566991614e-05,
+      "loss": 0.5228,
+      "step": 347
+    },
+    {
+      "epoch": 0.1650853889943074,
+      "grad_norm": 4.104083061218262,
+      "learning_rate": 1.9856194194300005e-05,
+      "loss": 0.5777,
+      "step": 348
+    },
+    {
+      "epoch": 0.16555977229601518,
+      "grad_norm": 4.0848798751831055,
+      "learning_rate": 1.9854893003594492e-05,
+      "loss": 0.5047,
+      "step": 349
+    },
+    {
+      "epoch": 0.16603415559772297,
+      "grad_norm": 3.6772310733795166,
+      "learning_rate": 1.9853585995643158e-05,
+      "loss": 0.5242,
+      "step": 350
+    },
+    {
+      "epoch": 0.16650853889943074,
+      "grad_norm": 3.716475009918213,
+      "learning_rate": 1.9852273171217518e-05,
+      "loss": 0.4722,
+      "step": 351
+    },
+    {
+      "epoch": 0.16698292220113853,
+      "grad_norm": 4.485456466674805,
+      "learning_rate": 1.9850954531092515e-05,
+      "loss": 0.4973,
+      "step": 352
+    },
+    {
+      "epoch": 0.1674573055028463,
+      "grad_norm": 3.9249846935272217,
+      "learning_rate": 1.9849630076046536e-05,
+      "loss": 0.506,
+      "step": 353
+    },
+    {
+      "epoch": 0.16793168880455409,
+      "grad_norm": 3.6390221118927,
+      "learning_rate": 1.9848299806861385e-05,
+      "loss": 0.4511,
+      "step": 354
+    },
+    {
+      "epoch": 0.16840607210626185,
+      "grad_norm": 4.2053937911987305,
+      "learning_rate": 1.984696372432231e-05,
+      "loss": 0.523,
+      "step": 355
+    },
+    {
+      "epoch": 0.16888045540796964,
+      "grad_norm": 3.690425157546997,
+      "learning_rate": 1.984562182921799e-05,
+      "loss": 0.5692,
+      "step": 356
+    },
+    {
+      "epoch": 0.1693548387096774,
+      "grad_norm": 3.777768611907959,
+      "learning_rate": 1.9844274122340534e-05,
+      "loss": 0.4881,
+      "step": 357
+    },
+    {
+      "epoch": 0.1698292220113852,
+      "grad_norm": 3.9081265926361084,
+      "learning_rate": 1.9842920604485474e-05,
+      "loss": 0.5767,
+      "step": 358
+    },
+    {
+      "epoch": 0.170303605313093,
+      "grad_norm": 4.124914169311523,
+      "learning_rate": 1.984156127645178e-05,
+      "loss": 0.5526,
+      "step": 359
+    },
+    {
+      "epoch": 0.17077798861480076,
+      "grad_norm": 3.433190107345581,
+      "learning_rate": 1.9840196139041853e-05,
+      "loss": 0.385,
+      "step": 360
+    },
+    {
+      "epoch": 0.17125237191650855,
+      "grad_norm": 3.6542704105377197,
+      "learning_rate": 1.9838825193061518e-05,
+      "loss": 0.4004,
+      "step": 361
+    },
+    {
+      "epoch": 0.17172675521821631,
+      "grad_norm": 4.064217567443848,
+      "learning_rate": 1.9837448439320027e-05,
+      "loss": 0.4996,
+      "step": 362
+    },
+    {
+      "epoch": 0.1722011385199241,
+      "grad_norm": 3.451735019683838,
+      "learning_rate": 1.9836065878630074e-05,
+      "loss": 0.4737,
+      "step": 363
+    },
+    {
+      "epoch": 0.17267552182163187,
+      "grad_norm": 3.747659206390381,
+      "learning_rate": 1.983467751180776e-05,
+      "loss": 0.5078,
+      "step": 364
+    },
+    {
+      "epoch": 0.17314990512333966,
+      "grad_norm": 3.898174524307251,
+      "learning_rate": 1.983328333967263e-05,
+      "loss": 0.5549,
+      "step": 365
+    },
+    {
+      "epoch": 0.17362428842504743,
+      "grad_norm": 3.5523483753204346,
+      "learning_rate": 1.983188336304765e-05,
+      "loss": 0.5202,
+      "step": 366
+    },
+    {
+      "epoch": 0.17409867172675522,
+      "grad_norm": 3.90126633644104,
+      "learning_rate": 1.9830477582759213e-05,
+      "loss": 0.5718,
+      "step": 367
+    },
+    {
+      "epoch": 0.174573055028463,
+      "grad_norm": 4.264011383056641,
+      "learning_rate": 1.9829065999637134e-05,
+      "loss": 0.561,
+      "step": 368
+    },
+    {
+      "epoch": 0.17504743833017078,
+      "grad_norm": 3.489223003387451,
+      "learning_rate": 1.982764861451466e-05,
+      "loss": 0.5515,
+      "step": 369
+    },
+    {
+      "epoch": 0.17552182163187854,
+      "grad_norm": 3.144134998321533,
+      "learning_rate": 1.9826225428228455e-05,
+      "loss": 0.3927,
+      "step": 370
+    },
+    {
+      "epoch": 0.17599620493358634,
+      "grad_norm": 3.7182862758636475,
+      "learning_rate": 1.9824796441618617e-05,
+      "loss": 0.5222,
+      "step": 371
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 3.7761785984039307,
+      "learning_rate": 1.9823361655528658e-05,
+      "loss": 0.5393,
+      "step": 372
+    },
+    {
+      "epoch": 0.1769449715370019,
+      "grad_norm": 3.673797369003296,
+      "learning_rate": 1.9821921070805522e-05,
+      "loss": 0.5508,
+      "step": 373
+    },
+    {
+      "epoch": 0.1774193548387097,
+      "grad_norm": 5.075558662414551,
+      "learning_rate": 1.9820474688299566e-05,
+      "loss": 0.4923,
+      "step": 374
+    },
+    {
+      "epoch": 0.17789373814041745,
+      "grad_norm": 2.8637001514434814,
+      "learning_rate": 1.9819022508864582e-05,
+      "loss": 0.4581,
+      "step": 375
+    },
+    {
+      "epoch": 0.17836812144212524,
+      "grad_norm": 3.7829291820526123,
+      "learning_rate": 1.9817564533357775e-05,
+      "loss": 0.4989,
+      "step": 376
+    },
+    {
+      "epoch": 0.178842504743833,
+      "grad_norm": 3.2799975872039795,
+      "learning_rate": 1.9816100762639773e-05,
+      "loss": 0.513,
+      "step": 377
+    },
+    {
+      "epoch": 0.1793168880455408,
+      "grad_norm": 3.540116310119629,
+      "learning_rate": 1.9814631197574626e-05,
+      "loss": 0.4851,
+      "step": 378
+    },
+    {
+      "epoch": 0.17979127134724857,
+      "grad_norm": 3.821983814239502,
+      "learning_rate": 1.98131558390298e-05,
+      "loss": 0.4854,
+      "step": 379
+    },
+    {
+      "epoch": 0.18026565464895636,
+      "grad_norm": 7.540534973144531,
+      "learning_rate": 1.981167468787619e-05,
+      "loss": 0.5211,
+      "step": 380
+    },
+    {
+      "epoch": 0.18074003795066412,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.98101877449881e-05,
+      "loss": 0.5135,
+      "step": 381
+    },
+    {
+      "epoch": 0.18121442125237192,
+      "grad_norm": 3.8135201930999756,
+      "learning_rate": 1.980869501124326e-05,
+      "loss": 0.551,
+      "step": 382
+    },
+    {
+      "epoch": 0.1816888045540797,
+      "grad_norm": 3.3921446800231934,
+      "learning_rate": 1.9807196487522818e-05,
+      "loss": 0.5174,
+      "step": 383
+    },
+    {
+      "epoch": 0.18216318785578747,
+      "grad_norm": 3.473004102706909,
+      "learning_rate": 1.9805692174711337e-05,
+      "loss": 0.5091,
+      "step": 384
+    },
+    {
+      "epoch": 0.18263757115749527,
+      "grad_norm": 3.449249505996704,
+      "learning_rate": 1.9804182073696793e-05,
+      "loss": 0.5027,
+      "step": 385
+    },
+    {
+      "epoch": 0.18311195445920303,
+      "grad_norm": 3.633528470993042,
+      "learning_rate": 1.980266618537059e-05,
+      "loss": 0.5025,
+      "step": 386
+    },
+    {
+      "epoch": 0.18358633776091082,
+      "grad_norm": 3.325308084487915,
+      "learning_rate": 1.980114451062754e-05,
+      "loss": 0.4739,
+      "step": 387
+    },
+    {
+      "epoch": 0.1840607210626186,
+      "grad_norm": 3.829240083694458,
+      "learning_rate": 1.979961705036587e-05,
+      "loss": 0.4921,
+      "step": 388
+    },
+    {
+      "epoch": 0.18453510436432638,
+      "grad_norm": 3.1751868724823,
+      "learning_rate": 1.9798083805487228e-05,
+      "loss": 0.4318,
+      "step": 389
+    },
+    {
+      "epoch": 0.18500948766603414,
+      "grad_norm": 3.521556854248047,
+      "learning_rate": 1.979654477689667e-05,
+      "loss": 0.4898,
+      "step": 390
+    },
+    {
+      "epoch": 0.18548387096774194,
+      "grad_norm": 3.7403318881988525,
+      "learning_rate": 1.979499996550267e-05,
+      "loss": 0.5121,
+      "step": 391
+    },
+    {
+      "epoch": 0.1859582542694497,
+      "grad_norm": 4.491858959197998,
+      "learning_rate": 1.9793449372217123e-05,
+      "loss": 0.5375,
+      "step": 392
+    },
+    {
+      "epoch": 0.1864326375711575,
+      "grad_norm": 3.4098191261291504,
+      "learning_rate": 1.9791892997955317e-05,
+      "loss": 0.5462,
+      "step": 393
+    },
+    {
+      "epoch": 0.1869070208728653,
+      "grad_norm": 3.658325672149658,
+      "learning_rate": 1.9790330843635967e-05,
+      "loss": 0.4567,
+      "step": 394
+    },
+    {
+      "epoch": 0.18738140417457305,
+      "grad_norm": 4.077053546905518,
+      "learning_rate": 1.97887629101812e-05,
+      "loss": 0.5194,
+      "step": 395
+    },
+    {
+      "epoch": 0.18785578747628084,
+      "grad_norm": 3.680041551589966,
+      "learning_rate": 1.9787189198516553e-05,
+      "loss": 0.4515,
+      "step": 396
+    },
+    {
+      "epoch": 0.1883301707779886,
+      "grad_norm": 3.7228457927703857,
+      "learning_rate": 1.9785609709570973e-05,
+      "loss": 0.4504,
+      "step": 397
+    },
+    {
+      "epoch": 0.1888045540796964,
+      "grad_norm": 4.01034688949585,
+      "learning_rate": 1.9784024444276812e-05,
+      "loss": 0.4637,
+      "step": 398
+    },
+    {
+      "epoch": 0.18927893738140417,
+      "grad_norm": 2.7363219261169434,
+      "learning_rate": 1.9782433403569836e-05,
+      "loss": 0.4619,
+      "step": 399
+    },
+    {
+      "epoch": 0.18975332068311196,
+      "grad_norm": 3.8097171783447266,
+      "learning_rate": 1.9780836588389225e-05,
+      "loss": 0.578,
+      "step": 400
+    },
+    {
+      "epoch": 0.19022770398481972,
+      "grad_norm": 3.741736650466919,
+      "learning_rate": 1.9779233999677563e-05,
+      "loss": 0.5731,
+      "step": 401
+    },
+    {
+      "epoch": 0.19070208728652752,
+      "grad_norm": 3.690082311630249,
+      "learning_rate": 1.9777625638380838e-05,
+      "loss": 0.5538,
+      "step": 402
+    },
+    {
+      "epoch": 0.19117647058823528,
+      "grad_norm": 3.4858877658843994,
+      "learning_rate": 1.9776011505448455e-05,
+      "loss": 0.4858,
+      "step": 403
+    },
+    {
+      "epoch": 0.19165085388994307,
+      "grad_norm": 3.8378117084503174,
+      "learning_rate": 1.977439160183322e-05,
+      "loss": 0.5408,
+      "step": 404
+    },
+    {
+      "epoch": 0.19212523719165087,
+      "grad_norm": 3.4993057250976562,
+      "learning_rate": 1.977276592849134e-05,
+      "loss": 0.4522,
+      "step": 405
+    },
+    {
+      "epoch": 0.19259962049335863,
+      "grad_norm": 3.624081611633301,
+      "learning_rate": 1.9771134486382436e-05,
+      "loss": 0.4739,
+      "step": 406
+    },
+    {
+      "epoch": 0.19307400379506642,
+      "grad_norm": 3.2720398902893066,
+      "learning_rate": 1.9769497276469538e-05,
+      "loss": 0.5003,
+      "step": 407
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 3.429849624633789,
+      "learning_rate": 1.9767854299719073e-05,
+      "loss": 0.5161,
+      "step": 408
+    },
+    {
+      "epoch": 0.19402277039848198,
+      "grad_norm": 3.6602742671966553,
+      "learning_rate": 1.976620555710087e-05,
+      "loss": 0.5546,
+      "step": 409
+    },
+    {
+      "epoch": 0.19449715370018975,
+      "grad_norm": 4.340588092803955,
+      "learning_rate": 1.9764551049588165e-05,
+      "loss": 0.5262,
+      "step": 410
+    },
+    {
+      "epoch": 0.19497153700189754,
+      "grad_norm": 3.39455509185791,
+      "learning_rate": 1.97628907781576e-05,
+      "loss": 0.4408,
+      "step": 411
+    },
+    {
+      "epoch": 0.1954459203036053,
+      "grad_norm": 3.521057605743408,
+      "learning_rate": 1.976122474378922e-05,
+      "loss": 0.4413,
+      "step": 412
+    },
+    {
+      "epoch": 0.1959203036053131,
+      "grad_norm": 4.244317531585693,
+      "learning_rate": 1.9759552947466462e-05,
+      "loss": 0.5525,
+      "step": 413
+    },
+    {
+      "epoch": 0.19639468690702086,
+      "grad_norm": 3.8756752014160156,
+      "learning_rate": 1.9757875390176177e-05,
+      "loss": 0.5267,
+      "step": 414
+    },
+    {
+      "epoch": 0.19686907020872865,
+      "grad_norm": 3.6845438480377197,
+      "learning_rate": 1.9756192072908605e-05,
+      "loss": 0.5128,
+      "step": 415
+    },
+    {
+      "epoch": 0.19734345351043645,
+      "grad_norm": 4.302131652832031,
+      "learning_rate": 1.9754502996657395e-05,
+      "loss": 0.5089,
+      "step": 416
+    },
+    {
+      "epoch": 0.1978178368121442,
+      "grad_norm": 3.6274499893188477,
+      "learning_rate": 1.975280816241959e-05,
+      "loss": 0.4639,
+      "step": 417
+    },
+    {
+      "epoch": 0.198292220113852,
+      "grad_norm": 3.127408742904663,
+      "learning_rate": 1.975110757119564e-05,
+      "loss": 0.4801,
+      "step": 418
+    },
+    {
+      "epoch": 0.19876660341555977,
+      "grad_norm": 9.724570274353027,
+      "learning_rate": 1.9749401223989376e-05,
+      "loss": 0.6008,
+      "step": 419
+    },
+    {
+      "epoch": 0.19924098671726756,
+      "grad_norm": 3.2324442863464355,
+      "learning_rate": 1.9747689121808045e-05,
+      "loss": 0.4913,
+      "step": 420
+    },
+    {
+      "epoch": 0.19971537001897532,
+      "grad_norm": 2.958690881729126,
+      "learning_rate": 1.9745971265662286e-05,
+      "loss": 0.3945,
+      "step": 421
+    },
+    {
+      "epoch": 0.20018975332068312,
+      "grad_norm": 3.4928455352783203,
+      "learning_rate": 1.9744247656566125e-05,
+      "loss": 0.4863,
+      "step": 422
+    },
+    {
+      "epoch": 0.20066413662239088,
+      "grad_norm": 3.0363330841064453,
+      "learning_rate": 1.9742518295536997e-05,
+      "loss": 0.4847,
+      "step": 423
+    },
+    {
+      "epoch": 0.20113851992409867,
+      "grad_norm": 3.126666307449341,
+      "learning_rate": 1.9740783183595726e-05,
+      "loss": 0.4281,
+      "step": 424
+    },
+    {
+      "epoch": 0.20161290322580644,
+      "grad_norm": 3.345747232437134,
+      "learning_rate": 1.9739042321766533e-05,
+      "loss": 0.4699,
+      "step": 425
+    },
+    {
+      "epoch": 0.20208728652751423,
+      "grad_norm": 3.1891050338745117,
+      "learning_rate": 1.973729571107703e-05,
+      "loss": 0.4362,
+      "step": 426
+    },
+    {
+      "epoch": 0.20256166982922202,
+      "grad_norm": 3.56644606590271,
+      "learning_rate": 1.973554335255822e-05,
+      "loss": 0.4374,
+      "step": 427
+    },
+    {
+      "epoch": 0.2030360531309298,
+      "grad_norm": 3.70436692237854,
+      "learning_rate": 1.9733785247244506e-05,
+      "loss": 0.4582,
+      "step": 428
+    },
+    {
+      "epoch": 0.20351043643263758,
+      "grad_norm": 3.7359845638275146,
+      "learning_rate": 1.9732021396173682e-05,
+      "loss": 0.5247,
+      "step": 429
+    },
+    {
+      "epoch": 0.20398481973434535,
+      "grad_norm": 3.910747528076172,
+      "learning_rate": 1.973025180038693e-05,
+      "loss": 0.5534,
+      "step": 430
+    },
+    {
+      "epoch": 0.20445920303605314,
+      "grad_norm": 3.243649482727051,
+      "learning_rate": 1.9728476460928828e-05,
+      "loss": 0.53,
+      "step": 431
+    },
+    {
+      "epoch": 0.2049335863377609,
+      "grad_norm": 3.5252304077148438,
+      "learning_rate": 1.9726695378847332e-05,
+      "loss": 0.5136,
+      "step": 432
+    },
+    {
+      "epoch": 0.2054079696394687,
+      "grad_norm": 3.3857786655426025,
+      "learning_rate": 1.972490855519381e-05,
+      "loss": 0.4836,
+      "step": 433
+    },
+    {
+      "epoch": 0.20588235294117646,
+      "grad_norm": 3.3901121616363525,
+      "learning_rate": 1.9723115991022997e-05,
+      "loss": 0.4585,
+      "step": 434
+    },
+    {
+      "epoch": 0.20635673624288425,
+      "grad_norm": 3.816561460494995,
+      "learning_rate": 1.972131768739303e-05,
+      "loss": 0.5757,
+      "step": 435
+    },
+    {
+      "epoch": 0.20683111954459202,
+      "grad_norm": 3.4764530658721924,
+      "learning_rate": 1.9719513645365426e-05,
+      "loss": 0.4892,
+      "step": 436
+    },
+    {
+      "epoch": 0.2073055028462998,
+      "grad_norm": 3.6275835037231445,
+      "learning_rate": 1.97177038660051e-05,
+      "loss": 0.4744,
+      "step": 437
+    },
+    {
+      "epoch": 0.2077798861480076,
+      "grad_norm": 3.73422908782959,
+      "learning_rate": 1.971588835038034e-05,
+      "loss": 0.6116,
+      "step": 438
+    },
+    {
+      "epoch": 0.20825426944971537,
+      "grad_norm": 3.751572370529175,
+      "learning_rate": 1.971406709956283e-05,
+      "loss": 0.534,
+      "step": 439
+    },
+    {
+      "epoch": 0.20872865275142316,
+      "grad_norm": 2.7028439044952393,
+      "learning_rate": 1.9712240114627637e-05,
+      "loss": 0.4065,
+      "step": 440
+    },
+    {
+      "epoch": 0.20920303605313093,
+      "grad_norm": 2.995476007461548,
+      "learning_rate": 1.971040739665321e-05,
+      "loss": 0.4091,
+      "step": 441
+    },
+    {
+      "epoch": 0.20967741935483872,
+      "grad_norm": 2.9523215293884277,
+      "learning_rate": 1.970856894672139e-05,
+      "loss": 0.465,
+      "step": 442
+    },
+    {
+      "epoch": 0.21015180265654648,
+      "grad_norm": 3.5105068683624268,
+      "learning_rate": 1.9706724765917384e-05,
+      "loss": 0.459,
+      "step": 443
+    },
+    {
+      "epoch": 0.21062618595825428,
+      "grad_norm": 2.8509361743927,
+      "learning_rate": 1.97048748553298e-05,
+      "loss": 0.4519,
+      "step": 444
+    },
+    {
+      "epoch": 0.21110056925996204,
+      "grad_norm": 3.4983131885528564,
+      "learning_rate": 1.9703019216050627e-05,
+      "loss": 0.5233,
+      "step": 445
+    },
+    {
+      "epoch": 0.21157495256166983,
+      "grad_norm": 3.1177706718444824,
+      "learning_rate": 1.970115784917523e-05,
+      "loss": 0.4847,
+      "step": 446
+    },
+    {
+      "epoch": 0.2120493358633776,
+      "grad_norm": 3.1251273155212402,
+      "learning_rate": 1.9699290755802344e-05,
+      "loss": 0.4727,
+      "step": 447
+    },
+    {
+      "epoch": 0.2125237191650854,
+      "grad_norm": 2.4966578483581543,
+      "learning_rate": 1.9697417937034106e-05,
+      "loss": 0.4165,
+      "step": 448
+    },
+    {
+      "epoch": 0.21299810246679318,
+      "grad_norm": 3.8288350105285645,
+      "learning_rate": 1.969553939397602e-05,
+      "loss": 0.553,
+      "step": 449
+    },
+    {
+      "epoch": 0.21347248576850095,
+      "grad_norm": 3.0275819301605225,
+      "learning_rate": 1.9693655127736974e-05,
+      "loss": 0.3849,
+      "step": 450
+    },
+    {
+      "epoch": 0.21394686907020874,
+      "grad_norm": 3.2554712295532227,
+      "learning_rate": 1.9691765139429227e-05,
+      "loss": 0.5852,
+      "step": 451
+    },
+    {
+      "epoch": 0.2144212523719165,
+      "grad_norm": 3.404175043106079,
+      "learning_rate": 1.968986943016842e-05,
+      "loss": 0.5296,
+      "step": 452
+    },
+    {
+      "epoch": 0.2148956356736243,
+      "grad_norm": 3.853860378265381,
+      "learning_rate": 1.9687968001073575e-05,
+      "loss": 0.6205,
+      "step": 453
+    },
+    {
+      "epoch": 0.21537001897533206,
+      "grad_norm": 3.2675604820251465,
+      "learning_rate": 1.9686060853267088e-05,
+      "loss": 0.5094,
+      "step": 454
+    },
+    {
+      "epoch": 0.21584440227703985,
+      "grad_norm": 4.281660079956055,
+      "learning_rate": 1.9684147987874725e-05,
+      "loss": 0.514,
+      "step": 455
+    },
+    {
+      "epoch": 0.21631878557874762,
+      "grad_norm": 3.5292105674743652,
+      "learning_rate": 1.9682229406025635e-05,
+      "loss": 0.4982,
+      "step": 456
+    },
+    {
+      "epoch": 0.2167931688804554,
+      "grad_norm": 3.585477352142334,
+      "learning_rate": 1.9680305108852335e-05,
+      "loss": 0.4612,
+      "step": 457
+    },
+    {
+      "epoch": 0.21726755218216318,
+      "grad_norm": 3.200887680053711,
+      "learning_rate": 1.9678375097490717e-05,
+      "loss": 0.5057,
+      "step": 458
+    },
+    {
+      "epoch": 0.21774193548387097,
+      "grad_norm": 3.3191189765930176,
+      "learning_rate": 1.9676439373080056e-05,
+      "loss": 0.4977,
+      "step": 459
+    },
+    {
+      "epoch": 0.21821631878557876,
+      "grad_norm": 3.9757373332977295,
+      "learning_rate": 1.9674497936762984e-05,
+      "loss": 0.5366,
+      "step": 460
+    },
+    {
+      "epoch": 0.21869070208728653,
+      "grad_norm": 3.3760838508605957,
+      "learning_rate": 1.9672550789685512e-05,
+      "loss": 0.4957,
+      "step": 461
+    },
+    {
+      "epoch": 0.21916508538899432,
+      "grad_norm": 3.371440887451172,
+      "learning_rate": 1.9670597932997023e-05,
+      "loss": 0.5172,
+      "step": 462
+    },
+    {
+      "epoch": 0.21963946869070208,
+      "grad_norm": 5.2982001304626465,
+      "learning_rate": 1.966863936785027e-05,
+      "loss": 0.5267,
+      "step": 463
+    },
+    {
+      "epoch": 0.22011385199240988,
+      "grad_norm": 3.9646644592285156,
+      "learning_rate": 1.966667509540137e-05,
+      "loss": 0.4803,
+      "step": 464
+    },
+    {
+      "epoch": 0.22058823529411764,
+      "grad_norm": 3.059048891067505,
+      "learning_rate": 1.9664705116809823e-05,
+      "loss": 0.4096,
+      "step": 465
+    },
+    {
+      "epoch": 0.22106261859582543,
+      "grad_norm": 3.714001178741455,
+      "learning_rate": 1.9662729433238477e-05,
+      "loss": 0.6027,
+      "step": 466
+    },
+    {
+      "epoch": 0.2215370018975332,
+      "grad_norm": 3.6565237045288086,
+      "learning_rate": 1.9660748045853567e-05,
+      "loss": 0.4968,
+      "step": 467
+    },
+    {
+      "epoch": 0.222011385199241,
+      "grad_norm": 3.092435836791992,
+      "learning_rate": 1.965876095582468e-05,
+      "loss": 0.4037,
+      "step": 468
+    },
+    {
+      "epoch": 0.22248576850094876,
+      "grad_norm": 3.3902924060821533,
+      "learning_rate": 1.965676816432478e-05,
+      "loss": 0.5345,
+      "step": 469
+    },
+    {
+      "epoch": 0.22296015180265655,
+      "grad_norm": 3.4829235076904297,
+      "learning_rate": 1.9654769672530186e-05,
+      "loss": 0.504,
+      "step": 470
+    },
+    {
+      "epoch": 0.22343453510436434,
+      "grad_norm": 3.3810994625091553,
+      "learning_rate": 1.9652765481620596e-05,
+      "loss": 0.4718,
+      "step": 471
+    },
+    {
+      "epoch": 0.2239089184060721,
+      "grad_norm": 4.006764888763428,
+      "learning_rate": 1.965075559277906e-05,
+      "loss": 0.5748,
+      "step": 472
+    },
+    {
+      "epoch": 0.2243833017077799,
+      "grad_norm": 3.318103313446045,
+      "learning_rate": 1.9648740007191994e-05,
+      "loss": 0.4585,
+      "step": 473
+    },
+    {
+      "epoch": 0.22485768500948766,
+      "grad_norm": 3.561944007873535,
+      "learning_rate": 1.9646718726049187e-05,
+      "loss": 0.5574,
+      "step": 474
+    },
+    {
+      "epoch": 0.22533206831119545,
+      "grad_norm": 3.0337014198303223,
+      "learning_rate": 1.964469175054377e-05,
+      "loss": 0.4363,
+      "step": 475
+    },
+    {
+      "epoch": 0.22580645161290322,
+      "grad_norm": 2.986750364303589,
+      "learning_rate": 1.964265908187225e-05,
+      "loss": 0.4626,
+      "step": 476
+    },
+    {
+      "epoch": 0.226280834914611,
+      "grad_norm": 3.6310508251190186,
+      "learning_rate": 1.9640620721234488e-05,
+      "loss": 0.4709,
+      "step": 477
+    },
+    {
+      "epoch": 0.22675521821631878,
+      "grad_norm": 3.44527530670166,
+      "learning_rate": 1.963857666983372e-05,
+      "loss": 0.4663,
+      "step": 478
+    },
+    {
+      "epoch": 0.22722960151802657,
+      "grad_norm": 4.240223407745361,
+      "learning_rate": 1.963652692887652e-05,
+      "loss": 0.6681,
+      "step": 479
+    },
+    {
+      "epoch": 0.22770398481973433,
+      "grad_norm": 3.381404161453247,
+      "learning_rate": 1.9634471499572826e-05,
+      "loss": 0.4429,
+      "step": 480
+    },
+    {
+      "epoch": 0.22817836812144213,
+      "grad_norm": 3.138699769973755,
+      "learning_rate": 1.9632410383135946e-05,
+      "loss": 0.4803,
+      "step": 481
+    },
+    {
+      "epoch": 0.2286527514231499,
+      "grad_norm": 3.806920051574707,
+      "learning_rate": 1.9630343580782538e-05,
+      "loss": 0.4943,
+      "step": 482
+    },
+    {
+      "epoch": 0.22912713472485768,
+      "grad_norm": 3.408698320388794,
+      "learning_rate": 1.9628271093732605e-05,
+      "loss": 0.5208,
+      "step": 483
+    },
+    {
+      "epoch": 0.22960151802656548,
+      "grad_norm": 2.962362051010132,
+      "learning_rate": 1.9626192923209524e-05,
+      "loss": 0.4449,
+      "step": 484
+    },
+    {
+      "epoch": 0.23007590132827324,
+      "grad_norm": 3.10517954826355,
+      "learning_rate": 1.9624109070440017e-05,
+      "loss": 0.4865,
+      "step": 485
+    },
+    {
+      "epoch": 0.23055028462998103,
+      "grad_norm": 3.3021395206451416,
+      "learning_rate": 1.9622019536654154e-05,
+      "loss": 0.5415,
+      "step": 486
+    },
+    {
+      "epoch": 0.2310246679316888,
+      "grad_norm": 3.686612606048584,
+      "learning_rate": 1.961992432308538e-05,
+      "loss": 0.4875,
+      "step": 487
+    },
+    {
+      "epoch": 0.2314990512333966,
+      "grad_norm": 3.50701904296875,
+      "learning_rate": 1.961782343097047e-05,
+      "loss": 0.459,
+      "step": 488
+    },
+    {
+      "epoch": 0.23197343453510436,
+      "grad_norm": 3.291677713394165,
+      "learning_rate": 1.9615716861549557e-05,
+      "loss": 0.5038,
+      "step": 489
+    },
+    {
+      "epoch": 0.23244781783681215,
+      "grad_norm": 3.3820300102233887,
+      "learning_rate": 1.9613604616066137e-05,
+      "loss": 0.4445,
+      "step": 490
+    },
+    {
+      "epoch": 0.2329222011385199,
+      "grad_norm": 3.5324206352233887,
+      "learning_rate": 1.9611486695767037e-05,
+      "loss": 0.5252,
+      "step": 491
+    },
+    {
+      "epoch": 0.2333965844402277,
+      "grad_norm": 3.8190603256225586,
+      "learning_rate": 1.9609363101902456e-05,
+      "loss": 0.4666,
+      "step": 492
+    },
+    {
+      "epoch": 0.23387096774193547,
+      "grad_norm": 4.703067779541016,
+      "learning_rate": 1.960723383572592e-05,
+      "loss": 0.4529,
+      "step": 493
+    },
+    {
+      "epoch": 0.23434535104364326,
+      "grad_norm": 3.1753973960876465,
+      "learning_rate": 1.960509889849432e-05,
+      "loss": 0.4731,
+      "step": 494
+    },
+    {
+      "epoch": 0.23481973434535106,
+      "grad_norm": 3.4943294525146484,
+      "learning_rate": 1.960295829146789e-05,
+      "loss": 0.5578,
+      "step": 495
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 3.4101996421813965,
+      "learning_rate": 1.9600812015910203e-05,
+      "loss": 0.5255,
+      "step": 496
+    },
+    {
+      "epoch": 0.2357685009487666,
+      "grad_norm": 3.5300912857055664,
+      "learning_rate": 1.9598660073088186e-05,
+      "loss": 0.5781,
+      "step": 497
+    },
+    {
+      "epoch": 0.23624288425047438,
+      "grad_norm": 3.204566478729248,
+      "learning_rate": 1.959650246427211e-05,
+      "loss": 0.5813,
+      "step": 498
+    },
+    {
+      "epoch": 0.23671726755218217,
+      "grad_norm": 3.192633628845215,
+      "learning_rate": 1.9594339190735594e-05,
+      "loss": 0.4062,
+      "step": 499
+    },
+    {
+      "epoch": 0.23719165085388993,
+      "grad_norm": 3.062784194946289,
+      "learning_rate": 1.9592170253755594e-05,
+      "loss": 0.4764,
+      "step": 500
+    },
+    {
+      "epoch": 0.23766603415559773,
+      "grad_norm": 2.8469433784484863,
+      "learning_rate": 1.9589995654612412e-05,
+      "loss": 0.4091,
+      "step": 501
+    },
+    {
+      "epoch": 0.2381404174573055,
+      "grad_norm": 3.5228500366210938,
+      "learning_rate": 1.958781539458969e-05,
+      "loss": 0.51,
+      "step": 502
+    },
+    {
+      "epoch": 0.23861480075901328,
+      "grad_norm": 3.872974395751953,
+      "learning_rate": 1.9585629474974413e-05,
+      "loss": 0.5438,
+      "step": 503
+    },
+    {
+      "epoch": 0.23908918406072105,
+      "grad_norm": 3.1505889892578125,
+      "learning_rate": 1.9583437897056915e-05,
+      "loss": 0.4181,
+      "step": 504
+    },
+    {
+      "epoch": 0.23956356736242884,
+      "grad_norm": 3.113068103790283,
+      "learning_rate": 1.958124066213086e-05,
+      "loss": 0.4512,
+      "step": 505
+    },
+    {
+      "epoch": 0.24003795066413663,
+      "grad_norm": 3.0535147190093994,
+      "learning_rate": 1.957903777149325e-05,
+      "loss": 0.3913,
+      "step": 506
+    },
+    {
+      "epoch": 0.2405123339658444,
+      "grad_norm": 3.152808427810669,
+      "learning_rate": 1.957682922644443e-05,
+      "loss": 0.4813,
+      "step": 507
+    },
+    {
+      "epoch": 0.2409867172675522,
+      "grad_norm": 3.2799031734466553,
+      "learning_rate": 1.957461502828809e-05,
+      "loss": 0.4559,
+      "step": 508
+    },
+    {
+      "epoch": 0.24146110056925996,
+      "grad_norm": 3.305764675140381,
+      "learning_rate": 1.957239517833124e-05,
+      "loss": 0.4553,
+      "step": 509
+    },
+    {
+      "epoch": 0.24193548387096775,
+      "grad_norm": 2.7665562629699707,
+      "learning_rate": 1.957016967788424e-05,
+      "loss": 0.4145,
+      "step": 510
+    },
+    {
+      "epoch": 0.2424098671726755,
+      "grad_norm": 3.1011545658111572,
+      "learning_rate": 1.9567938528260778e-05,
+      "loss": 0.5109,
+      "step": 511
+    },
+    {
+      "epoch": 0.2428842504743833,
+      "grad_norm": 3.0363271236419678,
+      "learning_rate": 1.9565701730777883e-05,
+      "loss": 0.4649,
+      "step": 512
+    },
+    {
+      "epoch": 0.24335863377609107,
+      "grad_norm": 3.1920406818389893,
+      "learning_rate": 1.9563459286755914e-05,
+      "loss": 0.5797,
+      "step": 513
+    },
+    {
+      "epoch": 0.24383301707779886,
+      "grad_norm": 3.131014347076416,
+      "learning_rate": 1.9561211197518564e-05,
+      "loss": 0.4634,
+      "step": 514
+    },
+    {
+      "epoch": 0.24430740037950663,
+      "grad_norm": 3.515669107437134,
+      "learning_rate": 1.9558957464392853e-05,
+      "loss": 0.5281,
+      "step": 515
+    },
+    {
+      "epoch": 0.24478178368121442,
+      "grad_norm": 3.187274932861328,
+      "learning_rate": 1.9556698088709144e-05,
+      "loss": 0.443,
+      "step": 516
+    },
+    {
+      "epoch": 0.2452561669829222,
+      "grad_norm": 3.458089828491211,
+      "learning_rate": 1.9554433071801117e-05,
+      "loss": 0.4375,
+      "step": 517
+    },
+    {
+      "epoch": 0.24573055028462998,
+      "grad_norm": 3.4186503887176514,
+      "learning_rate": 1.9552162415005797e-05,
+      "loss": 0.5475,
+      "step": 518
+    },
+    {
+      "epoch": 0.24620493358633777,
+      "grad_norm": 3.319234848022461,
+      "learning_rate": 1.954988611966352e-05,
+      "loss": 0.4545,
+      "step": 519
+    },
+    {
+      "epoch": 0.24667931688804554,
+      "grad_norm": 3.491206169128418,
+      "learning_rate": 1.9547604187117974e-05,
+      "loss": 0.4993,
+      "step": 520
+    },
+    {
+      "epoch": 0.24715370018975333,
+      "grad_norm": 3.1041345596313477,
+      "learning_rate": 1.954531661871615e-05,
+      "loss": 0.4866,
+      "step": 521
+    },
+    {
+      "epoch": 0.2476280834914611,
+      "grad_norm": 4.381217956542969,
+      "learning_rate": 1.954302341580838e-05,
+      "loss": 0.5897,
+      "step": 522
+    },
+    {
+      "epoch": 0.24810246679316889,
+      "grad_norm": 3.8608601093292236,
+      "learning_rate": 1.9540724579748323e-05,
+      "loss": 0.4814,
+      "step": 523
+    },
+    {
+      "epoch": 0.24857685009487665,
+      "grad_norm": 3.7535743713378906,
+      "learning_rate": 1.953842011189295e-05,
+      "loss": 0.5214,
+      "step": 524
+    },
+    {
+      "epoch": 0.24905123339658444,
+      "grad_norm": 3.2677531242370605,
+      "learning_rate": 1.9536110013602578e-05,
+      "loss": 0.3946,
+      "step": 525
+    },
+    {
+      "epoch": 0.2495256166982922,
+      "grad_norm": 3.734286069869995,
+      "learning_rate": 1.9533794286240828e-05,
+      "loss": 0.4848,
+      "step": 526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.121314764022827,
+      "learning_rate": 1.953147293117465e-05,
+      "loss": 0.4554,
+      "step": 527
+    },
+    {
+      "epoch": 0.2504743833017078,
+      "grad_norm": 4.03399658203125,
+      "learning_rate": 1.9529145949774322e-05,
+      "loss": 0.5657,
+      "step": 528
+    },
+    {
+      "epoch": 0.2509487666034156,
+      "grad_norm": 3.444110631942749,
+      "learning_rate": 1.952681334341343e-05,
+      "loss": 0.4767,
+      "step": 529
+    },
+    {
+      "epoch": 0.2514231499051233,
+      "grad_norm": 3.272854804992676,
+      "learning_rate": 1.9524475113468897e-05,
+      "loss": 0.5218,
+      "step": 530
+    },
+    {
+      "epoch": 0.2518975332068311,
+      "grad_norm": 3.146151542663574,
+      "learning_rate": 1.9522131261320952e-05,
+      "loss": 0.47,
+      "step": 531
+    },
+    {
+      "epoch": 0.2523719165085389,
+      "grad_norm": 3.1734254360198975,
+      "learning_rate": 1.9519781788353148e-05,
+      "loss": 0.4846,
+      "step": 532
+    },
+    {
+      "epoch": 0.2528462998102467,
+      "grad_norm": 4.805948257446289,
+      "learning_rate": 1.9517426695952358e-05,
+      "loss": 0.4689,
+      "step": 533
+    },
+    {
+      "epoch": 0.25332068311195444,
+      "grad_norm": 3.5787010192871094,
+      "learning_rate": 1.9515065985508766e-05,
+      "loss": 0.4505,
+      "step": 534
+    },
+    {
+      "epoch": 0.25379506641366223,
+      "grad_norm": 3.29832124710083,
+      "learning_rate": 1.9512699658415882e-05,
+      "loss": 0.4994,
+      "step": 535
+    },
+    {
+      "epoch": 0.25426944971537,
+      "grad_norm": 3.159043312072754,
+      "learning_rate": 1.951032771607052e-05,
+      "loss": 0.5383,
+      "step": 536
+    },
+    {
+      "epoch": 0.2547438330170778,
+      "grad_norm": 3.565232515335083,
+      "learning_rate": 1.9507950159872814e-05,
+      "loss": 0.4824,
+      "step": 537
+    },
+    {
+      "epoch": 0.25521821631878555,
+      "grad_norm": 3.3344407081604004,
+      "learning_rate": 1.9505566991226214e-05,
+      "loss": 0.4797,
+      "step": 538
+    },
+    {
+      "epoch": 0.25569259962049334,
+      "grad_norm": 3.4726908206939697,
+      "learning_rate": 1.9503178211537483e-05,
+      "loss": 0.5105,
+      "step": 539
+    },
+    {
+      "epoch": 0.25616698292220114,
+      "grad_norm": 3.1351394653320312,
+      "learning_rate": 1.9500783822216693e-05,
+      "loss": 0.4824,
+      "step": 540
+    },
+    {
+      "epoch": 0.25664136622390893,
+      "grad_norm": 3.049128532409668,
+      "learning_rate": 1.9498383824677223e-05,
+      "loss": 0.518,
+      "step": 541
+    },
+    {
+      "epoch": 0.2571157495256167,
+      "grad_norm": 3.9116055965423584,
+      "learning_rate": 1.9495978220335774e-05,
+      "loss": 0.4997,
+      "step": 542
+    },
+    {
+      "epoch": 0.25759013282732446,
+      "grad_norm": 3.9851927757263184,
+      "learning_rate": 1.949356701061235e-05,
+      "loss": 0.4598,
+      "step": 543
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 3.269645929336548,
+      "learning_rate": 1.9491150196930258e-05,
+      "loss": 0.4539,
+      "step": 544
+    },
+    {
+      "epoch": 0.25853889943074004,
+      "grad_norm": 3.270271062850952,
+      "learning_rate": 1.9488727780716125e-05,
+      "loss": 0.4371,
+      "step": 545
+    },
+    {
+      "epoch": 0.25901328273244784,
+      "grad_norm": 3.121974468231201,
+      "learning_rate": 1.948629976339988e-05,
+      "loss": 0.4952,
+      "step": 546
+    },
+    {
+      "epoch": 0.2594876660341556,
+      "grad_norm": 2.9817652702331543,
+      "learning_rate": 1.9483866146414756e-05,
+      "loss": 0.4515,
+      "step": 547
+    },
+    {
+      "epoch": 0.25996204933586337,
+      "grad_norm": 3.0989949703216553,
+      "learning_rate": 1.9481426931197293e-05,
+      "loss": 0.4139,
+      "step": 548
+    },
+    {
+      "epoch": 0.26043643263757116,
+      "grad_norm": 4.340046405792236,
+      "learning_rate": 1.9478982119187333e-05,
+      "loss": 0.6325,
+      "step": 549
+    },
+    {
+      "epoch": 0.26091081593927895,
+      "grad_norm": 3.455524444580078,
+      "learning_rate": 1.9476531711828027e-05,
+      "loss": 0.5801,
+      "step": 550
+    },
+    {
+      "epoch": 0.26138519924098674,
+      "grad_norm": 3.185102939605713,
+      "learning_rate": 1.9474075710565825e-05,
+      "loss": 0.5409,
+      "step": 551
+    },
+    {
+      "epoch": 0.2618595825426945,
+      "grad_norm": 3.527089834213257,
+      "learning_rate": 1.9471614116850482e-05,
+      "loss": 0.44,
+      "step": 552
+    },
+    {
+      "epoch": 0.2623339658444023,
+      "grad_norm": 3.273226499557495,
+      "learning_rate": 1.946914693213505e-05,
+      "loss": 0.4909,
+      "step": 553
+    },
+    {
+      "epoch": 0.26280834914611007,
+      "grad_norm": 3.4487481117248535,
+      "learning_rate": 1.946667415787589e-05,
+      "loss": 0.4712,
+      "step": 554
+    },
+    {
+      "epoch": 0.26328273244781786,
+      "grad_norm": 3.9401209354400635,
+      "learning_rate": 1.9464195795532648e-05,
+      "loss": 0.4435,
+      "step": 555
+    },
+    {
+      "epoch": 0.2637571157495256,
+      "grad_norm": 3.349949836730957,
+      "learning_rate": 1.946171184656828e-05,
+      "loss": 0.4909,
+      "step": 556
+    },
+    {
+      "epoch": 0.2642314990512334,
+      "grad_norm": 4.190546989440918,
+      "learning_rate": 1.9459222312449036e-05,
+      "loss": 0.5163,
+      "step": 557
+    },
+    {
+      "epoch": 0.2647058823529412,
+      "grad_norm": 3.7838876247406006,
+      "learning_rate": 1.9456727194644465e-05,
+      "loss": 0.5356,
+      "step": 558
+    },
+    {
+      "epoch": 0.265180265654649,
+      "grad_norm": 3.253169536590576,
+      "learning_rate": 1.945422649462741e-05,
+      "loss": 0.5241,
+      "step": 559
+    },
+    {
+      "epoch": 0.2656546489563567,
+      "grad_norm": 3.7193899154663086,
+      "learning_rate": 1.9451720213874007e-05,
+      "loss": 0.4999,
+      "step": 560
+    },
+    {
+      "epoch": 0.2661290322580645,
+      "grad_norm": 3.82112717628479,
+      "learning_rate": 1.9449208353863693e-05,
+      "loss": 0.4958,
+      "step": 561
+    },
+    {
+      "epoch": 0.2666034155597723,
+      "grad_norm": 3.6052136421203613,
+      "learning_rate": 1.944669091607919e-05,
+      "loss": 0.5644,
+      "step": 562
+    },
+    {
+      "epoch": 0.2670777988614801,
+      "grad_norm": 3.2630624771118164,
+      "learning_rate": 1.9444167902006516e-05,
+      "loss": 0.4038,
+      "step": 563
+    },
+    {
+      "epoch": 0.2675521821631879,
+      "grad_norm": 3.3602817058563232,
+      "learning_rate": 1.944163931313499e-05,
+      "loss": 0.4928,
+      "step": 564
+    },
+    {
+      "epoch": 0.2680265654648956,
+      "grad_norm": 2.9780235290527344,
+      "learning_rate": 1.94391051509572e-05,
+      "loss": 0.4451,
+      "step": 565
+    },
+    {
+      "epoch": 0.2685009487666034,
+      "grad_norm": 3.799266815185547,
+      "learning_rate": 1.9436565416969045e-05,
+      "loss": 0.5159,
+      "step": 566
+    },
+    {
+      "epoch": 0.2689753320683112,
+      "grad_norm": 3.299086093902588,
+      "learning_rate": 1.9434020112669706e-05,
+      "loss": 0.4683,
+      "step": 567
+    },
+    {
+      "epoch": 0.269449715370019,
+      "grad_norm": 3.3280222415924072,
+      "learning_rate": 1.9431469239561646e-05,
+      "loss": 0.4656,
+      "step": 568
+    },
+    {
+      "epoch": 0.26992409867172673,
+      "grad_norm": 3.7164220809936523,
+      "learning_rate": 1.9428912799150624e-05,
+      "loss": 0.4551,
+      "step": 569
+    },
+    {
+      "epoch": 0.2703984819734345,
+      "grad_norm": 3.125343084335327,
+      "learning_rate": 1.9426350792945676e-05,
+      "loss": 0.4851,
+      "step": 570
+    },
+    {
+      "epoch": 0.2708728652751423,
+      "grad_norm": 2.9615533351898193,
+      "learning_rate": 1.9423783222459135e-05,
+      "loss": 0.4318,
+      "step": 571
+    },
+    {
+      "epoch": 0.2713472485768501,
+      "grad_norm": 2.992379903793335,
+      "learning_rate": 1.942121008920661e-05,
+      "loss": 0.4402,
+      "step": 572
+    },
+    {
+      "epoch": 0.2718216318785579,
+      "grad_norm": 4.136736869812012,
+      "learning_rate": 1.9418631394706998e-05,
+      "loss": 0.5978,
+      "step": 573
+    },
+    {
+      "epoch": 0.27229601518026564,
+      "grad_norm": 3.9404451847076416,
+      "learning_rate": 1.941604714048247e-05,
+      "loss": 0.5606,
+      "step": 574
+    },
+    {
+      "epoch": 0.27277039848197343,
+      "grad_norm": 3.232971668243408,
+      "learning_rate": 1.9413457328058495e-05,
+      "loss": 0.5066,
+      "step": 575
+    },
+    {
+      "epoch": 0.2732447817836812,
+      "grad_norm": 3.407205820083618,
+      "learning_rate": 1.941086195896381e-05,
+      "loss": 0.5584,
+      "step": 576
+    },
+    {
+      "epoch": 0.273719165085389,
+      "grad_norm": 3.10275936126709,
+      "learning_rate": 1.940826103473043e-05,
+      "loss": 0.4627,
+      "step": 577
+    },
+    {
+      "epoch": 0.27419354838709675,
+      "grad_norm": 3.237243413925171,
+      "learning_rate": 1.9405654556893667e-05,
+      "loss": 0.4897,
+      "step": 578
+    },
+    {
+      "epoch": 0.27466793168880455,
+      "grad_norm": 3.3360302448272705,
+      "learning_rate": 1.9403042526992087e-05,
+      "loss": 0.5143,
+      "step": 579
+    },
+    {
+      "epoch": 0.27514231499051234,
+      "grad_norm": 3.802823305130005,
+      "learning_rate": 1.9400424946567552e-05,
+      "loss": 0.492,
+      "step": 580
+    },
+    {
+      "epoch": 0.27561669829222013,
+      "grad_norm": 2.7487895488739014,
+      "learning_rate": 1.9397801817165192e-05,
+      "loss": 0.4175,
+      "step": 581
+    },
+    {
+      "epoch": 0.27609108159392787,
+      "grad_norm": 3.0970826148986816,
+      "learning_rate": 1.9395173140333413e-05,
+      "loss": 0.4156,
+      "step": 582
+    },
+    {
+      "epoch": 0.27656546489563566,
+      "grad_norm": 4.566722393035889,
+      "learning_rate": 1.93925389176239e-05,
+      "loss": 0.5653,
+      "step": 583
+    },
+    {
+      "epoch": 0.27703984819734345,
+      "grad_norm": 2.8217339515686035,
+      "learning_rate": 1.9389899150591605e-05,
+      "loss": 0.4451,
+      "step": 584
+    },
+    {
+      "epoch": 0.27751423149905124,
+      "grad_norm": 3.9668426513671875,
+      "learning_rate": 1.938725384079476e-05,
+      "loss": 0.5351,
+      "step": 585
+    },
+    {
+      "epoch": 0.27798861480075904,
+      "grad_norm": 2.601659059524536,
+      "learning_rate": 1.9384602989794868e-05,
+      "loss": 0.3897,
+      "step": 586
+    },
+    {
+      "epoch": 0.2784629981024668,
+      "grad_norm": 4.612437725067139,
+      "learning_rate": 1.938194659915669e-05,
+      "loss": 0.436,
+      "step": 587
+    },
+    {
+      "epoch": 0.27893738140417457,
+      "grad_norm": 3.368626356124878,
+      "learning_rate": 1.9379284670448274e-05,
+      "loss": 0.3943,
+      "step": 588
+    },
+    {
+      "epoch": 0.27941176470588236,
+      "grad_norm": 3.076282262802124,
+      "learning_rate": 1.937661720524093e-05,
+      "loss": 0.4004,
+      "step": 589
+    },
+    {
+      "epoch": 0.27988614800759015,
+      "grad_norm": 2.78485107421875,
+      "learning_rate": 1.9373944205109236e-05,
+      "loss": 0.35,
+      "step": 590
+    },
+    {
+      "epoch": 0.2803605313092979,
+      "grad_norm": 3.2779622077941895,
+      "learning_rate": 1.9371265671631038e-05,
+      "loss": 0.4068,
+      "step": 591
+    },
+    {
+      "epoch": 0.2808349146110057,
+      "grad_norm": 2.9968180656433105,
+      "learning_rate": 1.9368581606387442e-05,
+      "loss": 0.4667,
+      "step": 592
+    },
+    {
+      "epoch": 0.2813092979127135,
+      "grad_norm": 3.2578439712524414,
+      "learning_rate": 1.9365892010962834e-05,
+      "loss": 0.4826,
+      "step": 593
+    },
+    {
+      "epoch": 0.28178368121442127,
+      "grad_norm": 3.467496395111084,
+      "learning_rate": 1.936319688694485e-05,
+      "loss": 0.503,
+      "step": 594
+    },
+    {
+      "epoch": 0.28225806451612906,
+      "grad_norm": 2.929417848587036,
+      "learning_rate": 1.9360496235924396e-05,
+      "loss": 0.4527,
+      "step": 595
+    },
+    {
+      "epoch": 0.2827324478178368,
+      "grad_norm": 3.066240072250366,
+      "learning_rate": 1.935779005949564e-05,
+      "loss": 0.4604,
+      "step": 596
+    },
+    {
+      "epoch": 0.2832068311195446,
+      "grad_norm": 2.88321852684021,
+      "learning_rate": 1.935507835925601e-05,
+      "loss": 0.4544,
+      "step": 597
+    },
+    {
+      "epoch": 0.2836812144212524,
+      "grad_norm": 2.9748549461364746,
+      "learning_rate": 1.93523611368062e-05,
+      "loss": 0.4272,
+      "step": 598
+    },
+    {
+      "epoch": 0.2841555977229602,
+      "grad_norm": 2.744473695755005,
+      "learning_rate": 1.9349638393750156e-05,
+      "loss": 0.4874,
+      "step": 599
+    },
+    {
+      "epoch": 0.2846299810246679,
+      "grad_norm": 3.6597301959991455,
+      "learning_rate": 1.9346910131695084e-05,
+      "loss": 0.4996,
+      "step": 600
+    },
+    {
+      "epoch": 0.2851043643263757,
+      "grad_norm": 2.5389652252197266,
+      "learning_rate": 1.9344176352251456e-05,
+      "loss": 0.3724,
+      "step": 601
+    },
+    {
+      "epoch": 0.2855787476280835,
+      "grad_norm": 3.2770135402679443,
+      "learning_rate": 1.934143705703299e-05,
+      "loss": 0.4549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2860531309297913,
+      "grad_norm": 2.7468864917755127,
+      "learning_rate": 1.933869224765667e-05,
+      "loss": 0.4371,
+      "step": 603
+    },
+    {
+      "epoch": 0.286527514231499,
+      "grad_norm": 2.662562370300293,
+      "learning_rate": 1.933594192574272e-05,
+      "loss": 0.4146,
+      "step": 604
+    },
+    {
+      "epoch": 0.2870018975332068,
+      "grad_norm": 2.997250556945801,
+      "learning_rate": 1.933318609291464e-05,
+      "loss": 0.485,
+      "step": 605
+    },
+    {
+      "epoch": 0.2874762808349146,
+      "grad_norm": 2.8720083236694336,
+      "learning_rate": 1.9330424750799165e-05,
+      "loss": 0.4197,
+      "step": 606
+    },
+    {
+      "epoch": 0.2879506641366224,
+      "grad_norm": 3.827880620956421,
+      "learning_rate": 1.9327657901026284e-05,
+      "loss": 0.4941,
+      "step": 607
+    },
+    {
+      "epoch": 0.2884250474383302,
+      "grad_norm": 3.460559368133545,
+      "learning_rate": 1.9324885545229248e-05,
+      "loss": 0.4842,
+      "step": 608
+    },
+    {
+      "epoch": 0.28889943074003793,
+      "grad_norm": 2.8307604789733887,
+      "learning_rate": 1.932210768504455e-05,
+      "loss": 0.4564,
+      "step": 609
+    },
+    {
+      "epoch": 0.2893738140417457,
+      "grad_norm": 3.346304416656494,
+      "learning_rate": 1.9319324322111928e-05,
+      "loss": 0.5572,
+      "step": 610
+    },
+    {
+      "epoch": 0.2898481973434535,
+      "grad_norm": 3.4799015522003174,
+      "learning_rate": 1.931653545807438e-05,
+      "loss": 0.542,
+      "step": 611
+    },
+    {
+      "epoch": 0.2903225806451613,
+      "grad_norm": 2.8453822135925293,
+      "learning_rate": 1.931374109457814e-05,
+      "loss": 0.3665,
+      "step": 612
+    },
+    {
+      "epoch": 0.29079696394686905,
+      "grad_norm": 3.6189804077148438,
+      "learning_rate": 1.9310941233272698e-05,
+      "loss": 0.4945,
+      "step": 613
+    },
+    {
+      "epoch": 0.29127134724857684,
+      "grad_norm": 3.2037627696990967,
+      "learning_rate": 1.9308135875810778e-05,
+      "loss": 0.4499,
+      "step": 614
+    },
+    {
+      "epoch": 0.29174573055028463,
+      "grad_norm": 3.068361520767212,
+      "learning_rate": 1.930532502384836e-05,
+      "loss": 0.4529,
+      "step": 615
+    },
+    {
+      "epoch": 0.2922201138519924,
+      "grad_norm": 2.5314414501190186,
+      "learning_rate": 1.9302508679044662e-05,
+      "loss": 0.3866,
+      "step": 616
+    },
+    {
+      "epoch": 0.2926944971537002,
+      "grad_norm": 3.3053672313690186,
+      "learning_rate": 1.929968684306214e-05,
+      "loss": 0.5091,
+      "step": 617
+    },
+    {
+      "epoch": 0.29316888045540795,
+      "grad_norm": 3.5351502895355225,
+      "learning_rate": 1.9296859517566505e-05,
+      "loss": 0.5416,
+      "step": 618
+    },
+    {
+      "epoch": 0.29364326375711575,
+      "grad_norm": 3.617424964904785,
+      "learning_rate": 1.929402670422669e-05,
+      "loss": 0.5325,
+      "step": 619
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 2.799778938293457,
+      "learning_rate": 1.9291188404714876e-05,
+      "loss": 0.514,
+      "step": 620
+    },
+    {
+      "epoch": 0.29459203036053133,
+      "grad_norm": 3.1185672283172607,
+      "learning_rate": 1.9288344620706493e-05,
+      "loss": 0.5391,
+      "step": 621
+    },
+    {
+      "epoch": 0.29506641366223907,
+      "grad_norm": 3.1492013931274414,
+      "learning_rate": 1.9285495353880187e-05,
+      "loss": 0.4585,
+      "step": 622
+    },
+    {
+      "epoch": 0.29554079696394686,
+      "grad_norm": 3.211419105529785,
+      "learning_rate": 1.928264060591786e-05,
+      "loss": 0.4983,
+      "step": 623
+    },
+    {
+      "epoch": 0.29601518026565465,
+      "grad_norm": 3.3846914768218994,
+      "learning_rate": 1.927978037850464e-05,
+      "loss": 0.5182,
+      "step": 624
+    },
+    {
+      "epoch": 0.29648956356736245,
+      "grad_norm": 2.692244052886963,
+      "learning_rate": 1.927691467332889e-05,
+      "loss": 0.4298,
+      "step": 625
+    },
+    {
+      "epoch": 0.2969639468690702,
+      "grad_norm": 2.660977363586426,
+      "learning_rate": 1.9274043492082205e-05,
+      "loss": 0.396,
+      "step": 626
+    },
+    {
+      "epoch": 0.297438330170778,
+      "grad_norm": 3.270251989364624,
+      "learning_rate": 1.9271166836459418e-05,
+      "loss": 0.499,
+      "step": 627
+    },
+    {
+      "epoch": 0.29791271347248577,
+      "grad_norm": 3.6545050144195557,
+      "learning_rate": 1.926828470815859e-05,
+      "loss": 0.5114,
+      "step": 628
+    },
+    {
+      "epoch": 0.29838709677419356,
+      "grad_norm": 3.5217530727386475,
+      "learning_rate": 1.9265397108881015e-05,
+      "loss": 0.4573,
+      "step": 629
+    },
+    {
+      "epoch": 0.29886148007590135,
+      "grad_norm": 3.2720742225646973,
+      "learning_rate": 1.9262504040331208e-05,
+      "loss": 0.4423,
+      "step": 630
+    },
+    {
+      "epoch": 0.2993358633776091,
+      "grad_norm": 3.0071029663085938,
+      "learning_rate": 1.9259605504216922e-05,
+      "loss": 0.4364,
+      "step": 631
+    },
+    {
+      "epoch": 0.2998102466793169,
+      "grad_norm": 4.148307800292969,
+      "learning_rate": 1.925670150224914e-05,
+      "loss": 0.4924,
+      "step": 632
+    },
+    {
+      "epoch": 0.3002846299810247,
+      "grad_norm": 3.19143009185791,
+      "learning_rate": 1.9253792036142052e-05,
+      "loss": 0.5972,
+      "step": 633
+    },
+    {
+      "epoch": 0.30075901328273247,
+      "grad_norm": 2.9321649074554443,
+      "learning_rate": 1.92508771076131e-05,
+      "loss": 0.3967,
+      "step": 634
+    },
+    {
+      "epoch": 0.3012333965844402,
+      "grad_norm": 3.0709781646728516,
+      "learning_rate": 1.9247956718382933e-05,
+      "loss": 0.5369,
+      "step": 635
+    },
+    {
+      "epoch": 0.301707779886148,
+      "grad_norm": 2.6908938884735107,
+      "learning_rate": 1.9245030870175427e-05,
+      "loss": 0.4135,
+      "step": 636
+    },
+    {
+      "epoch": 0.3021821631878558,
+      "grad_norm": 2.776761054992676,
+      "learning_rate": 1.9242099564717683e-05,
+      "loss": 0.519,
+      "step": 637
+    },
+    {
+      "epoch": 0.3026565464895636,
+      "grad_norm": 3.2145349979400635,
+      "learning_rate": 1.9239162803740016e-05,
+      "loss": 0.4982,
+      "step": 638
+    },
+    {
+      "epoch": 0.3031309297912713,
+      "grad_norm": 2.7437076568603516,
+      "learning_rate": 1.9236220588975976e-05,
+      "loss": 0.4159,
+      "step": 639
+    },
+    {
+      "epoch": 0.3036053130929791,
+      "grad_norm": 3.104196310043335,
+      "learning_rate": 1.9233272922162318e-05,
+      "loss": 0.493,
+      "step": 640
+    },
+    {
+      "epoch": 0.3040796963946869,
+      "grad_norm": 3.3658387660980225,
+      "learning_rate": 1.9230319805039022e-05,
+      "loss": 0.5201,
+      "step": 641
+    },
+    {
+      "epoch": 0.3045540796963947,
+      "grad_norm": 2.732059955596924,
+      "learning_rate": 1.922736123934928e-05,
+      "loss": 0.4961,
+      "step": 642
+    },
+    {
+      "epoch": 0.3050284629981025,
+      "grad_norm": 3.038705825805664,
+      "learning_rate": 1.922439722683951e-05,
+      "loss": 0.4819,
+      "step": 643
+    },
+    {
+      "epoch": 0.3055028462998102,
+      "grad_norm": 2.702758312225342,
+      "learning_rate": 1.9221427769259333e-05,
+      "loss": 0.4632,
+      "step": 644
+    },
+    {
+      "epoch": 0.305977229601518,
+      "grad_norm": 3.554279327392578,
+      "learning_rate": 1.9218452868361597e-05,
+      "loss": 0.4135,
+      "step": 645
+    },
+    {
+      "epoch": 0.3064516129032258,
+      "grad_norm": 3.9966790676116943,
+      "learning_rate": 1.921547252590235e-05,
+      "loss": 0.537,
+      "step": 646
+    },
+    {
+      "epoch": 0.3069259962049336,
+      "grad_norm": 3.105196475982666,
+      "learning_rate": 1.9212486743640864e-05,
+      "loss": 0.474,
+      "step": 647
+    },
+    {
+      "epoch": 0.30740037950664134,
+      "grad_norm": 3.0190954208374023,
+      "learning_rate": 1.9209495523339614e-05,
+      "loss": 0.4715,
+      "step": 648
+    },
+    {
+      "epoch": 0.30787476280834913,
+      "grad_norm": 3.061038017272949,
+      "learning_rate": 1.920649886676429e-05,
+      "loss": 0.5493,
+      "step": 649
+    },
+    {
+      "epoch": 0.3083491461100569,
+      "grad_norm": 3.0604310035705566,
+      "learning_rate": 1.9203496775683787e-05,
+      "loss": 0.4818,
+      "step": 650
+    },
+    {
+      "epoch": 0.3088235294117647,
+      "grad_norm": 3.209660768508911,
+      "learning_rate": 1.9200489251870207e-05,
+      "loss": 0.3923,
+      "step": 651
+    },
+    {
+      "epoch": 0.3092979127134725,
+      "grad_norm": 3.1261727809906006,
+      "learning_rate": 1.9197476297098868e-05,
+      "loss": 0.4662,
+      "step": 652
+    },
+    {
+      "epoch": 0.30977229601518025,
+      "grad_norm": 3.9394774436950684,
+      "learning_rate": 1.919445791314828e-05,
+      "loss": 0.4693,
+      "step": 653
+    },
+    {
+      "epoch": 0.31024667931688804,
+      "grad_norm": 2.708280086517334,
+      "learning_rate": 1.9191434101800174e-05,
+      "loss": 0.4855,
+      "step": 654
+    },
+    {
+      "epoch": 0.31072106261859583,
+      "grad_norm": 3.1161880493164062,
+      "learning_rate": 1.9188404864839465e-05,
+      "loss": 0.5513,
+      "step": 655
+    },
+    {
+      "epoch": 0.3111954459203036,
+      "grad_norm": 3.3281567096710205,
+      "learning_rate": 1.918537020405429e-05,
+      "loss": 0.4403,
+      "step": 656
+    },
+    {
+      "epoch": 0.31166982922201136,
+      "grad_norm": 3.432615041732788,
+      "learning_rate": 1.9182330121235978e-05,
+      "loss": 0.5654,
+      "step": 657
+    },
+    {
+      "epoch": 0.31214421252371916,
+      "grad_norm": 3.0545363426208496,
+      "learning_rate": 1.917928461817906e-05,
+      "loss": 0.4361,
+      "step": 658
+    },
+    {
+      "epoch": 0.31261859582542695,
+      "grad_norm": 2.9954404830932617,
+      "learning_rate": 1.917623369668126e-05,
+      "loss": 0.4687,
+      "step": 659
+    },
+    {
+      "epoch": 0.31309297912713474,
+      "grad_norm": 3.5407347679138184,
+      "learning_rate": 1.9173177358543512e-05,
+      "loss": 0.5126,
+      "step": 660
+    },
+    {
+      "epoch": 0.3135673624288425,
+      "grad_norm": 3.2592506408691406,
+      "learning_rate": 1.9170115605569945e-05,
+      "loss": 0.5198,
+      "step": 661
+    },
+    {
+      "epoch": 0.31404174573055027,
+      "grad_norm": 2.86240291595459,
+      "learning_rate": 1.9167048439567876e-05,
+      "loss": 0.4064,
+      "step": 662
+    },
+    {
+      "epoch": 0.31451612903225806,
+      "grad_norm": 2.8558406829833984,
+      "learning_rate": 1.9163975862347824e-05,
+      "loss": 0.4397,
+      "step": 663
+    },
+    {
+      "epoch": 0.31499051233396586,
+      "grad_norm": 2.9497201442718506,
+      "learning_rate": 1.9160897875723505e-05,
+      "loss": 0.447,
+      "step": 664
+    },
+    {
+      "epoch": 0.31546489563567365,
+      "grad_norm": 2.7483532428741455,
+      "learning_rate": 1.915781448151182e-05,
+      "loss": 0.4017,
+      "step": 665
+    },
+    {
+      "epoch": 0.3159392789373814,
+      "grad_norm": 3.7263011932373047,
+      "learning_rate": 1.9154725681532867e-05,
+      "loss": 0.5402,
+      "step": 666
+    },
+    {
+      "epoch": 0.3164136622390892,
+      "grad_norm": 3.7040700912475586,
+      "learning_rate": 1.9151631477609932e-05,
+      "loss": 0.4836,
+      "step": 667
+    },
+    {
+      "epoch": 0.31688804554079697,
+      "grad_norm": 2.6552894115448,
+      "learning_rate": 1.9148531871569496e-05,
+      "loss": 0.4124,
+      "step": 668
+    },
+    {
+      "epoch": 0.31736242884250476,
+      "grad_norm": 3.1435623168945312,
+      "learning_rate": 1.9145426865241224e-05,
+      "loss": 0.4807,
+      "step": 669
+    },
+    {
+      "epoch": 0.3178368121442125,
+      "grad_norm": 2.9633023738861084,
+      "learning_rate": 1.9142316460457974e-05,
+      "loss": 0.4609,
+      "step": 670
+    },
+    {
+      "epoch": 0.3183111954459203,
+      "grad_norm": 3.0346839427948,
+      "learning_rate": 1.9139200659055785e-05,
+      "loss": 0.4028,
+      "step": 671
+    },
+    {
+      "epoch": 0.3187855787476281,
+      "grad_norm": 3.140936851501465,
+      "learning_rate": 1.913607946287388e-05,
+      "loss": 0.4631,
+      "step": 672
+    },
+    {
+      "epoch": 0.3192599620493359,
+      "grad_norm": 2.8173251152038574,
+      "learning_rate": 1.9132952873754675e-05,
+      "loss": 0.4232,
+      "step": 673
+    },
+    {
+      "epoch": 0.31973434535104367,
+      "grad_norm": 2.9139256477355957,
+      "learning_rate": 1.9129820893543766e-05,
+      "loss": 0.4401,
+      "step": 674
+    },
+    {
+      "epoch": 0.3202087286527514,
+      "grad_norm": 3.9275453090667725,
+      "learning_rate": 1.912668352408992e-05,
+      "loss": 0.5251,
+      "step": 675
+    },
+    {
+      "epoch": 0.3206831119544592,
+      "grad_norm": 2.808434247970581,
+      "learning_rate": 1.9123540767245107e-05,
+      "loss": 0.4207,
+      "step": 676
+    },
+    {
+      "epoch": 0.321157495256167,
+      "grad_norm": 2.854140043258667,
+      "learning_rate": 1.912039262486446e-05,
+      "loss": 0.4516,
+      "step": 677
+    },
+    {
+      "epoch": 0.3216318785578748,
+      "grad_norm": 3.1188645362854004,
+      "learning_rate": 1.9117239098806296e-05,
+      "loss": 0.473,
+      "step": 678
+    },
+    {
+      "epoch": 0.3221062618595825,
+      "grad_norm": 2.9418070316314697,
+      "learning_rate": 1.911408019093211e-05,
+      "loss": 0.4666,
+      "step": 679
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 7.523387908935547,
+      "learning_rate": 1.9110915903106577e-05,
+      "loss": 0.4657,
+      "step": 680
+    },
+    {
+      "epoch": 0.3230550284629981,
+      "grad_norm": 3.0443685054779053,
+      "learning_rate": 1.9107746237197542e-05,
+      "loss": 0.5032,
+      "step": 681
+    },
+    {
+      "epoch": 0.3235294117647059,
+      "grad_norm": 3.1826181411743164,
+      "learning_rate": 1.910457119507603e-05,
+      "loss": 0.4716,
+      "step": 682
+    },
+    {
+      "epoch": 0.32400379506641364,
+      "grad_norm": 2.9390838146209717,
+      "learning_rate": 1.9101390778616232e-05,
+      "loss": 0.3951,
+      "step": 683
+    },
+    {
+      "epoch": 0.32447817836812143,
+      "grad_norm": 3.393476963043213,
+      "learning_rate": 1.909820498969552e-05,
+      "loss": 0.5147,
+      "step": 684
+    },
+    {
+      "epoch": 0.3249525616698292,
+      "grad_norm": 3.6798181533813477,
+      "learning_rate": 1.9095013830194432e-05,
+      "loss": 0.4758,
+      "step": 685
+    },
+    {
+      "epoch": 0.325426944971537,
+      "grad_norm": 3.2658169269561768,
+      "learning_rate": 1.9091817301996684e-05,
+      "loss": 0.4902,
+      "step": 686
+    },
+    {
+      "epoch": 0.3259013282732448,
+      "grad_norm": 3.256455898284912,
+      "learning_rate": 1.9088615406989146e-05,
+      "loss": 0.5238,
+      "step": 687
+    },
+    {
+      "epoch": 0.32637571157495254,
+      "grad_norm": 3.187185287475586,
+      "learning_rate": 1.908540814706187e-05,
+      "loss": 0.4681,
+      "step": 688
+    },
+    {
+      "epoch": 0.32685009487666034,
+      "grad_norm": 2.9497554302215576,
+      "learning_rate": 1.9082195524108068e-05,
+      "loss": 0.4091,
+      "step": 689
+    },
+    {
+      "epoch": 0.32732447817836813,
+      "grad_norm": 3.060279607772827,
+      "learning_rate": 1.907897754002412e-05,
+      "loss": 0.504,
+      "step": 690
+    },
+    {
+      "epoch": 0.3277988614800759,
+      "grad_norm": 2.83272123336792,
+      "learning_rate": 1.9075754196709574e-05,
+      "loss": 0.4661,
+      "step": 691
+    },
+    {
+      "epoch": 0.32827324478178366,
+      "grad_norm": 3.2421908378601074,
+      "learning_rate": 1.9072525496067128e-05,
+      "loss": 0.3761,
+      "step": 692
+    },
+    {
+      "epoch": 0.32874762808349145,
+      "grad_norm": 3.3531394004821777,
+      "learning_rate": 1.9069291440002665e-05,
+      "loss": 0.553,
+      "step": 693
+    },
+    {
+      "epoch": 0.32922201138519924,
+      "grad_norm": 3.3758304119110107,
+      "learning_rate": 1.9066052030425206e-05,
+      "loss": 0.5256,
+      "step": 694
+    },
+    {
+      "epoch": 0.32969639468690703,
+      "grad_norm": 3.2974579334259033,
+      "learning_rate": 1.9062807269246945e-05,
+      "loss": 0.4929,
+      "step": 695
+    },
+    {
+      "epoch": 0.3301707779886148,
+      "grad_norm": 3.126060962677002,
+      "learning_rate": 1.9059557158383234e-05,
+      "loss": 0.4732,
+      "step": 696
+    },
+    {
+      "epoch": 0.33064516129032256,
+      "grad_norm": 3.069438934326172,
+      "learning_rate": 1.9056301699752578e-05,
+      "loss": 0.4659,
+      "step": 697
+    },
+    {
+      "epoch": 0.33111954459203036,
+      "grad_norm": 3.131999969482422,
+      "learning_rate": 1.9053040895276652e-05,
+      "loss": 0.486,
+      "step": 698
+    },
+    {
+      "epoch": 0.33159392789373815,
+      "grad_norm": 3.255291700363159,
+      "learning_rate": 1.904977474688026e-05,
+      "loss": 0.4963,
+      "step": 699
+    },
+    {
+      "epoch": 0.33206831119544594,
+      "grad_norm": 3.3611719608306885,
+      "learning_rate": 1.9046503256491395e-05,
+      "loss": 0.5017,
+      "step": 700
+    },
+    {
+      "epoch": 0.3325426944971537,
+      "grad_norm": 2.538780689239502,
+      "learning_rate": 1.904322642604117e-05,
+      "loss": 0.405,
+      "step": 701
+    },
+    {
+      "epoch": 0.33301707779886147,
+      "grad_norm": 2.767956495285034,
+      "learning_rate": 1.9039944257463875e-05,
+      "loss": 0.3662,
+      "step": 702
+    },
+    {
+      "epoch": 0.33349146110056926,
+      "grad_norm": 2.7074410915374756,
+      "learning_rate": 1.903665675269694e-05,
+      "loss": 0.4577,
+      "step": 703
+    },
+    {
+      "epoch": 0.33396584440227706,
+      "grad_norm": 3.3300936222076416,
+      "learning_rate": 1.9033363913680944e-05,
+      "loss": 0.4979,
+      "step": 704
+    },
+    {
+      "epoch": 0.3344402277039848,
+      "grad_norm": 2.8957359790802,
+      "learning_rate": 1.9030065742359618e-05,
+      "loss": 0.359,
+      "step": 705
+    },
+    {
+      "epoch": 0.3349146110056926,
+      "grad_norm": 3.4609487056732178,
+      "learning_rate": 1.9026762240679843e-05,
+      "loss": 0.5508,
+      "step": 706
+    },
+    {
+      "epoch": 0.3353889943074004,
+      "grad_norm": 3.0877790451049805,
+      "learning_rate": 1.902345341059164e-05,
+      "loss": 0.5167,
+      "step": 707
+    },
+    {
+      "epoch": 0.33586337760910817,
+      "grad_norm": 3.2066214084625244,
+      "learning_rate": 1.9020139254048174e-05,
+      "loss": 0.4389,
+      "step": 708
+    },
+    {
+      "epoch": 0.33633776091081596,
+      "grad_norm": 3.8554136753082275,
+      "learning_rate": 1.9016819773005774e-05,
+      "loss": 0.5015,
+      "step": 709
+    },
+    {
+      "epoch": 0.3368121442125237,
+      "grad_norm": 3.0794146060943604,
+      "learning_rate": 1.901349496942388e-05,
+      "loss": 0.4256,
+      "step": 710
+    },
+    {
+      "epoch": 0.3372865275142315,
+      "grad_norm": 3.269186496734619,
+      "learning_rate": 1.9010164845265103e-05,
+      "loss": 0.4864,
+      "step": 711
+    },
+    {
+      "epoch": 0.3377609108159393,
+      "grad_norm": 3.30047607421875,
+      "learning_rate": 1.9006829402495174e-05,
+      "loss": 0.5635,
+      "step": 712
+    },
+    {
+      "epoch": 0.3382352941176471,
+      "grad_norm": 3.331386089324951,
+      "learning_rate": 1.9003488643082978e-05,
+      "loss": 0.5245,
+      "step": 713
+    },
+    {
+      "epoch": 0.3387096774193548,
+      "grad_norm": 3.2732012271881104,
+      "learning_rate": 1.9000142569000524e-05,
+      "loss": 0.4657,
+      "step": 714
+    },
+    {
+      "epoch": 0.3391840607210626,
+      "grad_norm": 2.6173768043518066,
+      "learning_rate": 1.8996791182222977e-05,
+      "loss": 0.4839,
+      "step": 715
+    },
+    {
+      "epoch": 0.3396584440227704,
+      "grad_norm": 2.7928354740142822,
+      "learning_rate": 1.899343448472862e-05,
+      "loss": 0.5419,
+      "step": 716
+    },
+    {
+      "epoch": 0.3401328273244782,
+      "grad_norm": 2.8685014247894287,
+      "learning_rate": 1.899007247849888e-05,
+      "loss": 0.3946,
+      "step": 717
+    },
+    {
+      "epoch": 0.340607210626186,
+      "grad_norm": 2.6877050399780273,
+      "learning_rate": 1.8986705165518318e-05,
+      "loss": 0.457,
+      "step": 718
+    },
+    {
+      "epoch": 0.3410815939278937,
+      "grad_norm": 3.0826516151428223,
+      "learning_rate": 1.898333254777462e-05,
+      "loss": 0.4636,
+      "step": 719
+    },
+    {
+      "epoch": 0.3415559772296015,
+      "grad_norm": 4.633234024047852,
+      "learning_rate": 1.897995462725862e-05,
+      "loss": 0.4727,
+      "step": 720
+    },
+    {
+      "epoch": 0.3420303605313093,
+      "grad_norm": 2.6432247161865234,
+      "learning_rate": 1.8976571405964258e-05,
+      "loss": 0.4449,
+      "step": 721
+    },
+    {
+      "epoch": 0.3425047438330171,
+      "grad_norm": 5.404833793640137,
+      "learning_rate": 1.8973182885888626e-05,
+      "loss": 0.4411,
+      "step": 722
+    },
+    {
+      "epoch": 0.34297912713472484,
+      "grad_norm": 2.931230306625366,
+      "learning_rate": 1.8969789069031927e-05,
+      "loss": 0.4184,
+      "step": 723
+    },
+    {
+      "epoch": 0.34345351043643263,
+      "grad_norm": 3.3786165714263916,
+      "learning_rate": 1.8966389957397503e-05,
+      "loss": 0.4311,
+      "step": 724
+    },
+    {
+      "epoch": 0.3439278937381404,
+      "grad_norm": 2.9587435722351074,
+      "learning_rate": 1.896298555299181e-05,
+      "loss": 0.4465,
+      "step": 725
+    },
+    {
+      "epoch": 0.3444022770398482,
+      "grad_norm": 2.8019826412200928,
+      "learning_rate": 1.895957585782444e-05,
+      "loss": 0.4498,
+      "step": 726
+    },
+    {
+      "epoch": 0.34487666034155595,
+      "grad_norm": 3.1993303298950195,
+      "learning_rate": 1.8956160873908097e-05,
+      "loss": 0.4221,
+      "step": 727
+    },
+    {
+      "epoch": 0.34535104364326374,
+      "grad_norm": 2.909705877304077,
+      "learning_rate": 1.895274060325862e-05,
+      "loss": 0.4789,
+      "step": 728
+    },
+    {
+      "epoch": 0.34582542694497154,
+      "grad_norm": 3.371626138687134,
+      "learning_rate": 1.8949315047894956e-05,
+      "loss": 0.5294,
+      "step": 729
+    },
+    {
+      "epoch": 0.34629981024667933,
+      "grad_norm": 3.704002618789673,
+      "learning_rate": 1.8945884209839172e-05,
+      "loss": 0.5202,
+      "step": 730
+    },
+    {
+      "epoch": 0.3467741935483871,
+      "grad_norm": 2.856421709060669,
+      "learning_rate": 1.8942448091116464e-05,
+      "loss": 0.4168,
+      "step": 731
+    },
+    {
+      "epoch": 0.34724857685009486,
+      "grad_norm": 3.03263783454895,
+      "learning_rate": 1.8939006693755138e-05,
+      "loss": 0.4592,
+      "step": 732
+    },
+    {
+      "epoch": 0.34772296015180265,
+      "grad_norm": 3.3531453609466553,
+      "learning_rate": 1.8935560019786618e-05,
+      "loss": 0.4228,
+      "step": 733
+    },
+    {
+      "epoch": 0.34819734345351044,
+      "grad_norm": 2.8518269062042236,
+      "learning_rate": 1.8932108071245435e-05,
+      "loss": 0.4481,
+      "step": 734
+    },
+    {
+      "epoch": 0.34867172675521824,
+      "grad_norm": 3.01810622215271,
+      "learning_rate": 1.8928650850169246e-05,
+      "loss": 0.5109,
+      "step": 735
+    },
+    {
+      "epoch": 0.349146110056926,
+      "grad_norm": 3.7409400939941406,
+      "learning_rate": 1.8925188358598815e-05,
+      "loss": 0.5323,
+      "step": 736
+    },
+    {
+      "epoch": 0.34962049335863377,
+      "grad_norm": 2.571120262145996,
+      "learning_rate": 1.892172059857801e-05,
+      "loss": 0.4073,
+      "step": 737
+    },
+    {
+      "epoch": 0.35009487666034156,
+      "grad_norm": 3.2061607837677,
+      "learning_rate": 1.8918247572153822e-05,
+      "loss": 0.5185,
+      "step": 738
+    },
+    {
+      "epoch": 0.35056925996204935,
+      "grad_norm": 2.933711528778076,
+      "learning_rate": 1.8914769281376345e-05,
+      "loss": 0.3978,
+      "step": 739
+    },
+    {
+      "epoch": 0.3510436432637571,
+      "grad_norm": 3.2886109352111816,
+      "learning_rate": 1.8911285728298778e-05,
+      "loss": 0.4505,
+      "step": 740
+    },
+    {
+      "epoch": 0.3515180265654649,
+      "grad_norm": 3.886824607849121,
+      "learning_rate": 1.8907796914977422e-05,
+      "loss": 0.5427,
+      "step": 741
+    },
+    {
+      "epoch": 0.3519924098671727,
+      "grad_norm": 3.2919833660125732,
+      "learning_rate": 1.8904302843471692e-05,
+      "loss": 0.409,
+      "step": 742
+    },
+    {
+      "epoch": 0.35246679316888047,
+      "grad_norm": 3.437626600265503,
+      "learning_rate": 1.8900803515844107e-05,
+      "loss": 0.4457,
+      "step": 743
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 3.2604973316192627,
+      "learning_rate": 1.8897298934160285e-05,
+      "loss": 0.4763,
+      "step": 744
+    },
+    {
+      "epoch": 0.353415559772296,
+      "grad_norm": 2.8597803115844727,
+      "learning_rate": 1.8893789100488945e-05,
+      "loss": 0.4451,
+      "step": 745
+    },
+    {
+      "epoch": 0.3538899430740038,
+      "grad_norm": 2.9492034912109375,
+      "learning_rate": 1.8890274016901905e-05,
+      "loss": 0.4771,
+      "step": 746
+    },
+    {
+      "epoch": 0.3543643263757116,
+      "grad_norm": 3.0947840213775635,
+      "learning_rate": 1.888675368547409e-05,
+      "loss": 0.4063,
+      "step": 747
+    },
+    {
+      "epoch": 0.3548387096774194,
+      "grad_norm": 2.7532737255096436,
+      "learning_rate": 1.888322810828351e-05,
+      "loss": 0.461,
+      "step": 748
+    },
+    {
+      "epoch": 0.3553130929791271,
+      "grad_norm": 3.891683340072632,
+      "learning_rate": 1.887969728741128e-05,
+      "loss": 0.4789,
+      "step": 749
+    },
+    {
+      "epoch": 0.3557874762808349,
+      "grad_norm": 3.2587778568267822,
+      "learning_rate": 1.8876161224941607e-05,
+      "loss": 0.4776,
+      "step": 750
+    },
+    {
+      "epoch": 0.3562618595825427,
+      "grad_norm": 2.7931885719299316,
+      "learning_rate": 1.8872619922961802e-05,
+      "loss": 0.4503,
+      "step": 751
+    },
+    {
+      "epoch": 0.3567362428842505,
+      "grad_norm": 3.225916862487793,
+      "learning_rate": 1.886907338356225e-05,
+      "loss": 0.4879,
+      "step": 752
+    },
+    {
+      "epoch": 0.3572106261859583,
+      "grad_norm": 2.8587422370910645,
+      "learning_rate": 1.8865521608836446e-05,
+      "loss": 0.4243,
+      "step": 753
+    },
+    {
+      "epoch": 0.357685009487666,
+      "grad_norm": 3.333045482635498,
+      "learning_rate": 1.8861964600880963e-05,
+      "loss": 0.4953,
+      "step": 754
+    },
+    {
+      "epoch": 0.3581593927893738,
+      "grad_norm": 2.7636733055114746,
+      "learning_rate": 1.885840236179547e-05,
+      "loss": 0.4484,
+      "step": 755
+    },
+    {
+      "epoch": 0.3586337760910816,
+      "grad_norm": 2.7610294818878174,
+      "learning_rate": 1.8854834893682722e-05,
+      "loss": 0.4055,
+      "step": 756
+    },
+    {
+      "epoch": 0.3591081593927894,
+      "grad_norm": 3.245123863220215,
+      "learning_rate": 1.8851262198648555e-05,
+      "loss": 0.5173,
+      "step": 757
+    },
+    {
+      "epoch": 0.35958254269449713,
+      "grad_norm": 2.8020386695861816,
+      "learning_rate": 1.88476842788019e-05,
+      "loss": 0.4727,
+      "step": 758
+    },
+    {
+      "epoch": 0.3600569259962049,
+      "grad_norm": 3.41428542137146,
+      "learning_rate": 1.8844101136254768e-05,
+      "loss": 0.4924,
+      "step": 759
+    },
+    {
+      "epoch": 0.3605313092979127,
+      "grad_norm": 2.779858350753784,
+      "learning_rate": 1.884051277312225e-05,
+      "loss": 0.4455,
+      "step": 760
+    },
+    {
+      "epoch": 0.3610056925996205,
+      "grad_norm": 2.965505838394165,
+      "learning_rate": 1.8836919191522526e-05,
+      "loss": 0.5216,
+      "step": 761
+    },
+    {
+      "epoch": 0.36148007590132825,
+      "grad_norm": 2.897125005722046,
+      "learning_rate": 1.8833320393576847e-05,
+      "loss": 0.4364,
+      "step": 762
+    },
+    {
+      "epoch": 0.36195445920303604,
+      "grad_norm": 3.3312103748321533,
+      "learning_rate": 1.8829716381409545e-05,
+      "loss": 0.541,
+      "step": 763
+    },
+    {
+      "epoch": 0.36242884250474383,
+      "grad_norm": 3.1323423385620117,
+      "learning_rate": 1.8826107157148042e-05,
+      "loss": 0.4879,
+      "step": 764
+    },
+    {
+      "epoch": 0.3629032258064516,
+      "grad_norm": 2.554779052734375,
+      "learning_rate": 1.882249272292282e-05,
+      "loss": 0.3459,
+      "step": 765
+    },
+    {
+      "epoch": 0.3633776091081594,
+      "grad_norm": 2.9526796340942383,
+      "learning_rate": 1.8818873080867445e-05,
+      "loss": 0.4802,
+      "step": 766
+    },
+    {
+      "epoch": 0.36385199240986715,
+      "grad_norm": 2.4592695236206055,
+      "learning_rate": 1.8815248233118558e-05,
+      "loss": 0.3779,
+      "step": 767
+    },
+    {
+      "epoch": 0.36432637571157495,
+      "grad_norm": 2.9448318481445312,
+      "learning_rate": 1.881161818181587e-05,
+      "loss": 0.4546,
+      "step": 768
+    },
+    {
+      "epoch": 0.36480075901328274,
+      "grad_norm": 3.2894225120544434,
+      "learning_rate": 1.8807982929102164e-05,
+      "loss": 0.4668,
+      "step": 769
+    },
+    {
+      "epoch": 0.36527514231499053,
+      "grad_norm": 2.9263877868652344,
+      "learning_rate": 1.8804342477123292e-05,
+      "loss": 0.4316,
+      "step": 770
+    },
+    {
+      "epoch": 0.36574952561669827,
+      "grad_norm": 2.8721306324005127,
+      "learning_rate": 1.880069682802818e-05,
+      "loss": 0.4152,
+      "step": 771
+    },
+    {
+      "epoch": 0.36622390891840606,
+      "grad_norm": 2.6576414108276367,
+      "learning_rate": 1.879704598396882e-05,
+      "loss": 0.3802,
+      "step": 772
+    },
+    {
+      "epoch": 0.36669829222011385,
+      "grad_norm": 2.7889490127563477,
+      "learning_rate": 1.879338994710026e-05,
+      "loss": 0.3833,
+      "step": 773
+    },
+    {
+      "epoch": 0.36717267552182165,
+      "grad_norm": 3.141726016998291,
+      "learning_rate": 1.8789728719580632e-05,
+      "loss": 0.4202,
+      "step": 774
+    },
+    {
+      "epoch": 0.36764705882352944,
+      "grad_norm": 3.9903249740600586,
+      "learning_rate": 1.878606230357112e-05,
+      "loss": 0.5108,
+      "step": 775
+    },
+    {
+      "epoch": 0.3681214421252372,
+      "grad_norm": 2.6597206592559814,
+      "learning_rate": 1.878239070123597e-05,
+      "loss": 0.4236,
+      "step": 776
+    },
+    {
+      "epoch": 0.36859582542694497,
+      "grad_norm": 3.9152090549468994,
+      "learning_rate": 1.8778713914742494e-05,
+      "loss": 0.5953,
+      "step": 777
+    },
+    {
+      "epoch": 0.36907020872865276,
+      "grad_norm": 2.510295867919922,
+      "learning_rate": 1.8775031946261065e-05,
+      "loss": 0.389,
+      "step": 778
+    },
+    {
+      "epoch": 0.36954459203036055,
+      "grad_norm": 2.6154842376708984,
+      "learning_rate": 1.877134479796511e-05,
+      "loss": 0.4015,
+      "step": 779
+    },
+    {
+      "epoch": 0.3700189753320683,
+      "grad_norm": 3.2904458045959473,
+      "learning_rate": 1.8767652472031118e-05,
+      "loss": 0.4731,
+      "step": 780
+    },
+    {
+      "epoch": 0.3704933586337761,
+      "grad_norm": 3.2095184326171875,
+      "learning_rate": 1.8763954970638628e-05,
+      "loss": 0.5158,
+      "step": 781
+    },
+    {
+      "epoch": 0.3709677419354839,
+      "grad_norm": 2.783874034881592,
+      "learning_rate": 1.8760252295970245e-05,
+      "loss": 0.4831,
+      "step": 782
+    },
+    {
+      "epoch": 0.37144212523719167,
+      "grad_norm": 2.9339380264282227,
+      "learning_rate": 1.8756544450211614e-05,
+      "loss": 0.5171,
+      "step": 783
+    },
+    {
+      "epoch": 0.3719165085388994,
+      "grad_norm": 3.8392632007598877,
+      "learning_rate": 1.875283143555145e-05,
+      "loss": 0.4464,
+      "step": 784
+    },
+    {
+      "epoch": 0.3723908918406072,
+      "grad_norm": 2.6166434288024902,
+      "learning_rate": 1.8749113254181498e-05,
+      "loss": 0.4343,
+      "step": 785
+    },
+    {
+      "epoch": 0.372865275142315,
+      "grad_norm": 2.91298246383667,
+      "learning_rate": 1.874538990829657e-05,
+      "loss": 0.5344,
+      "step": 786
+    },
+    {
+      "epoch": 0.3733396584440228,
+      "grad_norm": 2.931877374649048,
+      "learning_rate": 1.874166140009452e-05,
+      "loss": 0.4732,
+      "step": 787
+    },
+    {
+      "epoch": 0.3738140417457306,
+      "grad_norm": 2.885974407196045,
+      "learning_rate": 1.8737927731776245e-05,
+      "loss": 0.4594,
+      "step": 788
+    },
+    {
+      "epoch": 0.3742884250474383,
+      "grad_norm": 3.037022352218628,
+      "learning_rate": 1.8734188905545697e-05,
+      "loss": 0.4378,
+      "step": 789
+    },
+    {
+      "epoch": 0.3747628083491461,
+      "grad_norm": 3.06512713432312,
+      "learning_rate": 1.8730444923609865e-05,
+      "loss": 0.4882,
+      "step": 790
+    },
+    {
+      "epoch": 0.3752371916508539,
+      "grad_norm": 3.42930269241333,
+      "learning_rate": 1.872669578817879e-05,
+      "loss": 0.5288,
+      "step": 791
+    },
+    {
+      "epoch": 0.3757115749525617,
+      "grad_norm": 2.3780622482299805,
+      "learning_rate": 1.872294150146554e-05,
+      "loss": 0.3665,
+      "step": 792
+    },
+    {
+      "epoch": 0.3761859582542694,
+      "grad_norm": 3.4545505046844482,
+      "learning_rate": 1.8719182065686242e-05,
+      "loss": 0.5605,
+      "step": 793
+    },
+    {
+      "epoch": 0.3766603415559772,
+      "grad_norm": 2.9837419986724854,
+      "learning_rate": 1.871541748306005e-05,
+      "loss": 0.4979,
+      "step": 794
+    },
+    {
+      "epoch": 0.377134724857685,
+      "grad_norm": 2.7663657665252686,
+      "learning_rate": 1.871164775580916e-05,
+      "loss": 0.4332,
+      "step": 795
+    },
+    {
+      "epoch": 0.3776091081593928,
+      "grad_norm": 3.1293842792510986,
+      "learning_rate": 1.8707872886158806e-05,
+      "loss": 0.4708,
+      "step": 796
+    },
+    {
+      "epoch": 0.3780834914611006,
+      "grad_norm": 2.830396890640259,
+      "learning_rate": 1.870409287633726e-05,
+      "loss": 0.5119,
+      "step": 797
+    },
+    {
+      "epoch": 0.37855787476280833,
+      "grad_norm": 2.5124335289001465,
+      "learning_rate": 1.8700307728575813e-05,
+      "loss": 0.3997,
+      "step": 798
+    },
+    {
+      "epoch": 0.3790322580645161,
+      "grad_norm": 2.8015406131744385,
+      "learning_rate": 1.8696517445108807e-05,
+      "loss": 0.4406,
+      "step": 799
+    },
+    {
+      "epoch": 0.3795066413662239,
+      "grad_norm": 2.6684601306915283,
+      "learning_rate": 1.8692722028173612e-05,
+      "loss": 0.4578,
+      "step": 800
+    },
+    {
+      "epoch": 0.3799810246679317,
+      "grad_norm": 2.6726903915405273,
+      "learning_rate": 1.868892148001062e-05,
+      "loss": 0.3832,
+      "step": 801
+    },
+    {
+      "epoch": 0.38045540796963945,
+      "grad_norm": 2.7872049808502197,
+      "learning_rate": 1.868511580286326e-05,
+      "loss": 0.4706,
+      "step": 802
+    },
+    {
+      "epoch": 0.38092979127134724,
+      "grad_norm": 2.7430388927459717,
+      "learning_rate": 1.8681304998977988e-05,
+      "loss": 0.4383,
+      "step": 803
+    },
+    {
+      "epoch": 0.38140417457305503,
+      "grad_norm": 2.5697414875030518,
+      "learning_rate": 1.8677489070604274e-05,
+      "loss": 0.4179,
+      "step": 804
+    },
+    {
+      "epoch": 0.3818785578747628,
+      "grad_norm": 2.9383411407470703,
+      "learning_rate": 1.8673668019994632e-05,
+      "loss": 0.4117,
+      "step": 805
+    },
+    {
+      "epoch": 0.38235294117647056,
+      "grad_norm": 3.752838373184204,
+      "learning_rate": 1.866984184940459e-05,
+      "loss": 0.5366,
+      "step": 806
+    },
+    {
+      "epoch": 0.38282732447817835,
+      "grad_norm": 4.9944024085998535,
+      "learning_rate": 1.866601056109269e-05,
+      "loss": 0.5161,
+      "step": 807
+    },
+    {
+      "epoch": 0.38330170777988615,
+      "grad_norm": 2.624593734741211,
+      "learning_rate": 1.8662174157320515e-05,
+      "loss": 0.3766,
+      "step": 808
+    },
+    {
+      "epoch": 0.38377609108159394,
+      "grad_norm": 2.987598419189453,
+      "learning_rate": 1.8658332640352647e-05,
+      "loss": 0.5166,
+      "step": 809
+    },
+    {
+      "epoch": 0.38425047438330173,
+      "grad_norm": 2.785848379135132,
+      "learning_rate": 1.8654486012456704e-05,
+      "loss": 0.4269,
+      "step": 810
+    },
+    {
+      "epoch": 0.38472485768500947,
+      "grad_norm": 2.538492202758789,
+      "learning_rate": 1.8650634275903304e-05,
+      "loss": 0.4338,
+      "step": 811
+    },
+    {
+      "epoch": 0.38519924098671726,
+      "grad_norm": 2.889648914337158,
+      "learning_rate": 1.864677743296609e-05,
+      "loss": 0.4407,
+      "step": 812
+    },
+    {
+      "epoch": 0.38567362428842505,
+      "grad_norm": 2.803192138671875,
+      "learning_rate": 1.8642915485921726e-05,
+      "loss": 0.4189,
+      "step": 813
+    },
+    {
+      "epoch": 0.38614800759013285,
+      "grad_norm": 2.4680614471435547,
+      "learning_rate": 1.8639048437049875e-05,
+      "loss": 0.3655,
+      "step": 814
+    },
+    {
+      "epoch": 0.3866223908918406,
+      "grad_norm": 3.2701048851013184,
+      "learning_rate": 1.8635176288633218e-05,
+      "loss": 0.4297,
+      "step": 815
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 3.7797579765319824,
+      "learning_rate": 1.8631299042957448e-05,
+      "loss": 0.5702,
+      "step": 816
+    },
+    {
+      "epoch": 0.38757115749525617,
+      "grad_norm": 3.297485113143921,
+      "learning_rate": 1.862741670231126e-05,
+      "loss": 0.5324,
+      "step": 817
+    },
+    {
+      "epoch": 0.38804554079696396,
+      "grad_norm": 2.463757276535034,
+      "learning_rate": 1.8623529268986366e-05,
+      "loss": 0.3748,
+      "step": 818
+    },
+    {
+      "epoch": 0.38851992409867175,
+      "grad_norm": 2.859805107116699,
+      "learning_rate": 1.861963674527748e-05,
+      "loss": 0.5214,
+      "step": 819
+    },
+    {
+      "epoch": 0.3889943074003795,
+      "grad_norm": 3.135437488555908,
+      "learning_rate": 1.8615739133482315e-05,
+      "loss": 0.4381,
+      "step": 820
+    },
+    {
+      "epoch": 0.3894686907020873,
+      "grad_norm": 3.0858075618743896,
+      "learning_rate": 1.8611836435901595e-05,
+      "loss": 0.4133,
+      "step": 821
+    },
+    {
+      "epoch": 0.3899430740037951,
+      "grad_norm": 2.585782527923584,
+      "learning_rate": 1.860792865483905e-05,
+      "loss": 0.3963,
+      "step": 822
+    },
+    {
+      "epoch": 0.39041745730550287,
+      "grad_norm": 3.3522262573242188,
+      "learning_rate": 1.8604015792601395e-05,
+      "loss": 0.478,
+      "step": 823
+    },
+    {
+      "epoch": 0.3908918406072106,
+      "grad_norm": 2.940312147140503,
+      "learning_rate": 1.860009785149836e-05,
+      "loss": 0.5005,
+      "step": 824
+    },
+    {
+      "epoch": 0.3913662239089184,
+      "grad_norm": 3.0102758407592773,
+      "learning_rate": 1.8596174833842664e-05,
+      "loss": 0.4627,
+      "step": 825
+    },
+    {
+      "epoch": 0.3918406072106262,
+      "grad_norm": 2.446089744567871,
+      "learning_rate": 1.8592246741950027e-05,
+      "loss": 0.3507,
+      "step": 826
+    },
+    {
+      "epoch": 0.392314990512334,
+      "grad_norm": 2.569380283355713,
+      "learning_rate": 1.858831357813916e-05,
+      "loss": 0.3988,
+      "step": 827
+    },
+    {
+      "epoch": 0.3927893738140417,
+      "grad_norm": 3.037994861602783,
+      "learning_rate": 1.8584375344731777e-05,
+      "loss": 0.3629,
+      "step": 828
+    },
+    {
+      "epoch": 0.3932637571157495,
+      "grad_norm": 3.2176268100738525,
+      "learning_rate": 1.8580432044052567e-05,
+      "loss": 0.4764,
+      "step": 829
+    },
+    {
+      "epoch": 0.3937381404174573,
+      "grad_norm": 2.4146931171417236,
+      "learning_rate": 1.8576483678429234e-05,
+      "loss": 0.3852,
+      "step": 830
+    },
+    {
+      "epoch": 0.3942125237191651,
+      "grad_norm": 3.2689833641052246,
+      "learning_rate": 1.8572530250192453e-05,
+      "loss": 0.444,
+      "step": 831
+    },
+    {
+      "epoch": 0.3946869070208729,
+      "grad_norm": 3.194239377975464,
+      "learning_rate": 1.8568571761675893e-05,
+      "loss": 0.4141,
+      "step": 832
+    },
+    {
+      "epoch": 0.3951612903225806,
+      "grad_norm": 2.9334359169006348,
+      "learning_rate": 1.8564608215216212e-05,
+      "loss": 0.4693,
+      "step": 833
+    },
+    {
+      "epoch": 0.3956356736242884,
+      "grad_norm": 2.812944173812866,
+      "learning_rate": 1.8560639613153056e-05,
+      "loss": 0.436,
+      "step": 834
+    },
+    {
+      "epoch": 0.3961100569259962,
+      "grad_norm": 2.986809730529785,
+      "learning_rate": 1.855666595782904e-05,
+      "loss": 0.4789,
+      "step": 835
+    },
+    {
+      "epoch": 0.396584440227704,
+      "grad_norm": 2.4972949028015137,
+      "learning_rate": 1.8552687251589786e-05,
+      "loss": 0.4695,
+      "step": 836
+    },
+    {
+      "epoch": 0.39705882352941174,
+      "grad_norm": 2.675363779067993,
+      "learning_rate": 1.8548703496783877e-05,
+      "loss": 0.4012,
+      "step": 837
+    },
+    {
+      "epoch": 0.39753320683111953,
+      "grad_norm": 2.956235885620117,
+      "learning_rate": 1.854471469576289e-05,
+      "loss": 0.503,
+      "step": 838
+    },
+    {
+      "epoch": 0.3980075901328273,
+      "grad_norm": 2.4831976890563965,
+      "learning_rate": 1.8540720850881372e-05,
+      "loss": 0.3945,
+      "step": 839
+    },
+    {
+      "epoch": 0.3984819734345351,
+      "grad_norm": 2.747809886932373,
+      "learning_rate": 1.8536721964496846e-05,
+      "loss": 0.431,
+      "step": 840
+    },
+    {
+      "epoch": 0.3989563567362429,
+      "grad_norm": 3.124937057495117,
+      "learning_rate": 1.853271803896982e-05,
+      "loss": 0.505,
+      "step": 841
+    },
+    {
+      "epoch": 0.39943074003795065,
+      "grad_norm": 3.363137722015381,
+      "learning_rate": 1.8528709076663772e-05,
+      "loss": 0.4563,
+      "step": 842
+    },
+    {
+      "epoch": 0.39990512333965844,
+      "grad_norm": 2.7867400646209717,
+      "learning_rate": 1.8524695079945154e-05,
+      "loss": 0.4487,
+      "step": 843
+    },
+    {
+      "epoch": 0.40037950664136623,
+      "grad_norm": 2.9408624172210693,
+      "learning_rate": 1.8520676051183377e-05,
+      "loss": 0.4014,
+      "step": 844
+    },
+    {
+      "epoch": 0.400853889943074,
+      "grad_norm": 3.224708318710327,
+      "learning_rate": 1.851665199275085e-05,
+      "loss": 0.4478,
+      "step": 845
+    },
+    {
+      "epoch": 0.40132827324478176,
+      "grad_norm": 2.7404086589813232,
+      "learning_rate": 1.8512622907022924e-05,
+      "loss": 0.4689,
+      "step": 846
+    },
+    {
+      "epoch": 0.40180265654648956,
+      "grad_norm": 2.713667154312134,
+      "learning_rate": 1.8508588796377936e-05,
+      "loss": 0.4486,
+      "step": 847
+    },
+    {
+      "epoch": 0.40227703984819735,
+      "grad_norm": 3.0665454864501953,
+      "learning_rate": 1.8504549663197175e-05,
+      "loss": 0.4629,
+      "step": 848
+    },
+    {
+      "epoch": 0.40275142314990514,
+      "grad_norm": 3.2429921627044678,
+      "learning_rate": 1.8500505509864903e-05,
+      "loss": 0.4945,
+      "step": 849
+    },
+    {
+      "epoch": 0.4032258064516129,
+      "grad_norm": 3.3584725856781006,
+      "learning_rate": 1.8496456338768345e-05,
+      "loss": 0.4229,
+      "step": 850
+    },
+    {
+      "epoch": 0.40370018975332067,
+      "grad_norm": 2.912458896636963,
+      "learning_rate": 1.8492402152297688e-05,
+      "loss": 0.4418,
+      "step": 851
+    },
+    {
+      "epoch": 0.40417457305502846,
+      "grad_norm": 2.935781478881836,
+      "learning_rate": 1.8488342952846074e-05,
+      "loss": 0.3275,
+      "step": 852
+    },
+    {
+      "epoch": 0.40464895635673626,
+      "grad_norm": 3.8504891395568848,
+      "learning_rate": 1.848427874280961e-05,
+      "loss": 0.413,
+      "step": 853
+    },
+    {
+      "epoch": 0.40512333965844405,
+      "grad_norm": 2.4456560611724854,
+      "learning_rate": 1.8480209524587363e-05,
+      "loss": 0.419,
+      "step": 854
+    },
+    {
+      "epoch": 0.4055977229601518,
+      "grad_norm": 3.666505813598633,
+      "learning_rate": 1.8476135300581347e-05,
+      "loss": 0.4197,
+      "step": 855
+    },
+    {
+      "epoch": 0.4060721062618596,
+      "grad_norm": 2.4932539463043213,
+      "learning_rate": 1.847205607319654e-05,
+      "loss": 0.3565,
+      "step": 856
+    },
+    {
+      "epoch": 0.40654648956356737,
+      "grad_norm": 2.7682738304138184,
+      "learning_rate": 1.8467971844840864e-05,
+      "loss": 0.4519,
+      "step": 857
+    },
+    {
+      "epoch": 0.40702087286527516,
+      "grad_norm": 2.8969831466674805,
+      "learning_rate": 1.8463882617925208e-05,
+      "loss": 0.4616,
+      "step": 858
+    },
+    {
+      "epoch": 0.4074952561669829,
+      "grad_norm": 3.2404887676239014,
+      "learning_rate": 1.8459788394863388e-05,
+      "loss": 0.5208,
+      "step": 859
+    },
+    {
+      "epoch": 0.4079696394686907,
+      "grad_norm": 2.699791669845581,
+      "learning_rate": 1.8455689178072197e-05,
+      "loss": 0.438,
+      "step": 860
+    },
+    {
+      "epoch": 0.4084440227703985,
+      "grad_norm": 3.718968152999878,
+      "learning_rate": 1.8451584969971358e-05,
+      "loss": 0.5273,
+      "step": 861
+    },
+    {
+      "epoch": 0.4089184060721063,
+      "grad_norm": 2.4565632343292236,
+      "learning_rate": 1.8447475772983542e-05,
+      "loss": 0.3949,
+      "step": 862
+    },
+    {
+      "epoch": 0.409392789373814,
+      "grad_norm": 2.7938921451568604,
+      "learning_rate": 1.8443361589534366e-05,
+      "loss": 0.3812,
+      "step": 863
+    },
+    {
+      "epoch": 0.4098671726755218,
+      "grad_norm": 2.8506662845611572,
+      "learning_rate": 1.84392424220524e-05,
+      "loss": 0.4237,
+      "step": 864
+    },
+    {
+      "epoch": 0.4103415559772296,
+      "grad_norm": 2.958847999572754,
+      "learning_rate": 1.8435118272969135e-05,
+      "loss": 0.5132,
+      "step": 865
+    },
+    {
+      "epoch": 0.4108159392789374,
+      "grad_norm": 3.2671656608581543,
+      "learning_rate": 1.8430989144719028e-05,
+      "loss": 0.4776,
+      "step": 866
+    },
+    {
+      "epoch": 0.4112903225806452,
+      "grad_norm": 2.673391103744507,
+      "learning_rate": 1.8426855039739454e-05,
+      "loss": 0.4367,
+      "step": 867
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 2.532644271850586,
+      "learning_rate": 1.8422715960470737e-05,
+      "loss": 0.4003,
+      "step": 868
+    },
+    {
+      "epoch": 0.4122390891840607,
+      "grad_norm": 2.612393856048584,
+      "learning_rate": 1.8418571909356138e-05,
+      "loss": 0.4027,
+      "step": 869
+    },
+    {
+      "epoch": 0.4127134724857685,
+      "grad_norm": 2.484384775161743,
+      "learning_rate": 1.8414422888841844e-05,
+      "loss": 0.4135,
+      "step": 870
+    },
+    {
+      "epoch": 0.4131878557874763,
+      "grad_norm": 3.7062361240386963,
+      "learning_rate": 1.8410268901376983e-05,
+      "loss": 0.4552,
+      "step": 871
+    },
+    {
+      "epoch": 0.41366223908918404,
+      "grad_norm": 2.7248425483703613,
+      "learning_rate": 1.8406109949413614e-05,
+      "loss": 0.4411,
+      "step": 872
+    },
+    {
+      "epoch": 0.41413662239089183,
+      "grad_norm": 2.9587390422821045,
+      "learning_rate": 1.8401946035406723e-05,
+      "loss": 0.5239,
+      "step": 873
+    },
+    {
+      "epoch": 0.4146110056925996,
+      "grad_norm": 3.329129457473755,
+      "learning_rate": 1.839777716181423e-05,
+      "loss": 0.4036,
+      "step": 874
+    },
+    {
+      "epoch": 0.4150853889943074,
+      "grad_norm": 2.5802624225616455,
+      "learning_rate": 1.8393603331096974e-05,
+      "loss": 0.4146,
+      "step": 875
+    },
+    {
+      "epoch": 0.4155597722960152,
+      "grad_norm": 3.270162582397461,
+      "learning_rate": 1.8389424545718733e-05,
+      "loss": 0.4765,
+      "step": 876
+    },
+    {
+      "epoch": 0.41603415559772294,
+      "grad_norm": 2.688945770263672,
+      "learning_rate": 1.8385240808146197e-05,
+      "loss": 0.4338,
+      "step": 877
+    },
+    {
+      "epoch": 0.41650853889943074,
+      "grad_norm": 2.7994351387023926,
+      "learning_rate": 1.838105212084899e-05,
+      "loss": 0.4464,
+      "step": 878
+    },
+    {
+      "epoch": 0.41698292220113853,
+      "grad_norm": 2.3464057445526123,
+      "learning_rate": 1.837685848629965e-05,
+      "loss": 0.3783,
+      "step": 879
+    },
+    {
+      "epoch": 0.4174573055028463,
+      "grad_norm": 3.1224400997161865,
+      "learning_rate": 1.837265990697364e-05,
+      "loss": 0.5672,
+      "step": 880
+    },
+    {
+      "epoch": 0.41793168880455406,
+      "grad_norm": 3.112441301345825,
+      "learning_rate": 1.8368456385349333e-05,
+      "loss": 0.4204,
+      "step": 881
+    },
+    {
+      "epoch": 0.41840607210626185,
+      "grad_norm": 3.453056812286377,
+      "learning_rate": 1.8364247923908033e-05,
+      "loss": 0.4365,
+      "step": 882
+    },
+    {
+      "epoch": 0.41888045540796964,
+      "grad_norm": 3.2397301197052,
+      "learning_rate": 1.8360034525133953e-05,
+      "loss": 0.4454,
+      "step": 883
+    },
+    {
+      "epoch": 0.41935483870967744,
+      "grad_norm": 2.897355794906616,
+      "learning_rate": 1.8355816191514216e-05,
+      "loss": 0.4647,
+      "step": 884
+    },
+    {
+      "epoch": 0.4198292220113852,
+      "grad_norm": 3.6154603958129883,
+      "learning_rate": 1.8351592925538865e-05,
+      "loss": 0.6002,
+      "step": 885
+    },
+    {
+      "epoch": 0.42030360531309297,
+      "grad_norm": 2.3313372135162354,
+      "learning_rate": 1.8347364729700857e-05,
+      "loss": 0.4218,
+      "step": 886
+    },
+    {
+      "epoch": 0.42077798861480076,
+      "grad_norm": 2.8984670639038086,
+      "learning_rate": 1.8343131606496046e-05,
+      "loss": 0.4025,
+      "step": 887
+    },
+    {
+      "epoch": 0.42125237191650855,
+      "grad_norm": 3.124525785446167,
+      "learning_rate": 1.8338893558423207e-05,
+      "loss": 0.441,
+      "step": 888
+    },
+    {
+      "epoch": 0.42172675521821634,
+      "grad_norm": 2.574047565460205,
+      "learning_rate": 1.833465058798402e-05,
+      "loss": 0.4107,
+      "step": 889
+    },
+    {
+      "epoch": 0.4222011385199241,
+      "grad_norm": 2.7786552906036377,
+      "learning_rate": 1.8330402697683067e-05,
+      "loss": 0.4836,
+      "step": 890
+    },
+    {
+      "epoch": 0.42267552182163187,
+      "grad_norm": 2.599820137023926,
+      "learning_rate": 1.832614989002783e-05,
+      "loss": 0.4446,
+      "step": 891
+    },
+    {
+      "epoch": 0.42314990512333966,
+      "grad_norm": 2.800710678100586,
+      "learning_rate": 1.8321892167528707e-05,
+      "loss": 0.4315,
+      "step": 892
+    },
+    {
+      "epoch": 0.42362428842504746,
+      "grad_norm": 2.935514450073242,
+      "learning_rate": 1.831762953269898e-05,
+      "loss": 0.4062,
+      "step": 893
+    },
+    {
+      "epoch": 0.4240986717267552,
+      "grad_norm": 2.597562551498413,
+      "learning_rate": 1.8313361988054853e-05,
+      "loss": 0.3958,
+      "step": 894
+    },
+    {
+      "epoch": 0.424573055028463,
+      "grad_norm": 2.7025997638702393,
+      "learning_rate": 1.8309089536115406e-05,
+      "loss": 0.4288,
+      "step": 895
+    },
+    {
+      "epoch": 0.4250474383301708,
+      "grad_norm": 2.58357310295105,
+      "learning_rate": 1.8304812179402626e-05,
+      "loss": 0.4734,
+      "step": 896
+    },
+    {
+      "epoch": 0.42552182163187857,
+      "grad_norm": 2.8267996311187744,
+      "learning_rate": 1.83005299204414e-05,
+      "loss": 0.4395,
+      "step": 897
+    },
+    {
+      "epoch": 0.42599620493358636,
+      "grad_norm": 2.934715986251831,
+      "learning_rate": 1.82962427617595e-05,
+      "loss": 0.4381,
+      "step": 898
+    },
+    {
+      "epoch": 0.4264705882352941,
+      "grad_norm": 2.855116128921509,
+      "learning_rate": 1.829195070588759e-05,
+      "loss": 0.4122,
+      "step": 899
+    },
+    {
+      "epoch": 0.4269449715370019,
+      "grad_norm": 2.7135651111602783,
+      "learning_rate": 1.8287653755359228e-05,
+      "loss": 0.4812,
+      "step": 900
+    },
+    {
+      "epoch": 0.4274193548387097,
+      "grad_norm": 2.9477922916412354,
+      "learning_rate": 1.8283351912710867e-05,
+      "loss": 0.4166,
+      "step": 901
+    },
+    {
+      "epoch": 0.4278937381404175,
+      "grad_norm": 3.2937138080596924,
+      "learning_rate": 1.827904518048184e-05,
+      "loss": 0.4541,
+      "step": 902
+    },
+    {
+      "epoch": 0.4283681214421252,
+      "grad_norm": 2.7012522220611572,
+      "learning_rate": 1.8274733561214368e-05,
+      "loss": 0.4235,
+      "step": 903
+    },
+    {
+      "epoch": 0.428842504743833,
+      "grad_norm": 3.3364474773406982,
+      "learning_rate": 1.8270417057453554e-05,
+      "loss": 0.4433,
+      "step": 904
+    },
+    {
+      "epoch": 0.4293168880455408,
+      "grad_norm": 2.75413179397583,
+      "learning_rate": 1.826609567174739e-05,
+      "loss": 0.4105,
+      "step": 905
+    },
+    {
+      "epoch": 0.4297912713472486,
+      "grad_norm": 2.27841854095459,
+      "learning_rate": 1.826176940664675e-05,
+      "loss": 0.3656,
+      "step": 906
+    },
+    {
+      "epoch": 0.43026565464895633,
+      "grad_norm": 3.102612018585205,
+      "learning_rate": 1.8257438264705382e-05,
+      "loss": 0.4927,
+      "step": 907
+    },
+    {
+      "epoch": 0.4307400379506641,
+      "grad_norm": 2.6770787239074707,
+      "learning_rate": 1.825310224847992e-05,
+      "loss": 0.4812,
+      "step": 908
+    },
+    {
+      "epoch": 0.4312144212523719,
+      "grad_norm": 3.230632781982422,
+      "learning_rate": 1.8248761360529864e-05,
+      "loss": 0.4326,
+      "step": 909
+    },
+    {
+      "epoch": 0.4316888045540797,
+      "grad_norm": 3.148620843887329,
+      "learning_rate": 1.8244415603417603e-05,
+      "loss": 0.4355,
+      "step": 910
+    },
+    {
+      "epoch": 0.4321631878557875,
+      "grad_norm": 3.1337029933929443,
+      "learning_rate": 1.8240064979708397e-05,
+      "loss": 0.5506,
+      "step": 911
+    },
+    {
+      "epoch": 0.43263757115749524,
+      "grad_norm": 2.7768473625183105,
+      "learning_rate": 1.8235709491970366e-05,
+      "loss": 0.4683,
+      "step": 912
+    },
+    {
+      "epoch": 0.43311195445920303,
+      "grad_norm": 2.6969683170318604,
+      "learning_rate": 1.8231349142774525e-05,
+      "loss": 0.4588,
+      "step": 913
+    },
+    {
+      "epoch": 0.4335863377609108,
+      "grad_norm": 2.7988765239715576,
+      "learning_rate": 1.8226983934694732e-05,
+      "loss": 0.399,
+      "step": 914
+    },
+    {
+      "epoch": 0.4340607210626186,
+      "grad_norm": 3.1512439250946045,
+      "learning_rate": 1.8222613870307735e-05,
+      "loss": 0.5061,
+      "step": 915
+    },
+    {
+      "epoch": 0.43453510436432635,
+      "grad_norm": 2.885368824005127,
+      "learning_rate": 1.8218238952193136e-05,
+      "loss": 0.4037,
+      "step": 916
+    },
+    {
+      "epoch": 0.43500948766603414,
+      "grad_norm": 2.4949498176574707,
+      "learning_rate": 1.8213859182933407e-05,
+      "loss": 0.3942,
+      "step": 917
+    },
+    {
+      "epoch": 0.43548387096774194,
+      "grad_norm": 2.3052561283111572,
+      "learning_rate": 1.820947456511388e-05,
+      "loss": 0.4208,
+      "step": 918
+    },
+    {
+      "epoch": 0.43595825426944973,
+      "grad_norm": 2.642266273498535,
+      "learning_rate": 1.8205085101322754e-05,
+      "loss": 0.465,
+      "step": 919
+    },
+    {
+      "epoch": 0.4364326375711575,
+      "grad_norm": 3.87248158454895,
+      "learning_rate": 1.8200690794151087e-05,
+      "loss": 0.5536,
+      "step": 920
+    },
+    {
+      "epoch": 0.43690702087286526,
+      "grad_norm": 2.5663421154022217,
+      "learning_rate": 1.819629164619279e-05,
+      "loss": 0.4441,
+      "step": 921
+    },
+    {
+      "epoch": 0.43738140417457305,
+      "grad_norm": 2.330423593521118,
+      "learning_rate": 1.8191887660044646e-05,
+      "loss": 0.36,
+      "step": 922
+    },
+    {
+      "epoch": 0.43785578747628084,
+      "grad_norm": 2.401970386505127,
+      "learning_rate": 1.8187478838306273e-05,
+      "loss": 0.3381,
+      "step": 923
+    },
+    {
+      "epoch": 0.43833017077798864,
+      "grad_norm": 3.270221710205078,
+      "learning_rate": 1.818306518358016e-05,
+      "loss": 0.5003,
+      "step": 924
+    },
+    {
+      "epoch": 0.4388045540796964,
+      "grad_norm": 2.4524645805358887,
+      "learning_rate": 1.817864669847165e-05,
+      "loss": 0.4074,
+      "step": 925
+    },
+    {
+      "epoch": 0.43927893738140417,
+      "grad_norm": 3.223022222518921,
+      "learning_rate": 1.817422338558892e-05,
+      "loss": 0.5338,
+      "step": 926
+    },
+    {
+      "epoch": 0.43975332068311196,
+      "grad_norm": 2.6757102012634277,
+      "learning_rate": 1.8169795247543014e-05,
+      "loss": 0.4344,
+      "step": 927
+    },
+    {
+      "epoch": 0.44022770398481975,
+      "grad_norm": 3.086233139038086,
+      "learning_rate": 1.8165362286947817e-05,
+      "loss": 0.4607,
+      "step": 928
+    },
+    {
+      "epoch": 0.4407020872865275,
+      "grad_norm": 2.754612922668457,
+      "learning_rate": 1.8160924506420064e-05,
+      "loss": 0.3898,
+      "step": 929
+    },
+    {
+      "epoch": 0.4411764705882353,
+      "grad_norm": 4.099813461303711,
+      "learning_rate": 1.8156481908579326e-05,
+      "loss": 0.4333,
+      "step": 930
+    },
+    {
+      "epoch": 0.4416508538899431,
+      "grad_norm": 2.578651189804077,
+      "learning_rate": 1.815203449604803e-05,
+      "loss": 0.3853,
+      "step": 931
+    },
+    {
+      "epoch": 0.44212523719165087,
+      "grad_norm": 2.7206499576568604,
+      "learning_rate": 1.8147582271451443e-05,
+      "loss": 0.4229,
+      "step": 932
+    },
+    {
+      "epoch": 0.44259962049335866,
+      "grad_norm": 3.497983694076538,
+      "learning_rate": 1.814312523741766e-05,
+      "loss": 0.5558,
+      "step": 933
+    },
+    {
+      "epoch": 0.4430740037950664,
+      "grad_norm": 3.077042579650879,
+      "learning_rate": 1.8138663396577633e-05,
+      "loss": 0.4388,
+      "step": 934
+    },
+    {
+      "epoch": 0.4435483870967742,
+      "grad_norm": 2.824192523956299,
+      "learning_rate": 1.813419675156514e-05,
+      "loss": 0.4736,
+      "step": 935
+    },
+    {
+      "epoch": 0.444022770398482,
+      "grad_norm": 2.8267297744750977,
+      "learning_rate": 1.8129725305016793e-05,
+      "loss": 0.4546,
+      "step": 936
+    },
+    {
+      "epoch": 0.4444971537001898,
+      "grad_norm": 2.9733405113220215,
+      "learning_rate": 1.8125249059572042e-05,
+      "loss": 0.5409,
+      "step": 937
+    },
+    {
+      "epoch": 0.4449715370018975,
+      "grad_norm": 3.0438060760498047,
+      "learning_rate": 1.8120768017873178e-05,
+      "loss": 0.4126,
+      "step": 938
+    },
+    {
+      "epoch": 0.4454459203036053,
+      "grad_norm": 2.6118757724761963,
+      "learning_rate": 1.8116282182565313e-05,
+      "loss": 0.3854,
+      "step": 939
+    },
+    {
+      "epoch": 0.4459203036053131,
+      "grad_norm": 2.909747838973999,
+      "learning_rate": 1.8111791556296386e-05,
+      "loss": 0.5156,
+      "step": 940
+    },
+    {
+      "epoch": 0.4463946869070209,
+      "grad_norm": 2.6982431411743164,
+      "learning_rate": 1.8107296141717175e-05,
+      "loss": 0.3597,
+      "step": 941
+    },
+    {
+      "epoch": 0.4468690702087287,
+      "grad_norm": 3.628803253173828,
+      "learning_rate": 1.8102795941481277e-05,
+      "loss": 0.5069,
+      "step": 942
+    },
+    {
+      "epoch": 0.4473434535104364,
+      "grad_norm": 2.5725972652435303,
+      "learning_rate": 1.8098290958245116e-05,
+      "loss": 0.3675,
+      "step": 943
+    },
+    {
+      "epoch": 0.4478178368121442,
+      "grad_norm": 3.47405743598938,
+      "learning_rate": 1.8093781194667935e-05,
+      "loss": 0.4533,
+      "step": 944
+    },
+    {
+      "epoch": 0.448292220113852,
+      "grad_norm": 2.7190101146698,
+      "learning_rate": 1.808926665341181e-05,
+      "loss": 0.3759,
+      "step": 945
+    },
+    {
+      "epoch": 0.4487666034155598,
+      "grad_norm": 4.446865558624268,
+      "learning_rate": 1.8084747337141622e-05,
+      "loss": 0.3144,
+      "step": 946
+    },
+    {
+      "epoch": 0.44924098671726753,
+      "grad_norm": 2.6146459579467773,
+      "learning_rate": 1.8080223248525087e-05,
+      "loss": 0.4736,
+      "step": 947
+    },
+    {
+      "epoch": 0.4497153700189753,
+      "grad_norm": 3.1823534965515137,
+      "learning_rate": 1.8075694390232725e-05,
+      "loss": 0.5324,
+      "step": 948
+    },
+    {
+      "epoch": 0.4501897533206831,
+      "grad_norm": 3.1742212772369385,
+      "learning_rate": 1.8071160764937875e-05,
+      "loss": 0.4744,
+      "step": 949
+    },
+    {
+      "epoch": 0.4506641366223909,
+      "grad_norm": 2.7472152709960938,
+      "learning_rate": 1.8066622375316695e-05,
+      "loss": 0.3879,
+      "step": 950
+    },
+    {
+      "epoch": 0.45113851992409865,
+      "grad_norm": 2.5368971824645996,
+      "learning_rate": 1.8062079224048146e-05,
+      "loss": 0.4705,
+      "step": 951
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 2.812864065170288,
+      "learning_rate": 1.805753131381401e-05,
+      "loss": 0.4219,
+      "step": 952
+    },
+    {
+      "epoch": 0.45208728652751423,
+      "grad_norm": 2.2089133262634277,
+      "learning_rate": 1.8052978647298873e-05,
+      "loss": 0.387,
+      "step": 953
+    },
+    {
+      "epoch": 0.452561669829222,
+      "grad_norm": 2.8962409496307373,
+      "learning_rate": 1.804842122719013e-05,
+      "loss": 0.5197,
+      "step": 954
+    },
+    {
+      "epoch": 0.4530360531309298,
+      "grad_norm": 2.630946397781372,
+      "learning_rate": 1.8043859056177976e-05,
+      "loss": 0.3808,
+      "step": 955
+    },
+    {
+      "epoch": 0.45351043643263755,
+      "grad_norm": 2.4841771125793457,
+      "learning_rate": 1.803929213695542e-05,
+      "loss": 0.4099,
+      "step": 956
+    },
+    {
+      "epoch": 0.45398481973434535,
+      "grad_norm": 2.3919498920440674,
+      "learning_rate": 1.803472047221827e-05,
+      "loss": 0.3677,
+      "step": 957
+    },
+    {
+      "epoch": 0.45445920303605314,
+      "grad_norm": 2.786252498626709,
+      "learning_rate": 1.8030144064665127e-05,
+      "loss": 0.4811,
+      "step": 958
+    },
+    {
+      "epoch": 0.45493358633776093,
+      "grad_norm": 2.7596981525421143,
+      "learning_rate": 1.802556291699741e-05,
+      "loss": 0.4529,
+      "step": 959
+    },
+    {
+      "epoch": 0.45540796963946867,
+      "grad_norm": 2.7790467739105225,
+      "learning_rate": 1.8020977031919315e-05,
+      "loss": 0.4452,
+      "step": 960
+    },
+    {
+      "epoch": 0.45588235294117646,
+      "grad_norm": 3.088778018951416,
+      "learning_rate": 1.801638641213785e-05,
+      "loss": 0.4797,
+      "step": 961
+    },
+    {
+      "epoch": 0.45635673624288425,
+      "grad_norm": 2.6321218013763428,
+      "learning_rate": 1.801179106036281e-05,
+      "loss": 0.4205,
+      "step": 962
+    },
+    {
+      "epoch": 0.45683111954459205,
+      "grad_norm": 2.7974629402160645,
+      "learning_rate": 1.8007190979306793e-05,
+      "loss": 0.4557,
+      "step": 963
+    },
+    {
+      "epoch": 0.4573055028462998,
+      "grad_norm": 2.396785259246826,
+      "learning_rate": 1.800258617168517e-05,
+      "loss": 0.4265,
+      "step": 964
+    },
+    {
+      "epoch": 0.4577798861480076,
+      "grad_norm": 2.8968255519866943,
+      "learning_rate": 1.799797664021612e-05,
+      "loss": 0.3469,
+      "step": 965
+    },
+    {
+      "epoch": 0.45825426944971537,
+      "grad_norm": 2.8797080516815186,
+      "learning_rate": 1.7993362387620602e-05,
+      "loss": 0.4216,
+      "step": 966
+    },
+    {
+      "epoch": 0.45872865275142316,
+      "grad_norm": 3.1519381999969482,
+      "learning_rate": 1.798874341662237e-05,
+      "loss": 0.5129,
+      "step": 967
+    },
+    {
+      "epoch": 0.45920303605313095,
+      "grad_norm": 2.6369104385375977,
+      "learning_rate": 1.7984119729947944e-05,
+      "loss": 0.4251,
+      "step": 968
+    },
+    {
+      "epoch": 0.4596774193548387,
+      "grad_norm": 2.4390084743499756,
+      "learning_rate": 1.797949133032665e-05,
+      "loss": 0.4245,
+      "step": 969
+    },
+    {
+      "epoch": 0.4601518026565465,
+      "grad_norm": 2.8855037689208984,
+      "learning_rate": 1.7974858220490586e-05,
+      "loss": 0.4015,
+      "step": 970
+    },
+    {
+      "epoch": 0.4606261859582543,
+      "grad_norm": 2.9386088848114014,
+      "learning_rate": 1.7970220403174626e-05,
+      "loss": 0.4222,
+      "step": 971
+    },
+    {
+      "epoch": 0.46110056925996207,
+      "grad_norm": 2.18912410736084,
+      "learning_rate": 1.796557788111643e-05,
+      "loss": 0.3247,
+      "step": 972
+    },
+    {
+      "epoch": 0.4615749525616698,
+      "grad_norm": 3.183178663253784,
+      "learning_rate": 1.796093065705644e-05,
+      "loss": 0.5049,
+      "step": 973
+    },
+    {
+      "epoch": 0.4620493358633776,
+      "grad_norm": 2.8676347732543945,
+      "learning_rate": 1.7956278733737855e-05,
+      "loss": 0.458,
+      "step": 974
+    },
+    {
+      "epoch": 0.4625237191650854,
+      "grad_norm": 5.36901330947876,
+      "learning_rate": 1.7951622113906663e-05,
+      "loss": 0.4597,
+      "step": 975
+    },
+    {
+      "epoch": 0.4629981024667932,
+      "grad_norm": 2.528803586959839,
+      "learning_rate": 1.7946960800311623e-05,
+      "loss": 0.3229,
+      "step": 976
+    },
+    {
+      "epoch": 0.463472485768501,
+      "grad_norm": 2.7239816188812256,
+      "learning_rate": 1.7942294795704265e-05,
+      "loss": 0.4284,
+      "step": 977
+    },
+    {
+      "epoch": 0.4639468690702087,
+      "grad_norm": 2.4139273166656494,
+      "learning_rate": 1.7937624102838878e-05,
+      "loss": 0.3585,
+      "step": 978
+    },
+    {
+      "epoch": 0.4644212523719165,
+      "grad_norm": 2.762474775314331,
+      "learning_rate": 1.793294872447253e-05,
+      "loss": 0.3938,
+      "step": 979
+    },
+    {
+      "epoch": 0.4648956356736243,
+      "grad_norm": 2.676497459411621,
+      "learning_rate": 1.792826866336505e-05,
+      "loss": 0.4188,
+      "step": 980
+    },
+    {
+      "epoch": 0.4653700189753321,
+      "grad_norm": 3.0633938312530518,
+      "learning_rate": 1.792358392227903e-05,
+      "loss": 0.3947,
+      "step": 981
+    },
+    {
+      "epoch": 0.4658444022770398,
+      "grad_norm": 2.3310940265655518,
+      "learning_rate": 1.791889450397983e-05,
+      "loss": 0.3666,
+      "step": 982
+    },
+    {
+      "epoch": 0.4663187855787476,
+      "grad_norm": 3.144381523132324,
+      "learning_rate": 1.7914200411235562e-05,
+      "loss": 0.4274,
+      "step": 983
+    },
+    {
+      "epoch": 0.4667931688804554,
+      "grad_norm": 2.8709566593170166,
+      "learning_rate": 1.7909501646817108e-05,
+      "loss": 0.4486,
+      "step": 984
+    },
+    {
+      "epoch": 0.4672675521821632,
+      "grad_norm": 2.4950621128082275,
+      "learning_rate": 1.79047982134981e-05,
+      "loss": 0.4079,
+      "step": 985
+    },
+    {
+      "epoch": 0.46774193548387094,
+      "grad_norm": 2.85103178024292,
+      "learning_rate": 1.7900090114054925e-05,
+      "loss": 0.3948,
+      "step": 986
+    },
+    {
+      "epoch": 0.46821631878557873,
+      "grad_norm": 2.7091755867004395,
+      "learning_rate": 1.7895377351266737e-05,
+      "loss": 0.4352,
+      "step": 987
+    },
+    {
+      "epoch": 0.4686907020872865,
+      "grad_norm": 2.969698905944824,
+      "learning_rate": 1.7890659927915418e-05,
+      "loss": 0.4333,
+      "step": 988
+    },
+    {
+      "epoch": 0.4691650853889943,
+      "grad_norm": 2.8901190757751465,
+      "learning_rate": 1.7885937846785633e-05,
+      "loss": 0.4314,
+      "step": 989
+    },
+    {
+      "epoch": 0.4696394686907021,
+      "grad_norm": 2.5206494331359863,
+      "learning_rate": 1.7881211110664767e-05,
+      "loss": 0.4896,
+      "step": 990
+    },
+    {
+      "epoch": 0.47011385199240985,
+      "grad_norm": 2.556368589401245,
+      "learning_rate": 1.7876479722342972e-05,
+      "loss": 0.4165,
+      "step": 991
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 2.626251459121704,
+      "learning_rate": 1.7871743684613137e-05,
+      "loss": 0.5195,
+      "step": 992
+    },
+    {
+      "epoch": 0.47106261859582543,
+      "grad_norm": 2.428295373916626,
+      "learning_rate": 1.78670030002709e-05,
+      "loss": 0.4136,
+      "step": 993
+    },
+    {
+      "epoch": 0.4715370018975332,
+      "grad_norm": 2.7496650218963623,
+      "learning_rate": 1.786225767211464e-05,
+      "loss": 0.4622,
+      "step": 994
+    },
+    {
+      "epoch": 0.47201138519924096,
+      "grad_norm": 2.3996787071228027,
+      "learning_rate": 1.7857507702945472e-05,
+      "loss": 0.3481,
+      "step": 995
+    },
+    {
+      "epoch": 0.47248576850094876,
+      "grad_norm": 2.776946544647217,
+      "learning_rate": 1.7852753095567266e-05,
+      "loss": 0.4187,
+      "step": 996
+    },
+    {
+      "epoch": 0.47296015180265655,
+      "grad_norm": 2.833527088165283,
+      "learning_rate": 1.7847993852786612e-05,
+      "loss": 0.4635,
+      "step": 997
+    },
+    {
+      "epoch": 0.47343453510436434,
+      "grad_norm": 2.589299201965332,
+      "learning_rate": 1.7843229977412844e-05,
+      "loss": 0.4279,
+      "step": 998
+    },
+    {
+      "epoch": 0.47390891840607213,
+      "grad_norm": 2.5228946208953857,
+      "learning_rate": 1.7838461472258035e-05,
+      "loss": 0.4273,
+      "step": 999
+    },
+    {
+      "epoch": 0.47438330170777987,
+      "grad_norm": 2.441437244415283,
+      "learning_rate": 1.7833688340136982e-05,
+      "loss": 0.3756,
+      "step": 1000
+    },
+    {
+      "epoch": 0.47485768500948766,
+      "grad_norm": 3.258878231048584,
+      "learning_rate": 1.782891058386722e-05,
+      "loss": 0.4789,
+      "step": 1001
+    },
+    {
+      "epoch": 0.47533206831119545,
+      "grad_norm": 2.807469367980957,
+      "learning_rate": 1.782412820626901e-05,
+      "loss": 0.4335,
+      "step": 1002
+    },
+    {
+      "epoch": 0.47580645161290325,
+      "grad_norm": 2.546851873397827,
+      "learning_rate": 1.7819341210165347e-05,
+      "loss": 0.3704,
+      "step": 1003
+    },
+    {
+      "epoch": 0.476280834914611,
+      "grad_norm": 2.9630987644195557,
+      "learning_rate": 1.781454959838194e-05,
+      "loss": 0.4367,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4767552182163188,
+      "grad_norm": 2.7907612323760986,
+      "learning_rate": 1.7809753373747235e-05,
+      "loss": 0.4591,
+      "step": 1005
+    },
+    {
+      "epoch": 0.47722960151802657,
+      "grad_norm": 3.0077595710754395,
+      "learning_rate": 1.7804952539092393e-05,
+      "loss": 0.4258,
+      "step": 1006
+    },
+    {
+      "epoch": 0.47770398481973436,
+      "grad_norm": 2.507061243057251,
+      "learning_rate": 1.78001470972513e-05,
+      "loss": 0.4579,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4781783681214421,
+      "grad_norm": 2.3923323154449463,
+      "learning_rate": 1.7795337051060562e-05,
+      "loss": 0.3826,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4786527514231499,
+      "grad_norm": 2.3353378772735596,
+      "learning_rate": 1.77905224033595e-05,
+      "loss": 0.3645,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4791271347248577,
+      "grad_norm": 2.9613916873931885,
+      "learning_rate": 1.7785703156990153e-05,
+      "loss": 0.4481,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4796015180265655,
+      "grad_norm": 3.13313889503479,
+      "learning_rate": 1.7780879314797278e-05,
+      "loss": 0.4703,
+      "step": 1011
+    },
+    {
+      "epoch": 0.48007590132827327,
+      "grad_norm": 3.504654884338379,
+      "learning_rate": 1.7776050879628338e-05,
+      "loss": 0.4535,
+      "step": 1012
+    },
+    {
+      "epoch": 0.480550284629981,
+      "grad_norm": 2.6365973949432373,
+      "learning_rate": 1.777121785433351e-05,
+      "loss": 0.3529,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4810246679316888,
+      "grad_norm": 3.4946820735931396,
+      "learning_rate": 1.7766380241765682e-05,
+      "loss": 0.4124,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4814990512333966,
+      "grad_norm": 2.5579440593719482,
+      "learning_rate": 1.776153804478045e-05,
+      "loss": 0.4045,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4819734345351044,
+      "grad_norm": 2.6704046726226807,
+      "learning_rate": 1.7756691266236114e-05,
+      "loss": 0.4014,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4824478178368121,
+      "grad_norm": 2.637950897216797,
+      "learning_rate": 1.7751839908993677e-05,
+      "loss": 0.4664,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4829222011385199,
+      "grad_norm": 2.9768013954162598,
+      "learning_rate": 1.7746983975916852e-05,
+      "loss": 0.5027,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4833965844402277,
+      "grad_norm": 2.8544719219207764,
+      "learning_rate": 1.7742123469872043e-05,
+      "loss": 0.479,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4838709677419355,
+      "grad_norm": 2.228623151779175,
+      "learning_rate": 1.7737258393728363e-05,
+      "loss": 0.3347,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4843453510436433,
+      "grad_norm": 2.553727865219116,
+      "learning_rate": 1.7732388750357618e-05,
+      "loss": 0.3965,
+      "step": 1021
+    },
+    {
+      "epoch": 0.484819734345351,
+      "grad_norm": 3.6200978755950928,
+      "learning_rate": 1.7727514542634308e-05,
+      "loss": 0.532,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4852941176470588,
+      "grad_norm": 2.9352922439575195,
+      "learning_rate": 1.772263577343563e-05,
+      "loss": 0.428,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4857685009487666,
+      "grad_norm": 2.7596585750579834,
+      "learning_rate": 1.7717752445641473e-05,
+      "loss": 0.4415,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4862428842504744,
+      "grad_norm": 2.538545608520508,
+      "learning_rate": 1.771286456213442e-05,
+      "loss": 0.3995,
+      "step": 1025
+    },
+    {
+      "epoch": 0.48671726755218214,
+      "grad_norm": 2.592538833618164,
+      "learning_rate": 1.7707972125799738e-05,
+      "loss": 0.3906,
+      "step": 1026
+    },
+    {
+      "epoch": 0.48719165085388993,
+      "grad_norm": 2.813682794570923,
+      "learning_rate": 1.770307513952538e-05,
+      "loss": 0.409,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4876660341555977,
+      "grad_norm": 2.8214876651763916,
+      "learning_rate": 1.7698173606201994e-05,
+      "loss": 0.4843,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4881404174573055,
+      "grad_norm": 2.7187728881835938,
+      "learning_rate": 1.7693267528722907e-05,
+      "loss": 0.4048,
+      "step": 1029
+    },
+    {
+      "epoch": 0.48861480075901326,
+      "grad_norm": 3.217090368270874,
+      "learning_rate": 1.7688356909984125e-05,
+      "loss": 0.5723,
+      "step": 1030
+    },
+    {
+      "epoch": 0.48908918406072105,
+      "grad_norm": 2.610750675201416,
+      "learning_rate": 1.7683441752884337e-05,
+      "loss": 0.3835,
+      "step": 1031
+    },
+    {
+      "epoch": 0.48956356736242884,
+      "grad_norm": 2.7182836532592773,
+      "learning_rate": 1.7678522060324918e-05,
+      "loss": 0.4672,
+      "step": 1032
+    },
+    {
+      "epoch": 0.49003795066413663,
+      "grad_norm": 11.401104927062988,
+      "learning_rate": 1.7673597835209904e-05,
+      "loss": 0.3387,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4905123339658444,
+      "grad_norm": 2.622774124145508,
+      "learning_rate": 1.766866908044602e-05,
+      "loss": 0.4282,
+      "step": 1034
+    },
+    {
+      "epoch": 0.49098671726755216,
+      "grad_norm": 2.8863613605499268,
+      "learning_rate": 1.7663735798942666e-05,
+      "loss": 0.4643,
+      "step": 1035
+    },
+    {
+      "epoch": 0.49146110056925996,
+      "grad_norm": 2.4292235374450684,
+      "learning_rate": 1.7658797993611907e-05,
+      "loss": 0.3813,
+      "step": 1036
+    },
+    {
+      "epoch": 0.49193548387096775,
+      "grad_norm": 2.83770751953125,
+      "learning_rate": 1.7653855667368474e-05,
+      "loss": 0.5212,
+      "step": 1037
+    },
+    {
+      "epoch": 0.49240986717267554,
+      "grad_norm": 2.406994581222534,
+      "learning_rate": 1.7648908823129788e-05,
+      "loss": 0.429,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4928842504743833,
+      "grad_norm": 2.8495097160339355,
+      "learning_rate": 1.7643957463815904e-05,
+      "loss": 0.4773,
+      "step": 1039
+    },
+    {
+      "epoch": 0.49335863377609107,
+      "grad_norm": 2.946979522705078,
+      "learning_rate": 1.7639001592349575e-05,
+      "loss": 0.4433,
+      "step": 1040
+    },
+    {
+      "epoch": 0.49383301707779886,
+      "grad_norm": 2.636714220046997,
+      "learning_rate": 1.7634041211656193e-05,
+      "loss": 0.3856,
+      "step": 1041
+    },
+    {
+      "epoch": 0.49430740037950666,
+      "grad_norm": 2.825824737548828,
+      "learning_rate": 1.7629076324663827e-05,
+      "loss": 0.4985,
+      "step": 1042
+    },
+    {
+      "epoch": 0.49478178368121445,
+      "grad_norm": 2.4185476303100586,
+      "learning_rate": 1.7624106934303202e-05,
+      "loss": 0.4149,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4952561669829222,
+      "grad_norm": 2.691755533218384,
+      "learning_rate": 1.7619133043507694e-05,
+      "loss": 0.4321,
+      "step": 1044
+    },
+    {
+      "epoch": 0.49573055028463,
+      "grad_norm": 2.816511392593384,
+      "learning_rate": 1.7614154655213343e-05,
+      "loss": 0.4777,
+      "step": 1045
+    },
+    {
+      "epoch": 0.49620493358633777,
+      "grad_norm": 2.25382924079895,
+      "learning_rate": 1.760917177235885e-05,
+      "loss": 0.4227,
+      "step": 1046
+    },
+    {
+      "epoch": 0.49667931688804556,
+      "grad_norm": 3.144374132156372,
+      "learning_rate": 1.7604184397885554e-05,
+      "loss": 0.4284,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4971537001897533,
+      "grad_norm": 2.518986940383911,
+      "learning_rate": 1.7599192534737453e-05,
+      "loss": 0.4029,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4976280834914611,
+      "grad_norm": 2.7416396141052246,
+      "learning_rate": 1.7594196185861198e-05,
+      "loss": 0.4459,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4981024667931689,
+      "grad_norm": 3.26064395904541,
+      "learning_rate": 1.7589195354206082e-05,
+      "loss": 0.3789,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4985768500948767,
+      "grad_norm": 3.0012292861938477,
+      "learning_rate": 1.7584190042724047e-05,
+      "loss": 0.5016,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4990512333965844,
+      "grad_norm": 3.3961806297302246,
+      "learning_rate": 1.7579180254369674e-05,
+      "loss": 0.4325,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4995256166982922,
+      "grad_norm": 2.4713120460510254,
+      "learning_rate": 1.75741659921002e-05,
+      "loss": 0.3976,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7608566284179688,
+      "learning_rate": 1.756914725887549e-05,
+      "loss": 0.4458,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5004743833017078,
+      "grad_norm": 2.3729867935180664,
+      "learning_rate": 1.7564124057658057e-05,
+      "loss": 0.3763,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5009487666034156,
+      "grad_norm": 2.746999979019165,
+      "learning_rate": 1.755909639141304e-05,
+      "loss": 0.4111,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5014231499051234,
+      "grad_norm": 2.6905953884124756,
+      "learning_rate": 1.755406426310822e-05,
+      "loss": 0.4423,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5018975332068312,
+      "grad_norm": 2.955181837081909,
+      "learning_rate": 1.754902767571402e-05,
+      "loss": 0.33,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5023719165085389,
+      "grad_norm": 2.381403684616089,
+      "learning_rate": 1.754398663220348e-05,
+      "loss": 0.4346,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5028462998102466,
+      "grad_norm": 2.5319292545318604,
+      "learning_rate": 1.7538941135552286e-05,
+      "loss": 0.4469,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5033206831119544,
+      "grad_norm": 2.7511696815490723,
+      "learning_rate": 1.7533891188738738e-05,
+      "loss": 0.4318,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5037950664136622,
+      "grad_norm": 2.6034233570098877,
+      "learning_rate": 1.7528836794743776e-05,
+      "loss": 0.3468,
+      "step": 1062
+    },
+    {
+      "epoch": 0.50426944971537,
+      "grad_norm": 2.702395439147949,
+      "learning_rate": 1.752377795655095e-05,
+      "loss": 0.4786,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5047438330170778,
+      "grad_norm": 2.741300582885742,
+      "learning_rate": 1.751871467714645e-05,
+      "loss": 0.3856,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5052182163187856,
+      "grad_norm": 2.5689239501953125,
+      "learning_rate": 1.751364695951908e-05,
+      "loss": 0.3626,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5056925996204934,
+      "grad_norm": 2.544797897338867,
+      "learning_rate": 1.7508574806660256e-05,
+      "loss": 0.4253,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5061669829222012,
+      "grad_norm": 2.194748640060425,
+      "learning_rate": 1.7503498221564026e-05,
+      "loss": 0.322,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5066413662239089,
+      "grad_norm": 2.4700746536254883,
+      "learning_rate": 1.7498417207227046e-05,
+      "loss": 0.3716,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5071157495256167,
+      "grad_norm": 3.115464687347412,
+      "learning_rate": 1.7493331766648588e-05,
+      "loss": 0.4376,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5075901328273245,
+      "grad_norm": 2.6482186317443848,
+      "learning_rate": 1.748824190283054e-05,
+      "loss": 0.3741,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5080645161290323,
+      "grad_norm": 2.6619374752044678,
+      "learning_rate": 1.7483147618777393e-05,
+      "loss": 0.4422,
+      "step": 1071
+    },
+    {
+      "epoch": 0.50853889943074,
+      "grad_norm": 2.94618558883667,
+      "learning_rate": 1.7478048917496255e-05,
+      "loss": 0.4338,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5090132827324478,
+      "grad_norm": 3.4346418380737305,
+      "learning_rate": 1.7472945801996842e-05,
+      "loss": 0.5294,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5094876660341556,
+      "grad_norm": 2.2107574939727783,
+      "learning_rate": 1.7467838275291467e-05,
+      "loss": 0.3451,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5099620493358634,
+      "grad_norm": 2.622593641281128,
+      "learning_rate": 1.746272634039506e-05,
+      "loss": 0.4888,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5104364326375711,
+      "grad_norm": 2.8133835792541504,
+      "learning_rate": 1.745761000032514e-05,
+      "loss": 0.45,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5109108159392789,
+      "grad_norm": 1.8900046348571777,
+      "learning_rate": 1.745248925810183e-05,
+      "loss": 0.3165,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5113851992409867,
+      "grad_norm": 2.4775540828704834,
+      "learning_rate": 1.744736411674786e-05,
+      "loss": 0.3891,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5118595825426945,
+      "grad_norm": 2.9898617267608643,
+      "learning_rate": 1.7442234579288543e-05,
+      "loss": 0.4028,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5123339658444023,
+      "grad_norm": 2.6482555866241455,
+      "learning_rate": 1.74371006487518e-05,
+      "loss": 0.445,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5128083491461101,
+      "grad_norm": 3.1871161460876465,
+      "learning_rate": 1.743196232816814e-05,
+      "loss": 0.3826,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5132827324478179,
+      "grad_norm": 2.722008466720581,
+      "learning_rate": 1.7426819620570655e-05,
+      "loss": 0.4419,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5137571157495257,
+      "grad_norm": 2.5276153087615967,
+      "learning_rate": 1.7421672528995043e-05,
+      "loss": 0.4635,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5142314990512334,
+      "grad_norm": 2.950514554977417,
+      "learning_rate": 1.7416521056479577e-05,
+      "loss": 0.3816,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5147058823529411,
+      "grad_norm": 2.382774829864502,
+      "learning_rate": 1.741136520606512e-05,
+      "loss": 0.3931,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5151802656546489,
+      "grad_norm": 2.6314048767089844,
+      "learning_rate": 1.7406204980795124e-05,
+      "loss": 0.4377,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5156546489563567,
+      "grad_norm": 3.256890296936035,
+      "learning_rate": 1.740104038371561e-05,
+      "loss": 0.5105,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 2.571377992630005,
+      "learning_rate": 1.7395871417875198e-05,
+      "loss": 0.4192,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5166034155597723,
+      "grad_norm": 2.945387601852417,
+      "learning_rate": 1.7390698086325072e-05,
+      "loss": 0.3837,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5170777988614801,
+      "grad_norm": 3.032877206802368,
+      "learning_rate": 1.7385520392118998e-05,
+      "loss": 0.4706,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5175521821631879,
+      "grad_norm": 3.0032753944396973,
+      "learning_rate": 1.7380338338313322e-05,
+      "loss": 0.4447,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5180265654648957,
+      "grad_norm": 2.387190103530884,
+      "learning_rate": 1.7375151927966954e-05,
+      "loss": 0.386,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5185009487666035,
+      "grad_norm": 3.4106404781341553,
+      "learning_rate": 1.7369961164141383e-05,
+      "loss": 0.5009,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5189753320683111,
+      "grad_norm": 2.958599805831909,
+      "learning_rate": 1.736476604990067e-05,
+      "loss": 0.4794,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5194497153700189,
+      "grad_norm": 2.795013189315796,
+      "learning_rate": 1.735956658831143e-05,
+      "loss": 0.4323,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5199240986717267,
+      "grad_norm": 2.330531120300293,
+      "learning_rate": 1.7354362782442864e-05,
+      "loss": 0.3815,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5203984819734345,
+      "grad_norm": 2.7681827545166016,
+      "learning_rate": 1.734915463536672e-05,
+      "loss": 0.4358,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5208728652751423,
+      "grad_norm": 2.7465672492980957,
+      "learning_rate": 1.7343942150157315e-05,
+      "loss": 0.4384,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5213472485768501,
+      "grad_norm": 2.6306185722351074,
+      "learning_rate": 1.7338725329891532e-05,
+      "loss": 0.4233,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5218216318785579,
+      "grad_norm": 2.5246400833129883,
+      "learning_rate": 1.7333504177648806e-05,
+      "loss": 0.3583,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5222960151802657,
+      "grad_norm": 2.543013334274292,
+      "learning_rate": 1.7328278696511135e-05,
+      "loss": 0.4036,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5227703984819735,
+      "grad_norm": 3.1978096961975098,
+      "learning_rate": 1.7323048889563062e-05,
+      "loss": 0.489,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5232447817836812,
+      "grad_norm": 2.8943567276000977,
+      "learning_rate": 1.73178147598917e-05,
+      "loss": 0.5043,
+      "step": 1103
+    },
+    {
+      "epoch": 0.523719165085389,
+      "grad_norm": 2.635658025741577,
+      "learning_rate": 1.7312576310586692e-05,
+      "loss": 0.4701,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5241935483870968,
+      "grad_norm": 2.4131529331207275,
+      "learning_rate": 1.730733354474025e-05,
+      "loss": 0.4144,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5246679316888045,
+      "grad_norm": 2.3183786869049072,
+      "learning_rate": 1.730208646544713e-05,
+      "loss": 0.4114,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5251423149905123,
+      "grad_norm": 2.472311496734619,
+      "learning_rate": 1.729683507580462e-05,
+      "loss": 0.3944,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5256166982922201,
+      "grad_norm": 2.4466824531555176,
+      "learning_rate": 1.7291579378912574e-05,
+      "loss": 0.4335,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5260910815939279,
+      "grad_norm": 2.289536237716675,
+      "learning_rate": 1.7286319377873374e-05,
+      "loss": 0.3746,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5265654648956357,
+      "grad_norm": 2.8903002738952637,
+      "learning_rate": 1.7281055075791946e-05,
+      "loss": 0.4114,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5270398481973435,
+      "grad_norm": 2.633312225341797,
+      "learning_rate": 1.7275786475775757e-05,
+      "loss": 0.3702,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5275142314990512,
+      "grad_norm": 3.3855140209198,
+      "learning_rate": 1.7270513580934805e-05,
+      "loss": 0.4827,
+      "step": 1112
+    },
+    {
+      "epoch": 0.527988614800759,
+      "grad_norm": 2.293537139892578,
+      "learning_rate": 1.7265236394381634e-05,
+      "loss": 0.3591,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5284629981024668,
+      "grad_norm": 2.2649781703948975,
+      "learning_rate": 1.725995491923131e-05,
+      "loss": 0.3496,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5289373814041746,
+      "grad_norm": 2.7540788650512695,
+      "learning_rate": 1.725466915860144e-05,
+      "loss": 0.4094,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 2.8190548419952393,
+      "learning_rate": 1.7249379115612154e-05,
+      "loss": 0.442,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5298861480075902,
+      "grad_norm": 2.757744789123535,
+      "learning_rate": 1.7244084793386108e-05,
+      "loss": 0.3825,
+      "step": 1117
+    },
+    {
+      "epoch": 0.530360531309298,
+      "grad_norm": 3.2028768062591553,
+      "learning_rate": 1.7238786195048493e-05,
+      "loss": 0.4394,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5308349146110057,
+      "grad_norm": 2.5741443634033203,
+      "learning_rate": 1.7233483323727018e-05,
+      "loss": 0.3955,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5313092979127134,
+      "grad_norm": 3.050013303756714,
+      "learning_rate": 1.7228176182551917e-05,
+      "loss": 0.4598,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5317836812144212,
+      "grad_norm": 2.7212207317352295,
+      "learning_rate": 1.722286477465594e-05,
+      "loss": 0.4926,
+      "step": 1121
+    },
+    {
+      "epoch": 0.532258064516129,
+      "grad_norm": 2.9433910846710205,
+      "learning_rate": 1.721754910317436e-05,
+      "loss": 0.4738,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5327324478178368,
+      "grad_norm": 2.578136444091797,
+      "learning_rate": 1.7212229171244966e-05,
+      "loss": 0.3956,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5332068311195446,
+      "grad_norm": 2.4337382316589355,
+      "learning_rate": 1.7206904982008058e-05,
+      "loss": 0.3345,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5336812144212524,
+      "grad_norm": 2.611060619354248,
+      "learning_rate": 1.7201576538606458e-05,
+      "loss": 0.4081,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5341555977229602,
+      "grad_norm": 2.9734725952148438,
+      "learning_rate": 1.7196243844185488e-05,
+      "loss": 0.4356,
+      "step": 1126
+    },
+    {
+      "epoch": 0.534629981024668,
+      "grad_norm": 2.1246142387390137,
+      "learning_rate": 1.7190906901892986e-05,
+      "loss": 0.412,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5351043643263758,
+      "grad_norm": 2.533374547958374,
+      "learning_rate": 1.7185565714879295e-05,
+      "loss": 0.4051,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5355787476280834,
+      "grad_norm": 2.476228713989258,
+      "learning_rate": 1.718022028629727e-05,
+      "loss": 0.4775,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5360531309297912,
+      "grad_norm": 2.6208913326263428,
+      "learning_rate": 1.7174870619302263e-05,
+      "loss": 0.3889,
+      "step": 1130
+    },
+    {
+      "epoch": 0.536527514231499,
+      "grad_norm": 2.2168285846710205,
+      "learning_rate": 1.7169516717052122e-05,
+      "loss": 0.3205,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5370018975332068,
+      "grad_norm": 2.6198856830596924,
+      "learning_rate": 1.7164158582707215e-05,
+      "loss": 0.3413,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5374762808349146,
+      "grad_norm": 2.9443211555480957,
+      "learning_rate": 1.715879621943038e-05,
+      "loss": 0.5137,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5379506641366224,
+      "grad_norm": 2.5952861309051514,
+      "learning_rate": 1.7153429630386985e-05,
+      "loss": 0.4471,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5384250474383302,
+      "grad_norm": 2.5865321159362793,
+      "learning_rate": 1.714805881874486e-05,
+      "loss": 0.3915,
+      "step": 1135
+    },
+    {
+      "epoch": 0.538899430740038,
+      "grad_norm": 2.849581003189087,
+      "learning_rate": 1.7142683787674353e-05,
+      "loss": 0.4406,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5393738140417458,
+      "grad_norm": 2.4955079555511475,
+      "learning_rate": 1.713730454034828e-05,
+      "loss": 0.4032,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5398481973434535,
+      "grad_norm": 2.804607391357422,
+      "learning_rate": 1.7131921079941965e-05,
+      "loss": 0.354,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5403225806451613,
+      "grad_norm": 2.3211519718170166,
+      "learning_rate": 1.7126533409633214e-05,
+      "loss": 0.3228,
+      "step": 1139
+    },
+    {
+      "epoch": 0.540796963946869,
+      "grad_norm": 2.563812732696533,
+      "learning_rate": 1.7121141532602306e-05,
+      "loss": 0.407,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5412713472485768,
+      "grad_norm": 2.5948188304901123,
+      "learning_rate": 1.7115745452032023e-05,
+      "loss": 0.3885,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5417457305502846,
+      "grad_norm": 2.253960371017456,
+      "learning_rate": 1.711034517110761e-05,
+      "loss": 0.4199,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5422201138519924,
+      "grad_norm": 2.6741251945495605,
+      "learning_rate": 1.7104940693016803e-05,
+      "loss": 0.4642,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5426944971537002,
+      "grad_norm": 2.9957103729248047,
+      "learning_rate": 1.709953202094981e-05,
+      "loss": 0.3952,
+      "step": 1144
+    },
+    {
+      "epoch": 0.543168880455408,
+      "grad_norm": 2.8041434288024902,
+      "learning_rate": 1.7094119158099318e-05,
+      "loss": 0.3991,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5436432637571158,
+      "grad_norm": 2.4647340774536133,
+      "learning_rate": 1.708870210766049e-05,
+      "loss": 0.3622,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5441176470588235,
+      "grad_norm": 3.1950418949127197,
+      "learning_rate": 1.708328087283095e-05,
+      "loss": 0.4606,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5445920303605313,
+      "grad_norm": 2.3248064517974854,
+      "learning_rate": 1.7077855456810803e-05,
+      "loss": 0.3685,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5450664136622391,
+      "grad_norm": 3.027299404144287,
+      "learning_rate": 1.7072425862802618e-05,
+      "loss": 0.4978,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5455407969639469,
+      "grad_norm": 3.0360865592956543,
+      "learning_rate": 1.706699209401143e-05,
+      "loss": 0.3587,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5460151802656547,
+      "grad_norm": 2.7599403858184814,
+      "learning_rate": 1.7061554153644743e-05,
+      "loss": 0.4213,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5464895635673624,
+      "grad_norm": 3.035600423812866,
+      "learning_rate": 1.7056112044912513e-05,
+      "loss": 0.4591,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5469639468690702,
+      "grad_norm": 2.4583334922790527,
+      "learning_rate": 1.705066577102717e-05,
+      "loss": 0.4092,
+      "step": 1153
+    },
+    {
+      "epoch": 0.547438330170778,
+      "grad_norm": 2.438873529434204,
+      "learning_rate": 1.704521533520359e-05,
+      "loss": 0.4513,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5479127134724858,
+      "grad_norm": 2.434335708618164,
+      "learning_rate": 1.703976074065911e-05,
+      "loss": 0.3763,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5483870967741935,
+      "grad_norm": 2.7896952629089355,
+      "learning_rate": 1.703430199061353e-05,
+      "loss": 0.4478,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5488614800759013,
+      "grad_norm": 3.330437183380127,
+      "learning_rate": 1.7028839088289092e-05,
+      "loss": 0.4884,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5493358633776091,
+      "grad_norm": 2.2704224586486816,
+      "learning_rate": 1.702337203691049e-05,
+      "loss": 0.3599,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5498102466793169,
+      "grad_norm": 2.6077935695648193,
+      "learning_rate": 1.7017900839704877e-05,
+      "loss": 0.439,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5502846299810247,
+      "grad_norm": 2.6503183841705322,
+      "learning_rate": 1.7012425499901842e-05,
+      "loss": 0.443,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5507590132827325,
+      "grad_norm": 2.3090269565582275,
+      "learning_rate": 1.7006946020733426e-05,
+      "loss": 0.4096,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5512333965844403,
+      "grad_norm": 2.1434450149536133,
+      "learning_rate": 1.700146240543411e-05,
+      "loss": 0.373,
+      "step": 1162
+    },
+    {
+      "epoch": 0.551707779886148,
+      "grad_norm": 2.830927610397339,
+      "learning_rate": 1.699597465724082e-05,
+      "loss": 0.4866,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5521821631878557,
+      "grad_norm": 2.427565336227417,
+      "learning_rate": 1.6990482779392918e-05,
+      "loss": 0.3733,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5526565464895635,
+      "grad_norm": 2.7859416007995605,
+      "learning_rate": 1.6984986775132202e-05,
+      "loss": 0.4309,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5531309297912713,
+      "grad_norm": 3.880078077316284,
+      "learning_rate": 1.6979486647702917e-05,
+      "loss": 0.6485,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5536053130929791,
+      "grad_norm": 2.7935032844543457,
+      "learning_rate": 1.6973982400351726e-05,
+      "loss": 0.4619,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5540796963946869,
+      "grad_norm": 2.788963794708252,
+      "learning_rate": 1.6968474036327733e-05,
+      "loss": 0.3934,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5545540796963947,
+      "grad_norm": 2.3804471492767334,
+      "learning_rate": 1.6962961558882476e-05,
+      "loss": 0.4006,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5550284629981025,
+      "grad_norm": 2.1218841075897217,
+      "learning_rate": 1.6957444971269907e-05,
+      "loss": 0.3862,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5555028462998103,
+      "grad_norm": 2.6136226654052734,
+      "learning_rate": 1.6951924276746425e-05,
+      "loss": 0.4073,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5559772296015181,
+      "grad_norm": 2.710028886795044,
+      "learning_rate": 1.694639947857083e-05,
+      "loss": 0.4464,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5564516129032258,
+      "grad_norm": 2.524308204650879,
+      "learning_rate": 1.6940870580004364e-05,
+      "loss": 0.4632,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5569259962049335,
+      "grad_norm": 2.7076687812805176,
+      "learning_rate": 1.6935337584310674e-05,
+      "loss": 0.4706,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5574003795066413,
+      "grad_norm": 2.7183334827423096,
+      "learning_rate": 1.6929800494755836e-05,
+      "loss": 0.4462,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5578747628083491,
+      "grad_norm": 2.5758368968963623,
+      "learning_rate": 1.692425931460834e-05,
+      "loss": 0.3369,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5583491461100569,
+      "grad_norm": 2.9988672733306885,
+      "learning_rate": 1.691871404713909e-05,
+      "loss": 0.5041,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5588235294117647,
+      "grad_norm": 2.278681516647339,
+      "learning_rate": 1.69131646956214e-05,
+      "loss": 0.4227,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5592979127134725,
+      "grad_norm": 2.4047083854675293,
+      "learning_rate": 1.6907611263331004e-05,
+      "loss": 0.3633,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5597722960151803,
+      "grad_norm": 2.6348397731781006,
+      "learning_rate": 1.6902053753546026e-05,
+      "loss": 0.4034,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5602466793168881,
+      "grad_norm": 2.6245360374450684,
+      "learning_rate": 1.6896492169547022e-05,
+      "loss": 0.4203,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5607210626185958,
+      "grad_norm": 2.4060757160186768,
+      "learning_rate": 1.6890926514616926e-05,
+      "loss": 0.3561,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5611954459203036,
+      "grad_norm": 4.168298244476318,
+      "learning_rate": 1.6885356792041107e-05,
+      "loss": 0.4976,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5616698292220114,
+      "grad_norm": 2.4673924446105957,
+      "learning_rate": 1.68797830051073e-05,
+      "loss": 0.4339,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5621442125237192,
+      "grad_norm": 2.879330635070801,
+      "learning_rate": 1.6874205157105667e-05,
+      "loss": 0.4091,
+      "step": 1185
+    },
+    {
+      "epoch": 0.562618595825427,
+      "grad_norm": 2.9739437103271484,
+      "learning_rate": 1.686862325132875e-05,
+      "loss": 0.379,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5630929791271347,
+      "grad_norm": 2.463998317718506,
+      "learning_rate": 1.68630372910715e-05,
+      "loss": 0.394,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5635673624288425,
+      "grad_norm": 3.4014320373535156,
+      "learning_rate": 1.685744727963125e-05,
+      "loss": 0.4995,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5640417457305503,
+      "grad_norm": 3.1018409729003906,
+      "learning_rate": 1.685185322030772e-05,
+      "loss": 0.3801,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5645161290322581,
+      "grad_norm": 3.028447151184082,
+      "learning_rate": 1.6846255116403044e-05,
+      "loss": 0.4118,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5649905123339658,
+      "grad_norm": 3.2124314308166504,
+      "learning_rate": 1.6840652971221714e-05,
+      "loss": 0.3829,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5654648956356736,
+      "grad_norm": 3.856614351272583,
+      "learning_rate": 1.683504678807063e-05,
+      "loss": 0.4024,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5659392789373814,
+      "grad_norm": 2.8994805812835693,
+      "learning_rate": 1.6829436570259064e-05,
+      "loss": 0.469,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5664136622390892,
+      "grad_norm": 2.968153953552246,
+      "learning_rate": 1.6823822321098667e-05,
+      "loss": 0.4357,
+      "step": 1194
+    },
+    {
+      "epoch": 0.566888045540797,
+      "grad_norm": 2.9536709785461426,
+      "learning_rate": 1.681820404390348e-05,
+      "loss": 0.4572,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5673624288425048,
+      "grad_norm": 2.574279546737671,
+      "learning_rate": 1.6812581741989915e-05,
+      "loss": 0.3608,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5678368121442126,
+      "grad_norm": 2.5156452655792236,
+      "learning_rate": 1.680695541867676e-05,
+      "loss": 0.402,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5683111954459203,
+      "grad_norm": 3.1304829120635986,
+      "learning_rate": 1.680132507728518e-05,
+      "loss": 0.3993,
+      "step": 1198
+    },
+    {
+      "epoch": 0.568785578747628,
+      "grad_norm": 3.0311906337738037,
+      "learning_rate": 1.679569072113871e-05,
+      "loss": 0.3883,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5692599620493358,
+      "grad_norm": 2.762141227722168,
+      "learning_rate": 1.6790052353563254e-05,
+      "loss": 0.3877,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5697343453510436,
+      "grad_norm": 2.4210193157196045,
+      "learning_rate": 1.678440997788708e-05,
+      "loss": 0.3758,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5702087286527514,
+      "grad_norm": 2.446246862411499,
+      "learning_rate": 1.6778763597440833e-05,
+      "loss": 0.3929,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5706831119544592,
+      "grad_norm": 2.542310953140259,
+      "learning_rate": 1.6773113215557514e-05,
+      "loss": 0.4038,
+      "step": 1203
+    },
+    {
+      "epoch": 0.571157495256167,
+      "grad_norm": 2.6096861362457275,
+      "learning_rate": 1.6767458835572487e-05,
+      "loss": 0.421,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5716318785578748,
+      "grad_norm": 2.5336506366729736,
+      "learning_rate": 1.6761800460823473e-05,
+      "loss": 0.4722,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5721062618595826,
+      "grad_norm": 2.657362461090088,
+      "learning_rate": 1.6756138094650563e-05,
+      "loss": 0.4266,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5725806451612904,
+      "grad_norm": 2.6421406269073486,
+      "learning_rate": 1.6750471740396187e-05,
+      "loss": 0.4215,
+      "step": 1207
+    },
+    {
+      "epoch": 0.573055028462998,
+      "grad_norm": 2.54229998588562,
+      "learning_rate": 1.6744801401405138e-05,
+      "loss": 0.4456,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5735294117647058,
+      "grad_norm": 3.0632123947143555,
+      "learning_rate": 1.6739127081024574e-05,
+      "loss": 0.3799,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5740037950664136,
+      "grad_norm": 2.753382921218872,
+      "learning_rate": 1.673344878260397e-05,
+      "loss": 0.4764,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5744781783681214,
+      "grad_norm": 2.06011700630188,
+      "learning_rate": 1.6727766509495186e-05,
+      "loss": 0.3336,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5749525616698292,
+      "grad_norm": 2.405019760131836,
+      "learning_rate": 1.6722080265052407e-05,
+      "loss": 0.4868,
+      "step": 1212
+    },
+    {
+      "epoch": 0.575426944971537,
+      "grad_norm": 2.274955987930298,
+      "learning_rate": 1.671639005263216e-05,
+      "loss": 0.3628,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5759013282732448,
+      "grad_norm": 3.2761285305023193,
+      "learning_rate": 1.671069587559333e-05,
+      "loss": 0.4588,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5763757115749526,
+      "grad_norm": 3.0697081089019775,
+      "learning_rate": 1.6704997737297134e-05,
+      "loss": 0.43,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5768500948766604,
+      "grad_norm": 2.4201908111572266,
+      "learning_rate": 1.6699295641107116e-05,
+      "loss": 0.3929,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5773244781783681,
+      "grad_norm": 2.5585038661956787,
+      "learning_rate": 1.6693589590389176e-05,
+      "loss": 0.4479,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5777988614800759,
+      "grad_norm": 2.488297939300537,
+      "learning_rate": 1.668787958851153e-05,
+      "loss": 0.4259,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5782732447817837,
+      "grad_norm": 2.3524911403656006,
+      "learning_rate": 1.6682165638844753e-05,
+      "loss": 0.4478,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5787476280834914,
+      "grad_norm": 2.2819252014160156,
+      "learning_rate": 1.6676447744761715e-05,
+      "loss": 0.3464,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5792220113851992,
+      "grad_norm": 2.497574806213379,
+      "learning_rate": 1.6670725909637645e-05,
+      "loss": 0.4008,
+      "step": 1221
+    },
+    {
+      "epoch": 0.579696394686907,
+      "grad_norm": 2.4774222373962402,
+      "learning_rate": 1.6665000136850076e-05,
+      "loss": 0.4056,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5801707779886148,
+      "grad_norm": 3.251617908477783,
+      "learning_rate": 1.665927042977888e-05,
+      "loss": 0.4725,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 5.06425666809082,
+      "learning_rate": 1.6653536791806248e-05,
+      "loss": 0.4183,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5811195445920304,
+      "grad_norm": 2.323410749435425,
+      "learning_rate": 1.6647799226316684e-05,
+      "loss": 0.3405,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5815939278937381,
+      "grad_norm": 2.2724153995513916,
+      "learning_rate": 1.6642057736697023e-05,
+      "loss": 0.3668,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5820683111954459,
+      "grad_norm": 2.63533878326416,
+      "learning_rate": 1.6636312326336402e-05,
+      "loss": 0.458,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5825426944971537,
+      "grad_norm": 3.030594825744629,
+      "learning_rate": 1.6630562998626287e-05,
+      "loss": 0.3992,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5830170777988615,
+      "grad_norm": 2.6020584106445312,
+      "learning_rate": 1.6624809756960445e-05,
+      "loss": 0.4317,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5834914611005693,
+      "grad_norm": 2.6600024700164795,
+      "learning_rate": 1.6619052604734958e-05,
+      "loss": 0.3952,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5839658444022771,
+      "grad_norm": 2.172214984893799,
+      "learning_rate": 1.661329154534822e-05,
+      "loss": 0.343,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5844402277039848,
+      "grad_norm": 2.652944803237915,
+      "learning_rate": 1.6607526582200918e-05,
+      "loss": 0.4041,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5849146110056926,
+      "grad_norm": 2.3528411388397217,
+      "learning_rate": 1.6601757718696065e-05,
+      "loss": 0.4106,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5853889943074004,
+      "grad_norm": 2.205167770385742,
+      "learning_rate": 1.6595984958238952e-05,
+      "loss": 0.3645,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5858633776091081,
+      "grad_norm": 2.9009509086608887,
+      "learning_rate": 1.6590208304237193e-05,
+      "loss": 0.5345,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5863377609108159,
+      "grad_norm": 3.4057090282440186,
+      "learning_rate": 1.6584427760100682e-05,
+      "loss": 0.4928,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5868121442125237,
+      "grad_norm": 2.6523916721343994,
+      "learning_rate": 1.657864332924162e-05,
+      "loss": 0.4055,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5872865275142315,
+      "grad_norm": 2.7750117778778076,
+      "learning_rate": 1.6572855015074502e-05,
+      "loss": 0.4551,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5877609108159393,
+      "grad_norm": 2.880966901779175,
+      "learning_rate": 1.656706282101611e-05,
+      "loss": 0.4637,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 2.484705924987793,
+      "learning_rate": 1.6561266750485517e-05,
+      "loss": 0.37,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5887096774193549,
+      "grad_norm": 2.381371259689331,
+      "learning_rate": 1.655546680690409e-05,
+      "loss": 0.4134,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5891840607210627,
+      "grad_norm": 2.261758327484131,
+      "learning_rate": 1.654966299369547e-05,
+      "loss": 0.3461,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5896584440227703,
+      "grad_norm": 2.374173402786255,
+      "learning_rate": 1.6543855314285598e-05,
+      "loss": 0.4009,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5901328273244781,
+      "grad_norm": 2.667855978012085,
+      "learning_rate": 1.6538043772102692e-05,
+      "loss": 0.4612,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5906072106261859,
+      "grad_norm": 2.694554328918457,
+      "learning_rate": 1.653222837057724e-05,
+      "loss": 0.4593,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5910815939278937,
+      "grad_norm": 2.515317678451538,
+      "learning_rate": 1.6526409113142022e-05,
+      "loss": 0.4236,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5915559772296015,
+      "grad_norm": 2.1223080158233643,
+      "learning_rate": 1.652058600323209e-05,
+      "loss": 0.3854,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5920303605313093,
+      "grad_norm": 2.624248504638672,
+      "learning_rate": 1.651475904428476e-05,
+      "loss": 0.3865,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5925047438330171,
+      "grad_norm": 2.6333069801330566,
+      "learning_rate": 1.6508928239739632e-05,
+      "loss": 0.4257,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5929791271347249,
+      "grad_norm": 2.5864479541778564,
+      "learning_rate": 1.6503093593038573e-05,
+      "loss": 0.415,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5934535104364327,
+      "grad_norm": 2.4937593936920166,
+      "learning_rate": 1.649725510762572e-05,
+      "loss": 0.4392,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5939278937381404,
+      "grad_norm": 2.5235211849212646,
+      "learning_rate": 1.6491412786947468e-05,
+      "loss": 0.3365,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5944022770398482,
+      "grad_norm": 2.3775665760040283,
+      "learning_rate": 1.6485566634452483e-05,
+      "loss": 0.3976,
+      "step": 1253
+    },
+    {
+      "epoch": 0.594876660341556,
+      "grad_norm": 2.5280470848083496,
+      "learning_rate": 1.6479716653591694e-05,
+      "loss": 0.4174,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5953510436432637,
+      "grad_norm": 2.390460968017578,
+      "learning_rate": 1.647386284781828e-05,
+      "loss": 0.3771,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5958254269449715,
+      "grad_norm": 2.5851383209228516,
+      "learning_rate": 1.6468005220587687e-05,
+      "loss": 0.4512,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5962998102466793,
+      "grad_norm": 2.5305631160736084,
+      "learning_rate": 1.646214377535762e-05,
+      "loss": 0.3618,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5967741935483871,
+      "grad_norm": 2.88232159614563,
+      "learning_rate": 1.6456278515588023e-05,
+      "loss": 0.4459,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5972485768500949,
+      "grad_norm": 2.761087656021118,
+      "learning_rate": 1.6450409444741112e-05,
+      "loss": 0.4103,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5977229601518027,
+      "grad_norm": 2.098811388015747,
+      "learning_rate": 1.6444536566281332e-05,
+      "loss": 0.3358,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5981973434535104,
+      "grad_norm": 2.363554000854492,
+      "learning_rate": 1.643865988367539e-05,
+      "loss": 0.3722,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5986717267552182,
+      "grad_norm": 2.7632856369018555,
+      "learning_rate": 1.6432779400392232e-05,
+      "loss": 0.4257,
+      "step": 1262
+    },
+    {
+      "epoch": 0.599146110056926,
+      "grad_norm": 2.5627601146698,
+      "learning_rate": 1.6426895119903046e-05,
+      "loss": 0.4035,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5996204933586338,
+      "grad_norm": 2.7176856994628906,
+      "learning_rate": 1.6421007045681273e-05,
+      "loss": 0.4467,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6000948766603416,
+      "grad_norm": 2.368619441986084,
+      "learning_rate": 1.6415115181202576e-05,
+      "loss": 0.3523,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6005692599620494,
+      "grad_norm": 2.4514405727386475,
+      "learning_rate": 1.6409219529944866e-05,
+      "loss": 0.3838,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6010436432637571,
+      "grad_norm": 2.8423192501068115,
+      "learning_rate": 1.640332009538829e-05,
+      "loss": 0.3943,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6015180265654649,
+      "grad_norm": 2.820495843887329,
+      "learning_rate": 1.639741688101523e-05,
+      "loss": 0.4247,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6019924098671727,
+      "grad_norm": 2.6144824028015137,
+      "learning_rate": 1.6391509890310285e-05,
+      "loss": 0.43,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6024667931688804,
+      "grad_norm": 2.3828113079071045,
+      "learning_rate": 1.63855991267603e-05,
+      "loss": 0.3844,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6029411764705882,
+      "grad_norm": 2.611931324005127,
+      "learning_rate": 1.637968459385434e-05,
+      "loss": 0.3651,
+      "step": 1271
+    },
+    {
+      "epoch": 0.603415559772296,
+      "grad_norm": 2.4862523078918457,
+      "learning_rate": 1.6373766295083693e-05,
+      "loss": 0.4257,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6038899430740038,
+      "grad_norm": 2.119983673095703,
+      "learning_rate": 1.636784423394187e-05,
+      "loss": 0.3721,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6043643263757116,
+      "grad_norm": 2.5551223754882812,
+      "learning_rate": 1.6361918413924614e-05,
+      "loss": 0.4151,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6048387096774194,
+      "grad_norm": 2.954251527786255,
+      "learning_rate": 1.635598883852987e-05,
+      "loss": 0.4131,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6053130929791272,
+      "grad_norm": 2.8312759399414062,
+      "learning_rate": 1.6350055511257806e-05,
+      "loss": 0.3901,
+      "step": 1276
+    },
+    {
+      "epoch": 0.605787476280835,
+      "grad_norm": 2.543302536010742,
+      "learning_rate": 1.6344118435610814e-05,
+      "loss": 0.3597,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6062618595825426,
+      "grad_norm": 2.338500738143921,
+      "learning_rate": 1.633817761509349e-05,
+      "loss": 0.3255,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6067362428842504,
+      "grad_norm": 2.9333157539367676,
+      "learning_rate": 1.6332233053212632e-05,
+      "loss": 0.4025,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6072106261859582,
+      "grad_norm": 2.388615608215332,
+      "learning_rate": 1.6326284753477267e-05,
+      "loss": 0.4042,
+      "step": 1280
+    },
+    {
+      "epoch": 0.607685009487666,
+      "grad_norm": 2.1982688903808594,
+      "learning_rate": 1.6320332719398612e-05,
+      "loss": 0.341,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6081593927893738,
+      "grad_norm": 2.6067326068878174,
+      "learning_rate": 1.6314376954490097e-05,
+      "loss": 0.441,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6086337760910816,
+      "grad_norm": 2.442051410675049,
+      "learning_rate": 1.6308417462267348e-05,
+      "loss": 0.4034,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6091081593927894,
+      "grad_norm": 2.5747456550598145,
+      "learning_rate": 1.6302454246248195e-05,
+      "loss": 0.4045,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6095825426944972,
+      "grad_norm": 3.0773766040802,
+      "learning_rate": 1.6296487309952666e-05,
+      "loss": 0.4241,
+      "step": 1285
+    },
+    {
+      "epoch": 0.610056925996205,
+      "grad_norm": 2.966850519180298,
+      "learning_rate": 1.6290516656902985e-05,
+      "loss": 0.4131,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6105313092979127,
+      "grad_norm": 2.5385637283325195,
+      "learning_rate": 1.6284542290623568e-05,
+      "loss": 0.3904,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6110056925996205,
+      "grad_norm": 2.2452964782714844,
+      "learning_rate": 1.6278564214641027e-05,
+      "loss": 0.3438,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6114800759013282,
+      "grad_norm": 3.1382503509521484,
+      "learning_rate": 1.6272582432484155e-05,
+      "loss": 0.4465,
+      "step": 1289
+    },
+    {
+      "epoch": 0.611954459203036,
+      "grad_norm": 2.5982370376586914,
+      "learning_rate": 1.6266596947683945e-05,
+      "loss": 0.4197,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6124288425047438,
+      "grad_norm": 3.1609013080596924,
+      "learning_rate": 1.626060776377357e-05,
+      "loss": 0.4191,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6129032258064516,
+      "grad_norm": 2.7382616996765137,
+      "learning_rate": 1.625461488428838e-05,
+      "loss": 0.4635,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6133776091081594,
+      "grad_norm": 2.5566318035125732,
+      "learning_rate": 1.6248618312765918e-05,
+      "loss": 0.4796,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6138519924098672,
+      "grad_norm": 2.541482448577881,
+      "learning_rate": 1.62426180527459e-05,
+      "loss": 0.4099,
+      "step": 1294
+    },
+    {
+      "epoch": 0.614326375711575,
+      "grad_norm": 3.1109683513641357,
+      "learning_rate": 1.6236614107770216e-05,
+      "loss": 0.4481,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6148007590132827,
+      "grad_norm": 2.3709452152252197,
+      "learning_rate": 1.623060648138294e-05,
+      "loss": 0.4273,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6152751423149905,
+      "grad_norm": 2.797222375869751,
+      "learning_rate": 1.622459517713031e-05,
+      "loss": 0.3849,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6157495256166983,
+      "grad_norm": 2.5726704597473145,
+      "learning_rate": 1.621858019856074e-05,
+      "loss": 0.4335,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6162239089184061,
+      "grad_norm": 2.80641508102417,
+      "learning_rate": 1.6212561549224818e-05,
+      "loss": 0.4747,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6166982922201139,
+      "grad_norm": 2.7815253734588623,
+      "learning_rate": 1.620653923267529e-05,
+      "loss": 0.4144,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6171726755218216,
+      "grad_norm": 2.2950119972229004,
+      "learning_rate": 1.6200513252467068e-05,
+      "loss": 0.4452,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6176470588235294,
+      "grad_norm": 2.441072463989258,
+      "learning_rate": 1.6194483612157232e-05,
+      "loss": 0.4684,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6181214421252372,
+      "grad_norm": 2.0913050174713135,
+      "learning_rate": 1.6188450315305012e-05,
+      "loss": 0.3706,
+      "step": 1303
+    },
+    {
+      "epoch": 0.618595825426945,
+      "grad_norm": 2.4034149646759033,
+      "learning_rate": 1.6182413365471815e-05,
+      "loss": 0.3506,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6190702087286527,
+      "grad_norm": 2.8475241661071777,
+      "learning_rate": 1.617637276622118e-05,
+      "loss": 0.4507,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6195445920303605,
+      "grad_norm": 2.499732255935669,
+      "learning_rate": 1.617032852111882e-05,
+      "loss": 0.368,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6200189753320683,
+      "grad_norm": 2.651540517807007,
+      "learning_rate": 1.6164280633732594e-05,
+      "loss": 0.4742,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6204933586337761,
+      "grad_norm": 2.2591097354888916,
+      "learning_rate": 1.6158229107632507e-05,
+      "loss": 0.3422,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6209677419354839,
+      "grad_norm": 2.7106447219848633,
+      "learning_rate": 1.6152173946390715e-05,
+      "loss": 0.4175,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6214421252371917,
+      "grad_norm": 3.118277072906494,
+      "learning_rate": 1.6146115153581523e-05,
+      "loss": 0.5013,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6219165085388995,
+      "grad_norm": 2.603504180908203,
+      "learning_rate": 1.6140052732781373e-05,
+      "loss": 0.3475,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6223908918406073,
+      "grad_norm": 2.1159276962280273,
+      "learning_rate": 1.6133986687568854e-05,
+      "loss": 0.3465,
+      "step": 1312
+    },
+    {
+      "epoch": 0.622865275142315,
+      "grad_norm": 2.8673932552337646,
+      "learning_rate": 1.612791702152469e-05,
+      "loss": 0.4357,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6233396584440227,
+      "grad_norm": 2.5344109535217285,
+      "learning_rate": 1.6121843738231748e-05,
+      "loss": 0.3729,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6238140417457305,
+      "grad_norm": 2.3215112686157227,
+      "learning_rate": 1.6115766841275027e-05,
+      "loss": 0.3495,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6242884250474383,
+      "grad_norm": 2.229944944381714,
+      "learning_rate": 1.6109686334241655e-05,
+      "loss": 0.3594,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6247628083491461,
+      "grad_norm": 3.3155102729797363,
+      "learning_rate": 1.6103602220720897e-05,
+      "loss": 0.3867,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6252371916508539,
+      "grad_norm": 2.2139594554901123,
+      "learning_rate": 1.609751450430415e-05,
+      "loss": 0.374,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6257115749525617,
+      "grad_norm": 2.4204013347625732,
+      "learning_rate": 1.6091423188584926e-05,
+      "loss": 0.3767,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6261859582542695,
+      "grad_norm": 2.2802765369415283,
+      "learning_rate": 1.608532827715887e-05,
+      "loss": 0.3397,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6266603415559773,
+      "grad_norm": 2.470691204071045,
+      "learning_rate": 1.607922977362375e-05,
+      "loss": 0.3941,
+      "step": 1321
+    },
+    {
+      "epoch": 0.627134724857685,
+      "grad_norm": 2.650733232498169,
+      "learning_rate": 1.607312768157945e-05,
+      "loss": 0.3187,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6276091081593927,
+      "grad_norm": 2.694547176361084,
+      "learning_rate": 1.606702200462798e-05,
+      "loss": 0.4531,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6280834914611005,
+      "grad_norm": 2.8122143745422363,
+      "learning_rate": 1.6060912746373453e-05,
+      "loss": 0.4155,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6285578747628083,
+      "grad_norm": 2.1507198810577393,
+      "learning_rate": 1.6054799910422106e-05,
+      "loss": 0.3287,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6290322580645161,
+      "grad_norm": 3.3852505683898926,
+      "learning_rate": 1.604868350038229e-05,
+      "loss": 0.4204,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6295066413662239,
+      "grad_norm": 2.315032482147217,
+      "learning_rate": 1.6042563519864452e-05,
+      "loss": 0.3035,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6299810246679317,
+      "grad_norm": 2.5337111949920654,
+      "learning_rate": 1.603643997248117e-05,
+      "loss": 0.4118,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6304554079696395,
+      "grad_norm": 2.549182653427124,
+      "learning_rate": 1.60303128618471e-05,
+      "loss": 0.3673,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6309297912713473,
+      "grad_norm": 2.420851230621338,
+      "learning_rate": 1.6024182191579024e-05,
+      "loss": 0.3595,
+      "step": 1330
+    },
+    {
+      "epoch": 0.631404174573055,
+      "grad_norm": 2.3073275089263916,
+      "learning_rate": 1.601804796529581e-05,
+      "loss": 0.3316,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6318785578747628,
+      "grad_norm": 2.652170181274414,
+      "learning_rate": 1.6011910186618433e-05,
+      "loss": 0.3742,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6323529411764706,
+      "grad_norm": 2.5604469776153564,
+      "learning_rate": 1.6005768859169965e-05,
+      "loss": 0.3906,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6328273244781784,
+      "grad_norm": 2.7032854557037354,
+      "learning_rate": 1.5999623986575565e-05,
+      "loss": 0.3651,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6333017077798861,
+      "grad_norm": 2.5799543857574463,
+      "learning_rate": 1.5993475572462498e-05,
+      "loss": 0.3985,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6337760910815939,
+      "grad_norm": 2.3732707500457764,
+      "learning_rate": 1.5987323620460106e-05,
+      "loss": 0.3936,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6342504743833017,
+      "grad_norm": 2.5276899337768555,
+      "learning_rate": 1.598116813419983e-05,
+      "loss": 0.3924,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6347248576850095,
+      "grad_norm": 2.2890419960021973,
+      "learning_rate": 1.597500911731519e-05,
+      "loss": 0.4306,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6351992409867173,
+      "grad_norm": 2.1135501861572266,
+      "learning_rate": 1.5968846573441794e-05,
+      "loss": 0.281,
+      "step": 1339
+    },
+    {
+      "epoch": 0.635673624288425,
+      "grad_norm": 2.5915725231170654,
+      "learning_rate": 1.596268050621733e-05,
+      "loss": 0.4047,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6361480075901328,
+      "grad_norm": 2.0635995864868164,
+      "learning_rate": 1.5956510919281564e-05,
+      "loss": 0.3489,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6366223908918406,
+      "grad_norm": 2.7874855995178223,
+      "learning_rate": 1.5950337816276347e-05,
+      "loss": 0.3854,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6370967741935484,
+      "grad_norm": 2.3711447715759277,
+      "learning_rate": 1.5944161200845595e-05,
+      "loss": 0.3981,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6375711574952562,
+      "grad_norm": 2.6661829948425293,
+      "learning_rate": 1.5937981076635313e-05,
+      "loss": 0.4276,
+      "step": 1344
+    },
+    {
+      "epoch": 0.638045540796964,
+      "grad_norm": 2.2618393898010254,
+      "learning_rate": 1.5931797447293553e-05,
+      "loss": 0.3637,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6385199240986718,
+      "grad_norm": 2.272256851196289,
+      "learning_rate": 1.5925610316470462e-05,
+      "loss": 0.3746,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6389943074003795,
+      "grad_norm": 2.7523257732391357,
+      "learning_rate": 1.5919419687818235e-05,
+      "loss": 0.4538,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6394686907020873,
+      "grad_norm": 2.0002548694610596,
+      "learning_rate": 1.5913225564991142e-05,
+      "loss": 0.352,
+      "step": 1348
+    },
+    {
+      "epoch": 0.639943074003795,
+      "grad_norm": 3.1848556995391846,
+      "learning_rate": 1.590702795164551e-05,
+      "loss": 0.4523,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6404174573055028,
+      "grad_norm": 2.4870452880859375,
+      "learning_rate": 1.5900826851439734e-05,
+      "loss": 0.4132,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6408918406072106,
+      "grad_norm": 2.299504041671753,
+      "learning_rate": 1.5894622268034258e-05,
+      "loss": 0.3974,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6413662239089184,
+      "grad_norm": 2.126984119415283,
+      "learning_rate": 1.588841420509159e-05,
+      "loss": 0.377,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6418406072106262,
+      "grad_norm": 2.595367193222046,
+      "learning_rate": 1.588220266627628e-05,
+      "loss": 0.375,
+      "step": 1353
+    },
+    {
+      "epoch": 0.642314990512334,
+      "grad_norm": 2.4395711421966553,
+      "learning_rate": 1.5875987655254947e-05,
+      "loss": 0.3947,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6427893738140418,
+      "grad_norm": 2.616209030151367,
+      "learning_rate": 1.5869769175696243e-05,
+      "loss": 0.3637,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6432637571157496,
+      "grad_norm": 2.404573917388916,
+      "learning_rate": 1.586354723127088e-05,
+      "loss": 0.3973,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6437381404174574,
+      "grad_norm": 2.426798105239868,
+      "learning_rate": 1.585732182565161e-05,
+      "loss": 0.3932,
+      "step": 1357
+    },
+    {
+      "epoch": 0.644212523719165,
+      "grad_norm": 2.2638189792633057,
+      "learning_rate": 1.5851092962513224e-05,
+      "loss": 0.3432,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6446869070208728,
+      "grad_norm": 3.1538186073303223,
+      "learning_rate": 1.5844860645532563e-05,
+      "loss": 0.3715,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 2.5172009468078613,
+      "learning_rate": 1.5838624878388498e-05,
+      "loss": 0.3756,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6456356736242884,
+      "grad_norm": 2.5488150119781494,
+      "learning_rate": 1.5832385664761943e-05,
+      "loss": 0.4415,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6461100569259962,
+      "grad_norm": 2.245295763015747,
+      "learning_rate": 1.582614300833585e-05,
+      "loss": 0.3377,
+      "step": 1362
+    },
+    {
+      "epoch": 0.646584440227704,
+      "grad_norm": 2.691495180130005,
+      "learning_rate": 1.5819896912795186e-05,
+      "loss": 0.3952,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 2.450929880142212,
+      "learning_rate": 1.5813647381826968e-05,
+      "loss": 0.436,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6475332068311196,
+      "grad_norm": 2.736772298812866,
+      "learning_rate": 1.5807394419120224e-05,
+      "loss": 0.4265,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6480075901328273,
+      "grad_norm": 2.9314868450164795,
+      "learning_rate": 1.5801138028366026e-05,
+      "loss": 0.353,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6484819734345351,
+      "grad_norm": 3.4517641067504883,
+      "learning_rate": 1.579487821325745e-05,
+      "loss": 0.4646,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6489563567362429,
+      "grad_norm": 2.4029881954193115,
+      "learning_rate": 1.5788614977489612e-05,
+      "loss": 0.3996,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6494307400379506,
+      "grad_norm": 2.305480718612671,
+      "learning_rate": 1.578234832475963e-05,
+      "loss": 0.3684,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6499051233396584,
+      "grad_norm": 2.9141790866851807,
+      "learning_rate": 1.5776078258766654e-05,
+      "loss": 0.4405,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6503795066413662,
+      "grad_norm": 2.265993356704712,
+      "learning_rate": 1.5769804783211837e-05,
+      "loss": 0.3726,
+      "step": 1371
+    },
+    {
+      "epoch": 0.650853889943074,
+      "grad_norm": 2.6375434398651123,
+      "learning_rate": 1.576352790179835e-05,
+      "loss": 0.3809,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6513282732447818,
+      "grad_norm": 2.3192057609558105,
+      "learning_rate": 1.5757247618231378e-05,
+      "loss": 0.4016,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6518026565464896,
+      "grad_norm": 2.4729015827178955,
+      "learning_rate": 1.5750963936218104e-05,
+      "loss": 0.4239,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6522770398481973,
+      "grad_norm": 2.1907765865325928,
+      "learning_rate": 1.574467685946773e-05,
+      "loss": 0.3452,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6527514231499051,
+      "grad_norm": 2.9924540519714355,
+      "learning_rate": 1.573838639169145e-05,
+      "loss": 0.5011,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6532258064516129,
+      "grad_norm": 2.7867164611816406,
+      "learning_rate": 1.5732092536602466e-05,
+      "loss": 0.4015,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6537001897533207,
+      "grad_norm": 1.9842467308044434,
+      "learning_rate": 1.572579529791598e-05,
+      "loss": 0.2994,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6541745730550285,
+      "grad_norm": 2.6319007873535156,
+      "learning_rate": 1.571949467934919e-05,
+      "loss": 0.4522,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6546489563567363,
+      "grad_norm": 2.6779913902282715,
+      "learning_rate": 1.5713190684621285e-05,
+      "loss": 0.3868,
+      "step": 1380
+    },
+    {
+      "epoch": 0.655123339658444,
+      "grad_norm": 2.5989339351654053,
+      "learning_rate": 1.5706883317453455e-05,
+      "loss": 0.3916,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6555977229601518,
+      "grad_norm": 2.8662238121032715,
+      "learning_rate": 1.5700572581568875e-05,
+      "loss": 0.4695,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6560721062618596,
+      "grad_norm": 2.4219655990600586,
+      "learning_rate": 1.569425848069271e-05,
+      "loss": 0.4111,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6565464895635673,
+      "grad_norm": 2.6834099292755127,
+      "learning_rate": 1.568794101855211e-05,
+      "loss": 0.4125,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6570208728652751,
+      "grad_norm": 2.470796585083008,
+      "learning_rate": 1.568162019887621e-05,
+      "loss": 0.4157,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6574952561669829,
+      "grad_norm": 2.3338913917541504,
+      "learning_rate": 1.567529602539613e-05,
+      "loss": 0.3847,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6579696394686907,
+      "grad_norm": 2.07490873336792,
+      "learning_rate": 1.5668968501844966e-05,
+      "loss": 0.3439,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6584440227703985,
+      "grad_norm": 2.4413368701934814,
+      "learning_rate": 1.5662637631957793e-05,
+      "loss": 0.4105,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6589184060721063,
+      "grad_norm": 2.3040366172790527,
+      "learning_rate": 1.565630341947166e-05,
+      "loss": 0.3204,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6593927893738141,
+      "grad_norm": 2.9284796714782715,
+      "learning_rate": 1.564996586812559e-05,
+      "loss": 0.4296,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6598671726755219,
+      "grad_norm": 2.683302164077759,
+      "learning_rate": 1.5643624981660573e-05,
+      "loss": 0.3965,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6603415559772297,
+      "grad_norm": 2.779085397720337,
+      "learning_rate": 1.563728076381958e-05,
+      "loss": 0.4836,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6608159392789373,
+      "grad_norm": 2.2810940742492676,
+      "learning_rate": 1.5630933218347536e-05,
+      "loss": 0.3765,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6612903225806451,
+      "grad_norm": 2.6218605041503906,
+      "learning_rate": 1.5624582348991327e-05,
+      "loss": 0.3954,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6617647058823529,
+      "grad_norm": 2.3885746002197266,
+      "learning_rate": 1.5618228159499823e-05,
+      "loss": 0.3817,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6622390891840607,
+      "grad_norm": 3.312493085861206,
+      "learning_rate": 1.5611870653623826e-05,
+      "loss": 0.4092,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6627134724857685,
+      "grad_norm": 3.737312078475952,
+      "learning_rate": 1.5605509835116115e-05,
+      "loss": 0.4746,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6631878557874763,
+      "grad_norm": 2.460756778717041,
+      "learning_rate": 1.5599145707731417e-05,
+      "loss": 0.415,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6636622390891841,
+      "grad_norm": 2.247875690460205,
+      "learning_rate": 1.5592778275226413e-05,
+      "loss": 0.3812,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6641366223908919,
+      "grad_norm": 2.569827079772949,
+      "learning_rate": 1.558640754135974e-05,
+      "loss": 0.4392,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6646110056925996,
+      "grad_norm": 2.2257590293884277,
+      "learning_rate": 1.558003350989197e-05,
+      "loss": 0.3763,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6650853889943074,
+      "grad_norm": 2.3620169162750244,
+      "learning_rate": 1.5573656184585643e-05,
+      "loss": 0.4499,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6655597722960152,
+      "grad_norm": 2.4123148918151855,
+      "learning_rate": 1.5567275569205216e-05,
+      "loss": 0.3642,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6660341555977229,
+      "grad_norm": 2.121957302093506,
+      "learning_rate": 1.556089166751712e-05,
+      "loss": 0.3786,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6665085388994307,
+      "grad_norm": 2.452988624572754,
+      "learning_rate": 1.555450448328969e-05,
+      "loss": 0.4058,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6669829222011385,
+      "grad_norm": 2.358628988265991,
+      "learning_rate": 1.554811402029323e-05,
+      "loss": 0.4026,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6674573055028463,
+      "grad_norm": 2.781123638153076,
+      "learning_rate": 1.554172028229997e-05,
+      "loss": 0.455,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6679316888045541,
+      "grad_norm": 3.405427932739258,
+      "learning_rate": 1.5535323273084062e-05,
+      "loss": 0.456,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6684060721062619,
+      "grad_norm": 2.7615528106689453,
+      "learning_rate": 1.55289229964216e-05,
+      "loss": 0.3644,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6688804554079696,
+      "grad_norm": 3.0365653038024902,
+      "learning_rate": 1.5522519456090604e-05,
+      "loss": 0.5515,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6693548387096774,
+      "grad_norm": 2.3187596797943115,
+      "learning_rate": 1.551611265587102e-05,
+      "loss": 0.3391,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6698292220113852,
+      "grad_norm": 2.320058584213257,
+      "learning_rate": 1.5509702599544723e-05,
+      "loss": 0.3337,
+      "step": 1412
+    },
+    {
+      "epoch": 0.670303605313093,
+      "grad_norm": 2.3811447620391846,
+      "learning_rate": 1.55032892908955e-05,
+      "loss": 0.3742,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6707779886148008,
+      "grad_norm": 2.2313289642333984,
+      "learning_rate": 1.5496872733709067e-05,
+      "loss": 0.3389,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6712523719165086,
+      "grad_norm": 2.9275193214416504,
+      "learning_rate": 1.5490452931773053e-05,
+      "loss": 0.4493,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6717267552182163,
+      "grad_norm": 2.3053927421569824,
+      "learning_rate": 1.5484029888877004e-05,
+      "loss": 0.3616,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6722011385199241,
+      "grad_norm": 2.178098201751709,
+      "learning_rate": 1.547760360881238e-05,
+      "loss": 0.3492,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6726755218216319,
+      "grad_norm": 2.574260950088501,
+      "learning_rate": 1.547117409537254e-05,
+      "loss": 0.4382,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6731499051233396,
+      "grad_norm": 2.7682528495788574,
+      "learning_rate": 1.546474135235278e-05,
+      "loss": 0.3529,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6736242884250474,
+      "grad_norm": 5.3996500968933105,
+      "learning_rate": 1.5458305383550275e-05,
+      "loss": 0.3958,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6740986717267552,
+      "grad_norm": 2.8296844959259033,
+      "learning_rate": 1.545186619276411e-05,
+      "loss": 0.4987,
+      "step": 1421
+    },
+    {
+      "epoch": 0.674573055028463,
+      "grad_norm": 2.638801097869873,
+      "learning_rate": 1.5445423783795283e-05,
+      "loss": 0.4888,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6750474383301708,
+      "grad_norm": 2.253319025039673,
+      "learning_rate": 1.5438978160446684e-05,
+      "loss": 0.3512,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6755218216318786,
+      "grad_norm": 2.3527333736419678,
+      "learning_rate": 1.543252932652309e-05,
+      "loss": 0.3054,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6759962049335864,
+      "grad_norm": 2.6959636211395264,
+      "learning_rate": 1.5426077285831195e-05,
+      "loss": 0.3844,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6764705882352942,
+      "grad_norm": 2.884432792663574,
+      "learning_rate": 1.5419622042179575e-05,
+      "loss": 0.4787,
+      "step": 1426
+    },
+    {
+      "epoch": 0.676944971537002,
+      "grad_norm": 3.0761313438415527,
+      "learning_rate": 1.541316359937869e-05,
+      "loss": 0.4422,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6774193548387096,
+      "grad_norm": 2.4010064601898193,
+      "learning_rate": 1.54067019612409e-05,
+      "loss": 0.4129,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6778937381404174,
+      "grad_norm": 2.568399429321289,
+      "learning_rate": 1.5400237131580443e-05,
+      "loss": 0.3751,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6783681214421252,
+      "grad_norm": 3.373952627182007,
+      "learning_rate": 1.539376911421344e-05,
+      "loss": 0.404,
+      "step": 1430
+    },
+    {
+      "epoch": 0.678842504743833,
+      "grad_norm": 2.3124148845672607,
+      "learning_rate": 1.5387297912957907e-05,
+      "loss": 0.3388,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6793168880455408,
+      "grad_norm": 2.5099921226501465,
+      "learning_rate": 1.5380823531633727e-05,
+      "loss": 0.3389,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6797912713472486,
+      "grad_norm": 2.23146915435791,
+      "learning_rate": 1.537434597406266e-05,
+      "loss": 0.3419,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6802656546489564,
+      "grad_norm": 2.684089183807373,
+      "learning_rate": 1.5367865244068346e-05,
+      "loss": 0.445,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6807400379506642,
+      "grad_norm": 2.252448320388794,
+      "learning_rate": 1.53613813454763e-05,
+      "loss": 0.3182,
+      "step": 1435
+    },
+    {
+      "epoch": 0.681214421252372,
+      "grad_norm": 2.7252156734466553,
+      "learning_rate": 1.5354894282113892e-05,
+      "loss": 0.4379,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6816888045540797,
+      "grad_norm": 2.7150869369506836,
+      "learning_rate": 1.5348404057810383e-05,
+      "loss": 0.405,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6821631878557874,
+      "grad_norm": 2.2281618118286133,
+      "learning_rate": 1.534191067639688e-05,
+      "loss": 0.3822,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6826375711574952,
+      "grad_norm": 2.0423622131347656,
+      "learning_rate": 1.5335414141706366e-05,
+      "loss": 0.3505,
+      "step": 1439
+    },
+    {
+      "epoch": 0.683111954459203,
+      "grad_norm": 2.7963783740997314,
+      "learning_rate": 1.5328914457573683e-05,
+      "loss": 0.4724,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6835863377609108,
+      "grad_norm": 2.374962568283081,
+      "learning_rate": 1.5322411627835526e-05,
+      "loss": 0.3721,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6840607210626186,
+      "grad_norm": 2.619636058807373,
+      "learning_rate": 1.531590565633045e-05,
+      "loss": 0.4549,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6845351043643264,
+      "grad_norm": 2.1424806118011475,
+      "learning_rate": 1.530939654689887e-05,
+      "loss": 0.3219,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6850094876660342,
+      "grad_norm": 2.061220407485962,
+      "learning_rate": 1.5302884303383046e-05,
+      "loss": 0.358,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6854838709677419,
+      "grad_norm": 3.31339955329895,
+      "learning_rate": 1.5296368929627097e-05,
+      "loss": 0.5536,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6859582542694497,
+      "grad_norm": 2.542325973510742,
+      "learning_rate": 1.528985042947697e-05,
+      "loss": 0.4265,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6864326375711575,
+      "grad_norm": 2.6769909858703613,
+      "learning_rate": 1.5283328806780488e-05,
+      "loss": 0.4561,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6869070208728653,
+      "grad_norm": 2.1639342308044434,
+      "learning_rate": 1.527680406538729e-05,
+      "loss": 0.3672,
+      "step": 1448
+    },
+    {
+      "epoch": 0.687381404174573,
+      "grad_norm": 2.8890748023986816,
+      "learning_rate": 1.5270276209148867e-05,
+      "loss": 0.3952,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6878557874762808,
+      "grad_norm": 2.939438819885254,
+      "learning_rate": 1.526374524191855e-05,
+      "loss": 0.4082,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6883301707779886,
+      "grad_norm": 2.6139535903930664,
+      "learning_rate": 1.52572111675515e-05,
+      "loss": 0.4213,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6888045540796964,
+      "grad_norm": 2.3408873081207275,
+      "learning_rate": 1.5250673989904728e-05,
+      "loss": 0.3499,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6892789373814042,
+      "grad_norm": 2.349148988723755,
+      "learning_rate": 1.524413371283705e-05,
+      "loss": 0.3571,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6897533206831119,
+      "grad_norm": 2.3099985122680664,
+      "learning_rate": 1.5237590340209139e-05,
+      "loss": 0.3955,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6902277039848197,
+      "grad_norm": 2.275421380996704,
+      "learning_rate": 1.5231043875883474e-05,
+      "loss": 0.3267,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6907020872865275,
+      "grad_norm": 2.5383238792419434,
+      "learning_rate": 1.5224494323724374e-05,
+      "loss": 0.355,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6911764705882353,
+      "grad_norm": 2.6643269062042236,
+      "learning_rate": 1.5217941687597976e-05,
+      "loss": 0.4296,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6916508538899431,
+      "grad_norm": 2.6603057384490967,
+      "learning_rate": 1.5211385971372233e-05,
+      "loss": 0.3736,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6921252371916509,
+      "grad_norm": 2.079129695892334,
+      "learning_rate": 1.5204827178916919e-05,
+      "loss": 0.3354,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6925996204933587,
+      "grad_norm": 2.422872304916382,
+      "learning_rate": 1.5198265314103626e-05,
+      "loss": 0.4841,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6930740037950665,
+      "grad_norm": 2.4735107421875,
+      "learning_rate": 1.5191700380805754e-05,
+      "loss": 0.3615,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6935483870967742,
+      "grad_norm": 2.0199763774871826,
+      "learning_rate": 1.5185132382898524e-05,
+      "loss": 0.2761,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6940227703984819,
+      "grad_norm": 2.425189733505249,
+      "learning_rate": 1.5178561324258956e-05,
+      "loss": 0.4204,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6944971537001897,
+      "grad_norm": 2.212663173675537,
+      "learning_rate": 1.5171987208765884e-05,
+      "loss": 0.35,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6949715370018975,
+      "grad_norm": 3.0007028579711914,
+      "learning_rate": 1.5165410040299942e-05,
+      "loss": 0.424,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6954459203036053,
+      "grad_norm": 2.555004835128784,
+      "learning_rate": 1.5158829822743567e-05,
+      "loss": 0.4113,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6959203036053131,
+      "grad_norm": 2.6518445014953613,
+      "learning_rate": 1.5152246559980994e-05,
+      "loss": 0.3866,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6963946869070209,
+      "grad_norm": 2.7971367835998535,
+      "learning_rate": 1.5145660255898262e-05,
+      "loss": 0.3921,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6968690702087287,
+      "grad_norm": 2.154005765914917,
+      "learning_rate": 1.5139070914383198e-05,
+      "loss": 0.334,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6973434535104365,
+      "grad_norm": 2.5318751335144043,
+      "learning_rate": 1.5132478539325428e-05,
+      "loss": 0.3787,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6978178368121443,
+      "grad_norm": 2.6897404193878174,
+      "learning_rate": 1.5125883134616363e-05,
+      "loss": 0.3783,
+      "step": 1471
+    },
+    {
+      "epoch": 0.698292220113852,
+      "grad_norm": 2.307647943496704,
+      "learning_rate": 1.5119284704149211e-05,
+      "loss": 0.3814,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6987666034155597,
+      "grad_norm": 2.4573044776916504,
+      "learning_rate": 1.5112683251818952e-05,
+      "loss": 0.42,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6992409867172675,
+      "grad_norm": 2.418071985244751,
+      "learning_rate": 1.5106078781522363e-05,
+      "loss": 0.3043,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6997153700189753,
+      "grad_norm": 2.5575947761535645,
+      "learning_rate": 1.509947129715799e-05,
+      "loss": 0.4268,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7001897533206831,
+      "grad_norm": 2.537424087524414,
+      "learning_rate": 1.5092860802626179e-05,
+      "loss": 0.4167,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7006641366223909,
+      "grad_norm": 2.9543752670288086,
+      "learning_rate": 1.5086247301829028e-05,
+      "loss": 0.407,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7011385199240987,
+      "grad_norm": 2.10909104347229,
+      "learning_rate": 1.507963079867043e-05,
+      "loss": 0.3587,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7016129032258065,
+      "grad_norm": 2.046889543533325,
+      "learning_rate": 1.5073011297056033e-05,
+      "loss": 0.3019,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7020872865275142,
+      "grad_norm": 2.1816864013671875,
+      "learning_rate": 1.5066388800893266e-05,
+      "loss": 0.349,
+      "step": 1480
+    },
+    {
+      "epoch": 0.702561669829222,
+      "grad_norm": 1.9779382944107056,
+      "learning_rate": 1.5059763314091326e-05,
+      "loss": 0.3138,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7030360531309298,
+      "grad_norm": 3.019711971282959,
+      "learning_rate": 1.505313484056117e-05,
+      "loss": 0.466,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7035104364326376,
+      "grad_norm": 2.450096607208252,
+      "learning_rate": 1.5046503384215521e-05,
+      "loss": 0.3534,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7039848197343453,
+      "grad_norm": 2.3313405513763428,
+      "learning_rate": 1.503986894896886e-05,
+      "loss": 0.4373,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7044592030360531,
+      "grad_norm": 2.4880897998809814,
+      "learning_rate": 1.5033231538737432e-05,
+      "loss": 0.4231,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7049335863377609,
+      "grad_norm": 2.491865873336792,
+      "learning_rate": 1.502659115743923e-05,
+      "loss": 0.4453,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7054079696394687,
+      "grad_norm": 2.641712188720703,
+      "learning_rate": 1.501994780899401e-05,
+      "loss": 0.4031,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 2.309870719909668,
+      "learning_rate": 1.5013301497323274e-05,
+      "loss": 0.3533,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7063567362428842,
+      "grad_norm": 2.5076959133148193,
+      "learning_rate": 1.5006652226350272e-05,
+      "loss": 0.3699,
+      "step": 1489
+    },
+    {
+      "epoch": 0.706831119544592,
+      "grad_norm": 2.0265212059020996,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.3459,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7073055028462998,
+      "grad_norm": 2.6799657344818115,
+      "learning_rate": 1.499334482219921e-05,
+      "loss": 0.4195,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7077798861480076,
+      "grad_norm": 2.5318443775177,
+      "learning_rate": 1.4986686696876381e-05,
+      "loss": 0.3655,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7082542694497154,
+      "grad_norm": 2.371875047683716,
+      "learning_rate": 1.498002562796174e-05,
+      "loss": 0.3385,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7087286527514232,
+      "grad_norm": 3.2066407203674316,
+      "learning_rate": 1.497336161938725e-05,
+      "loss": 0.3932,
+      "step": 1494
+    },
+    {
+      "epoch": 0.709203036053131,
+      "grad_norm": 2.56129789352417,
+      "learning_rate": 1.4966694675086611e-05,
+      "loss": 0.3833,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 1.934271216392517,
+      "learning_rate": 1.4960024798995252e-05,
+      "loss": 0.3025,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7101518026565465,
+      "grad_norm": 2.200591564178467,
+      "learning_rate": 1.4953351995050336e-05,
+      "loss": 0.3285,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7106261859582542,
+      "grad_norm": 2.4421536922454834,
+      "learning_rate": 1.4946676267190751e-05,
+      "loss": 0.3932,
+      "step": 1498
+    },
+    {
+      "epoch": 0.711100569259962,
+      "grad_norm": 2.842991590499878,
+      "learning_rate": 1.4939997619357116e-05,
+      "loss": 0.4339,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7115749525616698,
+      "grad_norm": 2.2917048931121826,
+      "learning_rate": 1.493331605549177e-05,
+      "loss": 0.3514,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7120493358633776,
+      "grad_norm": 2.323843002319336,
+      "learning_rate": 1.4926631579538775e-05,
+      "loss": 0.3419,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7125237191650854,
+      "grad_norm": 2.5014545917510986,
+      "learning_rate": 1.4919944195443908e-05,
+      "loss": 0.46,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7129981024667932,
+      "grad_norm": 2.414332389831543,
+      "learning_rate": 1.4913253907154666e-05,
+      "loss": 0.3509,
+      "step": 1503
+    },
+    {
+      "epoch": 0.713472485768501,
+      "grad_norm": 2.879897117614746,
+      "learning_rate": 1.4906560718620265e-05,
+      "loss": 0.318,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7139468690702088,
+      "grad_norm": 2.6578876972198486,
+      "learning_rate": 1.4899864633791623e-05,
+      "loss": 0.3298,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7144212523719166,
+      "grad_norm": 2.390251636505127,
+      "learning_rate": 1.4893165656621371e-05,
+      "loss": 0.3778,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7148956356736242,
+      "grad_norm": 3.521209478378296,
+      "learning_rate": 1.4886463791063854e-05,
+      "loss": 0.4522,
+      "step": 1507
+    },
+    {
+      "epoch": 0.715370018975332,
+      "grad_norm": 2.280897617340088,
+      "learning_rate": 1.4879759041075117e-05,
+      "loss": 0.3888,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7158444022770398,
+      "grad_norm": 2.56290340423584,
+      "learning_rate": 1.4873051410612905e-05,
+      "loss": 0.3578,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7163187855787476,
+      "grad_norm": 2.607135534286499,
+      "learning_rate": 1.486634090363666e-05,
+      "loss": 0.4305,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7167931688804554,
+      "grad_norm": 2.685330629348755,
+      "learning_rate": 1.4859627524107538e-05,
+      "loss": 0.4462,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7172675521821632,
+      "grad_norm": 2.119518280029297,
+      "learning_rate": 1.4852911275988379e-05,
+      "loss": 0.3338,
+      "step": 1512
+    },
+    {
+      "epoch": 0.717741935483871,
+      "grad_norm": 2.616091251373291,
+      "learning_rate": 1.4846192163243713e-05,
+      "loss": 0.387,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7182163187855788,
+      "grad_norm": 2.2375452518463135,
+      "learning_rate": 1.483947018983977e-05,
+      "loss": 0.3961,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7186907020872866,
+      "grad_norm": 2.1125147342681885,
+      "learning_rate": 1.483274535974446e-05,
+      "loss": 0.3652,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7191650853889943,
+      "grad_norm": 2.8368351459503174,
+      "learning_rate": 1.482601767692739e-05,
+      "loss": 0.4423,
+      "step": 1516
+    },
+    {
+      "epoch": 0.719639468690702,
+      "grad_norm": 2.951655864715576,
+      "learning_rate": 1.4819287145359837e-05,
+      "loss": 0.4042,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7201138519924098,
+      "grad_norm": 2.661517381668091,
+      "learning_rate": 1.481255376901477e-05,
+      "loss": 0.3976,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7205882352941176,
+      "grad_norm": 2.2923014163970947,
+      "learning_rate": 1.4805817551866839e-05,
+      "loss": 0.354,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7210626185958254,
+      "grad_norm": 2.025585174560547,
+      "learning_rate": 1.479907849789236e-05,
+      "loss": 0.3351,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7215370018975332,
+      "grad_norm": 2.3796160221099854,
+      "learning_rate": 1.4792336611069335e-05,
+      "loss": 0.4202,
+      "step": 1521
+    },
+    {
+      "epoch": 0.722011385199241,
+      "grad_norm": 2.3229618072509766,
+      "learning_rate": 1.478559189537743e-05,
+      "loss": 0.3969,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7224857685009488,
+      "grad_norm": 2.162994861602783,
+      "learning_rate": 1.4778844354797985e-05,
+      "loss": 0.302,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7229601518026565,
+      "grad_norm": 2.562293767929077,
+      "learning_rate": 1.4772093993314005e-05,
+      "loss": 0.3537,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7234345351043643,
+      "grad_norm": 2.323808431625366,
+      "learning_rate": 1.4765340814910163e-05,
+      "loss": 0.3961,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7239089184060721,
+      "grad_norm": 2.3642516136169434,
+      "learning_rate": 1.4758584823572792e-05,
+      "loss": 0.3491,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7243833017077799,
+      "grad_norm": 2.2448556423187256,
+      "learning_rate": 1.4751826023289889e-05,
+      "loss": 0.3941,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7248576850094877,
+      "grad_norm": 2.7379980087280273,
+      "learning_rate": 1.4745064418051107e-05,
+      "loss": 0.4637,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7253320683111955,
+      "grad_norm": 2.4983396530151367,
+      "learning_rate": 1.4738300011847752e-05,
+      "loss": 0.3884,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7258064516129032,
+      "grad_norm": 2.730942964553833,
+      "learning_rate": 1.4731532808672785e-05,
+      "loss": 0.4037,
+      "step": 1530
+    },
+    {
+      "epoch": 0.726280834914611,
+      "grad_norm": 2.594155788421631,
+      "learning_rate": 1.4724762812520825e-05,
+      "loss": 0.4064,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7267552182163188,
+      "grad_norm": 2.7857630252838135,
+      "learning_rate": 1.4717990027388129e-05,
+      "loss": 0.3294,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7272296015180265,
+      "grad_norm": 2.5247364044189453,
+      "learning_rate": 1.4711214457272611e-05,
+      "loss": 0.4019,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7277039848197343,
+      "grad_norm": 2.9784631729125977,
+      "learning_rate": 1.4704436106173813e-05,
+      "loss": 0.4528,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7281783681214421,
+      "grad_norm": 2.051332950592041,
+      "learning_rate": 1.4697654978092935e-05,
+      "loss": 0.3219,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7286527514231499,
+      "grad_norm": 2.4342894554138184,
+      "learning_rate": 1.4690871077032808e-05,
+      "loss": 0.4505,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7291271347248577,
+      "grad_norm": 2.270163059234619,
+      "learning_rate": 1.4684084406997903e-05,
+      "loss": 0.3858,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7296015180265655,
+      "grad_norm": 2.4693052768707275,
+      "learning_rate": 1.4677294971994325e-05,
+      "loss": 0.3749,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7300759013282733,
+      "grad_norm": 2.4099597930908203,
+      "learning_rate": 1.4670502776029804e-05,
+      "loss": 0.328,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7305502846299811,
+      "grad_norm": 3.063075304031372,
+      "learning_rate": 1.4663707823113717e-05,
+      "loss": 0.4566,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7310246679316889,
+      "grad_norm": 1.9556933641433716,
+      "learning_rate": 1.4656910117257049e-05,
+      "loss": 0.3263,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7314990512333965,
+      "grad_norm": 2.0928752422332764,
+      "learning_rate": 1.4650109662472422e-05,
+      "loss": 0.314,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7319734345351043,
+      "grad_norm": 2.1672606468200684,
+      "learning_rate": 1.4643306462774071e-05,
+      "loss": 0.2977,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7324478178368121,
+      "grad_norm": 2.769050121307373,
+      "learning_rate": 1.4636500522177868e-05,
+      "loss": 0.3666,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7329222011385199,
+      "grad_norm": 2.6004836559295654,
+      "learning_rate": 1.4629691844701288e-05,
+      "loss": 0.3975,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7333965844402277,
+      "grad_norm": 2.554579019546509,
+      "learning_rate": 1.462288043436342e-05,
+      "loss": 0.4653,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7338709677419355,
+      "grad_norm": 2.2475826740264893,
+      "learning_rate": 1.461606629518498e-05,
+      "loss": 0.4302,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7343453510436433,
+      "grad_norm": 2.2661335468292236,
+      "learning_rate": 1.460924943118828e-05,
+      "loss": 0.3663,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7348197343453511,
+      "grad_norm": 3.354585647583008,
+      "learning_rate": 1.4602429846397254e-05,
+      "loss": 0.385,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7352941176470589,
+      "grad_norm": 2.5863516330718994,
+      "learning_rate": 1.4595607544837435e-05,
+      "loss": 0.4031,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7357685009487666,
+      "grad_norm": 2.85939359664917,
+      "learning_rate": 1.4588782530535955e-05,
+      "loss": 0.4245,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7362428842504743,
+      "grad_norm": 2.234726667404175,
+      "learning_rate": 1.4581954807521555e-05,
+      "loss": 0.314,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7367172675521821,
+      "grad_norm": 2.442960262298584,
+      "learning_rate": 1.4575124379824574e-05,
+      "loss": 0.3825,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7371916508538899,
+      "grad_norm": 3.2431604862213135,
+      "learning_rate": 1.4568291251476944e-05,
+      "loss": 0.4371,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7376660341555977,
+      "grad_norm": 2.6331787109375,
+      "learning_rate": 1.4561455426512192e-05,
+      "loss": 0.4168,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7381404174573055,
+      "grad_norm": 2.5787758827209473,
+      "learning_rate": 1.455461690896544e-05,
+      "loss": 0.3975,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7386148007590133,
+      "grad_norm": 2.2691023349761963,
+      "learning_rate": 1.45477757028734e-05,
+      "loss": 0.3795,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7390891840607211,
+      "grad_norm": 2.3956029415130615,
+      "learning_rate": 1.4540931812274359e-05,
+      "loss": 0.393,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7395635673624289,
+      "grad_norm": 2.2934021949768066,
+      "learning_rate": 1.4534085241208206e-05,
+      "loss": 0.3465,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7400379506641366,
+      "grad_norm": 2.020655632019043,
+      "learning_rate": 1.4527235993716402e-05,
+      "loss": 0.2692,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7405123339658444,
+      "grad_norm": 2.299783945083618,
+      "learning_rate": 1.4520384073841991e-05,
+      "loss": 0.3394,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7409867172675522,
+      "grad_norm": 2.876800298690796,
+      "learning_rate": 1.4513529485629591e-05,
+      "loss": 0.3533,
+      "step": 1562
+    },
+    {
+      "epoch": 0.74146110056926,
+      "grad_norm": 2.770765542984009,
+      "learning_rate": 1.4506672233125398e-05,
+      "loss": 0.3855,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7419354838709677,
+      "grad_norm": 2.479526996612549,
+      "learning_rate": 1.4499812320377182e-05,
+      "loss": 0.4417,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7424098671726755,
+      "grad_norm": 2.534569025039673,
+      "learning_rate": 1.4492949751434282e-05,
+      "loss": 0.346,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7428842504743833,
+      "grad_norm": 2.4352223873138428,
+      "learning_rate": 1.4486084530347604e-05,
+      "loss": 0.3262,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7433586337760911,
+      "grad_norm": 2.7315893173217773,
+      "learning_rate": 1.4479216661169618e-05,
+      "loss": 0.3896,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7438330170777988,
+      "grad_norm": 2.1865360736846924,
+      "learning_rate": 1.4472346147954356e-05,
+      "loss": 0.4102,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7443074003795066,
+      "grad_norm": 1.9840688705444336,
+      "learning_rate": 1.4465472994757429e-05,
+      "loss": 0.2967,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7447817836812144,
+      "grad_norm": 2.338481903076172,
+      "learning_rate": 1.4458597205635973e-05,
+      "loss": 0.3746,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7452561669829222,
+      "grad_norm": 2.240323543548584,
+      "learning_rate": 1.445171878464871e-05,
+      "loss": 0.3847,
+      "step": 1571
+    },
+    {
+      "epoch": 0.74573055028463,
+      "grad_norm": 2.151249408721924,
+      "learning_rate": 1.4444837735855896e-05,
+      "loss": 0.3707,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7462049335863378,
+      "grad_norm": 2.2230353355407715,
+      "learning_rate": 1.4437954063319352e-05,
+      "loss": 0.3636,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7466793168880456,
+      "grad_norm": 2.0856575965881348,
+      "learning_rate": 1.4431067771102443e-05,
+      "loss": 0.2867,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7471537001897534,
+      "grad_norm": 2.0492308139801025,
+      "learning_rate": 1.442417886327007e-05,
+      "loss": 0.3734,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7476280834914611,
+      "grad_norm": 2.09089994430542,
+      "learning_rate": 1.4417287343888698e-05,
+      "loss": 0.3346,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7481024667931688,
+      "grad_norm": 2.4338037967681885,
+      "learning_rate": 1.4410393217026317e-05,
+      "loss": 0.3388,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7485768500948766,
+      "grad_norm": 2.734534740447998,
+      "learning_rate": 1.4403496486752465e-05,
+      "loss": 0.4177,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7490512333965844,
+      "grad_norm": 3.0047781467437744,
+      "learning_rate": 1.4396597157138205e-05,
+      "loss": 0.3927,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7495256166982922,
+      "grad_norm": 2.755629062652588,
+      "learning_rate": 1.4389695232256151e-05,
+      "loss": 0.4591,
+      "step": 1580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1415135860443115,
+      "learning_rate": 1.4382790716180446e-05,
+      "loss": 0.2856,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7504743833017078,
+      "grad_norm": 2.5595362186431885,
+      "learning_rate": 1.4375883612986744e-05,
+      "loss": 0.3146,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7509487666034156,
+      "grad_norm": 3.115945339202881,
+      "learning_rate": 1.4368973926752248e-05,
+      "loss": 0.4559,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7514231499051234,
+      "grad_norm": 2.1239240169525146,
+      "learning_rate": 1.4362061661555675e-05,
+      "loss": 0.3491,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7518975332068312,
+      "grad_norm": 2.0848984718322754,
+      "learning_rate": 1.435514682147727e-05,
+      "loss": 0.3327,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7523719165085389,
+      "grad_norm": 2.5240261554718018,
+      "learning_rate": 1.4348229410598791e-05,
+      "loss": 0.3467,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7528462998102466,
+      "grad_norm": 2.0627050399780273,
+      "learning_rate": 1.4341309433003518e-05,
+      "loss": 0.3109,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7533206831119544,
+      "grad_norm": 2.349839210510254,
+      "learning_rate": 1.4334386892776246e-05,
+      "loss": 0.3552,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7537950664136622,
+      "grad_norm": 3.1062917709350586,
+      "learning_rate": 1.4327461794003284e-05,
+      "loss": 0.4663,
+      "step": 1589
+    },
+    {
+      "epoch": 0.75426944971537,
+      "grad_norm": 2.2856526374816895,
+      "learning_rate": 1.4320534140772447e-05,
+      "loss": 0.3425,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7547438330170778,
+      "grad_norm": 2.193957567214966,
+      "learning_rate": 1.4313603937173058e-05,
+      "loss": 0.3448,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7552182163187856,
+      "grad_norm": 2.4013428688049316,
+      "learning_rate": 1.4306671187295948e-05,
+      "loss": 0.3795,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7556925996204934,
+      "grad_norm": 2.1507675647735596,
+      "learning_rate": 1.4299735895233457e-05,
+      "loss": 0.307,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7561669829222012,
+      "grad_norm": 2.046177864074707,
+      "learning_rate": 1.4292798065079413e-05,
+      "loss": 0.3221,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7566413662239089,
+      "grad_norm": 1.7904236316680908,
+      "learning_rate": 1.428585770092915e-05,
+      "loss": 0.3009,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7571157495256167,
+      "grad_norm": 2.811075210571289,
+      "learning_rate": 1.4278914806879494e-05,
+      "loss": 0.4872,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7575901328273245,
+      "grad_norm": 2.26833176612854,
+      "learning_rate": 1.4271969387028773e-05,
+      "loss": 0.3738,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7580645161290323,
+      "grad_norm": 2.7194955348968506,
+      "learning_rate": 1.4265021445476794e-05,
+      "loss": 0.4348,
+      "step": 1598
+    },
+    {
+      "epoch": 0.75853889943074,
+      "grad_norm": 2.353123664855957,
+      "learning_rate": 1.4258070986324859e-05,
+      "loss": 0.3861,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7590132827324478,
+      "grad_norm": 4.482807636260986,
+      "learning_rate": 1.4251118013675758e-05,
+      "loss": 0.3265,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7594876660341556,
+      "grad_norm": 2.567732810974121,
+      "learning_rate": 1.424416253163376e-05,
+      "loss": 0.4312,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7599620493358634,
+      "grad_norm": 2.412712335586548,
+      "learning_rate": 1.4237204544304616e-05,
+      "loss": 0.4085,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7604364326375711,
+      "grad_norm": 2.1643178462982178,
+      "learning_rate": 1.423024405579556e-05,
+      "loss": 0.3099,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7609108159392789,
+      "grad_norm": 2.2639448642730713,
+      "learning_rate": 1.4223281070215297e-05,
+      "loss": 0.3819,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7613851992409867,
+      "grad_norm": 2.341059446334839,
+      "learning_rate": 1.421631559167401e-05,
+      "loss": 0.4073,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7618595825426945,
+      "grad_norm": 2.276641845703125,
+      "learning_rate": 1.4209347624283352e-05,
+      "loss": 0.4333,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7623339658444023,
+      "grad_norm": 2.5862913131713867,
+      "learning_rate": 1.4202377172156443e-05,
+      "loss": 0.3642,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7628083491461101,
+      "grad_norm": 2.1797101497650146,
+      "learning_rate": 1.4195404239407873e-05,
+      "loss": 0.35,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7632827324478179,
+      "grad_norm": 2.534806251525879,
+      "learning_rate": 1.4188428830153698e-05,
+      "loss": 0.4342,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7637571157495257,
+      "grad_norm": 2.4096391201019287,
+      "learning_rate": 1.4181450948511431e-05,
+      "loss": 0.3694,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7642314990512334,
+      "grad_norm": 2.1816611289978027,
+      "learning_rate": 1.4174470598600048e-05,
+      "loss": 0.3174,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7647058823529411,
+      "grad_norm": 2.178921699523926,
+      "learning_rate": 1.4167487784539973e-05,
+      "loss": 0.3492,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7651802656546489,
+      "grad_norm": 2.4309422969818115,
+      "learning_rate": 1.4160502510453103e-05,
+      "loss": 0.3701,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7656546489563567,
+      "grad_norm": 2.0590929985046387,
+      "learning_rate": 1.4153514780462767e-05,
+      "loss": 0.305,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7661290322580645,
+      "grad_norm": 2.736320972442627,
+      "learning_rate": 1.4146524598693758e-05,
+      "loss": 0.4535,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7666034155597723,
+      "grad_norm": 2.418962001800537,
+      "learning_rate": 1.4139531969272313e-05,
+      "loss": 0.3975,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7670777988614801,
+      "grad_norm": 2.187948703765869,
+      "learning_rate": 1.4132536896326105e-05,
+      "loss": 0.3178,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7675521821631879,
+      "grad_norm": 2.1725316047668457,
+      "learning_rate": 1.4125539383984264e-05,
+      "loss": 0.3138,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7680265654648957,
+      "grad_norm": 2.384047508239746,
+      "learning_rate": 1.4118539436377345e-05,
+      "loss": 0.3081,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7685009487666035,
+      "grad_norm": 2.3366310596466064,
+      "learning_rate": 1.4111537057637353e-05,
+      "loss": 0.335,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7689753320683111,
+      "grad_norm": 2.152684211730957,
+      "learning_rate": 1.4104532251897722e-05,
+      "loss": 0.3145,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7694497153700189,
+      "grad_norm": 3.2571208477020264,
+      "learning_rate": 1.409752502329332e-05,
+      "loss": 0.297,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7699240986717267,
+      "grad_norm": 2.3170127868652344,
+      "learning_rate": 1.409051537596044e-05,
+      "loss": 0.3196,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7703984819734345,
+      "grad_norm": 2.442521810531616,
+      "learning_rate": 1.4083503314036813e-05,
+      "loss": 0.318,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7708728652751423,
+      "grad_norm": 3.705756902694702,
+      "learning_rate": 1.4076488841661588e-05,
+      "loss": 0.3793,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7713472485768501,
+      "grad_norm": 2.7555148601531982,
+      "learning_rate": 1.406947196297534e-05,
+      "loss": 0.3927,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7718216318785579,
+      "grad_norm": 2.6326348781585693,
+      "learning_rate": 1.4062452682120056e-05,
+      "loss": 0.4274,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7722960151802657,
+      "grad_norm": 2.276700258255005,
+      "learning_rate": 1.4055431003239156e-05,
+      "loss": 0.3913,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7727703984819735,
+      "grad_norm": 2.765024423599243,
+      "learning_rate": 1.4048406930477465e-05,
+      "loss": 0.4786,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7732447817836812,
+      "grad_norm": 2.5013036727905273,
+      "learning_rate": 1.4041380467981225e-05,
+      "loss": 0.4325,
+      "step": 1630
+    },
+    {
+      "epoch": 0.773719165085389,
+      "grad_norm": 2.752293348312378,
+      "learning_rate": 1.4034351619898088e-05,
+      "loss": 0.3946,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 2.66115665435791,
+      "learning_rate": 1.4027320390377106e-05,
+      "loss": 0.4486,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7746679316888045,
+      "grad_norm": 2.023422956466675,
+      "learning_rate": 1.4020286783568753e-05,
+      "loss": 0.3411,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7751423149905123,
+      "grad_norm": 2.1489474773406982,
+      "learning_rate": 1.4013250803624894e-05,
+      "loss": 0.3351,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7756166982922201,
+      "grad_norm": 2.9013800621032715,
+      "learning_rate": 1.4006212454698798e-05,
+      "loss": 0.4143,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7760910815939279,
+      "grad_norm": 2.3680756092071533,
+      "learning_rate": 1.3999171740945132e-05,
+      "loss": 0.3711,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7765654648956357,
+      "grad_norm": 2.0781219005584717,
+      "learning_rate": 1.3992128666519961e-05,
+      "loss": 0.3542,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7770398481973435,
+      "grad_norm": 1.984551191329956,
+      "learning_rate": 1.3985083235580743e-05,
+      "loss": 0.315,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7775142314990512,
+      "grad_norm": 1.8971257209777832,
+      "learning_rate": 1.3978035452286325e-05,
+      "loss": 0.2914,
+      "step": 1639
+    },
+    {
+      "epoch": 0.777988614800759,
+      "grad_norm": 2.8102481365203857,
+      "learning_rate": 1.3970985320796943e-05,
+      "loss": 0.3875,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7784629981024668,
+      "grad_norm": 2.2260053157806396,
+      "learning_rate": 1.396393284527422e-05,
+      "loss": 0.3428,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7789373814041746,
+      "grad_norm": 2.3798251152038574,
+      "learning_rate": 1.3956878029881167e-05,
+      "loss": 0.3244,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7794117647058824,
+      "grad_norm": 2.634979009628296,
+      "learning_rate": 1.3949820878782166e-05,
+      "loss": 0.317,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7798861480075902,
+      "grad_norm": 2.1837527751922607,
+      "learning_rate": 1.3942761396142982e-05,
+      "loss": 0.3588,
+      "step": 1644
+    },
+    {
+      "epoch": 0.780360531309298,
+      "grad_norm": 2.22882080078125,
+      "learning_rate": 1.3935699586130767e-05,
+      "loss": 0.3373,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7808349146110057,
+      "grad_norm": 2.3200697898864746,
+      "learning_rate": 1.3928635452914028e-05,
+      "loss": 0.3661,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7813092979127134,
+      "grad_norm": 2.158055543899536,
+      "learning_rate": 1.3921569000662658e-05,
+      "loss": 0.3346,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7817836812144212,
+      "grad_norm": 2.5481321811676025,
+      "learning_rate": 1.3914500233547909e-05,
+      "loss": 0.3935,
+      "step": 1648
+    },
+    {
+      "epoch": 0.782258064516129,
+      "grad_norm": 2.6886203289031982,
+      "learning_rate": 1.3907429155742414e-05,
+      "loss": 0.4086,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7827324478178368,
+      "grad_norm": 2.635003089904785,
+      "learning_rate": 1.3900355771420151e-05,
+      "loss": 0.4449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7832068311195446,
+      "grad_norm": 2.5500478744506836,
+      "learning_rate": 1.3893280084756468e-05,
+      "loss": 0.3725,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7836812144212524,
+      "grad_norm": 2.0035715103149414,
+      "learning_rate": 1.3886202099928083e-05,
+      "loss": 0.3206,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7841555977229602,
+      "grad_norm": 2.1694624423980713,
+      "learning_rate": 1.387912182111305e-05,
+      "loss": 0.3578,
+      "step": 1653
+    },
+    {
+      "epoch": 0.784629981024668,
+      "grad_norm": 2.505841016769409,
+      "learning_rate": 1.3872039252490796e-05,
+      "loss": 0.395,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7851043643263758,
+      "grad_norm": 1.9906131029129028,
+      "learning_rate": 1.386495439824208e-05,
+      "loss": 0.3343,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7855787476280834,
+      "grad_norm": 2.116349458694458,
+      "learning_rate": 1.3857867262549032e-05,
+      "loss": 0.3224,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7860531309297912,
+      "grad_norm": 2.3191020488739014,
+      "learning_rate": 1.3850777849595114e-05,
+      "loss": 0.4305,
+      "step": 1657
+    },
+    {
+      "epoch": 0.786527514231499,
+      "grad_norm": 2.867265224456787,
+      "learning_rate": 1.3843686163565138e-05,
+      "loss": 0.4917,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7870018975332068,
+      "grad_norm": 2.6565215587615967,
+      "learning_rate": 1.3836592208645252e-05,
+      "loss": 0.3741,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7874762808349146,
+      "grad_norm": 2.156074047088623,
+      "learning_rate": 1.382949598902295e-05,
+      "loss": 0.352,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7879506641366224,
+      "grad_norm": 2.667464017868042,
+      "learning_rate": 1.3822397508887063e-05,
+      "loss": 0.4249,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7884250474383302,
+      "grad_norm": 2.4767210483551025,
+      "learning_rate": 1.381529677242775e-05,
+      "loss": 0.3748,
+      "step": 1662
+    },
+    {
+      "epoch": 0.788899430740038,
+      "grad_norm": 2.4702491760253906,
+      "learning_rate": 1.38081937838365e-05,
+      "loss": 0.3526,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7893738140417458,
+      "grad_norm": 2.1653172969818115,
+      "learning_rate": 1.3801088547306149e-05,
+      "loss": 0.393,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7898481973434535,
+      "grad_norm": 2.3579320907592773,
+      "learning_rate": 1.379398106703084e-05,
+      "loss": 0.4005,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7903225806451613,
+      "grad_norm": 1.9658737182617188,
+      "learning_rate": 1.3786871347206053e-05,
+      "loss": 0.2901,
+      "step": 1666
+    },
+    {
+      "epoch": 0.790796963946869,
+      "grad_norm": 2.903672695159912,
+      "learning_rate": 1.377975939202858e-05,
+      "loss": 0.4421,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7912713472485768,
+      "grad_norm": 3.15793514251709,
+      "learning_rate": 1.377264520569654e-05,
+      "loss": 0.3416,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7917457305502846,
+      "grad_norm": 2.4560675621032715,
+      "learning_rate": 1.3765528792409368e-05,
+      "loss": 0.3839,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7922201138519924,
+      "grad_norm": 2.24422550201416,
+      "learning_rate": 1.3758410156367812e-05,
+      "loss": 0.33,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7926944971537002,
+      "grad_norm": 2.6136462688446045,
+      "learning_rate": 1.375128930177393e-05,
+      "loss": 0.3796,
+      "step": 1671
+    },
+    {
+      "epoch": 0.793168880455408,
+      "grad_norm": 2.220956325531006,
+      "learning_rate": 1.3744166232831093e-05,
+      "loss": 0.343,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7936432637571158,
+      "grad_norm": 2.2082831859588623,
+      "learning_rate": 1.3737040953743981e-05,
+      "loss": 0.3344,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7941176470588235,
+      "grad_norm": 2.327711582183838,
+      "learning_rate": 1.3729913468718574e-05,
+      "loss": 0.3889,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7945920303605313,
+      "grad_norm": 2.5623528957366943,
+      "learning_rate": 1.3722783781962155e-05,
+      "loss": 0.3582,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7950664136622391,
+      "grad_norm": 2.1435022354125977,
+      "learning_rate": 1.3715651897683306e-05,
+      "loss": 0.3027,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7955407969639469,
+      "grad_norm": 2.650327205657959,
+      "learning_rate": 1.3708517820091912e-05,
+      "loss": 0.3932,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7960151802656547,
+      "grad_norm": 2.224015235900879,
+      "learning_rate": 1.3701381553399147e-05,
+      "loss": 0.3401,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7964895635673624,
+      "grad_norm": 2.3124001026153564,
+      "learning_rate": 1.3694243101817475e-05,
+      "loss": 0.3646,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7969639468690702,
+      "grad_norm": 2.3588428497314453,
+      "learning_rate": 1.3687102469560656e-05,
+      "loss": 0.3372,
+      "step": 1680
+    },
+    {
+      "epoch": 0.797438330170778,
+      "grad_norm": 2.7870490550994873,
+      "learning_rate": 1.3679959660843736e-05,
+      "loss": 0.47,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7979127134724858,
+      "grad_norm": 2.235140562057495,
+      "learning_rate": 1.3672814679883044e-05,
+      "loss": 0.3261,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7983870967741935,
+      "grad_norm": 2.4782466888427734,
+      "learning_rate": 1.3665667530896189e-05,
+      "loss": 0.4113,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7988614800759013,
+      "grad_norm": 2.6865344047546387,
+      "learning_rate": 1.3658518218102064e-05,
+      "loss": 0.4226,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7993358633776091,
+      "grad_norm": 2.3782565593719482,
+      "learning_rate": 1.3651366745720837e-05,
+      "loss": 0.3715,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7998102466793169,
+      "grad_norm": 2.305004119873047,
+      "learning_rate": 1.3644213117973954e-05,
+      "loss": 0.4011,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8002846299810247,
+      "grad_norm": 2.5552866458892822,
+      "learning_rate": 1.3637057339084125e-05,
+      "loss": 0.3444,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8007590132827325,
+      "grad_norm": 2.5058846473693848,
+      "learning_rate": 1.3629899413275342e-05,
+      "loss": 0.3723,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8012333965844403,
+      "grad_norm": 2.0107834339141846,
+      "learning_rate": 1.3622739344772853e-05,
+      "loss": 0.2992,
+      "step": 1689
+    },
+    {
+      "epoch": 0.801707779886148,
+      "grad_norm": 1.9664660692214966,
+      "learning_rate": 1.361557713780318e-05,
+      "loss": 0.3447,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8021821631878557,
+      "grad_norm": 2.476154088973999,
+      "learning_rate": 1.3608412796594096e-05,
+      "loss": 0.4195,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8026565464895635,
+      "grad_norm": 2.669809103012085,
+      "learning_rate": 1.360124632537465e-05,
+      "loss": 0.4169,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8031309297912713,
+      "grad_norm": 2.68930983543396,
+      "learning_rate": 1.3594077728375129e-05,
+      "loss": 0.418,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8036053130929791,
+      "grad_norm": 2.4470651149749756,
+      "learning_rate": 1.3586907009827093e-05,
+      "loss": 0.4022,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8040796963946869,
+      "grad_norm": 2.488619565963745,
+      "learning_rate": 1.3579734173963343e-05,
+      "loss": 0.3691,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8045540796963947,
+      "grad_norm": 2.0860743522644043,
+      "learning_rate": 1.3572559225017932e-05,
+      "loss": 0.3586,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8050284629981025,
+      "grad_norm": 2.9916765689849854,
+      "learning_rate": 1.3565382167226162e-05,
+      "loss": 0.3821,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8055028462998103,
+      "grad_norm": 2.3840274810791016,
+      "learning_rate": 1.3558203004824581e-05,
+      "loss": 0.3548,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8059772296015181,
+      "grad_norm": 5.072951316833496,
+      "learning_rate": 1.3551021742050974e-05,
+      "loss": 0.3889,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8064516129032258,
+      "grad_norm": 1.943625569343567,
+      "learning_rate": 1.3543838383144374e-05,
+      "loss": 0.3432,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8069259962049335,
+      "grad_norm": 1.8006517887115479,
+      "learning_rate": 1.3536652932345043e-05,
+      "loss": 0.2821,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8074003795066413,
+      "grad_norm": 2.0171427726745605,
+      "learning_rate": 1.3529465393894485e-05,
+      "loss": 0.3287,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8078747628083491,
+      "grad_norm": 2.133181095123291,
+      "learning_rate": 1.3522275772035426e-05,
+      "loss": 0.3074,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8083491461100569,
+      "grad_norm": 1.9669580459594727,
+      "learning_rate": 1.3515084071011837e-05,
+      "loss": 0.3211,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8088235294117647,
+      "grad_norm": 2.351128101348877,
+      "learning_rate": 1.3507890295068902e-05,
+      "loss": 0.3666,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8092979127134725,
+      "grad_norm": 2.2970938682556152,
+      "learning_rate": 1.3500694448453038e-05,
+      "loss": 0.3401,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8097722960151803,
+      "grad_norm": 2.5522122383117676,
+      "learning_rate": 1.3493496535411885e-05,
+      "loss": 0.3661,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8102466793168881,
+      "grad_norm": 3.098694324493408,
+      "learning_rate": 1.3486296560194292e-05,
+      "loss": 0.3709,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8107210626185958,
+      "grad_norm": 2.7200305461883545,
+      "learning_rate": 1.347909452705034e-05,
+      "loss": 0.4361,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8111954459203036,
+      "grad_norm": 2.124351978302002,
+      "learning_rate": 1.3471890440231319e-05,
+      "loss": 0.411,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8116698292220114,
+      "grad_norm": 2.3741860389709473,
+      "learning_rate": 1.3464684303989723e-05,
+      "loss": 0.36,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8121442125237192,
+      "grad_norm": 2.1634681224823,
+      "learning_rate": 1.3457476122579268e-05,
+      "loss": 0.3456,
+      "step": 1712
+    },
+    {
+      "epoch": 0.812618595825427,
+      "grad_norm": 2.8045156002044678,
+      "learning_rate": 1.3450265900254868e-05,
+      "loss": 0.472,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8130929791271347,
+      "grad_norm": 3.158280372619629,
+      "learning_rate": 1.3443053641272656e-05,
+      "loss": 0.4005,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8135673624288425,
+      "grad_norm": 2.369152784347534,
+      "learning_rate": 1.3435839349889945e-05,
+      "loss": 0.3169,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8140417457305503,
+      "grad_norm": 2.4343760013580322,
+      "learning_rate": 1.3428623030365267e-05,
+      "loss": 0.3438,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8145161290322581,
+      "grad_norm": 2.882737636566162,
+      "learning_rate": 1.342140468695834e-05,
+      "loss": 0.3749,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8149905123339658,
+      "grad_norm": 2.584279775619507,
+      "learning_rate": 1.3414184323930082e-05,
+      "loss": 0.3517,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8154648956356736,
+      "grad_norm": 2.2734382152557373,
+      "learning_rate": 1.3406961945542605e-05,
+      "loss": 0.2909,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8159392789373814,
+      "grad_norm": 2.0772268772125244,
+      "learning_rate": 1.3399737556059203e-05,
+      "loss": 0.3331,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8164136622390892,
+      "grad_norm": 2.1270837783813477,
+      "learning_rate": 1.3392511159744364e-05,
+      "loss": 0.286,
+      "step": 1721
+    },
+    {
+      "epoch": 0.816888045540797,
+      "grad_norm": 2.7782678604125977,
+      "learning_rate": 1.3385282760863758e-05,
+      "loss": 0.3716,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8173624288425048,
+      "grad_norm": 2.4752237796783447,
+      "learning_rate": 1.3378052363684238e-05,
+      "loss": 0.3222,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8178368121442126,
+      "grad_norm": 2.3941702842712402,
+      "learning_rate": 1.3370819972473832e-05,
+      "loss": 0.3384,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8183111954459203,
+      "grad_norm": 2.5002641677856445,
+      "learning_rate": 1.3363585591501751e-05,
+      "loss": 0.3646,
+      "step": 1725
+    },
+    {
+      "epoch": 0.818785578747628,
+      "grad_norm": 2.297147750854492,
+      "learning_rate": 1.3356349225038384e-05,
+      "loss": 0.3964,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8192599620493358,
+      "grad_norm": 2.366065740585327,
+      "learning_rate": 1.334911087735528e-05,
+      "loss": 0.3453,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8197343453510436,
+      "grad_norm": 2.4925296306610107,
+      "learning_rate": 1.3341870552725166e-05,
+      "loss": 0.3305,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8202087286527514,
+      "grad_norm": 2.4606728553771973,
+      "learning_rate": 1.3334628255421932e-05,
+      "loss": 0.3727,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8206831119544592,
+      "grad_norm": 2.2296173572540283,
+      "learning_rate": 1.3327383989720639e-05,
+      "loss": 0.3335,
+      "step": 1730
+    },
+    {
+      "epoch": 0.821157495256167,
+      "grad_norm": 2.3122830390930176,
+      "learning_rate": 1.3320137759897505e-05,
+      "loss": 0.3168,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8216318785578748,
+      "grad_norm": 2.9335269927978516,
+      "learning_rate": 1.3312889570229901e-05,
+      "loss": 0.391,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8221062618595826,
+      "grad_norm": 1.8690437078475952,
+      "learning_rate": 1.3305639424996369e-05,
+      "loss": 0.3191,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8225806451612904,
+      "grad_norm": 2.195049524307251,
+      "learning_rate": 1.3298387328476594e-05,
+      "loss": 0.3354,
+      "step": 1734
+    },
+    {
+      "epoch": 0.823055028462998,
+      "grad_norm": 2.337287187576294,
+      "learning_rate": 1.3291133284951418e-05,
+      "loss": 0.3784,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 2.4185826778411865,
+      "learning_rate": 1.328387729870283e-05,
+      "loss": 0.3549,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8240037950664136,
+      "grad_norm": 2.838695526123047,
+      "learning_rate": 1.3276619374013965e-05,
+      "loss": 0.4295,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8244781783681214,
+      "grad_norm": 2.432915687561035,
+      "learning_rate": 1.3269359515169113e-05,
+      "loss": 0.2935,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8249525616698292,
+      "grad_norm": 3.295152187347412,
+      "learning_rate": 1.326209772645369e-05,
+      "loss": 0.473,
+      "step": 1739
+    },
+    {
+      "epoch": 0.825426944971537,
+      "grad_norm": 3.3393542766571045,
+      "learning_rate": 1.3254834012154259e-05,
+      "loss": 0.2849,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8259013282732448,
+      "grad_norm": 2.1521685123443604,
+      "learning_rate": 1.3247568376558515e-05,
+      "loss": 0.3756,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8263757115749526,
+      "grad_norm": 2.7531650066375732,
+      "learning_rate": 1.3240300823955297e-05,
+      "loss": 0.3388,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8268500948766604,
+      "grad_norm": 2.021456241607666,
+      "learning_rate": 1.3233031358634566e-05,
+      "loss": 0.3245,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8273244781783681,
+      "grad_norm": 2.3515677452087402,
+      "learning_rate": 1.3225759984887416e-05,
+      "loss": 0.412,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8277988614800759,
+      "grad_norm": 1.9360193014144897,
+      "learning_rate": 1.3218486707006069e-05,
+      "loss": 0.3299,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8282732447817837,
+      "grad_norm": 1.9744075536727905,
+      "learning_rate": 1.3211211529283867e-05,
+      "loss": 0.3216,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8287476280834914,
+      "grad_norm": 2.255706548690796,
+      "learning_rate": 1.3203934456015275e-05,
+      "loss": 0.3296,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8292220113851992,
+      "grad_norm": 2.0415632724761963,
+      "learning_rate": 1.3196655491495877e-05,
+      "loss": 0.2996,
+      "step": 1748
+    },
+    {
+      "epoch": 0.829696394686907,
+      "grad_norm": 2.2202017307281494,
+      "learning_rate": 1.3189374640022372e-05,
+      "loss": 0.3678,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8301707779886148,
+      "grad_norm": 2.691748857498169,
+      "learning_rate": 1.3182091905892581e-05,
+      "loss": 0.4671,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8306451612903226,
+      "grad_norm": 2.277745008468628,
+      "learning_rate": 1.3174807293405427e-05,
+      "loss": 0.3606,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8311195445920304,
+      "grad_norm": 2.287468910217285,
+      "learning_rate": 1.3167520806860943e-05,
+      "loss": 0.4041,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8315939278937381,
+      "grad_norm": 2.397675037384033,
+      "learning_rate": 1.3160232450560268e-05,
+      "loss": 0.4329,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8320683111954459,
+      "grad_norm": 2.1399128437042236,
+      "learning_rate": 1.3152942228805651e-05,
+      "loss": 0.3077,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8325426944971537,
+      "grad_norm": 2.231016159057617,
+      "learning_rate": 1.314565014590044e-05,
+      "loss": 0.3496,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8330170777988615,
+      "grad_norm": 1.9128320217132568,
+      "learning_rate": 1.3138356206149069e-05,
+      "loss": 0.2968,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8334914611005693,
+      "grad_norm": 2.082197666168213,
+      "learning_rate": 1.3131060413857087e-05,
+      "loss": 0.323,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8339658444022771,
+      "grad_norm": 2.4280385971069336,
+      "learning_rate": 1.3123762773331127e-05,
+      "loss": 0.3908,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8344402277039848,
+      "grad_norm": 2.383655071258545,
+      "learning_rate": 1.3116463288878914e-05,
+      "loss": 0.3721,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8349146110056926,
+      "grad_norm": 2.2262542247772217,
+      "learning_rate": 1.3109161964809256e-05,
+      "loss": 0.368,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8353889943074004,
+      "grad_norm": 2.4055254459381104,
+      "learning_rate": 1.310185880543206e-05,
+      "loss": 0.3321,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8358633776091081,
+      "grad_norm": 2.6039791107177734,
+      "learning_rate": 1.3094553815058304e-05,
+      "loss": 0.3922,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8363377609108159,
+      "grad_norm": 1.9806139469146729,
+      "learning_rate": 1.3087246998000054e-05,
+      "loss": 0.3628,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8368121442125237,
+      "grad_norm": 2.162916660308838,
+      "learning_rate": 1.307993835857045e-05,
+      "loss": 0.3516,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8372865275142315,
+      "grad_norm": 2.237438917160034,
+      "learning_rate": 1.307262790108371e-05,
+      "loss": 0.3383,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8377609108159393,
+      "grad_norm": 2.004232406616211,
+      "learning_rate": 1.3065315629855124e-05,
+      "loss": 0.3077,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8382352941176471,
+      "grad_norm": 2.4037442207336426,
+      "learning_rate": 1.3058001549201056e-05,
+      "loss": 0.3527,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 2.4244558811187744,
+      "learning_rate": 1.3050685663438931e-05,
+      "loss": 0.3533,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8391840607210627,
+      "grad_norm": 2.3665969371795654,
+      "learning_rate": 1.3043367976887251e-05,
+      "loss": 0.3269,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8396584440227703,
+      "grad_norm": 2.266602039337158,
+      "learning_rate": 1.3036048493865567e-05,
+      "loss": 0.3345,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8401328273244781,
+      "grad_norm": 2.609055757522583,
+      "learning_rate": 1.3028727218694503e-05,
+      "loss": 0.3333,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8406072106261859,
+      "grad_norm": 2.1717774868011475,
+      "learning_rate": 1.3021404155695728e-05,
+      "loss": 0.3405,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8410815939278937,
+      "grad_norm": 2.0103063583374023,
+      "learning_rate": 1.301407930919198e-05,
+      "loss": 0.2846,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8415559772296015,
+      "grad_norm": 2.3277881145477295,
+      "learning_rate": 1.300675268350704e-05,
+      "loss": 0.331,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8420303605313093,
+      "grad_norm": 2.5206899642944336,
+      "learning_rate": 1.2999424282965747e-05,
+      "loss": 0.3298,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8425047438330171,
+      "grad_norm": 2.3512508869171143,
+      "learning_rate": 1.299209411189398e-05,
+      "loss": 0.3856,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8429791271347249,
+      "grad_norm": 2.2688114643096924,
+      "learning_rate": 1.2984762174618664e-05,
+      "loss": 0.3016,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8434535104364327,
+      "grad_norm": 2.555619239807129,
+      "learning_rate": 1.2977428475467773e-05,
+      "loss": 0.317,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8439278937381404,
+      "grad_norm": 2.4550106525421143,
+      "learning_rate": 1.2970093018770318e-05,
+      "loss": 0.3679,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8444022770398482,
+      "grad_norm": 2.240664005279541,
+      "learning_rate": 1.2962755808856341e-05,
+      "loss": 0.3785,
+      "step": 1780
+    },
+    {
+      "epoch": 0.844876660341556,
+      "grad_norm": 2.236607313156128,
+      "learning_rate": 1.295541685005693e-05,
+      "loss": 0.3545,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8453510436432637,
+      "grad_norm": 2.1329894065856934,
+      "learning_rate": 1.2948076146704195e-05,
+      "loss": 0.3287,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8458254269449715,
+      "grad_norm": 2.1090683937072754,
+      "learning_rate": 1.2940733703131287e-05,
+      "loss": 0.3241,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8462998102466793,
+      "grad_norm": 2.2232136726379395,
+      "learning_rate": 1.293338952367237e-05,
+      "loss": 0.3764,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8467741935483871,
+      "grad_norm": 2.0039751529693604,
+      "learning_rate": 1.2926043612662646e-05,
+      "loss": 0.3344,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8472485768500949,
+      "grad_norm": 2.570286273956299,
+      "learning_rate": 1.291869597443833e-05,
+      "loss": 0.3779,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8477229601518027,
+      "grad_norm": 2.4321069717407227,
+      "learning_rate": 1.2911346613336666e-05,
+      "loss": 0.4621,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8481973434535104,
+      "grad_norm": 2.2266476154327393,
+      "learning_rate": 1.2903995533695904e-05,
+      "loss": 0.3687,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8486717267552182,
+      "grad_norm": 1.9736994504928589,
+      "learning_rate": 1.2896642739855311e-05,
+      "loss": 0.3805,
+      "step": 1789
+    },
+    {
+      "epoch": 0.849146110056926,
+      "grad_norm": 3.032970666885376,
+      "learning_rate": 1.2889288236155177e-05,
+      "loss": 0.403,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8496204933586338,
+      "grad_norm": 2.4022488594055176,
+      "learning_rate": 1.2881932026936785e-05,
+      "loss": 0.405,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8500948766603416,
+      "grad_norm": 2.144449234008789,
+      "learning_rate": 1.2874574116542439e-05,
+      "loss": 0.3194,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8505692599620494,
+      "grad_norm": 2.102857828140259,
+      "learning_rate": 1.2867214509315434e-05,
+      "loss": 0.339,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8510436432637571,
+      "grad_norm": 2.4417030811309814,
+      "learning_rate": 1.2859853209600075e-05,
+      "loss": 0.3896,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8515180265654649,
+      "grad_norm": 2.425147533416748,
+      "learning_rate": 1.2852490221741669e-05,
+      "loss": 0.2993,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8519924098671727,
+      "grad_norm": 2.761335849761963,
+      "learning_rate": 1.284512555008651e-05,
+      "loss": 0.3798,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8524667931688804,
+      "grad_norm": 2.5409250259399414,
+      "learning_rate": 1.2837759198981894e-05,
+      "loss": 0.3064,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8529411764705882,
+      "grad_norm": 2.241783380508423,
+      "learning_rate": 1.28303911727761e-05,
+      "loss": 0.3359,
+      "step": 1798
+    },
+    {
+      "epoch": 0.853415559772296,
+      "grad_norm": 2.344348907470703,
+      "learning_rate": 1.2823021475818408e-05,
+      "loss": 0.3305,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8538899430740038,
+      "grad_norm": 2.3390181064605713,
+      "learning_rate": 1.2815650112459075e-05,
+      "loss": 0.3343,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8543643263757116,
+      "grad_norm": 1.938061237335205,
+      "learning_rate": 1.2808277087049338e-05,
+      "loss": 0.2944,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8548387096774194,
+      "grad_norm": 2.33855938911438,
+      "learning_rate": 1.2800902403941429e-05,
+      "loss": 0.4172,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8553130929791272,
+      "grad_norm": 2.146927833557129,
+      "learning_rate": 1.279352606748855e-05,
+      "loss": 0.3548,
+      "step": 1803
+    },
+    {
+      "epoch": 0.855787476280835,
+      "grad_norm": 2.442716598510742,
+      "learning_rate": 1.278614808204487e-05,
+      "loss": 0.3444,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8562618595825426,
+      "grad_norm": 2.412533760070801,
+      "learning_rate": 1.277876845196555e-05,
+      "loss": 0.4137,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8567362428842504,
+      "grad_norm": 2.463116407394409,
+      "learning_rate": 1.2771387181606709e-05,
+      "loss": 0.4336,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8572106261859582,
+      "grad_norm": 2.2430520057678223,
+      "learning_rate": 1.2764004275325443e-05,
+      "loss": 0.2363,
+      "step": 1807
+    },
+    {
+      "epoch": 0.857685009487666,
+      "grad_norm": 2.2711448669433594,
+      "learning_rate": 1.2756619737479801e-05,
+      "loss": 0.4087,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8581593927893738,
+      "grad_norm": 2.3729405403137207,
+      "learning_rate": 1.2749233572428805e-05,
+      "loss": 0.3526,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8586337760910816,
+      "grad_norm": 2.3605003356933594,
+      "learning_rate": 1.2741845784532436e-05,
+      "loss": 0.3231,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8591081593927894,
+      "grad_norm": 2.177626848220825,
+      "learning_rate": 1.2734456378151636e-05,
+      "loss": 0.3287,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8595825426944972,
+      "grad_norm": 2.164935350418091,
+      "learning_rate": 1.272706535764829e-05,
+      "loss": 0.2948,
+      "step": 1812
+    },
+    {
+      "epoch": 0.860056925996205,
+      "grad_norm": 2.3316850662231445,
+      "learning_rate": 1.2719672727385249e-05,
+      "loss": 0.3226,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8605313092979127,
+      "grad_norm": 2.731200933456421,
+      "learning_rate": 1.271227849172631e-05,
+      "loss": 0.358,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8610056925996205,
+      "grad_norm": 3.0866591930389404,
+      "learning_rate": 1.270488265503622e-05,
+      "loss": 0.4681,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8614800759013282,
+      "grad_norm": 2.4756972789764404,
+      "learning_rate": 1.2697485221680663e-05,
+      "loss": 0.3664,
+      "step": 1816
+    },
+    {
+      "epoch": 0.861954459203036,
+      "grad_norm": 2.456561803817749,
+      "learning_rate": 1.269008619602627e-05,
+      "loss": 0.334,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8624288425047438,
+      "grad_norm": 2.1399877071380615,
+      "learning_rate": 1.2682685582440614e-05,
+      "loss": 0.3091,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8629032258064516,
+      "grad_norm": 2.5643470287323,
+      "learning_rate": 1.2675283385292212e-05,
+      "loss": 0.4395,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8633776091081594,
+      "grad_norm": 2.543917179107666,
+      "learning_rate": 1.2667879608950494e-05,
+      "loss": 0.3285,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8638519924098672,
+      "grad_norm": 2.08056902885437,
+      "learning_rate": 1.2660474257785844e-05,
+      "loss": 0.3604,
+      "step": 1821
+    },
+    {
+      "epoch": 0.864326375711575,
+      "grad_norm": 2.5211234092712402,
+      "learning_rate": 1.2653067336169566e-05,
+      "loss": 0.326,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8648007590132827,
+      "grad_norm": 1.8964157104492188,
+      "learning_rate": 1.2645658848473894e-05,
+      "loss": 0.2578,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8652751423149905,
+      "grad_norm": 1.9530812501907349,
+      "learning_rate": 1.2638248799071985e-05,
+      "loss": 0.3031,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8657495256166983,
+      "grad_norm": 2.3713808059692383,
+      "learning_rate": 1.2630837192337913e-05,
+      "loss": 0.3138,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8662239089184061,
+      "grad_norm": 2.222252607345581,
+      "learning_rate": 1.2623424032646683e-05,
+      "loss": 0.3455,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8666982922201139,
+      "grad_norm": 2.9525506496429443,
+      "learning_rate": 1.2616009324374205e-05,
+      "loss": 0.4101,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8671726755218216,
+      "grad_norm": 2.140566825866699,
+      "learning_rate": 1.2608593071897311e-05,
+      "loss": 0.3621,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8676470588235294,
+      "grad_norm": 2.7032320499420166,
+      "learning_rate": 1.2601175279593737e-05,
+      "loss": 0.4193,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8681214421252372,
+      "grad_norm": 2.723836660385132,
+      "learning_rate": 1.2593755951842134e-05,
+      "loss": 0.4037,
+      "step": 1830
+    },
+    {
+      "epoch": 0.868595825426945,
+      "grad_norm": 2.102358102798462,
+      "learning_rate": 1.2586335093022064e-05,
+      "loss": 0.3605,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8690702087286527,
+      "grad_norm": 2.0517046451568604,
+      "learning_rate": 1.2578912707513977e-05,
+      "loss": 0.3291,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8695445920303605,
+      "grad_norm": 2.466214418411255,
+      "learning_rate": 1.257148879969924e-05,
+      "loss": 0.3382,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8700189753320683,
+      "grad_norm": 2.1797330379486084,
+      "learning_rate": 1.2564063373960109e-05,
+      "loss": 0.3483,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8704933586337761,
+      "grad_norm": 2.182875394821167,
+      "learning_rate": 1.2556636434679744e-05,
+      "loss": 0.3307,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8709677419354839,
+      "grad_norm": 1.9578689336776733,
+      "learning_rate": 1.254920798624219e-05,
+      "loss": 0.284,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8714421252371917,
+      "grad_norm": 2.1100683212280273,
+      "learning_rate": 1.2541778033032383e-05,
+      "loss": 0.3402,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8719165085388995,
+      "grad_norm": 2.199645757675171,
+      "learning_rate": 1.2534346579436158e-05,
+      "loss": 0.3736,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8723908918406073,
+      "grad_norm": 2.371626853942871,
+      "learning_rate": 1.2526913629840228e-05,
+      "loss": 0.4038,
+      "step": 1839
+    },
+    {
+      "epoch": 0.872865275142315,
+      "grad_norm": 2.2260429859161377,
+      "learning_rate": 1.2519479188632184e-05,
+      "loss": 0.3058,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8733396584440227,
+      "grad_norm": 2.0778400897979736,
+      "learning_rate": 1.2512043260200506e-05,
+      "loss": 0.34,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8738140417457305,
+      "grad_norm": 1.6490452289581299,
+      "learning_rate": 1.2504605848934552e-05,
+      "loss": 0.2596,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8742884250474383,
+      "grad_norm": 2.222475051879883,
+      "learning_rate": 1.2497166959224546e-05,
+      "loss": 0.2611,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8747628083491461,
+      "grad_norm": 2.1947946548461914,
+      "learning_rate": 1.2489726595461598e-05,
+      "loss": 0.3622,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8752371916508539,
+      "grad_norm": 2.6763131618499756,
+      "learning_rate": 1.2482284762037675e-05,
+      "loss": 0.3425,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8757115749525617,
+      "grad_norm": 1.7346967458724976,
+      "learning_rate": 1.247484146334562e-05,
+      "loss": 0.243,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8761859582542695,
+      "grad_norm": 2.4290475845336914,
+      "learning_rate": 1.246739670377914e-05,
+      "loss": 0.4306,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8766603415559773,
+      "grad_norm": 2.096940517425537,
+      "learning_rate": 1.2459950487732804e-05,
+      "loss": 0.3498,
+      "step": 1848
+    },
+    {
+      "epoch": 0.877134724857685,
+      "grad_norm": 1.9826620817184448,
+      "learning_rate": 1.2452502819602035e-05,
+      "loss": 0.2844,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8776091081593927,
+      "grad_norm": 2.250736951828003,
+      "learning_rate": 1.2445053703783118e-05,
+      "loss": 0.3541,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8780834914611005,
+      "grad_norm": 2.070547342300415,
+      "learning_rate": 1.2437603144673198e-05,
+      "loss": 0.2969,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8785578747628083,
+      "grad_norm": 2.3527519702911377,
+      "learning_rate": 1.2430151146670261e-05,
+      "loss": 0.3719,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8790322580645161,
+      "grad_norm": 2.063654661178589,
+      "learning_rate": 1.242269771417315e-05,
+      "loss": 0.315,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8795066413662239,
+      "grad_norm": 2.292386531829834,
+      "learning_rate": 1.2415242851581552e-05,
+      "loss": 0.3539,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8799810246679317,
+      "grad_norm": 2.2170569896698,
+      "learning_rate": 1.2407786563296e-05,
+      "loss": 0.2982,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8804554079696395,
+      "grad_norm": 2.0592644214630127,
+      "learning_rate": 1.2400328853717862e-05,
+      "loss": 0.3065,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8809297912713473,
+      "grad_norm": 1.8755029439926147,
+      "learning_rate": 1.2392869727249358e-05,
+      "loss": 0.3065,
+      "step": 1857
+    },
+    {
+      "epoch": 0.881404174573055,
+      "grad_norm": 1.9304600954055786,
+      "learning_rate": 1.2385409188293528e-05,
+      "loss": 0.3204,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8818785578747628,
+      "grad_norm": 2.4600729942321777,
+      "learning_rate": 1.2377947241254263e-05,
+      "loss": 0.3721,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 2.047233819961548,
+      "learning_rate": 1.2370483890536271e-05,
+      "loss": 0.3081,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8828273244781784,
+      "grad_norm": 1.9848384857177734,
+      "learning_rate": 1.2363019140545096e-05,
+      "loss": 0.2854,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8833017077798861,
+      "grad_norm": 2.357391595840454,
+      "learning_rate": 1.2355552995687104e-05,
+      "loss": 0.4203,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8837760910815939,
+      "grad_norm": 1.8160470724105835,
+      "learning_rate": 1.2348085460369488e-05,
+      "loss": 0.236,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8842504743833017,
+      "grad_norm": 2.142580509185791,
+      "learning_rate": 1.234061653900026e-05,
+      "loss": 0.3978,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8847248576850095,
+      "grad_norm": 2.4461870193481445,
+      "learning_rate": 1.2333146235988251e-05,
+      "loss": 0.3631,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8851992409867173,
+      "grad_norm": 2.1813673973083496,
+      "learning_rate": 1.2325674555743106e-05,
+      "loss": 0.3192,
+      "step": 1866
+    },
+    {
+      "epoch": 0.885673624288425,
+      "grad_norm": 2.124764919281006,
+      "learning_rate": 1.2318201502675285e-05,
+      "loss": 0.3734,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8861480075901328,
+      "grad_norm": 1.9253191947937012,
+      "learning_rate": 1.2310727081196054e-05,
+      "loss": 0.3054,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8866223908918406,
+      "grad_norm": 1.8508185148239136,
+      "learning_rate": 1.2303251295717495e-05,
+      "loss": 0.2451,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8870967741935484,
+      "grad_norm": 2.3445682525634766,
+      "learning_rate": 1.2295774150652486e-05,
+      "loss": 0.3997,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8875711574952562,
+      "grad_norm": 2.4285199642181396,
+      "learning_rate": 1.2288295650414716e-05,
+      "loss": 0.3243,
+      "step": 1871
+    },
+    {
+      "epoch": 0.888045540796964,
+      "grad_norm": 1.8846771717071533,
+      "learning_rate": 1.2280815799418666e-05,
+      "loss": 0.2579,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8885199240986718,
+      "grad_norm": 3.0957839488983154,
+      "learning_rate": 1.2273334602079621e-05,
+      "loss": 0.4514,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8889943074003795,
+      "grad_norm": 2.224541425704956,
+      "learning_rate": 1.2265852062813652e-05,
+      "loss": 0.3709,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8894686907020873,
+      "grad_norm": 2.923872470855713,
+      "learning_rate": 1.2258368186037638e-05,
+      "loss": 0.4537,
+      "step": 1875
+    },
+    {
+      "epoch": 0.889943074003795,
+      "grad_norm": 2.0478696823120117,
+      "learning_rate": 1.2250882976169228e-05,
+      "loss": 0.3222,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8904174573055028,
+      "grad_norm": 2.2573540210723877,
+      "learning_rate": 1.2243396437626866e-05,
+      "loss": 0.37,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8908918406072106,
+      "grad_norm": 2.3689069747924805,
+      "learning_rate": 1.2235908574829792e-05,
+      "loss": 0.3659,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8913662239089184,
+      "grad_norm": 2.4283788204193115,
+      "learning_rate": 1.2228419392198008e-05,
+      "loss": 0.3417,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8918406072106262,
+      "grad_norm": 2.9532792568206787,
+      "learning_rate": 1.2220928894152313e-05,
+      "loss": 0.2853,
+      "step": 1880
+    },
+    {
+      "epoch": 0.892314990512334,
+      "grad_norm": 2.0899689197540283,
+      "learning_rate": 1.2213437085114263e-05,
+      "loss": 0.3315,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8927893738140418,
+      "grad_norm": 2.3636410236358643,
+      "learning_rate": 1.2205943969506207e-05,
+      "loss": 0.4148,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8932637571157496,
+      "grad_norm": 2.356285810470581,
+      "learning_rate": 1.2198449551751255e-05,
+      "loss": 0.3218,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8937381404174574,
+      "grad_norm": 2.254668951034546,
+      "learning_rate": 1.219095383627329e-05,
+      "loss": 0.3474,
+      "step": 1884
+    },
+    {
+      "epoch": 0.894212523719165,
+      "grad_norm": 2.1036298274993896,
+      "learning_rate": 1.2183456827496951e-05,
+      "loss": 0.3559,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8946869070208728,
+      "grad_norm": 2.2919163703918457,
+      "learning_rate": 1.2175958529847654e-05,
+      "loss": 0.3481,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8951612903225806,
+      "grad_norm": 2.041938066482544,
+      "learning_rate": 1.216845894775157e-05,
+      "loss": 0.2727,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8956356736242884,
+      "grad_norm": 2.564687728881836,
+      "learning_rate": 1.2160958085635628e-05,
+      "loss": 0.4146,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8961100569259962,
+      "grad_norm": 2.3934693336486816,
+      "learning_rate": 1.2153455947927509e-05,
+      "loss": 0.2986,
+      "step": 1889
+    },
+    {
+      "epoch": 0.896584440227704,
+      "grad_norm": 2.548827886581421,
+      "learning_rate": 1.2145952539055654e-05,
+      "loss": 0.4007,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8970588235294118,
+      "grad_norm": 2.382176399230957,
+      "learning_rate": 1.213844786344925e-05,
+      "loss": 0.3523,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8975332068311196,
+      "grad_norm": 2.39910626411438,
+      "learning_rate": 1.2130941925538237e-05,
+      "loss": 0.3493,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8980075901328273,
+      "grad_norm": 2.387073278427124,
+      "learning_rate": 1.2123434729753287e-05,
+      "loss": 0.3532,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8984819734345351,
+      "grad_norm": 2.265078544616699,
+      "learning_rate": 1.211592628052583e-05,
+      "loss": 0.3608,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8989563567362429,
+      "grad_norm": 2.8354647159576416,
+      "learning_rate": 1.2108416582288027e-05,
+      "loss": 0.3634,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8994307400379506,
+      "grad_norm": 2.2116317749023438,
+      "learning_rate": 1.210090563947278e-05,
+      "loss": 0.31,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8999051233396584,
+      "grad_norm": 2.5344135761260986,
+      "learning_rate": 1.2093393456513724e-05,
+      "loss": 0.3346,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9003795066413662,
+      "grad_norm": 2.046262502670288,
+      "learning_rate": 1.2085880037845223e-05,
+      "loss": 0.3562,
+      "step": 1898
+    },
+    {
+      "epoch": 0.900853889943074,
+      "grad_norm": 1.9915024042129517,
+      "learning_rate": 1.2078365387902379e-05,
+      "loss": 0.3242,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9013282732447818,
+      "grad_norm": 2.071885108947754,
+      "learning_rate": 1.2070849511121014e-05,
+      "loss": 0.3272,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9018026565464896,
+      "grad_norm": 2.1949217319488525,
+      "learning_rate": 1.2063332411937672e-05,
+      "loss": 0.3419,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9022770398481973,
+      "grad_norm": 2.311913013458252,
+      "learning_rate": 1.2055814094789625e-05,
+      "loss": 0.3776,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9027514231499051,
+      "grad_norm": 2.1613712310791016,
+      "learning_rate": 1.2048294564114859e-05,
+      "loss": 0.3538,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 2.3244266510009766,
+      "learning_rate": 1.2040773824352088e-05,
+      "loss": 0.366,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9037001897533207,
+      "grad_norm": 3.7160706520080566,
+      "learning_rate": 1.2033251879940716e-05,
+      "loss": 0.4384,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9041745730550285,
+      "grad_norm": 1.9680372476577759,
+      "learning_rate": 1.2025728735320878e-05,
+      "loss": 0.2941,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9046489563567363,
+      "grad_norm": 1.8036983013153076,
+      "learning_rate": 1.2018204394933417e-05,
+      "loss": 0.2192,
+      "step": 1907
+    },
+    {
+      "epoch": 0.905123339658444,
+      "grad_norm": 2.109776496887207,
+      "learning_rate": 1.201067886321987e-05,
+      "loss": 0.3297,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9055977229601518,
+      "grad_norm": 2.50654935836792,
+      "learning_rate": 1.2003152144622493e-05,
+      "loss": 0.4017,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9060721062618596,
+      "grad_norm": 2.427208185195923,
+      "learning_rate": 1.1995624243584219e-05,
+      "loss": 0.3476,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9065464895635673,
+      "grad_norm": 3.9568138122558594,
+      "learning_rate": 1.1988095164548706e-05,
+      "loss": 0.3642,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9070208728652751,
+      "grad_norm": 2.389507293701172,
+      "learning_rate": 1.198056491196029e-05,
+      "loss": 0.4409,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9074952561669829,
+      "grad_norm": 2.369760751724243,
+      "learning_rate": 1.1973033490264e-05,
+      "loss": 0.3742,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9079696394686907,
+      "grad_norm": 2.787485361099243,
+      "learning_rate": 1.1965500903905571e-05,
+      "loss": 0.3947,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9084440227703985,
+      "grad_norm": 2.2195568084716797,
+      "learning_rate": 1.1957967157331404e-05,
+      "loss": 0.3326,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9089184060721063,
+      "grad_norm": 2.2224323749542236,
+      "learning_rate": 1.1950432254988604e-05,
+      "loss": 0.4156,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9093927893738141,
+      "grad_norm": 2.9090452194213867,
+      "learning_rate": 1.1942896201324938e-05,
+      "loss": 0.4651,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9098671726755219,
+      "grad_norm": 2.5390193462371826,
+      "learning_rate": 1.1935359000788873e-05,
+      "loss": 0.3556,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9103415559772297,
+      "grad_norm": 2.103929281234741,
+      "learning_rate": 1.192782065782954e-05,
+      "loss": 0.3225,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9108159392789373,
+      "grad_norm": 2.0904929637908936,
+      "learning_rate": 1.1920281176896752e-05,
+      "loss": 0.2711,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9112903225806451,
+      "grad_norm": 2.126173257827759,
+      "learning_rate": 1.1912740562440988e-05,
+      "loss": 0.2951,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9117647058823529,
+      "grad_norm": 1.8137263059616089,
+      "learning_rate": 1.1905198818913393e-05,
+      "loss": 0.2783,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9122390891840607,
+      "grad_norm": 2.573413133621216,
+      "learning_rate": 1.1897655950765789e-05,
+      "loss": 0.3153,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9127134724857685,
+      "grad_norm": 1.9147273302078247,
+      "learning_rate": 1.189011196245066e-05,
+      "loss": 0.3062,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9131878557874763,
+      "grad_norm": 2.3643062114715576,
+      "learning_rate": 1.1882566858421137e-05,
+      "loss": 0.4424,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9136622390891841,
+      "grad_norm": 2.178414821624756,
+      "learning_rate": 1.1875020643131028e-05,
+      "loss": 0.3204,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9141366223908919,
+      "grad_norm": 2.598742961883545,
+      "learning_rate": 1.1867473321034786e-05,
+      "loss": 0.4002,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9146110056925996,
+      "grad_norm": 2.4741811752319336,
+      "learning_rate": 1.1859924896587528e-05,
+      "loss": 0.3557,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9150853889943074,
+      "grad_norm": 2.3986568450927734,
+      "learning_rate": 1.1852375374245003e-05,
+      "loss": 0.3971,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9155597722960152,
+      "grad_norm": 1.920324683189392,
+      "learning_rate": 1.1844824758463626e-05,
+      "loss": 0.2705,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9160341555977229,
+      "grad_norm": 2.353745937347412,
+      "learning_rate": 1.183727305370045e-05,
+      "loss": 0.3579,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9165085388994307,
+      "grad_norm": 2.8409640789031982,
+      "learning_rate": 1.1829720264413169e-05,
+      "loss": 0.3981,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9169829222011385,
+      "grad_norm": 2.5938303470611572,
+      "learning_rate": 1.1822166395060124e-05,
+      "loss": 0.3382,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9174573055028463,
+      "grad_norm": 2.2392194271087646,
+      "learning_rate": 1.1814611450100286e-05,
+      "loss": 0.3603,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9179316888045541,
+      "grad_norm": 2.44313907623291,
+      "learning_rate": 1.1807055433993266e-05,
+      "loss": 0.3935,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9184060721062619,
+      "grad_norm": 2.418811559677124,
+      "learning_rate": 1.1799498351199303e-05,
+      "loss": 0.3081,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9188804554079696,
+      "grad_norm": 2.3761041164398193,
+      "learning_rate": 1.179194020617927e-05,
+      "loss": 0.3717,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9193548387096774,
+      "grad_norm": 2.0884335041046143,
+      "learning_rate": 1.1784381003394669e-05,
+      "loss": 0.3512,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9198292220113852,
+      "grad_norm": 1.974050760269165,
+      "learning_rate": 1.1776820747307615e-05,
+      "loss": 0.3303,
+      "step": 1939
+    },
+    {
+      "epoch": 0.920303605313093,
+      "grad_norm": 2.162907600402832,
+      "learning_rate": 1.1769259442380862e-05,
+      "loss": 0.3726,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9207779886148008,
+      "grad_norm": 2.2310690879821777,
+      "learning_rate": 1.1761697093077762e-05,
+      "loss": 0.3117,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9212523719165086,
+      "grad_norm": 2.247523307800293,
+      "learning_rate": 1.1754133703862302e-05,
+      "loss": 0.3046,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9217267552182163,
+      "grad_norm": 2.2049553394317627,
+      "learning_rate": 1.1746569279199073e-05,
+      "loss": 0.2433,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9222011385199241,
+      "grad_norm": 2.9283275604248047,
+      "learning_rate": 1.1739003823553286e-05,
+      "loss": 0.3686,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9226755218216319,
+      "grad_norm": 1.9516762495040894,
+      "learning_rate": 1.1731437341390748e-05,
+      "loss": 0.2654,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9231499051233396,
+      "grad_norm": 2.1916255950927734,
+      "learning_rate": 1.172386983717788e-05,
+      "loss": 0.3597,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9236242884250474,
+      "grad_norm": 2.2968626022338867,
+      "learning_rate": 1.1716301315381706e-05,
+      "loss": 0.3295,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9240986717267552,
+      "grad_norm": 2.341371774673462,
+      "learning_rate": 1.170873178046985e-05,
+      "loss": 0.3365,
+      "step": 1948
+    },
+    {
+      "epoch": 0.924573055028463,
+      "grad_norm": 2.3758347034454346,
+      "learning_rate": 1.1701161236910532e-05,
+      "loss": 0.3714,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9250474383301708,
+      "grad_norm": 2.347499132156372,
+      "learning_rate": 1.1693589689172566e-05,
+      "loss": 0.3345,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9255218216318786,
+      "grad_norm": 2.4397614002227783,
+      "learning_rate": 1.1686017141725367e-05,
+      "loss": 0.295,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9259962049335864,
+      "grad_norm": 1.9351602792739868,
+      "learning_rate": 1.167844359903894e-05,
+      "loss": 0.2817,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9264705882352942,
+      "grad_norm": 2.63478684425354,
+      "learning_rate": 1.1670869065583856e-05,
+      "loss": 0.4525,
+      "step": 1953
+    },
+    {
+      "epoch": 0.926944971537002,
+      "grad_norm": 3.3862996101379395,
+      "learning_rate": 1.1663293545831302e-05,
+      "loss": 0.3677,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9274193548387096,
+      "grad_norm": 2.883392333984375,
+      "learning_rate": 1.1655717044253025e-05,
+      "loss": 0.41,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9278937381404174,
+      "grad_norm": 2.115570068359375,
+      "learning_rate": 1.164813956532136e-05,
+      "loss": 0.3187,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9283681214421252,
+      "grad_norm": 1.8709588050842285,
+      "learning_rate": 1.1640561113509222e-05,
+      "loss": 0.2728,
+      "step": 1957
+    },
+    {
+      "epoch": 0.928842504743833,
+      "grad_norm": 2.141753911972046,
+      "learning_rate": 1.1632981693290089e-05,
+      "loss": 0.2978,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9293168880455408,
+      "grad_norm": 2.3228015899658203,
+      "learning_rate": 1.1625401309138025e-05,
+      "loss": 0.3375,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9297912713472486,
+      "grad_norm": 2.2550671100616455,
+      "learning_rate": 1.161781996552765e-05,
+      "loss": 0.3075,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9302656546489564,
+      "grad_norm": 2.196171998977661,
+      "learning_rate": 1.1610237666934158e-05,
+      "loss": 0.3969,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9307400379506642,
+      "grad_norm": 3.4995615482330322,
+      "learning_rate": 1.1602654417833305e-05,
+      "loss": 0.349,
+      "step": 1962
+    },
+    {
+      "epoch": 0.931214421252372,
+      "grad_norm": 2.3877415657043457,
+      "learning_rate": 1.1595070222701408e-05,
+      "loss": 0.3497,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9316888045540797,
+      "grad_norm": 2.260338306427002,
+      "learning_rate": 1.1587485086015346e-05,
+      "loss": 0.2895,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9321631878557874,
+      "grad_norm": 2.212602138519287,
+      "learning_rate": 1.1579899012252543e-05,
+      "loss": 0.3157,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9326375711574952,
+      "grad_norm": 1.9429450035095215,
+      "learning_rate": 1.1572312005890986e-05,
+      "loss": 0.3059,
+      "step": 1966
+    },
+    {
+      "epoch": 0.933111954459203,
+      "grad_norm": 2.017580509185791,
+      "learning_rate": 1.1564724071409213e-05,
+      "loss": 0.2694,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9335863377609108,
+      "grad_norm": 3.207867383956909,
+      "learning_rate": 1.1557135213286303e-05,
+      "loss": 0.3742,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9340607210626186,
+      "grad_norm": 2.253946542739868,
+      "learning_rate": 1.1549545436001888e-05,
+      "loss": 0.3292,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9345351043643264,
+      "grad_norm": 2.34022855758667,
+      "learning_rate": 1.1541954744036131e-05,
+      "loss": 0.388,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9350094876660342,
+      "grad_norm": 2.2387092113494873,
+      "learning_rate": 1.153436314186975e-05,
+      "loss": 0.344,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9354838709677419,
+      "grad_norm": 2.4496212005615234,
+      "learning_rate": 1.152677063398399e-05,
+      "loss": 0.3724,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9359582542694497,
+      "grad_norm": 2.5698015689849854,
+      "learning_rate": 1.1519177224860632e-05,
+      "loss": 0.3821,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9364326375711575,
+      "grad_norm": 1.8094062805175781,
+      "learning_rate": 1.151158291898199e-05,
+      "loss": 0.3154,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9369070208728653,
+      "grad_norm": 2.1138343811035156,
+      "learning_rate": 1.1503987720830908e-05,
+      "loss": 0.347,
+      "step": 1975
+    },
+    {
+      "epoch": 0.937381404174573,
+      "grad_norm": 2.825183153152466,
+      "learning_rate": 1.1496391634890758e-05,
+      "loss": 0.4074,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9378557874762808,
+      "grad_norm": 2.231473922729492,
+      "learning_rate": 1.1488794665645434e-05,
+      "loss": 0.3262,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9383301707779886,
+      "grad_norm": 2.0554375648498535,
+      "learning_rate": 1.1481196817579352e-05,
+      "loss": 0.2908,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9388045540796964,
+      "grad_norm": 2.4389944076538086,
+      "learning_rate": 1.1473598095177443e-05,
+      "loss": 0.2931,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9392789373814042,
+      "grad_norm": 3.037283420562744,
+      "learning_rate": 1.1465998502925161e-05,
+      "loss": 0.4337,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9397533206831119,
+      "grad_norm": 2.06538987159729,
+      "learning_rate": 1.1458398045308471e-05,
+      "loss": 0.2837,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9402277039848197,
+      "grad_norm": 2.118652820587158,
+      "learning_rate": 1.1450796726813844e-05,
+      "loss": 0.3299,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9407020872865275,
+      "grad_norm": 2.2635416984558105,
+      "learning_rate": 1.1443194551928267e-05,
+      "loss": 0.3512,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 2.088435173034668,
+      "learning_rate": 1.1435591525139228e-05,
+      "loss": 0.3404,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9416508538899431,
+      "grad_norm": 2.3585917949676514,
+      "learning_rate": 1.1427987650934717e-05,
+      "loss": 0.3526,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9421252371916509,
+      "grad_norm": 1.858132243156433,
+      "learning_rate": 1.1420382933803225e-05,
+      "loss": 0.2791,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9425996204933587,
+      "grad_norm": 2.3248212337493896,
+      "learning_rate": 1.1412777378233742e-05,
+      "loss": 0.3954,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9430740037950665,
+      "grad_norm": 2.187311887741089,
+      "learning_rate": 1.1405170988715752e-05,
+      "loss": 0.3678,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9435483870967742,
+      "grad_norm": 2.151273488998413,
+      "learning_rate": 1.1397563769739235e-05,
+      "loss": 0.3197,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9440227703984819,
+      "grad_norm": 2.1821095943450928,
+      "learning_rate": 1.1389955725794655e-05,
+      "loss": 0.3604,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9444971537001897,
+      "grad_norm": 2.963883876800537,
+      "learning_rate": 1.138234686137296e-05,
+      "loss": 0.3858,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9449715370018975,
+      "grad_norm": 2.512237310409546,
+      "learning_rate": 1.1374737180965593e-05,
+      "loss": 0.3751,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9454459203036053,
+      "grad_norm": 2.063708543777466,
+      "learning_rate": 1.1367126689064473e-05,
+      "loss": 0.315,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9459203036053131,
+      "grad_norm": 2.0706629753112793,
+      "learning_rate": 1.1359515390161996e-05,
+      "loss": 0.3446,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9463946869070209,
+      "grad_norm": 2.1223249435424805,
+      "learning_rate": 1.1351903288751038e-05,
+      "loss": 0.3453,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9468690702087287,
+      "grad_norm": 2.1654293537139893,
+      "learning_rate": 1.1344290389324949e-05,
+      "loss": 0.2958,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9473434535104365,
+      "grad_norm": 2.032048225402832,
+      "learning_rate": 1.1336676696377542e-05,
+      "loss": 0.3034,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9478178368121443,
+      "grad_norm": 2.133802890777588,
+      "learning_rate": 1.1329062214403106e-05,
+      "loss": 0.3172,
+      "step": 1998
+    },
+    {
+      "epoch": 0.948292220113852,
+      "grad_norm": 2.323798656463623,
+      "learning_rate": 1.13214469478964e-05,
+      "loss": 0.3271,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9487666034155597,
+      "grad_norm": 2.347034454345703,
+      "learning_rate": 1.1313830901352634e-05,
+      "loss": 0.3382,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9492409867172675,
+      "grad_norm": 2.104706048965454,
+      "learning_rate": 1.130621407926749e-05,
+      "loss": 0.3714,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9497153700189753,
+      "grad_norm": 2.216869354248047,
+      "learning_rate": 1.12985964861371e-05,
+      "loss": 0.3052,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9501897533206831,
+      "grad_norm": 2.0643813610076904,
+      "learning_rate": 1.1290978126458054e-05,
+      "loss": 0.2865,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9506641366223909,
+      "grad_norm": 1.9597926139831543,
+      "learning_rate": 1.1283359004727397e-05,
+      "loss": 0.3136,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9511385199240987,
+      "grad_norm": 2.2669873237609863,
+      "learning_rate": 1.1275739125442618e-05,
+      "loss": 0.3544,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9516129032258065,
+      "grad_norm": 1.7317811250686646,
+      "learning_rate": 1.1268118493101654e-05,
+      "loss": 0.2356,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9520872865275142,
+      "grad_norm": 2.301215887069702,
+      "learning_rate": 1.1260497112202895e-05,
+      "loss": 0.3188,
+      "step": 2007
+    },
+    {
+      "epoch": 0.952561669829222,
+      "grad_norm": 1.8703101873397827,
+      "learning_rate": 1.1252874987245163e-05,
+      "loss": 0.2773,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9530360531309298,
+      "grad_norm": 2.490936517715454,
+      "learning_rate": 1.1245252122727726e-05,
+      "loss": 0.363,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9535104364326376,
+      "grad_norm": 2.491692066192627,
+      "learning_rate": 1.123762852315028e-05,
+      "loss": 0.3159,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9539848197343453,
+      "grad_norm": 2.416424036026001,
+      "learning_rate": 1.1230004193012964e-05,
+      "loss": 0.3757,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9544592030360531,
+      "grad_norm": 2.583618402481079,
+      "learning_rate": 1.1222379136816347e-05,
+      "loss": 0.3097,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9549335863377609,
+      "grad_norm": 2.5698044300079346,
+      "learning_rate": 1.1214753359061418e-05,
+      "loss": 0.3456,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9554079696394687,
+      "grad_norm": 2.5504226684570312,
+      "learning_rate": 1.1207126864249604e-05,
+      "loss": 0.3817,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9558823529411765,
+      "grad_norm": 2.0663864612579346,
+      "learning_rate": 1.1199499656882747e-05,
+      "loss": 0.2627,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9563567362428842,
+      "grad_norm": 2.0620248317718506,
+      "learning_rate": 1.1191871741463112e-05,
+      "loss": 0.3134,
+      "step": 2016
+    },
+    {
+      "epoch": 0.956831119544592,
+      "grad_norm": 2.855698585510254,
+      "learning_rate": 1.1184243122493381e-05,
+      "loss": 0.3152,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9573055028462998,
+      "grad_norm": 1.9541351795196533,
+      "learning_rate": 1.1176613804476655e-05,
+      "loss": 0.3101,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9577798861480076,
+      "grad_norm": 2.0520036220550537,
+      "learning_rate": 1.1168983791916442e-05,
+      "loss": 0.2803,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9582542694497154,
+      "grad_norm": 2.2084736824035645,
+      "learning_rate": 1.1161353089316664e-05,
+      "loss": 0.3272,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9587286527514232,
+      "grad_norm": 2.3713338375091553,
+      "learning_rate": 1.115372170118165e-05,
+      "loss": 0.3432,
+      "step": 2021
+    },
+    {
+      "epoch": 0.959203036053131,
+      "grad_norm": 2.585407257080078,
+      "learning_rate": 1.1146089632016132e-05,
+      "loss": 0.4149,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9596774193548387,
+      "grad_norm": 2.2623908519744873,
+      "learning_rate": 1.1138456886325247e-05,
+      "loss": 0.3292,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9601518026565465,
+      "grad_norm": 1.627692461013794,
+      "learning_rate": 1.1130823468614525e-05,
+      "loss": 0.2295,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9606261859582542,
+      "grad_norm": 2.19572114944458,
+      "learning_rate": 1.1123189383389901e-05,
+      "loss": 0.3843,
+      "step": 2025
+    },
+    {
+      "epoch": 0.961100569259962,
+      "grad_norm": 2.066941499710083,
+      "learning_rate": 1.1115554635157698e-05,
+      "loss": 0.3033,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9615749525616698,
+      "grad_norm": 2.1408028602600098,
+      "learning_rate": 1.1107919228424631e-05,
+      "loss": 0.3627,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9620493358633776,
+      "grad_norm": 2.3943800926208496,
+      "learning_rate": 1.110028316769781e-05,
+      "loss": 0.3059,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9625237191650854,
+      "grad_norm": 2.0022356510162354,
+      "learning_rate": 1.1092646457484721e-05,
+      "loss": 0.3247,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9629981024667932,
+      "grad_norm": 2.043557643890381,
+      "learning_rate": 1.1085009102293243e-05,
+      "loss": 0.2876,
+      "step": 2030
+    },
+    {
+      "epoch": 0.963472485768501,
+      "grad_norm": 2.535658836364746,
+      "learning_rate": 1.1077371106631625e-05,
+      "loss": 0.3627,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9639468690702088,
+      "grad_norm": 1.851802945137024,
+      "learning_rate": 1.1069732475008504e-05,
+      "loss": 0.2793,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9644212523719166,
+      "grad_norm": 2.064366340637207,
+      "learning_rate": 1.1062093211932891e-05,
+      "loss": 0.3435,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9648956356736242,
+      "grad_norm": 1.589677333831787,
+      "learning_rate": 1.1054453321914162e-05,
+      "loss": 0.2602,
+      "step": 2034
+    },
+    {
+      "epoch": 0.965370018975332,
+      "grad_norm": 2.559872627258301,
+      "learning_rate": 1.1046812809462073e-05,
+      "loss": 0.3856,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9658444022770398,
+      "grad_norm": 2.204385995864868,
+      "learning_rate": 1.1039171679086739e-05,
+      "loss": 0.336,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9663187855787476,
+      "grad_norm": 2.258500337600708,
+      "learning_rate": 1.1031529935298651e-05,
+      "loss": 0.3388,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9667931688804554,
+      "grad_norm": 2.894205331802368,
+      "learning_rate": 1.1023887582608645e-05,
+      "loss": 0.368,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9672675521821632,
+      "grad_norm": 3.6994946002960205,
+      "learning_rate": 1.1016244625527933e-05,
+      "loss": 0.3644,
+      "step": 2039
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 2.2392802238464355,
+      "learning_rate": 1.1008601068568074e-05,
+      "loss": 0.3416,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9682163187855788,
+      "grad_norm": 2.4265706539154053,
+      "learning_rate": 1.1000956916240985e-05,
+      "loss": 0.3332,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9686907020872866,
+      "grad_norm": 1.7920218706130981,
+      "learning_rate": 1.0993312173058934e-05,
+      "loss": 0.2237,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9691650853889943,
+      "grad_norm": 2.4036221504211426,
+      "learning_rate": 1.0985666843534534e-05,
+      "loss": 0.3508,
+      "step": 2043
+    },
+    {
+      "epoch": 0.969639468690702,
+      "grad_norm": 2.5761871337890625,
+      "learning_rate": 1.097802093218075e-05,
+      "loss": 0.3949,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9701138519924098,
+      "grad_norm": 2.6531662940979004,
+      "learning_rate": 1.0970374443510891e-05,
+      "loss": 0.3703,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9705882352941176,
+      "grad_norm": 1.9129854440689087,
+      "learning_rate": 1.0962727382038598e-05,
+      "loss": 0.2939,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9710626185958254,
+      "grad_norm": 2.6273584365844727,
+      "learning_rate": 1.0955079752277859e-05,
+      "loss": 0.418,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9715370018975332,
+      "grad_norm": 2.4543769359588623,
+      "learning_rate": 1.0947431558742991e-05,
+      "loss": 0.2984,
+      "step": 2048
+    },
+    {
+      "epoch": 0.972011385199241,
+      "grad_norm": 2.275509834289551,
+      "learning_rate": 1.0939782805948653e-05,
+      "loss": 0.3299,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9724857685009488,
+      "grad_norm": 2.039682626724243,
+      "learning_rate": 1.0932133498409821e-05,
+      "loss": 0.3081,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9729601518026565,
+      "grad_norm": 2.2457706928253174,
+      "learning_rate": 1.0924483640641808e-05,
+      "loss": 0.3862,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9734345351043643,
+      "grad_norm": 2.2409019470214844,
+      "learning_rate": 1.0916833237160252e-05,
+      "loss": 0.3165,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9739089184060721,
+      "grad_norm": 1.9504499435424805,
+      "learning_rate": 1.0909182292481108e-05,
+      "loss": 0.3161,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9743833017077799,
+      "grad_norm": 2.0921568870544434,
+      "learning_rate": 1.0901530811120655e-05,
+      "loss": 0.3617,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9748576850094877,
+      "grad_norm": 1.8210068941116333,
+      "learning_rate": 1.0893878797595484e-05,
+      "loss": 0.2962,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9753320683111955,
+      "grad_norm": 1.7719886302947998,
+      "learning_rate": 1.0886226256422501e-05,
+      "loss": 0.2733,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9758064516129032,
+      "grad_norm": 2.5448601245880127,
+      "learning_rate": 1.0878573192118933e-05,
+      "loss": 0.2718,
+      "step": 2057
+    },
+    {
+      "epoch": 0.976280834914611,
+      "grad_norm": 1.8214176893234253,
+      "learning_rate": 1.0870919609202301e-05,
+      "loss": 0.2664,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9767552182163188,
+      "grad_norm": 2.0862033367156982,
+      "learning_rate": 1.0863265512190444e-05,
+      "loss": 0.3164,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9772296015180265,
+      "grad_norm": 1.6361185312271118,
+      "learning_rate": 1.0855610905601495e-05,
+      "loss": 0.217,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9777039848197343,
+      "grad_norm": 2.2029919624328613,
+      "learning_rate": 1.0847955793953899e-05,
+      "loss": 0.3134,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9781783681214421,
+      "grad_norm": 2.551486015319824,
+      "learning_rate": 1.0840300181766383e-05,
+      "loss": 0.4083,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9786527514231499,
+      "grad_norm": 2.5572237968444824,
+      "learning_rate": 1.0832644073557987e-05,
+      "loss": 0.4408,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9791271347248577,
+      "grad_norm": 2.1839964389801025,
+      "learning_rate": 1.0824987473848037e-05,
+      "loss": 0.362,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9796015180265655,
+      "grad_norm": 2.5323362350463867,
+      "learning_rate": 1.0817330387156142e-05,
+      "loss": 0.3513,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9800759013282733,
+      "grad_norm": 2.2569549083709717,
+      "learning_rate": 1.0809672818002209e-05,
+      "loss": 0.3858,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9805502846299811,
+      "grad_norm": 2.198728561401367,
+      "learning_rate": 1.080201477090642e-05,
+      "loss": 0.36,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9810246679316889,
+      "grad_norm": 2.281818389892578,
+      "learning_rate": 1.079435625038925e-05,
+      "loss": 0.3084,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9814990512333965,
+      "grad_norm": 2.4596810340881348,
+      "learning_rate": 1.0786697260971449e-05,
+      "loss": 0.3617,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9819734345351043,
+      "grad_norm": 2.167909622192383,
+      "learning_rate": 1.0779037807174032e-05,
+      "loss": 0.3021,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9824478178368121,
+      "grad_norm": 1.8683514595031738,
+      "learning_rate": 1.0771377893518314e-05,
+      "loss": 0.2457,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9829222011385199,
+      "grad_norm": 2.5630767345428467,
+      "learning_rate": 1.0763717524525854e-05,
+      "loss": 0.3483,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9833965844402277,
+      "grad_norm": 2.544394016265869,
+      "learning_rate": 1.0756056704718498e-05,
+      "loss": 0.3901,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9838709677419355,
+      "grad_norm": 2.037989377975464,
+      "learning_rate": 1.0748395438618353e-05,
+      "loss": 0.3368,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9843453510436433,
+      "grad_norm": 1.889345407485962,
+      "learning_rate": 1.074073373074778e-05,
+      "loss": 0.2303,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9848197343453511,
+      "grad_norm": 2.1819489002227783,
+      "learning_rate": 1.0733071585629423e-05,
+      "loss": 0.31,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9852941176470589,
+      "grad_norm": 2.640437126159668,
+      "learning_rate": 1.0725409007786161e-05,
+      "loss": 0.4647,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9857685009487666,
+      "grad_norm": 2.5758280754089355,
+      "learning_rate": 1.0717746001741139e-05,
+      "loss": 0.3326,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9862428842504743,
+      "grad_norm": 1.8008846044540405,
+      "learning_rate": 1.0710082572017753e-05,
+      "loss": 0.2491,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9867172675521821,
+      "grad_norm": 2.1113157272338867,
+      "learning_rate": 1.0702418723139654e-05,
+      "loss": 0.3475,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9871916508538899,
+      "grad_norm": 2.860546827316284,
+      "learning_rate": 1.0694754459630732e-05,
+      "loss": 0.3269,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9876660341555977,
+      "grad_norm": 2.7859246730804443,
+      "learning_rate": 1.0687089786015126e-05,
+      "loss": 0.3732,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9881404174573055,
+      "grad_norm": 2.2017409801483154,
+      "learning_rate": 1.0679424706817221e-05,
+      "loss": 0.3555,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9886148007590133,
+      "grad_norm": 1.9931174516677856,
+      "learning_rate": 1.0671759226561631e-05,
+      "loss": 0.3141,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9890891840607211,
+      "grad_norm": 2.228123664855957,
+      "learning_rate": 1.0664093349773222e-05,
+      "loss": 0.3647,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9895635673624289,
+      "grad_norm": 2.191070795059204,
+      "learning_rate": 1.0656427080977077e-05,
+      "loss": 0.3575,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9900379506641366,
+      "grad_norm": 2.1509921550750732,
+      "learning_rate": 1.0648760424698521e-05,
+      "loss": 0.3169,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9905123339658444,
+      "grad_norm": 1.9772475957870483,
+      "learning_rate": 1.0641093385463108e-05,
+      "loss": 0.3271,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9909867172675522,
+      "grad_norm": 2.0178329944610596,
+      "learning_rate": 1.0633425967796614e-05,
+      "loss": 0.3091,
+      "step": 2089
+    },
+    {
+      "epoch": 0.99146110056926,
+      "grad_norm": 3.558840751647949,
+      "learning_rate": 1.0625758176225038e-05,
+      "loss": 0.3339,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9919354838709677,
+      "grad_norm": 2.146780014038086,
+      "learning_rate": 1.0618090015274603e-05,
+      "loss": 0.298,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9924098671726755,
+      "grad_norm": 2.24959659576416,
+      "learning_rate": 1.0610421489471748e-05,
+      "loss": 0.383,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9928842504743833,
+      "grad_norm": 2.0573537349700928,
+      "learning_rate": 1.0602752603343127e-05,
+      "loss": 0.3116,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9933586337760911,
+      "grad_norm": 2.7625508308410645,
+      "learning_rate": 1.059508336141561e-05,
+      "loss": 0.3006,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9938330170777988,
+      "grad_norm": 2.205559492111206,
+      "learning_rate": 1.0587413768216273e-05,
+      "loss": 0.3381,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9943074003795066,
+      "grad_norm": 2.6509761810302734,
+      "learning_rate": 1.05797438282724e-05,
+      "loss": 0.3785,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9947817836812144,
+      "grad_norm": 2.5124902725219727,
+      "learning_rate": 1.0572073546111485e-05,
+      "loss": 0.3773,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9952561669829222,
+      "grad_norm": 2.2544424533843994,
+      "learning_rate": 1.0564402926261216e-05,
+      "loss": 0.3657,
+      "step": 2098
+    },
+    {
+      "epoch": 0.99573055028463,
+      "grad_norm": 1.9105042219161987,
+      "learning_rate": 1.0556731973249486e-05,
+      "loss": 0.2686,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9962049335863378,
+      "grad_norm": 1.9611090421676636,
+      "learning_rate": 1.0549060691604376e-05,
+      "loss": 0.3119,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9966793168880456,
+      "grad_norm": 2.110646963119507,
+      "learning_rate": 1.0541389085854177e-05,
+      "loss": 0.3229,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9971537001897534,
+      "grad_norm": 2.017954111099243,
+      "learning_rate": 1.0533717160527357e-05,
+      "loss": 0.2829,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9976280834914611,
+      "grad_norm": 1.9769190549850464,
+      "learning_rate": 1.0526044920152578e-05,
+      "loss": 0.2776,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9981024667931688,
+      "grad_norm": 2.1241579055786133,
+      "learning_rate": 1.051837236925869e-05,
+      "loss": 0.3468,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9985768500948766,
+      "grad_norm": 2.512434720993042,
+      "learning_rate": 1.051069951237472e-05,
+      "loss": 0.2764,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9990512333965844,
+      "grad_norm": 2.3992347717285156,
+      "learning_rate": 1.0503026354029882e-05,
+      "loss": 0.3268,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9995256166982922,
+      "grad_norm": 2.2171542644500732,
+      "learning_rate": 1.0495352898753563e-05,
+      "loss": 0.329,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9980618953704834,
+      "learning_rate": 1.0487679151075332e-05,
+      "loss": 0.3509,
+      "step": 2108
+    },
+    {
+      "epoch": 1.0004743833017078,
+      "grad_norm": 2.0853278636932373,
+      "learning_rate": 1.0480005115524923e-05,
+      "loss": 0.2764,
+      "step": 2109
+    },
+    {
+      "epoch": 1.0009487666034156,
+      "grad_norm": 1.904092788696289,
+      "learning_rate": 1.0472330796632244e-05,
+      "loss": 0.26,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0014231499051234,
+      "grad_norm": 2.1483476161956787,
+      "learning_rate": 1.0464656198927373e-05,
+      "loss": 0.2206,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0018975332068312,
+      "grad_norm": 1.696155309677124,
+      "learning_rate": 1.0456981326940541e-05,
+      "loss": 0.1763,
+      "step": 2112
+    },
+    {
+      "epoch": 1.002371916508539,
+      "grad_norm": 1.9953058958053589,
+      "learning_rate": 1.0449306185202155e-05,
+      "loss": 0.1922,
+      "step": 2113
+    },
+    {
+      "epoch": 1.0028462998102468,
+      "grad_norm": 1.8557651042938232,
+      "learning_rate": 1.0441630778242775e-05,
+      "loss": 0.2421,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0033206831119545,
+      "grad_norm": 1.7121390104293823,
+      "learning_rate": 1.0433955110593115e-05,
+      "loss": 0.2016,
+      "step": 2115
+    },
+    {
+      "epoch": 1.0037950664136623,
+      "grad_norm": 1.799052357673645,
+      "learning_rate": 1.042627918678405e-05,
+      "loss": 0.1668,
+      "step": 2116
+    },
+    {
+      "epoch": 1.00426944971537,
+      "grad_norm": 1.8632630109786987,
+      "learning_rate": 1.04186030113466e-05,
+      "loss": 0.2101,
+      "step": 2117
+    },
+    {
+      "epoch": 1.0047438330170777,
+      "grad_norm": 2.4004900455474854,
+      "learning_rate": 1.0410926588811931e-05,
+      "loss": 0.2088,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0052182163187855,
+      "grad_norm": 1.4884165525436401,
+      "learning_rate": 1.0403249923711365e-05,
+      "loss": 0.1716,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0056925996204933,
+      "grad_norm": 1.6962333917617798,
+      "learning_rate": 1.0395573020576357e-05,
+      "loss": 0.1693,
+      "step": 2120
+    },
+    {
+      "epoch": 1.006166982922201,
+      "grad_norm": 1.9333925247192383,
+      "learning_rate": 1.038789588393851e-05,
+      "loss": 0.1512,
+      "step": 2121
+    },
+    {
+      "epoch": 1.0066413662239089,
+      "grad_norm": 2.6267874240875244,
+      "learning_rate": 1.0380218518329564e-05,
+      "loss": 0.2462,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0071157495256167,
+      "grad_norm": 1.9119246006011963,
+      "learning_rate": 1.037254092828139e-05,
+      "loss": 0.1687,
+      "step": 2123
+    },
+    {
+      "epoch": 1.0075901328273245,
+      "grad_norm": 2.043225049972534,
+      "learning_rate": 1.0364863118325988e-05,
+      "loss": 0.225,
+      "step": 2124
+    },
+    {
+      "epoch": 1.0080645161290323,
+      "grad_norm": 2.4549620151519775,
+      "learning_rate": 1.0357185092995499e-05,
+      "loss": 0.2173,
+      "step": 2125
+    },
+    {
+      "epoch": 1.00853889943074,
+      "grad_norm": 1.932741403579712,
+      "learning_rate": 1.0349506856822184e-05,
+      "loss": 0.2033,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0090132827324478,
+      "grad_norm": 2.0784912109375,
+      "learning_rate": 1.0341828414338431e-05,
+      "loss": 0.1974,
+      "step": 2127
+    },
+    {
+      "epoch": 1.0094876660341556,
+      "grad_norm": 1.9300061464309692,
+      "learning_rate": 1.0334149770076747e-05,
+      "loss": 0.1846,
+      "step": 2128
+    },
+    {
+      "epoch": 1.0099620493358634,
+      "grad_norm": 1.8968406915664673,
+      "learning_rate": 1.0326470928569758e-05,
+      "loss": 0.2126,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0104364326375712,
+      "grad_norm": 2.385936975479126,
+      "learning_rate": 1.0318791894350217e-05,
+      "loss": 0.2034,
+      "step": 2130
+    },
+    {
+      "epoch": 1.010910815939279,
+      "grad_norm": 2.1592063903808594,
+      "learning_rate": 1.0311112671950969e-05,
+      "loss": 0.2947,
+      "step": 2131
+    },
+    {
+      "epoch": 1.0113851992409868,
+      "grad_norm": 1.7248077392578125,
+      "learning_rate": 1.030343326590499e-05,
+      "loss": 0.158,
+      "step": 2132
+    },
+    {
+      "epoch": 1.0118595825426946,
+      "grad_norm": 2.433410167694092,
+      "learning_rate": 1.0295753680745359e-05,
+      "loss": 0.226,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0123339658444024,
+      "grad_norm": 1.5119600296020508,
+      "learning_rate": 1.0288073921005258e-05,
+      "loss": 0.1824,
+      "step": 2134
+    },
+    {
+      "epoch": 1.01280834914611,
+      "grad_norm": 1.8853365182876587,
+      "learning_rate": 1.028039399121797e-05,
+      "loss": 0.2013,
+      "step": 2135
+    },
+    {
+      "epoch": 1.0132827324478177,
+      "grad_norm": 1.7288734912872314,
+      "learning_rate": 1.0272713895916884e-05,
+      "loss": 0.1868,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0137571157495255,
+      "grad_norm": 1.6384727954864502,
+      "learning_rate": 1.0265033639635483e-05,
+      "loss": 0.178,
+      "step": 2137
+    },
+    {
+      "epoch": 1.0142314990512333,
+      "grad_norm": 1.747512936592102,
+      "learning_rate": 1.0257353226907349e-05,
+      "loss": 0.1839,
+      "step": 2138
+    },
+    {
+      "epoch": 1.0147058823529411,
+      "grad_norm": 2.3721799850463867,
+      "learning_rate": 1.0249672662266148e-05,
+      "loss": 0.1944,
+      "step": 2139
+    },
+    {
+      "epoch": 1.015180265654649,
+      "grad_norm": 2.0248069763183594,
+      "learning_rate": 1.0241991950245648e-05,
+      "loss": 0.2256,
+      "step": 2140
+    },
+    {
+      "epoch": 1.0156546489563567,
+      "grad_norm": 2.1707992553710938,
+      "learning_rate": 1.0234311095379694e-05,
+      "loss": 0.1732,
+      "step": 2141
+    },
+    {
+      "epoch": 1.0161290322580645,
+      "grad_norm": 2.033766746520996,
+      "learning_rate": 1.0226630102202221e-05,
+      "loss": 0.1923,
+      "step": 2142
+    },
+    {
+      "epoch": 1.0166034155597723,
+      "grad_norm": 1.76777184009552,
+      "learning_rate": 1.0218948975247238e-05,
+      "loss": 0.1628,
+      "step": 2143
+    },
+    {
+      "epoch": 1.01707779886148,
+      "grad_norm": 1.886703372001648,
+      "learning_rate": 1.021126771904884e-05,
+      "loss": 0.1511,
+      "step": 2144
+    },
+    {
+      "epoch": 1.0175521821631879,
+      "grad_norm": 1.8403862714767456,
+      "learning_rate": 1.0203586338141202e-05,
+      "loss": 0.1826,
+      "step": 2145
+    },
+    {
+      "epoch": 1.0180265654648957,
+      "grad_norm": 2.151848793029785,
+      "learning_rate": 1.0195904837058563e-05,
+      "loss": 0.2328,
+      "step": 2146
+    },
+    {
+      "epoch": 1.0185009487666035,
+      "grad_norm": 1.8675678968429565,
+      "learning_rate": 1.0188223220335238e-05,
+      "loss": 0.1969,
+      "step": 2147
+    },
+    {
+      "epoch": 1.0189753320683113,
+      "grad_norm": 2.2719991207122803,
+      "learning_rate": 1.0180541492505605e-05,
+      "loss": 0.2024,
+      "step": 2148
+    },
+    {
+      "epoch": 1.019449715370019,
+      "grad_norm": 3.2726147174835205,
+      "learning_rate": 1.0172859658104117e-05,
+      "loss": 0.2026,
+      "step": 2149
+    },
+    {
+      "epoch": 1.0199240986717268,
+      "grad_norm": 1.7774338722229004,
+      "learning_rate": 1.0165177721665284e-05,
+      "loss": 0.1555,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0203984819734346,
+      "grad_norm": 1.7717763185501099,
+      "learning_rate": 1.0157495687723675e-05,
+      "loss": 0.1982,
+      "step": 2151
+    },
+    {
+      "epoch": 1.0208728652751424,
+      "grad_norm": 1.9706634283065796,
+      "learning_rate": 1.0149813560813924e-05,
+      "loss": 0.1834,
+      "step": 2152
+    },
+    {
+      "epoch": 1.02134724857685,
+      "grad_norm": 1.683834433555603,
+      "learning_rate": 1.0142131345470704e-05,
+      "loss": 0.1814,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0218216318785578,
+      "grad_norm": 1.992972731590271,
+      "learning_rate": 1.0134449046228764e-05,
+      "loss": 0.2124,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0222960151802656,
+      "grad_norm": 1.861265778541565,
+      "learning_rate": 1.0126766667622878e-05,
+      "loss": 0.2043,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0227703984819734,
+      "grad_norm": 2.0793933868408203,
+      "learning_rate": 1.0119084214187882e-05,
+      "loss": 0.2147,
+      "step": 2156
+    },
+    {
+      "epoch": 1.0232447817836812,
+      "grad_norm": 1.6746715307235718,
+      "learning_rate": 1.0111401690458655e-05,
+      "loss": 0.2066,
+      "step": 2157
+    },
+    {
+      "epoch": 1.023719165085389,
+      "grad_norm": 1.7119837999343872,
+      "learning_rate": 1.0103719100970115e-05,
+      "loss": 0.161,
+      "step": 2158
+    },
+    {
+      "epoch": 1.0241935483870968,
+      "grad_norm": 2.014586925506592,
+      "learning_rate": 1.0096036450257214e-05,
+      "loss": 0.1854,
+      "step": 2159
+    },
+    {
+      "epoch": 1.0246679316888045,
+      "grad_norm": 1.4165558815002441,
+      "learning_rate": 1.0088353742854943e-05,
+      "loss": 0.1476,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0251423149905123,
+      "grad_norm": 1.4397838115692139,
+      "learning_rate": 1.0080670983298335e-05,
+      "loss": 0.1388,
+      "step": 2161
+    },
+    {
+      "epoch": 1.0256166982922201,
+      "grad_norm": 2.0614216327667236,
+      "learning_rate": 1.0072988176122445e-05,
+      "loss": 0.2106,
+      "step": 2162
+    },
+    {
+      "epoch": 1.026091081593928,
+      "grad_norm": 2.1988351345062256,
+      "learning_rate": 1.0065305325862354e-05,
+      "loss": 0.1489,
+      "step": 2163
+    },
+    {
+      "epoch": 1.0265654648956357,
+      "grad_norm": 1.5647263526916504,
+      "learning_rate": 1.0057622437053178e-05,
+      "loss": 0.1348,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0270398481973435,
+      "grad_norm": 1.8337029218673706,
+      "learning_rate": 1.0049939514230045e-05,
+      "loss": 0.1559,
+      "step": 2165
+    },
+    {
+      "epoch": 1.0275142314990513,
+      "grad_norm": 1.588962435722351,
+      "learning_rate": 1.0042256561928115e-05,
+      "loss": 0.1418,
+      "step": 2166
+    },
+    {
+      "epoch": 1.027988614800759,
+      "grad_norm": 1.8072073459625244,
+      "learning_rate": 1.003457358468255e-05,
+      "loss": 0.1858,
+      "step": 2167
+    },
+    {
+      "epoch": 1.0284629981024669,
+      "grad_norm": 2.0580496788024902,
+      "learning_rate": 1.0026890587028544e-05,
+      "loss": 0.2082,
+      "step": 2168
+    },
+    {
+      "epoch": 1.0289373814041747,
+      "grad_norm": 2.101830244064331,
+      "learning_rate": 1.0019207573501287e-05,
+      "loss": 0.2083,
+      "step": 2169
+    },
+    {
+      "epoch": 1.0294117647058822,
+      "grad_norm": 1.8399953842163086,
+      "learning_rate": 1.001152454863599e-05,
+      "loss": 0.1713,
+      "step": 2170
+    },
+    {
+      "epoch": 1.02988614800759,
+      "grad_norm": 1.6996338367462158,
+      "learning_rate": 1.000384151696787e-05,
+      "loss": 0.1551,
+      "step": 2171
+    },
+    {
+      "epoch": 1.0303605313092978,
+      "grad_norm": 1.8706146478652954,
+      "learning_rate": 9.996158483032137e-06,
+      "loss": 0.2203,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0308349146110056,
+      "grad_norm": 2.099688768386841,
+      "learning_rate": 9.988475451364013e-06,
+      "loss": 0.2117,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0313092979127134,
+      "grad_norm": 2.176435947418213,
+      "learning_rate": 9.980792426498717e-06,
+      "loss": 0.2196,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0317836812144212,
+      "grad_norm": 2.1011524200439453,
+      "learning_rate": 9.97310941297146e-06,
+      "loss": 0.1895,
+      "step": 2175
+    },
+    {
+      "epoch": 1.032258064516129,
+      "grad_norm": 1.816577434539795,
+      "learning_rate": 9.965426415317451e-06,
+      "loss": 0.1829,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0327324478178368,
+      "grad_norm": 2.274826765060425,
+      "learning_rate": 9.95774343807189e-06,
+      "loss": 0.167,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0332068311195446,
+      "grad_norm": 1.8276028633117676,
+      "learning_rate": 9.950060485769958e-06,
+      "loss": 0.1813,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0336812144212524,
+      "grad_norm": 1.7824639081954956,
+      "learning_rate": 9.942377562946825e-06,
+      "loss": 0.1577,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0341555977229602,
+      "grad_norm": 1.696703314781189,
+      "learning_rate": 9.934694674137648e-06,
+      "loss": 0.1729,
+      "step": 2180
+    },
+    {
+      "epoch": 1.034629981024668,
+      "grad_norm": 1.851335048675537,
+      "learning_rate": 9.927011823877559e-06,
+      "loss": 0.2252,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0351043643263758,
+      "grad_norm": 1.9400562047958374,
+      "learning_rate": 9.919329016701668e-06,
+      "loss": 0.2241,
+      "step": 2182
+    },
+    {
+      "epoch": 1.0355787476280836,
+      "grad_norm": 1.9708709716796875,
+      "learning_rate": 9.91164625714506e-06,
+      "loss": 0.2549,
+      "step": 2183
+    },
+    {
+      "epoch": 1.0360531309297913,
+      "grad_norm": 1.7907429933547974,
+      "learning_rate": 9.90396354974279e-06,
+      "loss": 0.1807,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0365275142314991,
+      "grad_norm": 1.606693148612976,
+      "learning_rate": 9.896280899029887e-06,
+      "loss": 0.1502,
+      "step": 2185
+    },
+    {
+      "epoch": 1.037001897533207,
+      "grad_norm": 2.3423125743865967,
+      "learning_rate": 9.888598309541347e-06,
+      "loss": 0.1645,
+      "step": 2186
+    },
+    {
+      "epoch": 1.0374762808349147,
+      "grad_norm": 1.9627526998519897,
+      "learning_rate": 9.88091578581212e-06,
+      "loss": 0.1701,
+      "step": 2187
+    },
+    {
+      "epoch": 1.0379506641366223,
+      "grad_norm": 1.9357240200042725,
+      "learning_rate": 9.873233332377125e-06,
+      "loss": 0.2047,
+      "step": 2188
+    },
+    {
+      "epoch": 1.03842504743833,
+      "grad_norm": 2.164787530899048,
+      "learning_rate": 9.865550953771237e-06,
+      "loss": 0.2643,
+      "step": 2189
+    },
+    {
+      "epoch": 1.0388994307400379,
+      "grad_norm": 2.26877498626709,
+      "learning_rate": 9.857868654529296e-06,
+      "loss": 0.2058,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0393738140417457,
+      "grad_norm": 1.9010010957717896,
+      "learning_rate": 9.850186439186083e-06,
+      "loss": 0.2138,
+      "step": 2191
+    },
+    {
+      "epoch": 1.0398481973434535,
+      "grad_norm": 1.56143057346344,
+      "learning_rate": 9.842504312276326e-06,
+      "loss": 0.1677,
+      "step": 2192
+    },
+    {
+      "epoch": 1.0403225806451613,
+      "grad_norm": 1.7175490856170654,
+      "learning_rate": 9.834822278334718e-06,
+      "loss": 0.2035,
+      "step": 2193
+    },
+    {
+      "epoch": 1.040796963946869,
+      "grad_norm": 1.5723693370819092,
+      "learning_rate": 9.827140341895885e-06,
+      "loss": 0.1521,
+      "step": 2194
+    },
+    {
+      "epoch": 1.0412713472485768,
+      "grad_norm": 1.6594966650009155,
+      "learning_rate": 9.819458507494395e-06,
+      "loss": 0.2055,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0417457305502846,
+      "grad_norm": 2.062875747680664,
+      "learning_rate": 9.811776779664767e-06,
+      "loss": 0.1937,
+      "step": 2196
+    },
+    {
+      "epoch": 1.0422201138519924,
+      "grad_norm": 4.114375591278076,
+      "learning_rate": 9.804095162941439e-06,
+      "loss": 0.3022,
+      "step": 2197
+    },
+    {
+      "epoch": 1.0426944971537002,
+      "grad_norm": 2.6715476512908936,
+      "learning_rate": 9.7964136618588e-06,
+      "loss": 0.2538,
+      "step": 2198
+    },
+    {
+      "epoch": 1.043168880455408,
+      "grad_norm": 2.0732991695404053,
+      "learning_rate": 9.788732280951158e-06,
+      "loss": 0.2196,
+      "step": 2199
+    },
+    {
+      "epoch": 1.0436432637571158,
+      "grad_norm": 2.420642375946045,
+      "learning_rate": 9.781051024752762e-06,
+      "loss": 0.2222,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0441176470588236,
+      "grad_norm": 1.8388843536376953,
+      "learning_rate": 9.773369897797784e-06,
+      "loss": 0.2209,
+      "step": 2201
+    },
+    {
+      "epoch": 1.0445920303605314,
+      "grad_norm": 1.8891000747680664,
+      "learning_rate": 9.76568890462031e-06,
+      "loss": 0.1758,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0450664136622392,
+      "grad_norm": 1.7700624465942383,
+      "learning_rate": 9.758008049754353e-06,
+      "loss": 0.1815,
+      "step": 2203
+    },
+    {
+      "epoch": 1.045540796963947,
+      "grad_norm": 2.6104068756103516,
+      "learning_rate": 9.750327337733852e-06,
+      "loss": 0.2194,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0460151802656545,
+      "grad_norm": 1.9325884580612183,
+      "learning_rate": 9.742646773092656e-06,
+      "loss": 0.1886,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0464895635673623,
+      "grad_norm": 2.4706602096557617,
+      "learning_rate": 9.734966360364519e-06,
+      "loss": 0.2789,
+      "step": 2206
+    },
+    {
+      "epoch": 1.0469639468690701,
+      "grad_norm": 2.077040195465088,
+      "learning_rate": 9.72728610408312e-06,
+      "loss": 0.2351,
+      "step": 2207
+    },
+    {
+      "epoch": 1.047438330170778,
+      "grad_norm": 1.9420323371887207,
+      "learning_rate": 9.719606008782031e-06,
+      "loss": 0.2085,
+      "step": 2208
+    },
+    {
+      "epoch": 1.0479127134724857,
+      "grad_norm": 1.9051023721694946,
+      "learning_rate": 9.711926078994744e-06,
+      "loss": 0.2164,
+      "step": 2209
+    },
+    {
+      "epoch": 1.0483870967741935,
+      "grad_norm": 1.6898927688598633,
+      "learning_rate": 9.704246319254643e-06,
+      "loss": 0.1834,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0488614800759013,
+      "grad_norm": 1.6880271434783936,
+      "learning_rate": 9.696566734095012e-06,
+      "loss": 0.1787,
+      "step": 2211
+    },
+    {
+      "epoch": 1.049335863377609,
+      "grad_norm": 1.8380780220031738,
+      "learning_rate": 9.688887328049034e-06,
+      "loss": 0.1836,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0498102466793169,
+      "grad_norm": 1.8687678575515747,
+      "learning_rate": 9.681208105649786e-06,
+      "loss": 0.2263,
+      "step": 2213
+    },
+    {
+      "epoch": 1.0502846299810247,
+      "grad_norm": 2.063650608062744,
+      "learning_rate": 9.673529071430242e-06,
+      "loss": 0.1881,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0507590132827325,
+      "grad_norm": 2.072175979614258,
+      "learning_rate": 9.665850229923258e-06,
+      "loss": 0.1931,
+      "step": 2215
+    },
+    {
+      "epoch": 1.0512333965844403,
+      "grad_norm": 2.0182647705078125,
+      "learning_rate": 9.658171585661572e-06,
+      "loss": 0.2125,
+      "step": 2216
+    },
+    {
+      "epoch": 1.051707779886148,
+      "grad_norm": 1.7542519569396973,
+      "learning_rate": 9.650493143177817e-06,
+      "loss": 0.1789,
+      "step": 2217
+    },
+    {
+      "epoch": 1.0521821631878558,
+      "grad_norm": 1.796041488647461,
+      "learning_rate": 9.642814907004505e-06,
+      "loss": 0.1936,
+      "step": 2218
+    },
+    {
+      "epoch": 1.0526565464895636,
+      "grad_norm": 2.394157648086548,
+      "learning_rate": 9.635136881674013e-06,
+      "loss": 0.2391,
+      "step": 2219
+    },
+    {
+      "epoch": 1.0531309297912714,
+      "grad_norm": 2.1615726947784424,
+      "learning_rate": 9.627459071718617e-06,
+      "loss": 0.2023,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0536053130929792,
+      "grad_norm": 2.2835464477539062,
+      "learning_rate": 9.619781481670437e-06,
+      "loss": 0.2261,
+      "step": 2221
+    },
+    {
+      "epoch": 1.054079696394687,
+      "grad_norm": 1.822164535522461,
+      "learning_rate": 9.612104116061491e-06,
+      "loss": 0.2148,
+      "step": 2222
+    },
+    {
+      "epoch": 1.0545540796963946,
+      "grad_norm": 1.8478764295578003,
+      "learning_rate": 9.604426979423645e-06,
+      "loss": 0.1656,
+      "step": 2223
+    },
+    {
+      "epoch": 1.0550284629981024,
+      "grad_norm": 2.215240001678467,
+      "learning_rate": 9.596750076288642e-06,
+      "loss": 0.2309,
+      "step": 2224
+    },
+    {
+      "epoch": 1.0555028462998102,
+      "grad_norm": 1.7701280117034912,
+      "learning_rate": 9.589073411188074e-06,
+      "loss": 0.2403,
+      "step": 2225
+    },
+    {
+      "epoch": 1.055977229601518,
+      "grad_norm": 1.9893231391906738,
+      "learning_rate": 9.581396988653404e-06,
+      "loss": 0.2326,
+      "step": 2226
+    },
+    {
+      "epoch": 1.0564516129032258,
+      "grad_norm": 1.8492276668548584,
+      "learning_rate": 9.573720813215954e-06,
+      "loss": 0.1864,
+      "step": 2227
+    },
+    {
+      "epoch": 1.0569259962049335,
+      "grad_norm": 1.7757240533828735,
+      "learning_rate": 9.566044889406885e-06,
+      "loss": 0.2256,
+      "step": 2228
+    },
+    {
+      "epoch": 1.0574003795066413,
+      "grad_norm": 1.9123115539550781,
+      "learning_rate": 9.558369221757229e-06,
+      "loss": 0.2317,
+      "step": 2229
+    },
+    {
+      "epoch": 1.0578747628083491,
+      "grad_norm": 2.3356640338897705,
+      "learning_rate": 9.550693814797847e-06,
+      "loss": 0.2255,
+      "step": 2230
+    },
+    {
+      "epoch": 1.058349146110057,
+      "grad_norm": 1.911953330039978,
+      "learning_rate": 9.543018673059462e-06,
+      "loss": 0.192,
+      "step": 2231
+    },
+    {
+      "epoch": 1.0588235294117647,
+      "grad_norm": 2.0029706954956055,
+      "learning_rate": 9.53534380107263e-06,
+      "loss": 0.2193,
+      "step": 2232
+    },
+    {
+      "epoch": 1.0592979127134725,
+      "grad_norm": 1.5258057117462158,
+      "learning_rate": 9.527669203367756e-06,
+      "loss": 0.1815,
+      "step": 2233
+    },
+    {
+      "epoch": 1.0597722960151803,
+      "grad_norm": 1.8638172149658203,
+      "learning_rate": 9.51999488447508e-06,
+      "loss": 0.1792,
+      "step": 2234
+    },
+    {
+      "epoch": 1.060246679316888,
+      "grad_norm": 2.022976875305176,
+      "learning_rate": 9.512320848924672e-06,
+      "loss": 0.2382,
+      "step": 2235
+    },
+    {
+      "epoch": 1.060721062618596,
+      "grad_norm": 1.830183744430542,
+      "learning_rate": 9.504647101246438e-06,
+      "loss": 0.207,
+      "step": 2236
+    },
+    {
+      "epoch": 1.0611954459203037,
+      "grad_norm": 1.8549655675888062,
+      "learning_rate": 9.49697364597012e-06,
+      "loss": 0.2103,
+      "step": 2237
+    },
+    {
+      "epoch": 1.0616698292220115,
+      "grad_norm": 2.056288719177246,
+      "learning_rate": 9.489300487625283e-06,
+      "loss": 0.2282,
+      "step": 2238
+    },
+    {
+      "epoch": 1.0621442125237193,
+      "grad_norm": 1.720819354057312,
+      "learning_rate": 9.481627630741315e-06,
+      "loss": 0.1979,
+      "step": 2239
+    },
+    {
+      "epoch": 1.0626185958254268,
+      "grad_norm": 1.9592931270599365,
+      "learning_rate": 9.473955079847426e-06,
+      "loss": 0.2263,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0630929791271346,
+      "grad_norm": 1.7493445873260498,
+      "learning_rate": 9.466282839472645e-06,
+      "loss": 0.1799,
+      "step": 2241
+    },
+    {
+      "epoch": 1.0635673624288424,
+      "grad_norm": 2.2582602500915527,
+      "learning_rate": 9.458610914145826e-06,
+      "loss": 0.1552,
+      "step": 2242
+    },
+    {
+      "epoch": 1.0640417457305502,
+      "grad_norm": 1.6923972368240356,
+      "learning_rate": 9.450939308395629e-06,
+      "loss": 0.1921,
+      "step": 2243
+    },
+    {
+      "epoch": 1.064516129032258,
+      "grad_norm": 1.8049591779708862,
+      "learning_rate": 9.443268026750521e-06,
+      "loss": 0.2018,
+      "step": 2244
+    },
+    {
+      "epoch": 1.0649905123339658,
+      "grad_norm": 1.8883256912231445,
+      "learning_rate": 9.435597073738787e-06,
+      "loss": 0.1884,
+      "step": 2245
+    },
+    {
+      "epoch": 1.0654648956356736,
+      "grad_norm": 2.0094857215881348,
+      "learning_rate": 9.427926453888518e-06,
+      "loss": 0.2109,
+      "step": 2246
+    },
+    {
+      "epoch": 1.0659392789373814,
+      "grad_norm": 1.975786566734314,
+      "learning_rate": 9.4202561717276e-06,
+      "loss": 0.1851,
+      "step": 2247
+    },
+    {
+      "epoch": 1.0664136622390892,
+      "grad_norm": 1.8108450174331665,
+      "learning_rate": 9.41258623178373e-06,
+      "loss": 0.1696,
+      "step": 2248
+    },
+    {
+      "epoch": 1.066888045540797,
+      "grad_norm": 2.1892595291137695,
+      "learning_rate": 9.404916638584394e-06,
+      "loss": 0.1848,
+      "step": 2249
+    },
+    {
+      "epoch": 1.0673624288425048,
+      "grad_norm": 2.6319334506988525,
+      "learning_rate": 9.397247396656875e-06,
+      "loss": 0.217,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0678368121442126,
+      "grad_norm": 1.9110881090164185,
+      "learning_rate": 9.389578510528256e-06,
+      "loss": 0.1634,
+      "step": 2251
+    },
+    {
+      "epoch": 1.0683111954459203,
+      "grad_norm": 1.8657715320587158,
+      "learning_rate": 9.381909984725399e-06,
+      "loss": 0.1685,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0687855787476281,
+      "grad_norm": 1.8046596050262451,
+      "learning_rate": 9.374241823774967e-06,
+      "loss": 0.1784,
+      "step": 2253
+    },
+    {
+      "epoch": 1.069259962049336,
+      "grad_norm": 1.8186533451080322,
+      "learning_rate": 9.36657403220339e-06,
+      "loss": 0.1822,
+      "step": 2254
+    },
+    {
+      "epoch": 1.0697343453510437,
+      "grad_norm": 2.0945181846618652,
+      "learning_rate": 9.358906614536895e-06,
+      "loss": 0.2373,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0702087286527515,
+      "grad_norm": 2.1683075428009033,
+      "learning_rate": 9.351239575301479e-06,
+      "loss": 0.2144,
+      "step": 2256
+    },
+    {
+      "epoch": 1.0706831119544593,
+      "grad_norm": 1.9169727563858032,
+      "learning_rate": 9.343572919022924e-06,
+      "loss": 0.2189,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0711574952561669,
+      "grad_norm": 1.7258111238479614,
+      "learning_rate": 9.335906650226783e-06,
+      "loss": 0.1937,
+      "step": 2258
+    },
+    {
+      "epoch": 1.0716318785578747,
+      "grad_norm": 2.077392816543579,
+      "learning_rate": 9.328240773438372e-06,
+      "loss": 0.1898,
+      "step": 2259
+    },
+    {
+      "epoch": 1.0721062618595825,
+      "grad_norm": 1.833132028579712,
+      "learning_rate": 9.320575293182782e-06,
+      "loss": 0.2103,
+      "step": 2260
+    },
+    {
+      "epoch": 1.0725806451612903,
+      "grad_norm": 1.7154309749603271,
+      "learning_rate": 9.312910213984876e-06,
+      "loss": 0.1733,
+      "step": 2261
+    },
+    {
+      "epoch": 1.073055028462998,
+      "grad_norm": 1.8700604438781738,
+      "learning_rate": 9.305245540369273e-06,
+      "loss": 0.2193,
+      "step": 2262
+    },
+    {
+      "epoch": 1.0735294117647058,
+      "grad_norm": 1.6208839416503906,
+      "learning_rate": 9.297581276860353e-06,
+      "loss": 0.1841,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0740037950664136,
+      "grad_norm": 1.4968012571334839,
+      "learning_rate": 9.28991742798225e-06,
+      "loss": 0.16,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0744781783681214,
+      "grad_norm": 1.8390696048736572,
+      "learning_rate": 9.282253998258865e-06,
+      "loss": 0.2012,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0749525616698292,
+      "grad_norm": 1.7694889307022095,
+      "learning_rate": 9.274590992213844e-06,
+      "loss": 0.1982,
+      "step": 2266
+    },
+    {
+      "epoch": 1.075426944971537,
+      "grad_norm": 1.4749209880828857,
+      "learning_rate": 9.26692841437058e-06,
+      "loss": 0.1581,
+      "step": 2267
+    },
+    {
+      "epoch": 1.0759013282732448,
+      "grad_norm": 1.6622833013534546,
+      "learning_rate": 9.259266269252221e-06,
+      "loss": 0.1919,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0763757115749526,
+      "grad_norm": 1.8222259283065796,
+      "learning_rate": 9.25160456138165e-06,
+      "loss": 0.1752,
+      "step": 2269
+    },
+    {
+      "epoch": 1.0768500948766604,
+      "grad_norm": 2.1095898151397705,
+      "learning_rate": 9.243943295281505e-06,
+      "loss": 0.1706,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0773244781783682,
+      "grad_norm": 1.9356722831726074,
+      "learning_rate": 9.236282475474146e-06,
+      "loss": 0.2063,
+      "step": 2271
+    },
+    {
+      "epoch": 1.077798861480076,
+      "grad_norm": 2.4067156314849854,
+      "learning_rate": 9.228622106481691e-06,
+      "loss": 0.188,
+      "step": 2272
+    },
+    {
+      "epoch": 1.0782732447817838,
+      "grad_norm": 3.1411545276641846,
+      "learning_rate": 9.22096219282597e-06,
+      "loss": 0.2492,
+      "step": 2273
+    },
+    {
+      "epoch": 1.0787476280834916,
+      "grad_norm": 1.8042513132095337,
+      "learning_rate": 9.213302739028555e-06,
+      "loss": 0.1861,
+      "step": 2274
+    },
+    {
+      "epoch": 1.0792220113851991,
+      "grad_norm": 1.792336106300354,
+      "learning_rate": 9.205643749610751e-06,
+      "loss": 0.1982,
+      "step": 2275
+    },
+    {
+      "epoch": 1.079696394686907,
+      "grad_norm": 3.020674228668213,
+      "learning_rate": 9.19798522909358e-06,
+      "loss": 0.1797,
+      "step": 2276
+    },
+    {
+      "epoch": 1.0801707779886147,
+      "grad_norm": 1.7047805786132812,
+      "learning_rate": 9.190327181997796e-06,
+      "loss": 0.1772,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0806451612903225,
+      "grad_norm": 2.018634080886841,
+      "learning_rate": 9.182669612843861e-06,
+      "loss": 0.1895,
+      "step": 2278
+    },
+    {
+      "epoch": 1.0811195445920303,
+      "grad_norm": 1.8824384212493896,
+      "learning_rate": 9.175012526151968e-06,
+      "loss": 0.1718,
+      "step": 2279
+    },
+    {
+      "epoch": 1.081593927893738,
+      "grad_norm": 1.511277437210083,
+      "learning_rate": 9.167355926442013e-06,
+      "loss": 0.1604,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0820683111954459,
+      "grad_norm": 1.670013427734375,
+      "learning_rate": 9.15969981823362e-06,
+      "loss": 0.1973,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0825426944971537,
+      "grad_norm": 1.8747272491455078,
+      "learning_rate": 9.152044206046106e-06,
+      "loss": 0.1461,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0830170777988615,
+      "grad_norm": 1.7594618797302246,
+      "learning_rate": 9.144389094398508e-06,
+      "loss": 0.1442,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0834914611005693,
+      "grad_norm": 2.149158477783203,
+      "learning_rate": 9.136734487809559e-06,
+      "loss": 0.2489,
+      "step": 2284
+    },
+    {
+      "epoch": 1.083965844402277,
+      "grad_norm": 1.746847152709961,
+      "learning_rate": 9.129080390797699e-06,
+      "loss": 0.1697,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0844402277039848,
+      "grad_norm": 1.7937593460083008,
+      "learning_rate": 9.12142680788107e-06,
+      "loss": 0.166,
+      "step": 2286
+    },
+    {
+      "epoch": 1.0849146110056926,
+      "grad_norm": 2.1846930980682373,
+      "learning_rate": 9.113773743577502e-06,
+      "loss": 0.1949,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0853889943074004,
+      "grad_norm": 2.1543169021606445,
+      "learning_rate": 9.106121202404521e-06,
+      "loss": 0.2145,
+      "step": 2288
+    },
+    {
+      "epoch": 1.0858633776091082,
+      "grad_norm": 1.6700936555862427,
+      "learning_rate": 9.098469188879348e-06,
+      "loss": 0.164,
+      "step": 2289
+    },
+    {
+      "epoch": 1.086337760910816,
+      "grad_norm": 1.5210555791854858,
+      "learning_rate": 9.090817707518893e-06,
+      "loss": 0.1588,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0868121442125238,
+      "grad_norm": 1.8209552764892578,
+      "learning_rate": 9.083166762839751e-06,
+      "loss": 0.2126,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0872865275142316,
+      "grad_norm": 1.954764485359192,
+      "learning_rate": 9.075516359358195e-06,
+      "loss": 0.1862,
+      "step": 2292
+    },
+    {
+      "epoch": 1.0877609108159392,
+      "grad_norm": 1.4143950939178467,
+      "learning_rate": 9.067866501590182e-06,
+      "loss": 0.1454,
+      "step": 2293
+    },
+    {
+      "epoch": 1.088235294117647,
+      "grad_norm": 2.105008363723755,
+      "learning_rate": 9.06021719405135e-06,
+      "loss": 0.2585,
+      "step": 2294
+    },
+    {
+      "epoch": 1.0887096774193548,
+      "grad_norm": 2.2465155124664307,
+      "learning_rate": 9.05256844125701e-06,
+      "loss": 0.2133,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0891840607210626,
+      "grad_norm": 2.175455331802368,
+      "learning_rate": 9.044920247722146e-06,
+      "loss": 0.2389,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0896584440227703,
+      "grad_norm": 1.8581993579864502,
+      "learning_rate": 9.037272617961405e-06,
+      "loss": 0.2035,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0901328273244781,
+      "grad_norm": 1.7983685731887817,
+      "learning_rate": 9.02962555648911e-06,
+      "loss": 0.2225,
+      "step": 2298
+    },
+    {
+      "epoch": 1.090607210626186,
+      "grad_norm": 1.98671293258667,
+      "learning_rate": 9.021979067819252e-06,
+      "loss": 0.2191,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0910815939278937,
+      "grad_norm": 1.4009077548980713,
+      "learning_rate": 9.014333156465467e-06,
+      "loss": 0.1482,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0915559772296015,
+      "grad_norm": 1.6995691061019897,
+      "learning_rate": 9.00668782694107e-06,
+      "loss": 0.1738,
+      "step": 2301
+    },
+    {
+      "epoch": 1.0920303605313093,
+      "grad_norm": 1.6954809427261353,
+      "learning_rate": 8.999043083759016e-06,
+      "loss": 0.1835,
+      "step": 2302
+    },
+    {
+      "epoch": 1.092504743833017,
+      "grad_norm": 1.936194896697998,
+      "learning_rate": 8.99139893143193e-06,
+      "loss": 0.2194,
+      "step": 2303
+    },
+    {
+      "epoch": 1.092979127134725,
+      "grad_norm": 1.9008138179779053,
+      "learning_rate": 8.983755374472069e-06,
+      "loss": 0.2544,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0934535104364327,
+      "grad_norm": 1.8978139162063599,
+      "learning_rate": 8.976112417391358e-06,
+      "loss": 0.2117,
+      "step": 2305
+    },
+    {
+      "epoch": 1.0939278937381405,
+      "grad_norm": 1.6767897605895996,
+      "learning_rate": 8.968470064701354e-06,
+      "loss": 0.1708,
+      "step": 2306
+    },
+    {
+      "epoch": 1.0944022770398483,
+      "grad_norm": 2.310898780822754,
+      "learning_rate": 8.960828320913263e-06,
+      "loss": 0.2313,
+      "step": 2307
+    },
+    {
+      "epoch": 1.094876660341556,
+      "grad_norm": 1.91811203956604,
+      "learning_rate": 8.953187190537929e-06,
+      "loss": 0.1858,
+      "step": 2308
+    },
+    {
+      "epoch": 1.0953510436432639,
+      "grad_norm": 1.711073398590088,
+      "learning_rate": 8.945546678085838e-06,
+      "loss": 0.183,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0958254269449714,
+      "grad_norm": 1.898019552230835,
+      "learning_rate": 8.937906788067114e-06,
+      "loss": 0.1654,
+      "step": 2310
+    },
+    {
+      "epoch": 1.0962998102466792,
+      "grad_norm": 1.8163573741912842,
+      "learning_rate": 8.9302675249915e-06,
+      "loss": 0.1461,
+      "step": 2311
+    },
+    {
+      "epoch": 1.096774193548387,
+      "grad_norm": 2.2861268520355225,
+      "learning_rate": 8.922628893368378e-06,
+      "loss": 0.2169,
+      "step": 2312
+    },
+    {
+      "epoch": 1.0972485768500948,
+      "grad_norm": 1.6889331340789795,
+      "learning_rate": 8.91499089770676e-06,
+      "loss": 0.1876,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0977229601518026,
+      "grad_norm": 1.7860161066055298,
+      "learning_rate": 8.90735354251528e-06,
+      "loss": 0.1808,
+      "step": 2314
+    },
+    {
+      "epoch": 1.0981973434535104,
+      "grad_norm": 1.4944920539855957,
+      "learning_rate": 8.899716832302193e-06,
+      "loss": 0.142,
+      "step": 2315
+    },
+    {
+      "epoch": 1.0986717267552182,
+      "grad_norm": 2.013535737991333,
+      "learning_rate": 8.89208077157537e-06,
+      "loss": 0.2055,
+      "step": 2316
+    },
+    {
+      "epoch": 1.099146110056926,
+      "grad_norm": 1.811508059501648,
+      "learning_rate": 8.884445364842304e-06,
+      "loss": 0.199,
+      "step": 2317
+    },
+    {
+      "epoch": 1.0996204933586338,
+      "grad_norm": 2.4114105701446533,
+      "learning_rate": 8.8768106166101e-06,
+      "loss": 0.2022,
+      "step": 2318
+    },
+    {
+      "epoch": 1.1000948766603416,
+      "grad_norm": 2.2680139541625977,
+      "learning_rate": 8.869176531385476e-06,
+      "loss": 0.2317,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1005692599620494,
+      "grad_norm": 1.6140555143356323,
+      "learning_rate": 8.861543113674758e-06,
+      "loss": 0.1574,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1010436432637571,
+      "grad_norm": 1.9578766822814941,
+      "learning_rate": 8.853910367983871e-06,
+      "loss": 0.1904,
+      "step": 2321
+    },
+    {
+      "epoch": 1.101518026565465,
+      "grad_norm": 2.459045886993408,
+      "learning_rate": 8.846278298818352e-06,
+      "loss": 0.1852,
+      "step": 2322
+    },
+    {
+      "epoch": 1.1019924098671727,
+      "grad_norm": 1.5891822576522827,
+      "learning_rate": 8.838646910683338e-06,
+      "loss": 0.1727,
+      "step": 2323
+    },
+    {
+      "epoch": 1.1024667931688805,
+      "grad_norm": 1.923692226409912,
+      "learning_rate": 8.831016208083563e-06,
+      "loss": 0.1874,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1029411764705883,
+      "grad_norm": 1.6137306690216064,
+      "learning_rate": 8.82338619552335e-06,
+      "loss": 0.1609,
+      "step": 2325
+    },
+    {
+      "epoch": 1.103415559772296,
+      "grad_norm": 2.0651705265045166,
+      "learning_rate": 8.815756877506622e-06,
+      "loss": 0.1954,
+      "step": 2326
+    },
+    {
+      "epoch": 1.103889943074004,
+      "grad_norm": 1.705381155014038,
+      "learning_rate": 8.808128258536893e-06,
+      "loss": 0.2241,
+      "step": 2327
+    },
+    {
+      "epoch": 1.1043643263757117,
+      "grad_norm": 1.756785273551941,
+      "learning_rate": 8.800500343117255e-06,
+      "loss": 0.1729,
+      "step": 2328
+    },
+    {
+      "epoch": 1.1048387096774193,
+      "grad_norm": 2.024177312850952,
+      "learning_rate": 8.7928731357504e-06,
+      "loss": 0.223,
+      "step": 2329
+    },
+    {
+      "epoch": 1.105313092979127,
+      "grad_norm": 1.7107657194137573,
+      "learning_rate": 8.785246640938584e-06,
+      "loss": 0.1882,
+      "step": 2330
+    },
+    {
+      "epoch": 1.1057874762808348,
+      "grad_norm": 2.2373104095458984,
+      "learning_rate": 8.777620863183658e-06,
+      "loss": 0.2527,
+      "step": 2331
+    },
+    {
+      "epoch": 1.1062618595825426,
+      "grad_norm": 1.823649525642395,
+      "learning_rate": 8.769995806987037e-06,
+      "loss": 0.2164,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1067362428842504,
+      "grad_norm": 1.7617485523223877,
+      "learning_rate": 8.762371476849722e-06,
+      "loss": 0.1942,
+      "step": 2333
+    },
+    {
+      "epoch": 1.1072106261859582,
+      "grad_norm": 1.7329638004302979,
+      "learning_rate": 8.754747877272279e-06,
+      "loss": 0.1631,
+      "step": 2334
+    },
+    {
+      "epoch": 1.107685009487666,
+      "grad_norm": 1.7115880250930786,
+      "learning_rate": 8.747125012754839e-06,
+      "loss": 0.174,
+      "step": 2335
+    },
+    {
+      "epoch": 1.1081593927893738,
+      "grad_norm": 1.9890222549438477,
+      "learning_rate": 8.739502887797108e-06,
+      "loss": 0.1825,
+      "step": 2336
+    },
+    {
+      "epoch": 1.1086337760910816,
+      "grad_norm": 1.9389824867248535,
+      "learning_rate": 8.731881506898348e-06,
+      "loss": 0.1788,
+      "step": 2337
+    },
+    {
+      "epoch": 1.1091081593927894,
+      "grad_norm": 1.6447457075119019,
+      "learning_rate": 8.724260874557384e-06,
+      "loss": 0.1562,
+      "step": 2338
+    },
+    {
+      "epoch": 1.1095825426944972,
+      "grad_norm": 1.9594886302947998,
+      "learning_rate": 8.716640995272607e-06,
+      "loss": 0.2101,
+      "step": 2339
+    },
+    {
+      "epoch": 1.110056925996205,
+      "grad_norm": 1.9339709281921387,
+      "learning_rate": 8.70902187354195e-06,
+      "loss": 0.2179,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1105313092979128,
+      "grad_norm": 1.7604994773864746,
+      "learning_rate": 8.701403513862901e-06,
+      "loss": 0.1864,
+      "step": 2341
+    },
+    {
+      "epoch": 1.1110056925996206,
+      "grad_norm": 1.8655658960342407,
+      "learning_rate": 8.69378592073251e-06,
+      "loss": 0.1842,
+      "step": 2342
+    },
+    {
+      "epoch": 1.1114800759013284,
+      "grad_norm": 1.4417341947555542,
+      "learning_rate": 8.68616909864737e-06,
+      "loss": 0.1523,
+      "step": 2343
+    },
+    {
+      "epoch": 1.1119544592030361,
+      "grad_norm": 1.832533836364746,
+      "learning_rate": 8.678553052103605e-06,
+      "loss": 0.1856,
+      "step": 2344
+    },
+    {
+      "epoch": 1.1124288425047437,
+      "grad_norm": 2.599240303039551,
+      "learning_rate": 8.670937785596897e-06,
+      "loss": 0.1896,
+      "step": 2345
+    },
+    {
+      "epoch": 1.1129032258064515,
+      "grad_norm": 1.5967415571212769,
+      "learning_rate": 8.663323303622462e-06,
+      "loss": 0.1631,
+      "step": 2346
+    },
+    {
+      "epoch": 1.1133776091081593,
+      "grad_norm": 1.8945375680923462,
+      "learning_rate": 8.655709610675056e-06,
+      "loss": 0.1659,
+      "step": 2347
+    },
+    {
+      "epoch": 1.113851992409867,
+      "grad_norm": 2.1004648208618164,
+      "learning_rate": 8.648096711248967e-06,
+      "loss": 0.2273,
+      "step": 2348
+    },
+    {
+      "epoch": 1.114326375711575,
+      "grad_norm": 1.748862624168396,
+      "learning_rate": 8.640484609838007e-06,
+      "loss": 0.1857,
+      "step": 2349
+    },
+    {
+      "epoch": 1.1148007590132827,
+      "grad_norm": 1.9971057176589966,
+      "learning_rate": 8.632873310935528e-06,
+      "loss": 0.1852,
+      "step": 2350
+    },
+    {
+      "epoch": 1.1152751423149905,
+      "grad_norm": 1.7587655782699585,
+      "learning_rate": 8.625262819034408e-06,
+      "loss": 0.1937,
+      "step": 2351
+    },
+    {
+      "epoch": 1.1157495256166983,
+      "grad_norm": 1.5481939315795898,
+      "learning_rate": 8.61765313862704e-06,
+      "loss": 0.1684,
+      "step": 2352
+    },
+    {
+      "epoch": 1.116223908918406,
+      "grad_norm": 1.8051375150680542,
+      "learning_rate": 8.610044274205352e-06,
+      "loss": 0.2023,
+      "step": 2353
+    },
+    {
+      "epoch": 1.1166982922201139,
+      "grad_norm": 2.1944355964660645,
+      "learning_rate": 8.602436230260768e-06,
+      "loss": 0.1854,
+      "step": 2354
+    },
+    {
+      "epoch": 1.1171726755218216,
+      "grad_norm": 2.0533533096313477,
+      "learning_rate": 8.59482901128425e-06,
+      "loss": 0.2236,
+      "step": 2355
+    },
+    {
+      "epoch": 1.1176470588235294,
+      "grad_norm": 2.361926555633545,
+      "learning_rate": 8.58722262176626e-06,
+      "loss": 0.2015,
+      "step": 2356
+    },
+    {
+      "epoch": 1.1181214421252372,
+      "grad_norm": 1.664730191230774,
+      "learning_rate": 8.579617066196777e-06,
+      "loss": 0.1824,
+      "step": 2357
+    },
+    {
+      "epoch": 1.118595825426945,
+      "grad_norm": 1.7999298572540283,
+      "learning_rate": 8.572012349065288e-06,
+      "loss": 0.1814,
+      "step": 2358
+    },
+    {
+      "epoch": 1.1190702087286528,
+      "grad_norm": 2.117183208465576,
+      "learning_rate": 8.564408474860774e-06,
+      "loss": 0.2215,
+      "step": 2359
+    },
+    {
+      "epoch": 1.1195445920303606,
+      "grad_norm": 1.658808946609497,
+      "learning_rate": 8.556805448071736e-06,
+      "loss": 0.1568,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1200189753320684,
+      "grad_norm": 2.2438066005706787,
+      "learning_rate": 8.549203273186156e-06,
+      "loss": 0.1864,
+      "step": 2361
+    },
+    {
+      "epoch": 1.1204933586337762,
+      "grad_norm": 1.6112077236175537,
+      "learning_rate": 8.541601954691534e-06,
+      "loss": 0.162,
+      "step": 2362
+    },
+    {
+      "epoch": 1.120967741935484,
+      "grad_norm": 1.4597880840301514,
+      "learning_rate": 8.534001497074842e-06,
+      "loss": 0.1589,
+      "step": 2363
+    },
+    {
+      "epoch": 1.1214421252371916,
+      "grad_norm": 1.9932173490524292,
+      "learning_rate": 8.52640190482256e-06,
+      "loss": 0.2256,
+      "step": 2364
+    },
+    {
+      "epoch": 1.1219165085388993,
+      "grad_norm": 1.9947072267532349,
+      "learning_rate": 8.518803182420651e-06,
+      "loss": 0.2217,
+      "step": 2365
+    },
+    {
+      "epoch": 1.1223908918406071,
+      "grad_norm": 1.7771036624908447,
+      "learning_rate": 8.511205334354566e-06,
+      "loss": 0.1938,
+      "step": 2366
+    },
+    {
+      "epoch": 1.122865275142315,
+      "grad_norm": 1.7289372682571411,
+      "learning_rate": 8.503608365109247e-06,
+      "loss": 0.1559,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1233396584440227,
+      "grad_norm": 1.8934062719345093,
+      "learning_rate": 8.496012279169097e-06,
+      "loss": 0.1868,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1238140417457305,
+      "grad_norm": 1.762656569480896,
+      "learning_rate": 8.488417081018015e-06,
+      "loss": 0.1465,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1242884250474383,
+      "grad_norm": 1.8933993577957153,
+      "learning_rate": 8.480822775139371e-06,
+      "loss": 0.2391,
+      "step": 2370
+    },
+    {
+      "epoch": 1.124762808349146,
+      "grad_norm": 1.5721724033355713,
+      "learning_rate": 8.473229366016014e-06,
+      "loss": 0.1591,
+      "step": 2371
+    },
+    {
+      "epoch": 1.125237191650854,
+      "grad_norm": 1.9659096002578735,
+      "learning_rate": 8.465636858130255e-06,
+      "loss": 0.1897,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1257115749525617,
+      "grad_norm": 2.4556431770324707,
+      "learning_rate": 8.45804525596387e-06,
+      "loss": 0.2287,
+      "step": 2373
+    },
+    {
+      "epoch": 1.1261859582542695,
+      "grad_norm": 1.9868581295013428,
+      "learning_rate": 8.450454563998117e-06,
+      "loss": 0.2112,
+      "step": 2374
+    },
+    {
+      "epoch": 1.1266603415559773,
+      "grad_norm": 1.849770188331604,
+      "learning_rate": 8.4428647867137e-06,
+      "loss": 0.1841,
+      "step": 2375
+    },
+    {
+      "epoch": 1.127134724857685,
+      "grad_norm": 1.5524128675460815,
+      "learning_rate": 8.435275928590789e-06,
+      "loss": 0.1456,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1276091081593929,
+      "grad_norm": 1.8452872037887573,
+      "learning_rate": 8.427687994109017e-06,
+      "loss": 0.1729,
+      "step": 2377
+    },
+    {
+      "epoch": 1.1280834914611007,
+      "grad_norm": 2.2624077796936035,
+      "learning_rate": 8.42010098774746e-06,
+      "loss": 0.2093,
+      "step": 2378
+    },
+    {
+      "epoch": 1.1285578747628084,
+      "grad_norm": 1.6121922731399536,
+      "learning_rate": 8.412514913984657e-06,
+      "loss": 0.157,
+      "step": 2379
+    },
+    {
+      "epoch": 1.129032258064516,
+      "grad_norm": 1.7655012607574463,
+      "learning_rate": 8.404929777298592e-06,
+      "loss": 0.1697,
+      "step": 2380
+    },
+    {
+      "epoch": 1.1295066413662238,
+      "grad_norm": 1.811368703842163,
+      "learning_rate": 8.3973455821667e-06,
+      "loss": 0.1537,
+      "step": 2381
+    },
+    {
+      "epoch": 1.1299810246679316,
+      "grad_norm": 1.8110166788101196,
+      "learning_rate": 8.389762333065847e-06,
+      "loss": 0.2188,
+      "step": 2382
+    },
+    {
+      "epoch": 1.1304554079696394,
+      "grad_norm": 1.7781898975372314,
+      "learning_rate": 8.382180034472353e-06,
+      "loss": 0.2,
+      "step": 2383
+    },
+    {
+      "epoch": 1.1309297912713472,
+      "grad_norm": 1.581636905670166,
+      "learning_rate": 8.374598690861978e-06,
+      "loss": 0.1607,
+      "step": 2384
+    },
+    {
+      "epoch": 1.131404174573055,
+      "grad_norm": 2.6846189498901367,
+      "learning_rate": 8.367018306709913e-06,
+      "loss": 0.1924,
+      "step": 2385
+    },
+    {
+      "epoch": 1.1318785578747628,
+      "grad_norm": 2.4494681358337402,
+      "learning_rate": 8.359438886490783e-06,
+      "loss": 0.2339,
+      "step": 2386
+    },
+    {
+      "epoch": 1.1323529411764706,
+      "grad_norm": 1.944273591041565,
+      "learning_rate": 8.351860434678641e-06,
+      "loss": 0.186,
+      "step": 2387
+    },
+    {
+      "epoch": 1.1328273244781784,
+      "grad_norm": 1.9705138206481934,
+      "learning_rate": 8.344282955746978e-06,
+      "loss": 0.186,
+      "step": 2388
+    },
+    {
+      "epoch": 1.1333017077798861,
+      "grad_norm": 1.6467750072479248,
+      "learning_rate": 8.336706454168701e-06,
+      "loss": 0.1791,
+      "step": 2389
+    },
+    {
+      "epoch": 1.133776091081594,
+      "grad_norm": 1.7570585012435913,
+      "learning_rate": 8.329130934416142e-06,
+      "loss": 0.1567,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1342504743833017,
+      "grad_norm": 1.4821341037750244,
+      "learning_rate": 8.321556400961067e-06,
+      "loss": 0.1465,
+      "step": 2391
+    },
+    {
+      "epoch": 1.1347248576850095,
+      "grad_norm": 2.2223289012908936,
+      "learning_rate": 8.313982858274634e-06,
+      "loss": 0.2004,
+      "step": 2392
+    },
+    {
+      "epoch": 1.1351992409867173,
+      "grad_norm": 1.931081771850586,
+      "learning_rate": 8.306410310827435e-06,
+      "loss": 0.1863,
+      "step": 2393
+    },
+    {
+      "epoch": 1.135673624288425,
+      "grad_norm": 1.7010200023651123,
+      "learning_rate": 8.298838763089471e-06,
+      "loss": 0.1667,
+      "step": 2394
+    },
+    {
+      "epoch": 1.136148007590133,
+      "grad_norm": 2.051765203475952,
+      "learning_rate": 8.291268219530153e-06,
+      "loss": 0.1978,
+      "step": 2395
+    },
+    {
+      "epoch": 1.1366223908918407,
+      "grad_norm": 1.797257900238037,
+      "learning_rate": 8.2836986846183e-06,
+      "loss": 0.1652,
+      "step": 2396
+    },
+    {
+      "epoch": 1.1370967741935485,
+      "grad_norm": 1.760908603668213,
+      "learning_rate": 8.276130162822124e-06,
+      "loss": 0.2005,
+      "step": 2397
+    },
+    {
+      "epoch": 1.1375711574952563,
+      "grad_norm": 1.7077192068099976,
+      "learning_rate": 8.268562658609254e-06,
+      "loss": 0.2053,
+      "step": 2398
+    },
+    {
+      "epoch": 1.1380455407969639,
+      "grad_norm": 2.186999797821045,
+      "learning_rate": 8.260996176446716e-06,
+      "loss": 0.2174,
+      "step": 2399
+    },
+    {
+      "epoch": 1.1385199240986716,
+      "grad_norm": 1.788298487663269,
+      "learning_rate": 8.253430720800928e-06,
+      "loss": 0.1705,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1389943074003794,
+      "grad_norm": 2.1071958541870117,
+      "learning_rate": 8.245866296137701e-06,
+      "loss": 0.1797,
+      "step": 2401
+    },
+    {
+      "epoch": 1.1394686907020872,
+      "grad_norm": 1.7165755033493042,
+      "learning_rate": 8.238302906922242e-06,
+      "loss": 0.2184,
+      "step": 2402
+    },
+    {
+      "epoch": 1.139943074003795,
+      "grad_norm": 1.7405943870544434,
+      "learning_rate": 8.230740557619142e-06,
+      "loss": 0.1772,
+      "step": 2403
+    },
+    {
+      "epoch": 1.1404174573055028,
+      "grad_norm": 2.0307819843292236,
+      "learning_rate": 8.223179252692385e-06,
+      "loss": 0.2016,
+      "step": 2404
+    },
+    {
+      "epoch": 1.1408918406072106,
+      "grad_norm": 1.6938095092773438,
+      "learning_rate": 8.215618996605336e-06,
+      "loss": 0.1615,
+      "step": 2405
+    },
+    {
+      "epoch": 1.1413662239089184,
+      "grad_norm": 1.7445240020751953,
+      "learning_rate": 8.208059793820731e-06,
+      "loss": 0.2164,
+      "step": 2406
+    },
+    {
+      "epoch": 1.1418406072106262,
+      "grad_norm": 2.3769569396972656,
+      "learning_rate": 8.200501648800698e-06,
+      "loss": 0.1711,
+      "step": 2407
+    },
+    {
+      "epoch": 1.142314990512334,
+      "grad_norm": 1.518480658531189,
+      "learning_rate": 8.192944566006737e-06,
+      "loss": 0.1561,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1427893738140418,
+      "grad_norm": 1.9066368341445923,
+      "learning_rate": 8.185388549899715e-06,
+      "loss": 0.2031,
+      "step": 2409
+    },
+    {
+      "epoch": 1.1432637571157496,
+      "grad_norm": 2.3878681659698486,
+      "learning_rate": 8.17783360493988e-06,
+      "loss": 0.2288,
+      "step": 2410
+    },
+    {
+      "epoch": 1.1437381404174574,
+      "grad_norm": 2.5900661945343018,
+      "learning_rate": 8.170279735586833e-06,
+      "loss": 0.1964,
+      "step": 2411
+    },
+    {
+      "epoch": 1.1442125237191652,
+      "grad_norm": 1.9122105836868286,
+      "learning_rate": 8.162726946299556e-06,
+      "loss": 0.1934,
+      "step": 2412
+    },
+    {
+      "epoch": 1.144686907020873,
+      "grad_norm": 2.024298667907715,
+      "learning_rate": 8.155175241536377e-06,
+      "loss": 0.2132,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1451612903225807,
+      "grad_norm": 1.898661732673645,
+      "learning_rate": 8.147624625754999e-06,
+      "loss": 0.2355,
+      "step": 2414
+    },
+    {
+      "epoch": 1.1456356736242883,
+      "grad_norm": 1.5798265933990479,
+      "learning_rate": 8.140075103412477e-06,
+      "loss": 0.1495,
+      "step": 2415
+    },
+    {
+      "epoch": 1.146110056925996,
+      "grad_norm": 1.6848498582839966,
+      "learning_rate": 8.132526678965215e-06,
+      "loss": 0.1671,
+      "step": 2416
+    },
+    {
+      "epoch": 1.146584440227704,
+      "grad_norm": 1.576760172843933,
+      "learning_rate": 8.124979356868976e-06,
+      "loss": 0.1631,
+      "step": 2417
+    },
+    {
+      "epoch": 1.1470588235294117,
+      "grad_norm": 2.446101427078247,
+      "learning_rate": 8.117433141578865e-06,
+      "loss": 0.253,
+      "step": 2418
+    },
+    {
+      "epoch": 1.1475332068311195,
+      "grad_norm": 1.735353946685791,
+      "learning_rate": 8.109888037549346e-06,
+      "loss": 0.1542,
+      "step": 2419
+    },
+    {
+      "epoch": 1.1480075901328273,
+      "grad_norm": 1.9680674076080322,
+      "learning_rate": 8.102344049234213e-06,
+      "loss": 0.2229,
+      "step": 2420
+    },
+    {
+      "epoch": 1.148481973434535,
+      "grad_norm": 1.939218521118164,
+      "learning_rate": 8.094801181086612e-06,
+      "loss": 0.188,
+      "step": 2421
+    },
+    {
+      "epoch": 1.1489563567362429,
+      "grad_norm": 1.6669310331344604,
+      "learning_rate": 8.087259437559017e-06,
+      "loss": 0.1843,
+      "step": 2422
+    },
+    {
+      "epoch": 1.1494307400379506,
+      "grad_norm": 1.6978830099105835,
+      "learning_rate": 8.079718823103251e-06,
+      "loss": 0.1643,
+      "step": 2423
+    },
+    {
+      "epoch": 1.1499051233396584,
+      "grad_norm": 1.4840154647827148,
+      "learning_rate": 8.072179342170461e-06,
+      "loss": 0.1328,
+      "step": 2424
+    },
+    {
+      "epoch": 1.1503795066413662,
+      "grad_norm": 1.5778673887252808,
+      "learning_rate": 8.06464099921113e-06,
+      "loss": 0.1554,
+      "step": 2425
+    },
+    {
+      "epoch": 1.150853889943074,
+      "grad_norm": 1.7717269659042358,
+      "learning_rate": 8.057103798675063e-06,
+      "loss": 0.1731,
+      "step": 2426
+    },
+    {
+      "epoch": 1.1513282732447818,
+      "grad_norm": 1.752498984336853,
+      "learning_rate": 8.0495677450114e-06,
+      "loss": 0.1794,
+      "step": 2427
+    },
+    {
+      "epoch": 1.1518026565464896,
+      "grad_norm": 1.9130120277404785,
+      "learning_rate": 8.042032842668598e-06,
+      "loss": 0.1863,
+      "step": 2428
+    },
+    {
+      "epoch": 1.1522770398481974,
+      "grad_norm": 1.86540687084198,
+      "learning_rate": 8.034499096094434e-06,
+      "loss": 0.1823,
+      "step": 2429
+    },
+    {
+      "epoch": 1.1527514231499052,
+      "grad_norm": 2.4312217235565186,
+      "learning_rate": 8.026966509736001e-06,
+      "loss": 0.1981,
+      "step": 2430
+    },
+    {
+      "epoch": 1.153225806451613,
+      "grad_norm": 1.7719708681106567,
+      "learning_rate": 8.019435088039714e-06,
+      "loss": 0.1733,
+      "step": 2431
+    },
+    {
+      "epoch": 1.1537001897533208,
+      "grad_norm": 2.0009617805480957,
+      "learning_rate": 8.011904835451298e-06,
+      "loss": 0.178,
+      "step": 2432
+    },
+    {
+      "epoch": 1.1541745730550286,
+      "grad_norm": 2.0233676433563232,
+      "learning_rate": 8.004375756415783e-06,
+      "loss": 0.1689,
+      "step": 2433
+    },
+    {
+      "epoch": 1.1546489563567364,
+      "grad_norm": 1.6531426906585693,
+      "learning_rate": 7.996847855377514e-06,
+      "loss": 0.1642,
+      "step": 2434
+    },
+    {
+      "epoch": 1.155123339658444,
+      "grad_norm": 2.1204326152801514,
+      "learning_rate": 7.989321136780131e-06,
+      "loss": 0.1987,
+      "step": 2435
+    },
+    {
+      "epoch": 1.1555977229601517,
+      "grad_norm": 1.7340539693832397,
+      "learning_rate": 7.981795605066585e-06,
+      "loss": 0.1949,
+      "step": 2436
+    },
+    {
+      "epoch": 1.1560721062618595,
+      "grad_norm": 1.8401095867156982,
+      "learning_rate": 7.974271264679122e-06,
+      "loss": 0.1563,
+      "step": 2437
+    },
+    {
+      "epoch": 1.1565464895635673,
+      "grad_norm": 1.637509822845459,
+      "learning_rate": 7.966748120059286e-06,
+      "loss": 0.1538,
+      "step": 2438
+    },
+    {
+      "epoch": 1.157020872865275,
+      "grad_norm": 2.007936954498291,
+      "learning_rate": 7.959226175647919e-06,
+      "loss": 0.1921,
+      "step": 2439
+    },
+    {
+      "epoch": 1.157495256166983,
+      "grad_norm": 1.9487900733947754,
+      "learning_rate": 7.951705435885143e-06,
+      "loss": 0.1798,
+      "step": 2440
+    },
+    {
+      "epoch": 1.1579696394686907,
+      "grad_norm": 1.8263933658599854,
+      "learning_rate": 7.944185905210377e-06,
+      "loss": 0.1991,
+      "step": 2441
+    },
+    {
+      "epoch": 1.1584440227703985,
+      "grad_norm": 2.152977466583252,
+      "learning_rate": 7.93666758806233e-06,
+      "loss": 0.1981,
+      "step": 2442
+    },
+    {
+      "epoch": 1.1589184060721063,
+      "grad_norm": 1.8188046216964722,
+      "learning_rate": 7.929150488878991e-06,
+      "loss": 0.1388,
+      "step": 2443
+    },
+    {
+      "epoch": 1.159392789373814,
+      "grad_norm": 1.7171995639801025,
+      "learning_rate": 7.921634612097623e-06,
+      "loss": 0.1689,
+      "step": 2444
+    },
+    {
+      "epoch": 1.1598671726755219,
+      "grad_norm": 1.9360337257385254,
+      "learning_rate": 7.914119962154779e-06,
+      "loss": 0.1858,
+      "step": 2445
+    },
+    {
+      "epoch": 1.1603415559772297,
+      "grad_norm": 1.8121434450149536,
+      "learning_rate": 7.906606543486278e-06,
+      "loss": 0.169,
+      "step": 2446
+    },
+    {
+      "epoch": 1.1608159392789374,
+      "grad_norm": 2.7344167232513428,
+      "learning_rate": 7.89909436052722e-06,
+      "loss": 0.2118,
+      "step": 2447
+    },
+    {
+      "epoch": 1.1612903225806452,
+      "grad_norm": 2.18343186378479,
+      "learning_rate": 7.891583417711975e-06,
+      "loss": 0.21,
+      "step": 2448
+    },
+    {
+      "epoch": 1.161764705882353,
+      "grad_norm": 1.4035104513168335,
+      "learning_rate": 7.884073719474174e-06,
+      "loss": 0.152,
+      "step": 2449
+    },
+    {
+      "epoch": 1.1622390891840606,
+      "grad_norm": 1.8150465488433838,
+      "learning_rate": 7.876565270246715e-06,
+      "loss": 0.1535,
+      "step": 2450
+    },
+    {
+      "epoch": 1.1627134724857684,
+      "grad_norm": 1.967027187347412,
+      "learning_rate": 7.869058074461766e-06,
+      "loss": 0.1987,
+      "step": 2451
+    },
+    {
+      "epoch": 1.1631878557874762,
+      "grad_norm": 1.906870722770691,
+      "learning_rate": 7.86155213655075e-06,
+      "loss": 0.1772,
+      "step": 2452
+    },
+    {
+      "epoch": 1.163662239089184,
+      "grad_norm": 1.788644790649414,
+      "learning_rate": 7.85404746094435e-06,
+      "loss": 0.1803,
+      "step": 2453
+    },
+    {
+      "epoch": 1.1641366223908918,
+      "grad_norm": 2.3639960289001465,
+      "learning_rate": 7.846544052072494e-06,
+      "loss": 0.1969,
+      "step": 2454
+    },
+    {
+      "epoch": 1.1646110056925996,
+      "grad_norm": 1.8786990642547607,
+      "learning_rate": 7.839041914364375e-06,
+      "loss": 0.2137,
+      "step": 2455
+    },
+    {
+      "epoch": 1.1650853889943074,
+      "grad_norm": 1.9279520511627197,
+      "learning_rate": 7.831541052248433e-06,
+      "loss": 0.1631,
+      "step": 2456
+    },
+    {
+      "epoch": 1.1655597722960152,
+      "grad_norm": 1.9649664163589478,
+      "learning_rate": 7.824041470152346e-06,
+      "loss": 0.1736,
+      "step": 2457
+    },
+    {
+      "epoch": 1.166034155597723,
+      "grad_norm": 2.580247402191162,
+      "learning_rate": 7.816543172503052e-06,
+      "loss": 0.2557,
+      "step": 2458
+    },
+    {
+      "epoch": 1.1665085388994307,
+      "grad_norm": 1.8188077211380005,
+      "learning_rate": 7.809046163726715e-06,
+      "loss": 0.1678,
+      "step": 2459
+    },
+    {
+      "epoch": 1.1669829222011385,
+      "grad_norm": 1.8238269090652466,
+      "learning_rate": 7.801550448248746e-06,
+      "loss": 0.2021,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1674573055028463,
+      "grad_norm": 1.9044618606567383,
+      "learning_rate": 7.794056030493793e-06,
+      "loss": 0.1528,
+      "step": 2461
+    },
+    {
+      "epoch": 1.1679316888045541,
+      "grad_norm": 1.5531914234161377,
+      "learning_rate": 7.78656291488574e-06,
+      "loss": 0.1613,
+      "step": 2462
+    },
+    {
+      "epoch": 1.168406072106262,
+      "grad_norm": 2.1100800037384033,
+      "learning_rate": 7.779071105847692e-06,
+      "loss": 0.2401,
+      "step": 2463
+    },
+    {
+      "epoch": 1.1688804554079697,
+      "grad_norm": 1.6972509622573853,
+      "learning_rate": 7.771580607801994e-06,
+      "loss": 0.2089,
+      "step": 2464
+    },
+    {
+      "epoch": 1.1693548387096775,
+      "grad_norm": 1.727967381477356,
+      "learning_rate": 7.76409142517021e-06,
+      "loss": 0.1693,
+      "step": 2465
+    },
+    {
+      "epoch": 1.1698292220113853,
+      "grad_norm": 2.0903284549713135,
+      "learning_rate": 7.756603562373134e-06,
+      "loss": 0.1983,
+      "step": 2466
+    },
+    {
+      "epoch": 1.170303605313093,
+      "grad_norm": 1.5043290853500366,
+      "learning_rate": 7.749117023830779e-06,
+      "loss": 0.1714,
+      "step": 2467
+    },
+    {
+      "epoch": 1.1707779886148009,
+      "grad_norm": 1.7228398323059082,
+      "learning_rate": 7.741631813962367e-06,
+      "loss": 0.1587,
+      "step": 2468
+    },
+    {
+      "epoch": 1.1712523719165087,
+      "grad_norm": 1.9864158630371094,
+      "learning_rate": 7.73414793718635e-06,
+      "loss": 0.1873,
+      "step": 2469
+    },
+    {
+      "epoch": 1.1717267552182162,
+      "grad_norm": 1.7577637434005737,
+      "learning_rate": 7.72666539792038e-06,
+      "loss": 0.1599,
+      "step": 2470
+    },
+    {
+      "epoch": 1.172201138519924,
+      "grad_norm": 1.9289382696151733,
+      "learning_rate": 7.719184200581334e-06,
+      "loss": 0.1762,
+      "step": 2471
+    },
+    {
+      "epoch": 1.1726755218216318,
+      "grad_norm": 1.6133570671081543,
+      "learning_rate": 7.711704349585287e-06,
+      "loss": 0.1568,
+      "step": 2472
+    },
+    {
+      "epoch": 1.1731499051233396,
+      "grad_norm": 1.882628083229065,
+      "learning_rate": 7.704225849347517e-06,
+      "loss": 0.1661,
+      "step": 2473
+    },
+    {
+      "epoch": 1.1736242884250474,
+      "grad_norm": 2.2073872089385986,
+      "learning_rate": 7.696748704282507e-06,
+      "loss": 0.1734,
+      "step": 2474
+    },
+    {
+      "epoch": 1.1740986717267552,
+      "grad_norm": 1.9427272081375122,
+      "learning_rate": 7.689272918803946e-06,
+      "loss": 0.1713,
+      "step": 2475
+    },
+    {
+      "epoch": 1.174573055028463,
+      "grad_norm": 1.8138381242752075,
+      "learning_rate": 7.681798497324717e-06,
+      "loss": 0.1864,
+      "step": 2476
+    },
+    {
+      "epoch": 1.1750474383301708,
+      "grad_norm": 2.078590154647827,
+      "learning_rate": 7.674325444256899e-06,
+      "loss": 0.1737,
+      "step": 2477
+    },
+    {
+      "epoch": 1.1755218216318786,
+      "grad_norm": 2.3086998462677,
+      "learning_rate": 7.666853764011752e-06,
+      "loss": 0.2696,
+      "step": 2478
+    },
+    {
+      "epoch": 1.1759962049335864,
+      "grad_norm": 2.107063055038452,
+      "learning_rate": 7.659383460999742e-06,
+      "loss": 0.1974,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1764705882352942,
+      "grad_norm": 1.7796244621276855,
+      "learning_rate": 7.651914539630515e-06,
+      "loss": 0.194,
+      "step": 2480
+    },
+    {
+      "epoch": 1.176944971537002,
+      "grad_norm": 2.3119170665740967,
+      "learning_rate": 7.644447004312903e-06,
+      "loss": 0.2256,
+      "step": 2481
+    },
+    {
+      "epoch": 1.1774193548387097,
+      "grad_norm": 1.837838053703308,
+      "learning_rate": 7.63698085945491e-06,
+      "loss": 0.1657,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1778937381404175,
+      "grad_norm": 1.628775954246521,
+      "learning_rate": 7.629516109463732e-06,
+      "loss": 0.1785,
+      "step": 2483
+    },
+    {
+      "epoch": 1.1783681214421253,
+      "grad_norm": 1.6993663311004639,
+      "learning_rate": 7.622052758745741e-06,
+      "loss": 0.1829,
+      "step": 2484
+    },
+    {
+      "epoch": 1.178842504743833,
+      "grad_norm": 2.071762800216675,
+      "learning_rate": 7.614590811706473e-06,
+      "loss": 0.2427,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1793168880455407,
+      "grad_norm": 1.9026820659637451,
+      "learning_rate": 7.607130272750647e-06,
+      "loss": 0.2126,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1797912713472485,
+      "grad_norm": 1.604719638824463,
+      "learning_rate": 7.59967114628214e-06,
+      "loss": 0.1347,
+      "step": 2487
+    },
+    {
+      "epoch": 1.1802656546489563,
+      "grad_norm": 1.5981990098953247,
+      "learning_rate": 7.592213436704004e-06,
+      "loss": 0.1716,
+      "step": 2488
+    },
+    {
+      "epoch": 1.180740037950664,
+      "grad_norm": 1.558847188949585,
+      "learning_rate": 7.58475714841845e-06,
+      "loss": 0.1558,
+      "step": 2489
+    },
+    {
+      "epoch": 1.1812144212523719,
+      "grad_norm": 1.6636430025100708,
+      "learning_rate": 7.577302285826851e-06,
+      "loss": 0.1693,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1816888045540797,
+      "grad_norm": 1.62824285030365,
+      "learning_rate": 7.569848853329742e-06,
+      "loss": 0.1649,
+      "step": 2491
+    },
+    {
+      "epoch": 1.1821631878557874,
+      "grad_norm": 1.6984463930130005,
+      "learning_rate": 7.562396855326805e-06,
+      "loss": 0.1345,
+      "step": 2492
+    },
+    {
+      "epoch": 1.1826375711574952,
+      "grad_norm": 2.12964129447937,
+      "learning_rate": 7.554946296216884e-06,
+      "loss": 0.1844,
+      "step": 2493
+    },
+    {
+      "epoch": 1.183111954459203,
+      "grad_norm": 1.8614979982376099,
+      "learning_rate": 7.547497180397968e-06,
+      "loss": 0.203,
+      "step": 2494
+    },
+    {
+      "epoch": 1.1835863377609108,
+      "grad_norm": 1.7780121564865112,
+      "learning_rate": 7.540049512267197e-06,
+      "loss": 0.199,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1840607210626186,
+      "grad_norm": 2.061229705810547,
+      "learning_rate": 7.532603296220862e-06,
+      "loss": 0.2002,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1845351043643264,
+      "grad_norm": 1.7762293815612793,
+      "learning_rate": 7.525158536654382e-06,
+      "loss": 0.1714,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1850094876660342,
+      "grad_norm": 1.525127649307251,
+      "learning_rate": 7.517715237962328e-06,
+      "loss": 0.1557,
+      "step": 2498
+    },
+    {
+      "epoch": 1.185483870967742,
+      "grad_norm": 1.9348129034042358,
+      "learning_rate": 7.510273404538404e-06,
+      "loss": 0.1971,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1859582542694498,
+      "grad_norm": 1.6467643976211548,
+      "learning_rate": 7.502833040775457e-06,
+      "loss": 0.1551,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1864326375711576,
+      "grad_norm": 1.4035449028015137,
+      "learning_rate": 7.4953941510654535e-06,
+      "loss": 0.1521,
+      "step": 2501
+    },
+    {
+      "epoch": 1.1869070208728654,
+      "grad_norm": 1.7778912782669067,
+      "learning_rate": 7.487956739799496e-06,
+      "loss": 0.1788,
+      "step": 2502
+    },
+    {
+      "epoch": 1.1873814041745732,
+      "grad_norm": 1.626543402671814,
+      "learning_rate": 7.480520811367817e-06,
+      "loss": 0.1382,
+      "step": 2503
+    },
+    {
+      "epoch": 1.187855787476281,
+      "grad_norm": 2.147493362426758,
+      "learning_rate": 7.473086370159776e-06,
+      "loss": 0.2429,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1883301707779885,
+      "grad_norm": 1.7285178899765015,
+      "learning_rate": 7.465653420563846e-06,
+      "loss": 0.1818,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1888045540796963,
+      "grad_norm": 1.7863154411315918,
+      "learning_rate": 7.45822196696762e-06,
+      "loss": 0.1791,
+      "step": 2506
+    },
+    {
+      "epoch": 1.189278937381404,
+      "grad_norm": 2.1906211376190186,
+      "learning_rate": 7.4507920137578146e-06,
+      "loss": 0.2307,
+      "step": 2507
+    },
+    {
+      "epoch": 1.189753320683112,
+      "grad_norm": 3.06583309173584,
+      "learning_rate": 7.443363565320259e-06,
+      "loss": 0.1962,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1902277039848197,
+      "grad_norm": 1.7111692428588867,
+      "learning_rate": 7.435936626039891e-06,
+      "loss": 0.1744,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1907020872865275,
+      "grad_norm": 3.7655081748962402,
+      "learning_rate": 7.428511200300765e-06,
+      "loss": 0.2345,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1911764705882353,
+      "grad_norm": 2.3031506538391113,
+      "learning_rate": 7.421087292486027e-06,
+      "loss": 0.208,
+      "step": 2511
+    },
+    {
+      "epoch": 1.191650853889943,
+      "grad_norm": 2.341597318649292,
+      "learning_rate": 7.41366490697794e-06,
+      "loss": 0.2249,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1921252371916509,
+      "grad_norm": 2.380882501602173,
+      "learning_rate": 7.406244048157867e-06,
+      "loss": 0.1982,
+      "step": 2513
+    },
+    {
+      "epoch": 1.1925996204933587,
+      "grad_norm": 2.25535249710083,
+      "learning_rate": 7.398824720406265e-06,
+      "loss": 0.2163,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1930740037950665,
+      "grad_norm": 1.5992552042007446,
+      "learning_rate": 7.391406928102695e-06,
+      "loss": 0.1696,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1935483870967742,
+      "grad_norm": 1.729461908340454,
+      "learning_rate": 7.383990675625797e-06,
+      "loss": 0.1747,
+      "step": 2516
+    },
+    {
+      "epoch": 1.194022770398482,
+      "grad_norm": 2.1388728618621826,
+      "learning_rate": 7.37657596735332e-06,
+      "loss": 0.1818,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1944971537001898,
+      "grad_norm": 1.8862483501434326,
+      "learning_rate": 7.369162807662087e-06,
+      "loss": 0.1946,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1949715370018976,
+      "grad_norm": 1.9275808334350586,
+      "learning_rate": 7.36175120092802e-06,
+      "loss": 0.2447,
+      "step": 2519
+    },
+    {
+      "epoch": 1.1954459203036052,
+      "grad_norm": 2.302499294281006,
+      "learning_rate": 7.354341151526107e-06,
+      "loss": 0.1424,
+      "step": 2520
+    },
+    {
+      "epoch": 1.195920303605313,
+      "grad_norm": 2.07778263092041,
+      "learning_rate": 7.3469326638304365e-06,
+      "loss": 0.2255,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1963946869070208,
+      "grad_norm": 1.843159794807434,
+      "learning_rate": 7.339525742214157e-06,
+      "loss": 0.1705,
+      "step": 2522
+    },
+    {
+      "epoch": 1.1968690702087286,
+      "grad_norm": 2.0549871921539307,
+      "learning_rate": 7.332120391049506e-06,
+      "loss": 0.2071,
+      "step": 2523
+    },
+    {
+      "epoch": 1.1973434535104364,
+      "grad_norm": 1.5816240310668945,
+      "learning_rate": 7.324716614707794e-06,
+      "loss": 0.1605,
+      "step": 2524
+    },
+    {
+      "epoch": 1.1978178368121442,
+      "grad_norm": 1.4971600770950317,
+      "learning_rate": 7.317314417559389e-06,
+      "loss": 0.1392,
+      "step": 2525
+    },
+    {
+      "epoch": 1.198292220113852,
+      "grad_norm": 1.5572508573532104,
+      "learning_rate": 7.309913803973734e-06,
+      "loss": 0.1518,
+      "step": 2526
+    },
+    {
+      "epoch": 1.1987666034155597,
+      "grad_norm": 2.0520272254943848,
+      "learning_rate": 7.302514778319341e-06,
+      "loss": 0.2069,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1992409867172675,
+      "grad_norm": 1.7996068000793457,
+      "learning_rate": 7.295117344963782e-06,
+      "loss": 0.2041,
+      "step": 2528
+    },
+    {
+      "epoch": 1.1997153700189753,
+      "grad_norm": 1.6315232515335083,
+      "learning_rate": 7.287721508273691e-06,
+      "loss": 0.1681,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2001897533206831,
+      "grad_norm": 1.8578884601593018,
+      "learning_rate": 7.280327272614753e-06,
+      "loss": 0.1671,
+      "step": 2530
+    },
+    {
+      "epoch": 1.200664136622391,
+      "grad_norm": 1.7982726097106934,
+      "learning_rate": 7.272934642351712e-06,
+      "loss": 0.1849,
+      "step": 2531
+    },
+    {
+      "epoch": 1.2011385199240987,
+      "grad_norm": 2.00294828414917,
+      "learning_rate": 7.265543621848368e-06,
+      "loss": 0.1644,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2016129032258065,
+      "grad_norm": 1.6262898445129395,
+      "learning_rate": 7.2581542154675654e-06,
+      "loss": 0.1612,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2020872865275143,
+      "grad_norm": 1.8584097623825073,
+      "learning_rate": 7.2507664275712e-06,
+      "loss": 0.1822,
+      "step": 2534
+    },
+    {
+      "epoch": 1.202561669829222,
+      "grad_norm": 2.0192692279815674,
+      "learning_rate": 7.243380262520203e-06,
+      "loss": 0.2231,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2030360531309299,
+      "grad_norm": 2.2712655067443848,
+      "learning_rate": 7.23599572467456e-06,
+      "loss": 0.2046,
+      "step": 2536
+    },
+    {
+      "epoch": 1.2035104364326377,
+      "grad_norm": 2.379418134689331,
+      "learning_rate": 7.228612818393292e-06,
+      "loss": 0.2186,
+      "step": 2537
+    },
+    {
+      "epoch": 1.2039848197343455,
+      "grad_norm": 1.9519847631454468,
+      "learning_rate": 7.221231548034451e-06,
+      "loss": 0.1875,
+      "step": 2538
+    },
+    {
+      "epoch": 1.2044592030360532,
+      "grad_norm": 1.7326889038085938,
+      "learning_rate": 7.2138519179551335e-06,
+      "loss": 0.153,
+      "step": 2539
+    },
+    {
+      "epoch": 1.2049335863377608,
+      "grad_norm": 1.4056307077407837,
+      "learning_rate": 7.206473932511455e-06,
+      "loss": 0.1415,
+      "step": 2540
+    },
+    {
+      "epoch": 1.2054079696394686,
+      "grad_norm": 1.8842997550964355,
+      "learning_rate": 7.199097596058573e-06,
+      "loss": 0.1948,
+      "step": 2541
+    },
+    {
+      "epoch": 1.2058823529411764,
+      "grad_norm": 1.9680218696594238,
+      "learning_rate": 7.1917229129506626e-06,
+      "loss": 0.1902,
+      "step": 2542
+    },
+    {
+      "epoch": 1.2063567362428842,
+      "grad_norm": 6.604353427886963,
+      "learning_rate": 7.1843498875409315e-06,
+      "loss": 0.1912,
+      "step": 2543
+    },
+    {
+      "epoch": 1.206831119544592,
+      "grad_norm": 1.8714125156402588,
+      "learning_rate": 7.176978524181595e-06,
+      "loss": 0.1609,
+      "step": 2544
+    },
+    {
+      "epoch": 1.2073055028462998,
+      "grad_norm": 1.6578891277313232,
+      "learning_rate": 7.169608827223902e-06,
+      "loss": 0.1785,
+      "step": 2545
+    },
+    {
+      "epoch": 1.2077798861480076,
+      "grad_norm": 1.9563171863555908,
+      "learning_rate": 7.16224080101811e-06,
+      "loss": 0.2015,
+      "step": 2546
+    },
+    {
+      "epoch": 1.2082542694497154,
+      "grad_norm": 1.5768934488296509,
+      "learning_rate": 7.154874449913492e-06,
+      "loss": 0.1484,
+      "step": 2547
+    },
+    {
+      "epoch": 1.2087286527514232,
+      "grad_norm": 1.6119853258132935,
+      "learning_rate": 7.147509778258334e-06,
+      "loss": 0.2121,
+      "step": 2548
+    },
+    {
+      "epoch": 1.209203036053131,
+      "grad_norm": 1.5381999015808105,
+      "learning_rate": 7.1401467903999285e-06,
+      "loss": 0.1711,
+      "step": 2549
+    },
+    {
+      "epoch": 1.2096774193548387,
+      "grad_norm": 1.6348845958709717,
+      "learning_rate": 7.1327854906845706e-06,
+      "loss": 0.1632,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2101518026565465,
+      "grad_norm": 1.7592006921768188,
+      "learning_rate": 7.125425883457564e-06,
+      "loss": 0.1786,
+      "step": 2551
+    },
+    {
+      "epoch": 1.2106261859582543,
+      "grad_norm": 1.6525636911392212,
+      "learning_rate": 7.118067973063216e-06,
+      "loss": 0.1704,
+      "step": 2552
+    },
+    {
+      "epoch": 1.2111005692599621,
+      "grad_norm": 1.930760383605957,
+      "learning_rate": 7.110711763844826e-06,
+      "loss": 0.1789,
+      "step": 2553
+    },
+    {
+      "epoch": 1.21157495256167,
+      "grad_norm": 1.6949514150619507,
+      "learning_rate": 7.10335726014469e-06,
+      "loss": 0.2061,
+      "step": 2554
+    },
+    {
+      "epoch": 1.2120493358633775,
+      "grad_norm": 1.494983434677124,
+      "learning_rate": 7.096004466304099e-06,
+      "loss": 0.1317,
+      "step": 2555
+    },
+    {
+      "epoch": 1.2125237191650853,
+      "grad_norm": 1.5275888442993164,
+      "learning_rate": 7.088653386663335e-06,
+      "loss": 0.1481,
+      "step": 2556
+    },
+    {
+      "epoch": 1.212998102466793,
+      "grad_norm": 2.201962471008301,
+      "learning_rate": 7.081304025561668e-06,
+      "loss": 0.2149,
+      "step": 2557
+    },
+    {
+      "epoch": 1.2134724857685009,
+      "grad_norm": 1.78719961643219,
+      "learning_rate": 7.073956387337357e-06,
+      "loss": 0.1635,
+      "step": 2558
+    },
+    {
+      "epoch": 1.2139468690702087,
+      "grad_norm": 1.9768500328063965,
+      "learning_rate": 7.066610476327632e-06,
+      "loss": 0.1727,
+      "step": 2559
+    },
+    {
+      "epoch": 1.2144212523719164,
+      "grad_norm": 2.0566301345825195,
+      "learning_rate": 7.059266296868715e-06,
+      "loss": 0.2103,
+      "step": 2560
+    },
+    {
+      "epoch": 1.2148956356736242,
+      "grad_norm": 1.8030273914337158,
+      "learning_rate": 7.051923853295805e-06,
+      "loss": 0.1743,
+      "step": 2561
+    },
+    {
+      "epoch": 1.215370018975332,
+      "grad_norm": 1.7541704177856445,
+      "learning_rate": 7.044583149943076e-06,
+      "loss": 0.191,
+      "step": 2562
+    },
+    {
+      "epoch": 1.2158444022770398,
+      "grad_norm": 1.8001971244812012,
+      "learning_rate": 7.037244191143662e-06,
+      "loss": 0.1788,
+      "step": 2563
+    },
+    {
+      "epoch": 1.2163187855787476,
+      "grad_norm": 1.620387315750122,
+      "learning_rate": 7.029906981229686e-06,
+      "loss": 0.1421,
+      "step": 2564
+    },
+    {
+      "epoch": 1.2167931688804554,
+      "grad_norm": 1.8180466890335083,
+      "learning_rate": 7.022571524532229e-06,
+      "loss": 0.1778,
+      "step": 2565
+    },
+    {
+      "epoch": 1.2172675521821632,
+      "grad_norm": 2.036728620529175,
+      "learning_rate": 7.015237825381338e-06,
+      "loss": 0.2076,
+      "step": 2566
+    },
+    {
+      "epoch": 1.217741935483871,
+      "grad_norm": 1.8897072076797485,
+      "learning_rate": 7.007905888106026e-06,
+      "loss": 0.1871,
+      "step": 2567
+    },
+    {
+      "epoch": 1.2182163187855788,
+      "grad_norm": 1.585066556930542,
+      "learning_rate": 7.000575717034256e-06,
+      "loss": 0.163,
+      "step": 2568
+    },
+    {
+      "epoch": 1.2186907020872866,
+      "grad_norm": 1.9996800422668457,
+      "learning_rate": 6.993247316492962e-06,
+      "loss": 0.1774,
+      "step": 2569
+    },
+    {
+      "epoch": 1.2191650853889944,
+      "grad_norm": 2.1412925720214844,
+      "learning_rate": 6.985920690808021e-06,
+      "loss": 0.1793,
+      "step": 2570
+    },
+    {
+      "epoch": 1.2196394686907022,
+      "grad_norm": 1.8670066595077515,
+      "learning_rate": 6.978595844304272e-06,
+      "loss": 0.1692,
+      "step": 2571
+    },
+    {
+      "epoch": 1.22011385199241,
+      "grad_norm": 1.7688610553741455,
+      "learning_rate": 6.971272781305503e-06,
+      "loss": 0.175,
+      "step": 2572
+    },
+    {
+      "epoch": 1.2205882352941178,
+      "grad_norm": 1.6328133344650269,
+      "learning_rate": 6.963951506134434e-06,
+      "loss": 0.1546,
+      "step": 2573
+    },
+    {
+      "epoch": 1.2210626185958255,
+      "grad_norm": 1.8396352529525757,
+      "learning_rate": 6.956632023112751e-06,
+      "loss": 0.1753,
+      "step": 2574
+    },
+    {
+      "epoch": 1.2215370018975331,
+      "grad_norm": 1.9134464263916016,
+      "learning_rate": 6.949314336561068e-06,
+      "loss": 0.1688,
+      "step": 2575
+    },
+    {
+      "epoch": 1.222011385199241,
+      "grad_norm": 1.5007470846176147,
+      "learning_rate": 6.941998450798946e-06,
+      "loss": 0.172,
+      "step": 2576
+    },
+    {
+      "epoch": 1.2224857685009487,
+      "grad_norm": 1.7491124868392944,
+      "learning_rate": 6.934684370144879e-06,
+      "loss": 0.2002,
+      "step": 2577
+    },
+    {
+      "epoch": 1.2229601518026565,
+      "grad_norm": 1.8310550451278687,
+      "learning_rate": 6.927372098916294e-06,
+      "loss": 0.1743,
+      "step": 2578
+    },
+    {
+      "epoch": 1.2234345351043643,
+      "grad_norm": 1.72085702419281,
+      "learning_rate": 6.9200616414295525e-06,
+      "loss": 0.1635,
+      "step": 2579
+    },
+    {
+      "epoch": 1.223908918406072,
+      "grad_norm": 1.723044514656067,
+      "learning_rate": 6.912753001999948e-06,
+      "loss": 0.1717,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2243833017077799,
+      "grad_norm": 1.7026135921478271,
+      "learning_rate": 6.9054461849417e-06,
+      "loss": 0.1621,
+      "step": 2581
+    },
+    {
+      "epoch": 1.2248576850094877,
+      "grad_norm": 1.7875248193740845,
+      "learning_rate": 6.898141194567944e-06,
+      "loss": 0.1824,
+      "step": 2582
+    },
+    {
+      "epoch": 1.2253320683111955,
+      "grad_norm": 1.8562390804290771,
+      "learning_rate": 6.890838035190747e-06,
+      "loss": 0.1861,
+      "step": 2583
+    },
+    {
+      "epoch": 1.2258064516129032,
+      "grad_norm": 1.7400999069213867,
+      "learning_rate": 6.88353671112109e-06,
+      "loss": 0.189,
+      "step": 2584
+    },
+    {
+      "epoch": 1.226280834914611,
+      "grad_norm": 1.8128386735916138,
+      "learning_rate": 6.876237226668876e-06,
+      "loss": 0.1825,
+      "step": 2585
+    },
+    {
+      "epoch": 1.2267552182163188,
+      "grad_norm": 1.9619581699371338,
+      "learning_rate": 6.868939586142917e-06,
+      "loss": 0.191,
+      "step": 2586
+    },
+    {
+      "epoch": 1.2272296015180266,
+      "grad_norm": 1.6164292097091675,
+      "learning_rate": 6.861643793850934e-06,
+      "loss": 0.1718,
+      "step": 2587
+    },
+    {
+      "epoch": 1.2277039848197344,
+      "grad_norm": 1.7258304357528687,
+      "learning_rate": 6.854349854099565e-06,
+      "loss": 0.1814,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2281783681214422,
+      "grad_norm": 2.029466390609741,
+      "learning_rate": 6.847057771194351e-06,
+      "loss": 0.1718,
+      "step": 2589
+    },
+    {
+      "epoch": 1.2286527514231498,
+      "grad_norm": 2.171515464782715,
+      "learning_rate": 6.839767549439733e-06,
+      "loss": 0.1829,
+      "step": 2590
+    },
+    {
+      "epoch": 1.2291271347248576,
+      "grad_norm": 1.7947821617126465,
+      "learning_rate": 6.832479193139062e-06,
+      "loss": 0.1668,
+      "step": 2591
+    },
+    {
+      "epoch": 1.2296015180265654,
+      "grad_norm": 1.5067108869552612,
+      "learning_rate": 6.8251927065945755e-06,
+      "loss": 0.1532,
+      "step": 2592
+    },
+    {
+      "epoch": 1.2300759013282732,
+      "grad_norm": 1.7323521375656128,
+      "learning_rate": 6.81790809410742e-06,
+      "loss": 0.1754,
+      "step": 2593
+    },
+    {
+      "epoch": 1.230550284629981,
+      "grad_norm": 1.8438886404037476,
+      "learning_rate": 6.8106253599776275e-06,
+      "loss": 0.173,
+      "step": 2594
+    },
+    {
+      "epoch": 1.2310246679316887,
+      "grad_norm": 1.911361813545227,
+      "learning_rate": 6.803344508504124e-06,
+      "loss": 0.1942,
+      "step": 2595
+    },
+    {
+      "epoch": 1.2314990512333965,
+      "grad_norm": 2.2166242599487305,
+      "learning_rate": 6.796065543984729e-06,
+      "loss": 0.1846,
+      "step": 2596
+    },
+    {
+      "epoch": 1.2319734345351043,
+      "grad_norm": 1.8407729864120483,
+      "learning_rate": 6.788788470716136e-06,
+      "loss": 0.1958,
+      "step": 2597
+    },
+    {
+      "epoch": 1.2324478178368121,
+      "grad_norm": 1.4400639533996582,
+      "learning_rate": 6.781513292993933e-06,
+      "loss": 0.1268,
+      "step": 2598
+    },
+    {
+      "epoch": 1.23292220113852,
+      "grad_norm": 1.7160131931304932,
+      "learning_rate": 6.774240015112583e-06,
+      "loss": 0.1838,
+      "step": 2599
+    },
+    {
+      "epoch": 1.2333965844402277,
+      "grad_norm": 1.9042142629623413,
+      "learning_rate": 6.766968641365437e-06,
+      "loss": 0.1965,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2338709677419355,
+      "grad_norm": 2.0009496212005615,
+      "learning_rate": 6.759699176044705e-06,
+      "loss": 0.1894,
+      "step": 2601
+    },
+    {
+      "epoch": 1.2343453510436433,
+      "grad_norm": 1.4664678573608398,
+      "learning_rate": 6.752431623441488e-06,
+      "loss": 0.1356,
+      "step": 2602
+    },
+    {
+      "epoch": 1.234819734345351,
+      "grad_norm": 2.177931547164917,
+      "learning_rate": 6.745165987845744e-06,
+      "loss": 0.1661,
+      "step": 2603
+    },
+    {
+      "epoch": 1.2352941176470589,
+      "grad_norm": 1.4767887592315674,
+      "learning_rate": 6.73790227354631e-06,
+      "loss": 0.161,
+      "step": 2604
+    },
+    {
+      "epoch": 1.2357685009487667,
+      "grad_norm": 1.9112201929092407,
+      "learning_rate": 6.73064048483089e-06,
+      "loss": 0.1683,
+      "step": 2605
+    },
+    {
+      "epoch": 1.2362428842504745,
+      "grad_norm": 1.6860960721969604,
+      "learning_rate": 6.7233806259860355e-06,
+      "loss": 0.183,
+      "step": 2606
+    },
+    {
+      "epoch": 1.2367172675521823,
+      "grad_norm": 1.788388967514038,
+      "learning_rate": 6.716122701297173e-06,
+      "loss": 0.1649,
+      "step": 2607
+    },
+    {
+      "epoch": 1.23719165085389,
+      "grad_norm": 1.9071511030197144,
+      "learning_rate": 6.708866715048586e-06,
+      "loss": 0.1961,
+      "step": 2608
+    },
+    {
+      "epoch": 1.2376660341555978,
+      "grad_norm": 1.8968346118927002,
+      "learning_rate": 6.70161267152341e-06,
+      "loss": 0.1793,
+      "step": 2609
+    },
+    {
+      "epoch": 1.2381404174573054,
+      "grad_norm": 1.6861655712127686,
+      "learning_rate": 6.694360575003637e-06,
+      "loss": 0.145,
+      "step": 2610
+    },
+    {
+      "epoch": 1.2386148007590132,
+      "grad_norm": 1.9473601579666138,
+      "learning_rate": 6.687110429770103e-06,
+      "loss": 0.1736,
+      "step": 2611
+    },
+    {
+      "epoch": 1.239089184060721,
+      "grad_norm": 1.6562045812606812,
+      "learning_rate": 6.679862240102499e-06,
+      "loss": 0.1866,
+      "step": 2612
+    },
+    {
+      "epoch": 1.2395635673624288,
+      "grad_norm": 1.8242155313491821,
+      "learning_rate": 6.672616010279362e-06,
+      "loss": 0.1956,
+      "step": 2613
+    },
+    {
+      "epoch": 1.2400379506641366,
+      "grad_norm": 1.7770357131958008,
+      "learning_rate": 6.6653717445780675e-06,
+      "loss": 0.1963,
+      "step": 2614
+    },
+    {
+      "epoch": 1.2405123339658444,
+      "grad_norm": 1.7528514862060547,
+      "learning_rate": 6.658129447274838e-06,
+      "loss": 0.176,
+      "step": 2615
+    },
+    {
+      "epoch": 1.2409867172675522,
+      "grad_norm": 1.6112436056137085,
+      "learning_rate": 6.650889122644723e-06,
+      "loss": 0.1464,
+      "step": 2616
+    },
+    {
+      "epoch": 1.24146110056926,
+      "grad_norm": 1.8748856782913208,
+      "learning_rate": 6.6436507749616195e-06,
+      "loss": 0.1667,
+      "step": 2617
+    },
+    {
+      "epoch": 1.2419354838709677,
+      "grad_norm": 1.466968297958374,
+      "learning_rate": 6.636414408498249e-06,
+      "loss": 0.1546,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2424098671726755,
+      "grad_norm": 1.6490691900253296,
+      "learning_rate": 6.629180027526174e-06,
+      "loss": 0.1427,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2428842504743833,
+      "grad_norm": 1.4727083444595337,
+      "learning_rate": 6.6219476363157685e-06,
+      "loss": 0.145,
+      "step": 2620
+    },
+    {
+      "epoch": 1.2433586337760911,
+      "grad_norm": 1.5241156816482544,
+      "learning_rate": 6.614717239136246e-06,
+      "loss": 0.1576,
+      "step": 2621
+    },
+    {
+      "epoch": 1.243833017077799,
+      "grad_norm": 1.842430830001831,
+      "learning_rate": 6.6074888402556405e-06,
+      "loss": 0.1815,
+      "step": 2622
+    },
+    {
+      "epoch": 1.2443074003795067,
+      "grad_norm": 2.016418933868408,
+      "learning_rate": 6.600262443940799e-06,
+      "loss": 0.227,
+      "step": 2623
+    },
+    {
+      "epoch": 1.2447817836812145,
+      "grad_norm": 1.7062424421310425,
+      "learning_rate": 6.593038054457402e-06,
+      "loss": 0.1792,
+      "step": 2624
+    },
+    {
+      "epoch": 1.2452561669829223,
+      "grad_norm": 1.7208788394927979,
+      "learning_rate": 6.58581567606992e-06,
+      "loss": 0.1858,
+      "step": 2625
+    },
+    {
+      "epoch": 1.2457305502846299,
+      "grad_norm": 1.6579339504241943,
+      "learning_rate": 6.578595313041665e-06,
+      "loss": 0.1828,
+      "step": 2626
+    },
+    {
+      "epoch": 1.2462049335863377,
+      "grad_norm": 1.732631802558899,
+      "learning_rate": 6.571376969634738e-06,
+      "loss": 0.1919,
+      "step": 2627
+    },
+    {
+      "epoch": 1.2466793168880455,
+      "grad_norm": 1.7477668523788452,
+      "learning_rate": 6.564160650110057e-06,
+      "loss": 0.1894,
+      "step": 2628
+    },
+    {
+      "epoch": 1.2471537001897532,
+      "grad_norm": 1.7273080348968506,
+      "learning_rate": 6.556946358727349e-06,
+      "loss": 0.194,
+      "step": 2629
+    },
+    {
+      "epoch": 1.247628083491461,
+      "grad_norm": 1.438499927520752,
+      "learning_rate": 6.5497340997451335e-06,
+      "loss": 0.1533,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2481024667931688,
+      "grad_norm": 1.645349383354187,
+      "learning_rate": 6.5425238774207345e-06,
+      "loss": 0.1597,
+      "step": 2631
+    },
+    {
+      "epoch": 1.2485768500948766,
+      "grad_norm": 2.0088460445404053,
+      "learning_rate": 6.535315696010278e-06,
+      "loss": 0.2085,
+      "step": 2632
+    },
+    {
+      "epoch": 1.2490512333965844,
+      "grad_norm": 2.179716110229492,
+      "learning_rate": 6.528109559768685e-06,
+      "loss": 0.1696,
+      "step": 2633
+    },
+    {
+      "epoch": 1.2495256166982922,
+      "grad_norm": 1.4562453031539917,
+      "learning_rate": 6.520905472949664e-06,
+      "loss": 0.145,
+      "step": 2634
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8796703815460205,
+      "learning_rate": 6.51370343980571e-06,
+      "loss": 0.1897,
+      "step": 2635
+    },
+    {
+      "epoch": 1.2504743833017078,
+      "grad_norm": 1.4412729740142822,
+      "learning_rate": 6.506503464588119e-06,
+      "loss": 0.1843,
+      "step": 2636
+    },
+    {
+      "epoch": 1.2509487666034156,
+      "grad_norm": 1.632536768913269,
+      "learning_rate": 6.499305551546964e-06,
+      "loss": 0.1567,
+      "step": 2637
+    },
+    {
+      "epoch": 1.2514231499051234,
+      "grad_norm": 1.5959841012954712,
+      "learning_rate": 6.492109704931101e-06,
+      "loss": 0.1611,
+      "step": 2638
+    },
+    {
+      "epoch": 1.2518975332068312,
+      "grad_norm": 1.659533977508545,
+      "learning_rate": 6.484915928988167e-06,
+      "loss": 0.1776,
+      "step": 2639
+    },
+    {
+      "epoch": 1.252371916508539,
+      "grad_norm": 2.8352959156036377,
+      "learning_rate": 6.4777242279645754e-06,
+      "loss": 0.1758,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2528462998102468,
+      "grad_norm": 1.6581546068191528,
+      "learning_rate": 6.470534606105519e-06,
+      "loss": 0.1767,
+      "step": 2641
+    },
+    {
+      "epoch": 1.2533206831119545,
+      "grad_norm": 2.0722496509552,
+      "learning_rate": 6.463347067654959e-06,
+      "loss": 0.1892,
+      "step": 2642
+    },
+    {
+      "epoch": 1.2537950664136623,
+      "grad_norm": 1.7481244802474976,
+      "learning_rate": 6.456161616855631e-06,
+      "loss": 0.1726,
+      "step": 2643
+    },
+    {
+      "epoch": 1.2542694497153701,
+      "grad_norm": 2.0864098072052,
+      "learning_rate": 6.448978257949028e-06,
+      "loss": 0.2157,
+      "step": 2644
+    },
+    {
+      "epoch": 1.254743833017078,
+      "grad_norm": 1.7608370780944824,
+      "learning_rate": 6.441796995175422e-06,
+      "loss": 0.1735,
+      "step": 2645
+    },
+    {
+      "epoch": 1.2552182163187855,
+      "grad_norm": 1.9006822109222412,
+      "learning_rate": 6.4346178327738405e-06,
+      "loss": 0.1814,
+      "step": 2646
+    },
+    {
+      "epoch": 1.2556925996204933,
+      "grad_norm": 1.6309512853622437,
+      "learning_rate": 6.42744077498207e-06,
+      "loss": 0.1562,
+      "step": 2647
+    },
+    {
+      "epoch": 1.256166982922201,
+      "grad_norm": 1.9142560958862305,
+      "learning_rate": 6.420265826036663e-06,
+      "loss": 0.1902,
+      "step": 2648
+    },
+    {
+      "epoch": 1.2566413662239089,
+      "grad_norm": 2.259213447570801,
+      "learning_rate": 6.41309299017291e-06,
+      "loss": 0.2249,
+      "step": 2649
+    },
+    {
+      "epoch": 1.2571157495256167,
+      "grad_norm": 1.8461946249008179,
+      "learning_rate": 6.405922271624874e-06,
+      "loss": 0.1775,
+      "step": 2650
+    },
+    {
+      "epoch": 1.2575901328273245,
+      "grad_norm": 2.4497241973876953,
+      "learning_rate": 6.398753674625353e-06,
+      "loss": 0.1998,
+      "step": 2651
+    },
+    {
+      "epoch": 1.2580645161290323,
+      "grad_norm": 1.6050024032592773,
+      "learning_rate": 6.391587203405903e-06,
+      "loss": 0.1655,
+      "step": 2652
+    },
+    {
+      "epoch": 1.25853889943074,
+      "grad_norm": 1.7544989585876465,
+      "learning_rate": 6.384422862196824e-06,
+      "loss": 0.1495,
+      "step": 2653
+    },
+    {
+      "epoch": 1.2590132827324478,
+      "grad_norm": 1.697287917137146,
+      "learning_rate": 6.37726065522715e-06,
+      "loss": 0.1454,
+      "step": 2654
+    },
+    {
+      "epoch": 1.2594876660341556,
+      "grad_norm": 3.1953728199005127,
+      "learning_rate": 6.37010058672466e-06,
+      "loss": 0.2283,
+      "step": 2655
+    },
+    {
+      "epoch": 1.2599620493358634,
+      "grad_norm": 1.9299657344818115,
+      "learning_rate": 6.362942660915875e-06,
+      "loss": 0.2082,
+      "step": 2656
+    },
+    {
+      "epoch": 1.2604364326375712,
+      "grad_norm": 1.5635534524917603,
+      "learning_rate": 6.3557868820260495e-06,
+      "loss": 0.1593,
+      "step": 2657
+    },
+    {
+      "epoch": 1.260910815939279,
+      "grad_norm": 2.0182600021362305,
+      "learning_rate": 6.348633254279166e-06,
+      "loss": 0.1987,
+      "step": 2658
+    },
+    {
+      "epoch": 1.2613851992409868,
+      "grad_norm": 1.7452116012573242,
+      "learning_rate": 6.341481781897939e-06,
+      "loss": 0.205,
+      "step": 2659
+    },
+    {
+      "epoch": 1.2618595825426944,
+      "grad_norm": 2.041940450668335,
+      "learning_rate": 6.334332469103814e-06,
+      "loss": 0.2476,
+      "step": 2660
+    },
+    {
+      "epoch": 1.2623339658444022,
+      "grad_norm": 1.419898509979248,
+      "learning_rate": 6.3271853201169594e-06,
+      "loss": 0.1227,
+      "step": 2661
+    },
+    {
+      "epoch": 1.26280834914611,
+      "grad_norm": 1.834674596786499,
+      "learning_rate": 6.320040339156267e-06,
+      "loss": 0.1672,
+      "step": 2662
+    },
+    {
+      "epoch": 1.2632827324478177,
+      "grad_norm": 1.685529351234436,
+      "learning_rate": 6.312897530439348e-06,
+      "loss": 0.1822,
+      "step": 2663
+    },
+    {
+      "epoch": 1.2637571157495255,
+      "grad_norm": 1.5215164422988892,
+      "learning_rate": 6.305756898182529e-06,
+      "loss": 0.1613,
+      "step": 2664
+    },
+    {
+      "epoch": 1.2642314990512333,
+      "grad_norm": 16.828428268432617,
+      "learning_rate": 6.298618446600856e-06,
+      "loss": 0.233,
+      "step": 2665
+    },
+    {
+      "epoch": 1.2647058823529411,
+      "grad_norm": 2.161356210708618,
+      "learning_rate": 6.29148217990809e-06,
+      "loss": 0.1777,
+      "step": 2666
+    },
+    {
+      "epoch": 1.265180265654649,
+      "grad_norm": 1.8170483112335205,
+      "learning_rate": 6.2843481023166975e-06,
+      "loss": 0.1898,
+      "step": 2667
+    },
+    {
+      "epoch": 1.2656546489563567,
+      "grad_norm": 1.768537998199463,
+      "learning_rate": 6.27721621803785e-06,
+      "loss": 0.1702,
+      "step": 2668
+    },
+    {
+      "epoch": 1.2661290322580645,
+      "grad_norm": 1.892989993095398,
+      "learning_rate": 6.270086531281428e-06,
+      "loss": 0.205,
+      "step": 2669
+    },
+    {
+      "epoch": 1.2666034155597723,
+      "grad_norm": 1.9306186437606812,
+      "learning_rate": 6.262959046256021e-06,
+      "loss": 0.2097,
+      "step": 2670
+    },
+    {
+      "epoch": 1.26707779886148,
+      "grad_norm": 1.8183523416519165,
+      "learning_rate": 6.255833767168907e-06,
+      "loss": 0.1883,
+      "step": 2671
+    },
+    {
+      "epoch": 1.2675521821631879,
+      "grad_norm": 1.4449431896209717,
+      "learning_rate": 6.248710698226074e-06,
+      "loss": 0.1465,
+      "step": 2672
+    },
+    {
+      "epoch": 1.2680265654648957,
+      "grad_norm": 1.5579372644424438,
+      "learning_rate": 6.241589843632192e-06,
+      "loss": 0.1579,
+      "step": 2673
+    },
+    {
+      "epoch": 1.2685009487666035,
+      "grad_norm": 1.6067235469818115,
+      "learning_rate": 6.234471207590636e-06,
+      "loss": 0.1597,
+      "step": 2674
+    },
+    {
+      "epoch": 1.2689753320683113,
+      "grad_norm": 1.6292030811309814,
+      "learning_rate": 6.227354794303461e-06,
+      "loss": 0.2124,
+      "step": 2675
+    },
+    {
+      "epoch": 1.269449715370019,
+      "grad_norm": 1.8330110311508179,
+      "learning_rate": 6.220240607971422e-06,
+      "loss": 0.1691,
+      "step": 2676
+    },
+    {
+      "epoch": 1.2699240986717268,
+      "grad_norm": 1.4492367506027222,
+      "learning_rate": 6.213128652793952e-06,
+      "loss": 0.1557,
+      "step": 2677
+    },
+    {
+      "epoch": 1.2703984819734346,
+      "grad_norm": 1.5724751949310303,
+      "learning_rate": 6.206018932969162e-06,
+      "loss": 0.142,
+      "step": 2678
+    },
+    {
+      "epoch": 1.2708728652751424,
+      "grad_norm": 1.4839202165603638,
+      "learning_rate": 6.1989114526938535e-06,
+      "loss": 0.1237,
+      "step": 2679
+    },
+    {
+      "epoch": 1.2713472485768502,
+      "grad_norm": 1.7914515733718872,
+      "learning_rate": 6.1918062161635005e-06,
+      "loss": 0.185,
+      "step": 2680
+    },
+    {
+      "epoch": 1.271821631878558,
+      "grad_norm": 1.844212532043457,
+      "learning_rate": 6.184703227572257e-06,
+      "loss": 0.1892,
+      "step": 2681
+    },
+    {
+      "epoch": 1.2722960151802656,
+      "grad_norm": 1.94633948802948,
+      "learning_rate": 6.1776024911129414e-06,
+      "loss": 0.1789,
+      "step": 2682
+    },
+    {
+      "epoch": 1.2727703984819734,
+      "grad_norm": 1.849686622619629,
+      "learning_rate": 6.170504010977053e-06,
+      "loss": 0.1831,
+      "step": 2683
+    },
+    {
+      "epoch": 1.2732447817836812,
+      "grad_norm": 1.323569893836975,
+      "learning_rate": 6.163407791354751e-06,
+      "loss": 0.1487,
+      "step": 2684
+    },
+    {
+      "epoch": 1.273719165085389,
+      "grad_norm": 2.2079055309295654,
+      "learning_rate": 6.156313836434864e-06,
+      "loss": 0.1744,
+      "step": 2685
+    },
+    {
+      "epoch": 1.2741935483870968,
+      "grad_norm": 1.6516516208648682,
+      "learning_rate": 6.149222150404889e-06,
+      "loss": 0.1772,
+      "step": 2686
+    },
+    {
+      "epoch": 1.2746679316888045,
+      "grad_norm": 1.9557101726531982,
+      "learning_rate": 6.142132737450971e-06,
+      "loss": 0.2113,
+      "step": 2687
+    },
+    {
+      "epoch": 1.2751423149905123,
+      "grad_norm": 1.9967963695526123,
+      "learning_rate": 6.135045601757921e-06,
+      "loss": 0.1954,
+      "step": 2688
+    },
+    {
+      "epoch": 1.2756166982922201,
+      "grad_norm": 1.9493542909622192,
+      "learning_rate": 6.127960747509207e-06,
+      "loss": 0.1692,
+      "step": 2689
+    },
+    {
+      "epoch": 1.276091081593928,
+      "grad_norm": 2.279658794403076,
+      "learning_rate": 6.120878178886951e-06,
+      "loss": 0.2131,
+      "step": 2690
+    },
+    {
+      "epoch": 1.2765654648956357,
+      "grad_norm": 1.827585220336914,
+      "learning_rate": 6.113797900071923e-06,
+      "loss": 0.1943,
+      "step": 2691
+    },
+    {
+      "epoch": 1.2770398481973435,
+      "grad_norm": 1.7768232822418213,
+      "learning_rate": 6.106719915243533e-06,
+      "loss": 0.1943,
+      "step": 2692
+    },
+    {
+      "epoch": 1.2775142314990513,
+      "grad_norm": 1.574704885482788,
+      "learning_rate": 6.099644228579852e-06,
+      "loss": 0.1617,
+      "step": 2693
+    },
+    {
+      "epoch": 1.277988614800759,
+      "grad_norm": 2.0819928646087646,
+      "learning_rate": 6.092570844257589e-06,
+      "loss": 0.2005,
+      "step": 2694
+    },
+    {
+      "epoch": 1.2784629981024667,
+      "grad_norm": 2.360511064529419,
+      "learning_rate": 6.08549976645209e-06,
+      "loss": 0.2153,
+      "step": 2695
+    },
+    {
+      "epoch": 1.2789373814041745,
+      "grad_norm": 1.7095462083816528,
+      "learning_rate": 6.078430999337346e-06,
+      "loss": 0.1612,
+      "step": 2696
+    },
+    {
+      "epoch": 1.2794117647058822,
+      "grad_norm": 2.0695948600769043,
+      "learning_rate": 6.071364547085974e-06,
+      "loss": 0.1911,
+      "step": 2697
+    },
+    {
+      "epoch": 1.27988614800759,
+      "grad_norm": 1.8348363637924194,
+      "learning_rate": 6.064300413869237e-06,
+      "loss": 0.1883,
+      "step": 2698
+    },
+    {
+      "epoch": 1.2803605313092978,
+      "grad_norm": 1.8135762214660645,
+      "learning_rate": 6.057238603857018e-06,
+      "loss": 0.1442,
+      "step": 2699
+    },
+    {
+      "epoch": 1.2808349146110056,
+      "grad_norm": 2.002474546432495,
+      "learning_rate": 6.050179121217839e-06,
+      "loss": 0.2274,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2813092979127134,
+      "grad_norm": 2.7125346660614014,
+      "learning_rate": 6.043121970118837e-06,
+      "loss": 0.1427,
+      "step": 2701
+    },
+    {
+      "epoch": 1.2817836812144212,
+      "grad_norm": 1.8385379314422607,
+      "learning_rate": 6.0360671547257825e-06,
+      "loss": 0.1609,
+      "step": 2702
+    },
+    {
+      "epoch": 1.282258064516129,
+      "grad_norm": 1.4639102220535278,
+      "learning_rate": 6.029014679203059e-06,
+      "loss": 0.1462,
+      "step": 2703
+    },
+    {
+      "epoch": 1.2827324478178368,
+      "grad_norm": 1.8368364572525024,
+      "learning_rate": 6.0219645477136764e-06,
+      "loss": 0.1943,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2832068311195446,
+      "grad_norm": 1.7069745063781738,
+      "learning_rate": 6.014916764419261e-06,
+      "loss": 0.1726,
+      "step": 2705
+    },
+    {
+      "epoch": 1.2836812144212524,
+      "grad_norm": 1.7310154438018799,
+      "learning_rate": 6.007871333480041e-06,
+      "loss": 0.1615,
+      "step": 2706
+    },
+    {
+      "epoch": 1.2841555977229602,
+      "grad_norm": 1.5262956619262695,
+      "learning_rate": 6.000828259054872e-06,
+      "loss": 0.1396,
+      "step": 2707
+    },
+    {
+      "epoch": 1.284629981024668,
+      "grad_norm": 1.7818677425384521,
+      "learning_rate": 5.993787545301204e-06,
+      "loss": 0.1845,
+      "step": 2708
+    },
+    {
+      "epoch": 1.2851043643263758,
+      "grad_norm": 2.1896657943725586,
+      "learning_rate": 5.986749196375108e-06,
+      "loss": 0.2015,
+      "step": 2709
+    },
+    {
+      "epoch": 1.2855787476280836,
+      "grad_norm": 2.082186698913574,
+      "learning_rate": 5.97971321643125e-06,
+      "loss": 0.2172,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2860531309297913,
+      "grad_norm": 2.232640504837036,
+      "learning_rate": 5.972679609622897e-06,
+      "loss": 0.2158,
+      "step": 2711
+    },
+    {
+      "epoch": 1.2865275142314991,
+      "grad_norm": 1.818824291229248,
+      "learning_rate": 5.965648380101916e-06,
+      "loss": 0.1672,
+      "step": 2712
+    },
+    {
+      "epoch": 1.287001897533207,
+      "grad_norm": 2.0493948459625244,
+      "learning_rate": 5.958619532018775e-06,
+      "loss": 0.1805,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2874762808349147,
+      "grad_norm": 2.046306610107422,
+      "learning_rate": 5.951593069522535e-06,
+      "loss": 0.1763,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2879506641366225,
+      "grad_norm": 1.797914743423462,
+      "learning_rate": 5.944568996760847e-06,
+      "loss": 0.2099,
+      "step": 2715
+    },
+    {
+      "epoch": 1.2884250474383303,
+      "grad_norm": 1.5372848510742188,
+      "learning_rate": 5.937547317879946e-06,
+      "loss": 0.1515,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2888994307400379,
+      "grad_norm": 1.811116337776184,
+      "learning_rate": 5.930528037024664e-06,
+      "loss": 0.151,
+      "step": 2717
+    },
+    {
+      "epoch": 1.2893738140417457,
+      "grad_norm": 1.7281832695007324,
+      "learning_rate": 5.923511158338415e-06,
+      "loss": 0.1729,
+      "step": 2718
+    },
+    {
+      "epoch": 1.2898481973434535,
+      "grad_norm": 1.5258969068527222,
+      "learning_rate": 5.916496685963191e-06,
+      "loss": 0.1412,
+      "step": 2719
+    },
+    {
+      "epoch": 1.2903225806451613,
+      "grad_norm": 1.6176496744155884,
+      "learning_rate": 5.909484624039563e-06,
+      "loss": 0.1712,
+      "step": 2720
+    },
+    {
+      "epoch": 1.290796963946869,
+      "grad_norm": 1.8939071893692017,
+      "learning_rate": 5.9024749767066835e-06,
+      "loss": 0.1806,
+      "step": 2721
+    },
+    {
+      "epoch": 1.2912713472485768,
+      "grad_norm": 1.8788526058197021,
+      "learning_rate": 5.89546774810228e-06,
+      "loss": 0.17,
+      "step": 2722
+    },
+    {
+      "epoch": 1.2917457305502846,
+      "grad_norm": 2.107088327407837,
+      "learning_rate": 5.888462942362647e-06,
+      "loss": 0.18,
+      "step": 2723
+    },
+    {
+      "epoch": 1.2922201138519924,
+      "grad_norm": 1.6402045488357544,
+      "learning_rate": 5.881460563622659e-06,
+      "loss": 0.1535,
+      "step": 2724
+    },
+    {
+      "epoch": 1.2926944971537002,
+      "grad_norm": 1.9502211809158325,
+      "learning_rate": 5.87446061601574e-06,
+      "loss": 0.1858,
+      "step": 2725
+    },
+    {
+      "epoch": 1.293168880455408,
+      "grad_norm": 1.5377591848373413,
+      "learning_rate": 5.867463103673898e-06,
+      "loss": 0.1621,
+      "step": 2726
+    },
+    {
+      "epoch": 1.2936432637571158,
+      "grad_norm": 2.0564045906066895,
+      "learning_rate": 5.8604680307276906e-06,
+      "loss": 0.1743,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2941176470588236,
+      "grad_norm": 1.4501993656158447,
+      "learning_rate": 5.853475401306241e-06,
+      "loss": 0.1646,
+      "step": 2728
+    },
+    {
+      "epoch": 1.2945920303605314,
+      "grad_norm": 2.006211042404175,
+      "learning_rate": 5.846485219537237e-06,
+      "loss": 0.2069,
+      "step": 2729
+    },
+    {
+      "epoch": 1.295066413662239,
+      "grad_norm": 2.675589084625244,
+      "learning_rate": 5.8394974895469015e-06,
+      "loss": 0.2098,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2955407969639468,
+      "grad_norm": 1.6392362117767334,
+      "learning_rate": 5.83251221546003e-06,
+      "loss": 0.1624,
+      "step": 2731
+    },
+    {
+      "epoch": 1.2960151802656545,
+      "grad_norm": 1.847262978553772,
+      "learning_rate": 5.825529401399956e-06,
+      "loss": 0.2169,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2964895635673623,
+      "grad_norm": 1.3665305376052856,
+      "learning_rate": 5.818549051488569e-06,
+      "loss": 0.1345,
+      "step": 2733
+    },
+    {
+      "epoch": 1.2969639468690701,
+      "grad_norm": 1.4444960355758667,
+      "learning_rate": 5.811571169846304e-06,
+      "loss": 0.1572,
+      "step": 2734
+    },
+    {
+      "epoch": 1.297438330170778,
+      "grad_norm": 1.631394624710083,
+      "learning_rate": 5.804595760592127e-06,
+      "loss": 0.1435,
+      "step": 2735
+    },
+    {
+      "epoch": 1.2979127134724857,
+      "grad_norm": 2.4109082221984863,
+      "learning_rate": 5.797622827843561e-06,
+      "loss": 0.2491,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2983870967741935,
+      "grad_norm": 1.8289345502853394,
+      "learning_rate": 5.790652375716653e-06,
+      "loss": 0.1797,
+      "step": 2737
+    },
+    {
+      "epoch": 1.2988614800759013,
+      "grad_norm": 1.665688157081604,
+      "learning_rate": 5.7836844083259954e-06,
+      "loss": 0.1651,
+      "step": 2738
+    },
+    {
+      "epoch": 1.299335863377609,
+      "grad_norm": 1.573738932609558,
+      "learning_rate": 5.776718929784707e-06,
+      "loss": 0.156,
+      "step": 2739
+    },
+    {
+      "epoch": 1.2998102466793169,
+      "grad_norm": 1.9333720207214355,
+      "learning_rate": 5.769755944204443e-06,
+      "loss": 0.1613,
+      "step": 2740
+    },
+    {
+      "epoch": 1.3002846299810247,
+      "grad_norm": 1.6634092330932617,
+      "learning_rate": 5.762795455695385e-06,
+      "loss": 0.198,
+      "step": 2741
+    },
+    {
+      "epoch": 1.3007590132827325,
+      "grad_norm": 1.8421812057495117,
+      "learning_rate": 5.755837468366241e-06,
+      "loss": 0.1715,
+      "step": 2742
+    },
+    {
+      "epoch": 1.3012333965844403,
+      "grad_norm": 1.5852497816085815,
+      "learning_rate": 5.748881986324245e-06,
+      "loss": 0.1821,
+      "step": 2743
+    },
+    {
+      "epoch": 1.301707779886148,
+      "grad_norm": 1.8755565881729126,
+      "learning_rate": 5.741929013675143e-06,
+      "loss": 0.182,
+      "step": 2744
+    },
+    {
+      "epoch": 1.3021821631878558,
+      "grad_norm": 1.483273983001709,
+      "learning_rate": 5.7349785545232115e-06,
+      "loss": 0.1509,
+      "step": 2745
+    },
+    {
+      "epoch": 1.3026565464895636,
+      "grad_norm": 1.5257030725479126,
+      "learning_rate": 5.728030612971231e-06,
+      "loss": 0.1342,
+      "step": 2746
+    },
+    {
+      "epoch": 1.3031309297912714,
+      "grad_norm": 1.8776187896728516,
+      "learning_rate": 5.721085193120507e-06,
+      "loss": 0.1959,
+      "step": 2747
+    },
+    {
+      "epoch": 1.3036053130929792,
+      "grad_norm": 1.5106158256530762,
+      "learning_rate": 5.714142299070856e-06,
+      "loss": 0.1563,
+      "step": 2748
+    },
+    {
+      "epoch": 1.304079696394687,
+      "grad_norm": 1.8498852252960205,
+      "learning_rate": 5.7072019349205925e-06,
+      "loss": 0.216,
+      "step": 2749
+    },
+    {
+      "epoch": 1.3045540796963948,
+      "grad_norm": 1.8534907102584839,
+      "learning_rate": 5.700264104766547e-06,
+      "loss": 0.1699,
+      "step": 2750
+    },
+    {
+      "epoch": 1.3050284629981026,
+      "grad_norm": 1.647162675857544,
+      "learning_rate": 5.6933288127040505e-06,
+      "loss": 0.1455,
+      "step": 2751
+    },
+    {
+      "epoch": 1.3055028462998102,
+      "grad_norm": 1.767333745956421,
+      "learning_rate": 5.686396062826946e-06,
+      "loss": 0.1483,
+      "step": 2752
+    },
+    {
+      "epoch": 1.305977229601518,
+      "grad_norm": 1.979466438293457,
+      "learning_rate": 5.679465859227561e-06,
+      "loss": 0.1747,
+      "step": 2753
+    },
+    {
+      "epoch": 1.3064516129032258,
+      "grad_norm": 1.9015872478485107,
+      "learning_rate": 5.6725382059967205e-06,
+      "loss": 0.2133,
+      "step": 2754
+    },
+    {
+      "epoch": 1.3069259962049335,
+      "grad_norm": 1.6040949821472168,
+      "learning_rate": 5.665613107223755e-06,
+      "loss": 0.158,
+      "step": 2755
+    },
+    {
+      "epoch": 1.3074003795066413,
+      "grad_norm": 1.9205068349838257,
+      "learning_rate": 5.658690566996483e-06,
+      "loss": 0.1565,
+      "step": 2756
+    },
+    {
+      "epoch": 1.3078747628083491,
+      "grad_norm": 1.864329218864441,
+      "learning_rate": 5.651770589401209e-06,
+      "loss": 0.1957,
+      "step": 2757
+    },
+    {
+      "epoch": 1.308349146110057,
+      "grad_norm": 2.1258440017700195,
+      "learning_rate": 5.644853178522734e-06,
+      "loss": 0.1696,
+      "step": 2758
+    },
+    {
+      "epoch": 1.3088235294117647,
+      "grad_norm": 1.630826711654663,
+      "learning_rate": 5.637938338444325e-06,
+      "loss": 0.1278,
+      "step": 2759
+    },
+    {
+      "epoch": 1.3092979127134725,
+      "grad_norm": 1.273641586303711,
+      "learning_rate": 5.631026073247752e-06,
+      "loss": 0.1296,
+      "step": 2760
+    },
+    {
+      "epoch": 1.3097722960151803,
+      "grad_norm": 1.7420580387115479,
+      "learning_rate": 5.624116387013259e-06,
+      "loss": 0.1666,
+      "step": 2761
+    },
+    {
+      "epoch": 1.310246679316888,
+      "grad_norm": 1.7752727270126343,
+      "learning_rate": 5.617209283819562e-06,
+      "loss": 0.1868,
+      "step": 2762
+    },
+    {
+      "epoch": 1.310721062618596,
+      "grad_norm": 1.8602519035339355,
+      "learning_rate": 5.61030476774385e-06,
+      "loss": 0.1966,
+      "step": 2763
+    },
+    {
+      "epoch": 1.3111954459203037,
+      "grad_norm": 1.8290077447891235,
+      "learning_rate": 5.603402842861797e-06,
+      "loss": 0.1749,
+      "step": 2764
+    },
+    {
+      "epoch": 1.3116698292220113,
+      "grad_norm": 1.944825291633606,
+      "learning_rate": 5.5965035132475395e-06,
+      "loss": 0.2062,
+      "step": 2765
+    },
+    {
+      "epoch": 1.312144212523719,
+      "grad_norm": 1.7076294422149658,
+      "learning_rate": 5.589606782973683e-06,
+      "loss": 0.1738,
+      "step": 2766
+    },
+    {
+      "epoch": 1.3126185958254268,
+      "grad_norm": 1.9629321098327637,
+      "learning_rate": 5.5827126561113045e-06,
+      "loss": 0.1418,
+      "step": 2767
+    },
+    {
+      "epoch": 1.3130929791271346,
+      "grad_norm": 1.9150692224502563,
+      "learning_rate": 5.575821136729929e-06,
+      "loss": 0.1978,
+      "step": 2768
+    },
+    {
+      "epoch": 1.3135673624288424,
+      "grad_norm": 1.7870982885360718,
+      "learning_rate": 5.568932228897563e-06,
+      "loss": 0.1764,
+      "step": 2769
+    },
+    {
+      "epoch": 1.3140417457305502,
+      "grad_norm": 2.10677170753479,
+      "learning_rate": 5.562045936680649e-06,
+      "loss": 0.2014,
+      "step": 2770
+    },
+    {
+      "epoch": 1.314516129032258,
+      "grad_norm": 2.482201337814331,
+      "learning_rate": 5.555162264144105e-06,
+      "loss": 0.2658,
+      "step": 2771
+    },
+    {
+      "epoch": 1.3149905123339658,
+      "grad_norm": 1.6711809635162354,
+      "learning_rate": 5.548281215351297e-06,
+      "loss": 0.1548,
+      "step": 2772
+    },
+    {
+      "epoch": 1.3154648956356736,
+      "grad_norm": 2.137512683868408,
+      "learning_rate": 5.54140279436403e-06,
+      "loss": 0.2021,
+      "step": 2773
+    },
+    {
+      "epoch": 1.3159392789373814,
+      "grad_norm": 1.648645281791687,
+      "learning_rate": 5.534527005242575e-06,
+      "loss": 0.1697,
+      "step": 2774
+    },
+    {
+      "epoch": 1.3164136622390892,
+      "grad_norm": 1.3213448524475098,
+      "learning_rate": 5.52765385204564e-06,
+      "loss": 0.1197,
+      "step": 2775
+    },
+    {
+      "epoch": 1.316888045540797,
+      "grad_norm": 1.4370523691177368,
+      "learning_rate": 5.520783338830386e-06,
+      "loss": 0.1534,
+      "step": 2776
+    },
+    {
+      "epoch": 1.3173624288425048,
+      "grad_norm": 1.695040225982666,
+      "learning_rate": 5.5139154696524025e-06,
+      "loss": 0.1957,
+      "step": 2777
+    },
+    {
+      "epoch": 1.3178368121442126,
+      "grad_norm": 1.5011173486709595,
+      "learning_rate": 5.5070502485657216e-06,
+      "loss": 0.157,
+      "step": 2778
+    },
+    {
+      "epoch": 1.3183111954459203,
+      "grad_norm": 1.3054955005645752,
+      "learning_rate": 5.500187679622819e-06,
+      "loss": 0.1141,
+      "step": 2779
+    },
+    {
+      "epoch": 1.3187855787476281,
+      "grad_norm": 1.8683662414550781,
+      "learning_rate": 5.4933277668746036e-06,
+      "loss": 0.1725,
+      "step": 2780
+    },
+    {
+      "epoch": 1.319259962049336,
+      "grad_norm": 1.2149678468704224,
+      "learning_rate": 5.486470514370415e-06,
+      "loss": 0.1239,
+      "step": 2781
+    },
+    {
+      "epoch": 1.3197343453510437,
+      "grad_norm": 1.8600114583969116,
+      "learning_rate": 5.479615926158013e-06,
+      "loss": 0.1712,
+      "step": 2782
+    },
+    {
+      "epoch": 1.3202087286527515,
+      "grad_norm": 1.914152979850769,
+      "learning_rate": 5.4727640062836e-06,
+      "loss": 0.1836,
+      "step": 2783
+    },
+    {
+      "epoch": 1.3206831119544593,
+      "grad_norm": 1.7095372676849365,
+      "learning_rate": 5.465914758791794e-06,
+      "loss": 0.17,
+      "step": 2784
+    },
+    {
+      "epoch": 1.321157495256167,
+      "grad_norm": 1.9404311180114746,
+      "learning_rate": 5.459068187725644e-06,
+      "loss": 0.1771,
+      "step": 2785
+    },
+    {
+      "epoch": 1.321631878557875,
+      "grad_norm": 1.4658271074295044,
+      "learning_rate": 5.452224297126607e-06,
+      "loss": 0.1525,
+      "step": 2786
+    },
+    {
+      "epoch": 1.3221062618595825,
+      "grad_norm": 1.9287587404251099,
+      "learning_rate": 5.445383091034564e-06,
+      "loss": 0.2087,
+      "step": 2787
+    },
+    {
+      "epoch": 1.3225806451612903,
+      "grad_norm": 1.8362351655960083,
+      "learning_rate": 5.438544573487811e-06,
+      "loss": 0.1848,
+      "step": 2788
+    },
+    {
+      "epoch": 1.323055028462998,
+      "grad_norm": 1.9103182554244995,
+      "learning_rate": 5.431708748523058e-06,
+      "loss": 0.1918,
+      "step": 2789
+    },
+    {
+      "epoch": 1.3235294117647058,
+      "grad_norm": 1.722016453742981,
+      "learning_rate": 5.424875620175427e-06,
+      "loss": 0.1852,
+      "step": 2790
+    },
+    {
+      "epoch": 1.3240037950664136,
+      "grad_norm": 1.559799313545227,
+      "learning_rate": 5.4180451924784475e-06,
+      "loss": 0.1548,
+      "step": 2791
+    },
+    {
+      "epoch": 1.3244781783681214,
+      "grad_norm": 2.0171709060668945,
+      "learning_rate": 5.4112174694640475e-06,
+      "loss": 0.2102,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3249525616698292,
+      "grad_norm": 1.49323570728302,
+      "learning_rate": 5.404392455162571e-06,
+      "loss": 0.1358,
+      "step": 2793
+    },
+    {
+      "epoch": 1.325426944971537,
+      "grad_norm": 1.6861432790756226,
+      "learning_rate": 5.397570153602747e-06,
+      "loss": 0.1531,
+      "step": 2794
+    },
+    {
+      "epoch": 1.3259013282732448,
+      "grad_norm": 2.1216015815734863,
+      "learning_rate": 5.39075056881172e-06,
+      "loss": 0.1996,
+      "step": 2795
+    },
+    {
+      "epoch": 1.3263757115749526,
+      "grad_norm": 1.6811705827713013,
+      "learning_rate": 5.383933704815025e-06,
+      "loss": 0.1685,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3268500948766604,
+      "grad_norm": 1.721708059310913,
+      "learning_rate": 5.377119565636584e-06,
+      "loss": 0.1591,
+      "step": 2797
+    },
+    {
+      "epoch": 1.3273244781783682,
+      "grad_norm": 1.8749873638153076,
+      "learning_rate": 5.370308155298716e-06,
+      "loss": 0.177,
+      "step": 2798
+    },
+    {
+      "epoch": 1.327798861480076,
+      "grad_norm": 1.5626192092895508,
+      "learning_rate": 5.363499477822132e-06,
+      "loss": 0.1364,
+      "step": 2799
+    },
+    {
+      "epoch": 1.3282732447817835,
+      "grad_norm": 1.7851414680480957,
+      "learning_rate": 5.35669353722593e-06,
+      "loss": 0.1724,
+      "step": 2800
+    },
+    {
+      "epoch": 1.3287476280834913,
+      "grad_norm": 1.9391238689422607,
+      "learning_rate": 5.3498903375275815e-06,
+      "loss": 0.161,
+      "step": 2801
+    },
+    {
+      "epoch": 1.3292220113851991,
+      "grad_norm": 2.0420241355895996,
+      "learning_rate": 5.3430898827429555e-06,
+      "loss": 0.1735,
+      "step": 2802
+    },
+    {
+      "epoch": 1.329696394686907,
+      "grad_norm": 1.4055426120758057,
+      "learning_rate": 5.336292176886287e-06,
+      "loss": 0.1211,
+      "step": 2803
+    },
+    {
+      "epoch": 1.3301707779886147,
+      "grad_norm": 1.658674716949463,
+      "learning_rate": 5.329497223970195e-06,
+      "loss": 0.1562,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3306451612903225,
+      "grad_norm": 2.3525173664093018,
+      "learning_rate": 5.32270502800568e-06,
+      "loss": 0.2096,
+      "step": 2805
+    },
+    {
+      "epoch": 1.3311195445920303,
+      "grad_norm": 1.7249343395233154,
+      "learning_rate": 5.3159155930021e-06,
+      "loss": 0.1569,
+      "step": 2806
+    },
+    {
+      "epoch": 1.331593927893738,
+      "grad_norm": 2.1588501930236816,
+      "learning_rate": 5.309128922967194e-06,
+      "loss": 0.2027,
+      "step": 2807
+    },
+    {
+      "epoch": 1.3320683111954459,
+      "grad_norm": 2.3709404468536377,
+      "learning_rate": 5.302345021907066e-06,
+      "loss": 0.2227,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3325426944971537,
+      "grad_norm": 1.7174904346466064,
+      "learning_rate": 5.295563893826191e-06,
+      "loss": 0.1723,
+      "step": 2809
+    },
+    {
+      "epoch": 1.3330170777988615,
+      "grad_norm": 1.9293659925460815,
+      "learning_rate": 5.288785542727397e-06,
+      "loss": 0.1562,
+      "step": 2810
+    },
+    {
+      "epoch": 1.3334914611005693,
+      "grad_norm": 1.5283536911010742,
+      "learning_rate": 5.282009972611873e-06,
+      "loss": 0.1329,
+      "step": 2811
+    },
+    {
+      "epoch": 1.333965844402277,
+      "grad_norm": 1.8217898607254028,
+      "learning_rate": 5.275237187479176e-06,
+      "loss": 0.1772,
+      "step": 2812
+    },
+    {
+      "epoch": 1.3344402277039848,
+      "grad_norm": 2.461566686630249,
+      "learning_rate": 5.268467191327214e-06,
+      "loss": 0.2084,
+      "step": 2813
+    },
+    {
+      "epoch": 1.3349146110056926,
+      "grad_norm": 1.7789011001586914,
+      "learning_rate": 5.261699988152249e-06,
+      "loss": 0.1837,
+      "step": 2814
+    },
+    {
+      "epoch": 1.3353889943074004,
+      "grad_norm": 1.8147854804992676,
+      "learning_rate": 5.254935581948897e-06,
+      "loss": 0.2358,
+      "step": 2815
+    },
+    {
+      "epoch": 1.3358633776091082,
+      "grad_norm": 1.6862436532974243,
+      "learning_rate": 5.248173976710111e-06,
+      "loss": 0.1419,
+      "step": 2816
+    },
+    {
+      "epoch": 1.336337760910816,
+      "grad_norm": 1.3814226388931274,
+      "learning_rate": 5.2414151764272116e-06,
+      "loss": 0.1232,
+      "step": 2817
+    },
+    {
+      "epoch": 1.3368121442125238,
+      "grad_norm": 1.4221547842025757,
+      "learning_rate": 5.23465918508984e-06,
+      "loss": 0.1346,
+      "step": 2818
+    },
+    {
+      "epoch": 1.3372865275142316,
+      "grad_norm": 2.043062686920166,
+      "learning_rate": 5.227906006686001e-06,
+      "loss": 0.1743,
+      "step": 2819
+    },
+    {
+      "epoch": 1.3377609108159394,
+      "grad_norm": 2.2925965785980225,
+      "learning_rate": 5.22115564520202e-06,
+      "loss": 0.1989,
+      "step": 2820
+    },
+    {
+      "epoch": 1.3382352941176472,
+      "grad_norm": 1.5758484601974487,
+      "learning_rate": 5.214408104622573e-06,
+      "loss": 0.154,
+      "step": 2821
+    },
+    {
+      "epoch": 1.3387096774193548,
+      "grad_norm": 1.688869833946228,
+      "learning_rate": 5.207663388930666e-06,
+      "loss": 0.158,
+      "step": 2822
+    },
+    {
+      "epoch": 1.3391840607210626,
+      "grad_norm": 1.5997257232666016,
+      "learning_rate": 5.200921502107638e-06,
+      "loss": 0.1363,
+      "step": 2823
+    },
+    {
+      "epoch": 1.3396584440227703,
+      "grad_norm": 2.4265429973602295,
+      "learning_rate": 5.194182448133163e-06,
+      "loss": 0.1848,
+      "step": 2824
+    },
+    {
+      "epoch": 1.3401328273244781,
+      "grad_norm": 1.8538546562194824,
+      "learning_rate": 5.187446230985229e-06,
+      "loss": 0.1652,
+      "step": 2825
+    },
+    {
+      "epoch": 1.340607210626186,
+      "grad_norm": 1.5969922542572021,
+      "learning_rate": 5.180712854640168e-06,
+      "loss": 0.1649,
+      "step": 2826
+    },
+    {
+      "epoch": 1.3410815939278937,
+      "grad_norm": 1.623705267906189,
+      "learning_rate": 5.173982323072615e-06,
+      "loss": 0.1676,
+      "step": 2827
+    },
+    {
+      "epoch": 1.3415559772296015,
+      "grad_norm": 1.4985768795013428,
+      "learning_rate": 5.167254640255542e-06,
+      "loss": 0.1731,
+      "step": 2828
+    },
+    {
+      "epoch": 1.3420303605313093,
+      "grad_norm": 1.6000432968139648,
+      "learning_rate": 5.160529810160235e-06,
+      "loss": 0.15,
+      "step": 2829
+    },
+    {
+      "epoch": 1.342504743833017,
+      "grad_norm": 1.7188128232955933,
+      "learning_rate": 5.153807836756288e-06,
+      "loss": 0.151,
+      "step": 2830
+    },
+    {
+      "epoch": 1.342979127134725,
+      "grad_norm": 2.0611462593078613,
+      "learning_rate": 5.147088724011622e-06,
+      "loss": 0.2002,
+      "step": 2831
+    },
+    {
+      "epoch": 1.3434535104364327,
+      "grad_norm": 1.8624550104141235,
+      "learning_rate": 5.14037247589246e-06,
+      "loss": 0.1607,
+      "step": 2832
+    },
+    {
+      "epoch": 1.3439278937381405,
+      "grad_norm": 1.335152506828308,
+      "learning_rate": 5.133659096363341e-06,
+      "loss": 0.118,
+      "step": 2833
+    },
+    {
+      "epoch": 1.3444022770398483,
+      "grad_norm": 1.6288925409317017,
+      "learning_rate": 5.126948589387104e-06,
+      "loss": 0.184,
+      "step": 2834
+    },
+    {
+      "epoch": 1.3448766603415558,
+      "grad_norm": 2.3706490993499756,
+      "learning_rate": 5.120240958924888e-06,
+      "loss": 0.2334,
+      "step": 2835
+    },
+    {
+      "epoch": 1.3453510436432636,
+      "grad_norm": 1.5236918926239014,
+      "learning_rate": 5.113536208936147e-06,
+      "loss": 0.1429,
+      "step": 2836
+    },
+    {
+      "epoch": 1.3458254269449714,
+      "grad_norm": 1.7775214910507202,
+      "learning_rate": 5.106834343378629e-06,
+      "loss": 0.1757,
+      "step": 2837
+    },
+    {
+      "epoch": 1.3462998102466792,
+      "grad_norm": 1.7919654846191406,
+      "learning_rate": 5.100135366208383e-06,
+      "loss": 0.1604,
+      "step": 2838
+    },
+    {
+      "epoch": 1.346774193548387,
+      "grad_norm": 2.141319990158081,
+      "learning_rate": 5.093439281379738e-06,
+      "loss": 0.2079,
+      "step": 2839
+    },
+    {
+      "epoch": 1.3472485768500948,
+      "grad_norm": 1.5796895027160645,
+      "learning_rate": 5.086746092845334e-06,
+      "loss": 0.1522,
+      "step": 2840
+    },
+    {
+      "epoch": 1.3477229601518026,
+      "grad_norm": 1.6436305046081543,
+      "learning_rate": 5.080055804556097e-06,
+      "loss": 0.15,
+      "step": 2841
+    },
+    {
+      "epoch": 1.3481973434535104,
+      "grad_norm": 2.3595945835113525,
+      "learning_rate": 5.073368420461229e-06,
+      "loss": 0.2145,
+      "step": 2842
+    },
+    {
+      "epoch": 1.3486717267552182,
+      "grad_norm": 2.0658891201019287,
+      "learning_rate": 5.066683944508235e-06,
+      "loss": 0.1851,
+      "step": 2843
+    },
+    {
+      "epoch": 1.349146110056926,
+      "grad_norm": 1.6966054439544678,
+      "learning_rate": 5.060002380642887e-06,
+      "loss": 0.1432,
+      "step": 2844
+    },
+    {
+      "epoch": 1.3496204933586338,
+      "grad_norm": 1.6572761535644531,
+      "learning_rate": 5.053323732809252e-06,
+      "loss": 0.1847,
+      "step": 2845
+    },
+    {
+      "epoch": 1.3500948766603416,
+      "grad_norm": 1.654388427734375,
+      "learning_rate": 5.046648004949667e-06,
+      "loss": 0.1683,
+      "step": 2846
+    },
+    {
+      "epoch": 1.3505692599620494,
+      "grad_norm": 1.6657936573028564,
+      "learning_rate": 5.0399752010047495e-06,
+      "loss": 0.2028,
+      "step": 2847
+    },
+    {
+      "epoch": 1.3510436432637571,
+      "grad_norm": 1.6080611944198608,
+      "learning_rate": 5.033305324913392e-06,
+      "loss": 0.1479,
+      "step": 2848
+    },
+    {
+      "epoch": 1.351518026565465,
+      "grad_norm": 1.4048763513565063,
+      "learning_rate": 5.0266383806127514e-06,
+      "loss": 0.1322,
+      "step": 2849
+    },
+    {
+      "epoch": 1.3519924098671727,
+      "grad_norm": 1.806875467300415,
+      "learning_rate": 5.019974372038265e-06,
+      "loss": 0.2062,
+      "step": 2850
+    },
+    {
+      "epoch": 1.3524667931688805,
+      "grad_norm": 1.8327056169509888,
+      "learning_rate": 5.0133133031236215e-06,
+      "loss": 0.1663,
+      "step": 2851
+    },
+    {
+      "epoch": 1.3529411764705883,
+      "grad_norm": 1.5724437236785889,
+      "learning_rate": 5.006655177800792e-06,
+      "loss": 0.1578,
+      "step": 2852
+    },
+    {
+      "epoch": 1.353415559772296,
+      "grad_norm": 1.6653856039047241,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.1481,
+      "step": 2853
+    },
+    {
+      "epoch": 1.353889943074004,
+      "grad_norm": 2.099308729171753,
+      "learning_rate": 4.993347773649732e-06,
+      "loss": 0.2375,
+      "step": 2854
+    },
+    {
+      "epoch": 1.3543643263757117,
+      "grad_norm": 1.8074994087219238,
+      "learning_rate": 4.9866985026767276e-06,
+      "loss": 0.1593,
+      "step": 2855
+    },
+    {
+      "epoch": 1.3548387096774195,
+      "grad_norm": 1.7318613529205322,
+      "learning_rate": 4.980052191005989e-06,
+      "loss": 0.1456,
+      "step": 2856
+    },
+    {
+      "epoch": 1.355313092979127,
+      "grad_norm": 1.5684703588485718,
+      "learning_rate": 4.973408842560772e-06,
+      "loss": 0.1847,
+      "step": 2857
+    },
+    {
+      "epoch": 1.3557874762808348,
+      "grad_norm": 2.352754831314087,
+      "learning_rate": 4.966768461262573e-06,
+      "loss": 0.1362,
+      "step": 2858
+    },
+    {
+      "epoch": 1.3562618595825426,
+      "grad_norm": 1.6529217958450317,
+      "learning_rate": 4.960131051031143e-06,
+      "loss": 0.1486,
+      "step": 2859
+    },
+    {
+      "epoch": 1.3567362428842504,
+      "grad_norm": 1.7900093793869019,
+      "learning_rate": 4.953496615784482e-06,
+      "loss": 0.1411,
+      "step": 2860
+    },
+    {
+      "epoch": 1.3572106261859582,
+      "grad_norm": 1.4834645986557007,
+      "learning_rate": 4.94686515943883e-06,
+      "loss": 0.143,
+      "step": 2861
+    },
+    {
+      "epoch": 1.357685009487666,
+      "grad_norm": 1.7863253355026245,
+      "learning_rate": 4.940236685908677e-06,
+      "loss": 0.1657,
+      "step": 2862
+    },
+    {
+      "epoch": 1.3581593927893738,
+      "grad_norm": 1.8910566568374634,
+      "learning_rate": 4.933611199106736e-06,
+      "loss": 0.1838,
+      "step": 2863
+    },
+    {
+      "epoch": 1.3586337760910816,
+      "grad_norm": 1.7281635999679565,
+      "learning_rate": 4.9269887029439686e-06,
+      "loss": 0.1884,
+      "step": 2864
+    },
+    {
+      "epoch": 1.3591081593927894,
+      "grad_norm": 1.703221321105957,
+      "learning_rate": 4.920369201329575e-06,
+      "loss": 0.1579,
+      "step": 2865
+    },
+    {
+      "epoch": 1.3595825426944972,
+      "grad_norm": 1.3635776042938232,
+      "learning_rate": 4.913752698170972e-06,
+      "loss": 0.137,
+      "step": 2866
+    },
+    {
+      "epoch": 1.360056925996205,
+      "grad_norm": 1.9177613258361816,
+      "learning_rate": 4.907139197373827e-06,
+      "loss": 0.1918,
+      "step": 2867
+    },
+    {
+      "epoch": 1.3605313092979128,
+      "grad_norm": 1.5520260334014893,
+      "learning_rate": 4.900528702842011e-06,
+      "loss": 0.1742,
+      "step": 2868
+    },
+    {
+      "epoch": 1.3610056925996206,
+      "grad_norm": 1.8986750841140747,
+      "learning_rate": 4.893921218477642e-06,
+      "loss": 0.1707,
+      "step": 2869
+    },
+    {
+      "epoch": 1.3614800759013281,
+      "grad_norm": 1.954786777496338,
+      "learning_rate": 4.8873167481810516e-06,
+      "loss": 0.1707,
+      "step": 2870
+    },
+    {
+      "epoch": 1.361954459203036,
+      "grad_norm": 1.7505497932434082,
+      "learning_rate": 4.880715295850791e-06,
+      "loss": 0.1829,
+      "step": 2871
+    },
+    {
+      "epoch": 1.3624288425047437,
+      "grad_norm": 1.406683087348938,
+      "learning_rate": 4.874116865383638e-06,
+      "loss": 0.1356,
+      "step": 2872
+    },
+    {
+      "epoch": 1.3629032258064515,
+      "grad_norm": 1.3671265840530396,
+      "learning_rate": 4.867521460674573e-06,
+      "loss": 0.1402,
+      "step": 2873
+    },
+    {
+      "epoch": 1.3633776091081593,
+      "grad_norm": 1.9116820096969604,
+      "learning_rate": 4.860929085616804e-06,
+      "loss": 0.1588,
+      "step": 2874
+    },
+    {
+      "epoch": 1.363851992409867,
+      "grad_norm": 1.5119657516479492,
+      "learning_rate": 4.85433974410174e-06,
+      "loss": 0.1318,
+      "step": 2875
+    },
+    {
+      "epoch": 1.364326375711575,
+      "grad_norm": 1.674631953239441,
+      "learning_rate": 4.8477534400190075e-06,
+      "loss": 0.1778,
+      "step": 2876
+    },
+    {
+      "epoch": 1.3648007590132827,
+      "grad_norm": 1.8323768377304077,
+      "learning_rate": 4.841170177256439e-06,
+      "loss": 0.2063,
+      "step": 2877
+    },
+    {
+      "epoch": 1.3652751423149905,
+      "grad_norm": 2.152963638305664,
+      "learning_rate": 4.834589959700061e-06,
+      "loss": 0.2062,
+      "step": 2878
+    },
+    {
+      "epoch": 1.3657495256166983,
+      "grad_norm": 1.483198881149292,
+      "learning_rate": 4.828012791234117e-06,
+      "loss": 0.1412,
+      "step": 2879
+    },
+    {
+      "epoch": 1.366223908918406,
+      "grad_norm": 1.3623414039611816,
+      "learning_rate": 4.821438675741044e-06,
+      "loss": 0.1237,
+      "step": 2880
+    },
+    {
+      "epoch": 1.3666982922201139,
+      "grad_norm": 1.7617155313491821,
+      "learning_rate": 4.814867617101479e-06,
+      "loss": 0.1496,
+      "step": 2881
+    },
+    {
+      "epoch": 1.3671726755218216,
+      "grad_norm": 1.808595061302185,
+      "learning_rate": 4.808299619194251e-06,
+      "loss": 0.1625,
+      "step": 2882
+    },
+    {
+      "epoch": 1.3676470588235294,
+      "grad_norm": 1.750730276107788,
+      "learning_rate": 4.80173468589638e-06,
+      "loss": 0.1531,
+      "step": 2883
+    },
+    {
+      "epoch": 1.3681214421252372,
+      "grad_norm": 1.9921399354934692,
+      "learning_rate": 4.795172821083084e-06,
+      "loss": 0.2081,
+      "step": 2884
+    },
+    {
+      "epoch": 1.368595825426945,
+      "grad_norm": 1.9898700714111328,
+      "learning_rate": 4.788614028627769e-06,
+      "loss": 0.1826,
+      "step": 2885
+    },
+    {
+      "epoch": 1.3690702087286528,
+      "grad_norm": 1.715350866317749,
+      "learning_rate": 4.782058312402027e-06,
+      "loss": 0.1736,
+      "step": 2886
+    },
+    {
+      "epoch": 1.3695445920303606,
+      "grad_norm": 1.990676760673523,
+      "learning_rate": 4.7755056762756255e-06,
+      "loss": 0.1491,
+      "step": 2887
+    },
+    {
+      "epoch": 1.3700189753320684,
+      "grad_norm": 1.4953407049179077,
+      "learning_rate": 4.768956124116526e-06,
+      "loss": 0.1616,
+      "step": 2888
+    },
+    {
+      "epoch": 1.3704933586337762,
+      "grad_norm": 1.76144278049469,
+      "learning_rate": 4.762409659790866e-06,
+      "loss": 0.1715,
+      "step": 2889
+    },
+    {
+      "epoch": 1.370967741935484,
+      "grad_norm": 1.379292368888855,
+      "learning_rate": 4.755866287162952e-06,
+      "loss": 0.1385,
+      "step": 2890
+    },
+    {
+      "epoch": 1.3714421252371918,
+      "grad_norm": 2.018620014190674,
+      "learning_rate": 4.74932601009528e-06,
+      "loss": 0.1926,
+      "step": 2891
+    },
+    {
+      "epoch": 1.3719165085388993,
+      "grad_norm": 2.3270931243896484,
+      "learning_rate": 4.742788832448501e-06,
+      "loss": 0.2674,
+      "step": 2892
+    },
+    {
+      "epoch": 1.3723908918406071,
+      "grad_norm": 1.6892091035842896,
+      "learning_rate": 4.736254758081454e-06,
+      "loss": 0.191,
+      "step": 2893
+    },
+    {
+      "epoch": 1.372865275142315,
+      "grad_norm": 1.5555603504180908,
+      "learning_rate": 4.729723790851135e-06,
+      "loss": 0.1301,
+      "step": 2894
+    },
+    {
+      "epoch": 1.3733396584440227,
+      "grad_norm": 1.7797441482543945,
+      "learning_rate": 4.723195934612711e-06,
+      "loss": 0.1627,
+      "step": 2895
+    },
+    {
+      "epoch": 1.3738140417457305,
+      "grad_norm": 2.268540859222412,
+      "learning_rate": 4.7166711932195155e-06,
+      "loss": 0.1985,
+      "step": 2896
+    },
+    {
+      "epoch": 1.3742884250474383,
+      "grad_norm": 1.6203911304473877,
+      "learning_rate": 4.7101495705230285e-06,
+      "loss": 0.1565,
+      "step": 2897
+    },
+    {
+      "epoch": 1.374762808349146,
+      "grad_norm": 1.598213791847229,
+      "learning_rate": 4.703631070372909e-06,
+      "loss": 0.179,
+      "step": 2898
+    },
+    {
+      "epoch": 1.375237191650854,
+      "grad_norm": 1.5291484594345093,
+      "learning_rate": 4.697115696616955e-06,
+      "loss": 0.1816,
+      "step": 2899
+    },
+    {
+      "epoch": 1.3757115749525617,
+      "grad_norm": 1.719058871269226,
+      "learning_rate": 4.690603453101134e-06,
+      "loss": 0.1626,
+      "step": 2900
+    },
+    {
+      "epoch": 1.3761859582542695,
+      "grad_norm": 1.810861587524414,
+      "learning_rate": 4.684094343669554e-06,
+      "loss": 0.1542,
+      "step": 2901
+    },
+    {
+      "epoch": 1.3766603415559773,
+      "grad_norm": 1.7553555965423584,
+      "learning_rate": 4.677588372164479e-06,
+      "loss": 0.1787,
+      "step": 2902
+    },
+    {
+      "epoch": 1.377134724857685,
+      "grad_norm": 1.9552857875823975,
+      "learning_rate": 4.6710855424263205e-06,
+      "loss": 0.1996,
+      "step": 2903
+    },
+    {
+      "epoch": 1.3776091081593929,
+      "grad_norm": 1.9977998733520508,
+      "learning_rate": 4.6645858582936345e-06,
+      "loss": 0.1963,
+      "step": 2904
+    },
+    {
+      "epoch": 1.3780834914611007,
+      "grad_norm": 1.5562477111816406,
+      "learning_rate": 4.658089323603123e-06,
+      "loss": 0.1483,
+      "step": 2905
+    },
+    {
+      "epoch": 1.3785578747628082,
+      "grad_norm": 1.7609916925430298,
+      "learning_rate": 4.651595942189624e-06,
+      "loss": 0.2095,
+      "step": 2906
+    },
+    {
+      "epoch": 1.379032258064516,
+      "grad_norm": 1.431341528892517,
+      "learning_rate": 4.645105717886112e-06,
+      "loss": 0.1309,
+      "step": 2907
+    },
+    {
+      "epoch": 1.3795066413662238,
+      "grad_norm": 1.580350399017334,
+      "learning_rate": 4.638618654523705e-06,
+      "loss": 0.1537,
+      "step": 2908
+    },
+    {
+      "epoch": 1.3799810246679316,
+      "grad_norm": 1.886465072631836,
+      "learning_rate": 4.632134755931653e-06,
+      "loss": 0.1535,
+      "step": 2909
+    },
+    {
+      "epoch": 1.3804554079696394,
+      "grad_norm": 1.5633918046951294,
+      "learning_rate": 4.625654025937342e-06,
+      "loss": 0.1684,
+      "step": 2910
+    },
+    {
+      "epoch": 1.3809297912713472,
+      "grad_norm": 1.7328919172286987,
+      "learning_rate": 4.619176468366274e-06,
+      "loss": 0.1585,
+      "step": 2911
+    },
+    {
+      "epoch": 1.381404174573055,
+      "grad_norm": 1.3697620630264282,
+      "learning_rate": 4.612702087042091e-06,
+      "loss": 0.1405,
+      "step": 2912
+    },
+    {
+      "epoch": 1.3818785578747628,
+      "grad_norm": 2.1742119789123535,
+      "learning_rate": 4.606230885786557e-06,
+      "loss": 0.1519,
+      "step": 2913
+    },
+    {
+      "epoch": 1.3823529411764706,
+      "grad_norm": 1.6288695335388184,
+      "learning_rate": 4.599762868419561e-06,
+      "loss": 0.1662,
+      "step": 2914
+    },
+    {
+      "epoch": 1.3828273244781784,
+      "grad_norm": 1.7532272338867188,
+      "learning_rate": 4.5932980387591054e-06,
+      "loss": 0.1819,
+      "step": 2915
+    },
+    {
+      "epoch": 1.3833017077798861,
+      "grad_norm": 1.4690742492675781,
+      "learning_rate": 4.586836400621313e-06,
+      "loss": 0.1624,
+      "step": 2916
+    },
+    {
+      "epoch": 1.383776091081594,
+      "grad_norm": 1.6084843873977661,
+      "learning_rate": 4.580377957820427e-06,
+      "loss": 0.1653,
+      "step": 2917
+    },
+    {
+      "epoch": 1.3842504743833017,
+      "grad_norm": 1.5763942003250122,
+      "learning_rate": 4.573922714168804e-06,
+      "loss": 0.1519,
+      "step": 2918
+    },
+    {
+      "epoch": 1.3847248576850095,
+      "grad_norm": 2.008824586868286,
+      "learning_rate": 4.567470673476912e-06,
+      "loss": 0.1915,
+      "step": 2919
+    },
+    {
+      "epoch": 1.3851992409867173,
+      "grad_norm": 1.5423541069030762,
+      "learning_rate": 4.561021839553323e-06,
+      "loss": 0.1571,
+      "step": 2920
+    },
+    {
+      "epoch": 1.385673624288425,
+      "grad_norm": 1.8469271659851074,
+      "learning_rate": 4.554576216204718e-06,
+      "loss": 0.1971,
+      "step": 2921
+    },
+    {
+      "epoch": 1.386148007590133,
+      "grad_norm": 2.0073492527008057,
+      "learning_rate": 4.548133807235893e-06,
+      "loss": 0.1826,
+      "step": 2922
+    },
+    {
+      "epoch": 1.3866223908918407,
+      "grad_norm": 1.9065463542938232,
+      "learning_rate": 4.541694616449729e-06,
+      "loss": 0.1774,
+      "step": 2923
+    },
+    {
+      "epoch": 1.3870967741935485,
+      "grad_norm": 1.4257770776748657,
+      "learning_rate": 4.535258647647225e-06,
+      "loss": 0.1407,
+      "step": 2924
+    },
+    {
+      "epoch": 1.3875711574952563,
+      "grad_norm": 1.6374725103378296,
+      "learning_rate": 4.5288259046274605e-06,
+      "loss": 0.1694,
+      "step": 2925
+    },
+    {
+      "epoch": 1.388045540796964,
+      "grad_norm": 1.9593760967254639,
+      "learning_rate": 4.5223963911876265e-06,
+      "loss": 0.1788,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3885199240986719,
+      "grad_norm": 1.881184697151184,
+      "learning_rate": 4.5159701111229995e-06,
+      "loss": 0.1875,
+      "step": 2927
+    },
+    {
+      "epoch": 1.3889943074003794,
+      "grad_norm": 1.786327838897705,
+      "learning_rate": 4.509547068226947e-06,
+      "loss": 0.171,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3894686907020872,
+      "grad_norm": 1.864014983177185,
+      "learning_rate": 4.503127266290935e-06,
+      "loss": 0.1636,
+      "step": 2929
+    },
+    {
+      "epoch": 1.389943074003795,
+      "grad_norm": 1.5407509803771973,
+      "learning_rate": 4.496710709104504e-06,
+      "loss": 0.1757,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3904174573055028,
+      "grad_norm": 1.7357945442199707,
+      "learning_rate": 4.49029740045528e-06,
+      "loss": 0.1297,
+      "step": 2931
+    },
+    {
+      "epoch": 1.3908918406072106,
+      "grad_norm": 1.6357675790786743,
+      "learning_rate": 4.48388734412898e-06,
+      "loss": 0.1654,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3913662239089184,
+      "grad_norm": 1.856079339981079,
+      "learning_rate": 4.477480543909396e-06,
+      "loss": 0.1842,
+      "step": 2933
+    },
+    {
+      "epoch": 1.3918406072106262,
+      "grad_norm": 2.02909779548645,
+      "learning_rate": 4.471077003578403e-06,
+      "loss": 0.1907,
+      "step": 2934
+    },
+    {
+      "epoch": 1.392314990512334,
+      "grad_norm": 1.560585379600525,
+      "learning_rate": 4.464676726915939e-06,
+      "loss": 0.1355,
+      "step": 2935
+    },
+    {
+      "epoch": 1.3927893738140418,
+      "grad_norm": 1.7451188564300537,
+      "learning_rate": 4.458279717700031e-06,
+      "loss": 0.172,
+      "step": 2936
+    },
+    {
+      "epoch": 1.3932637571157496,
+      "grad_norm": 1.3806904554367065,
+      "learning_rate": 4.451885979706767e-06,
+      "loss": 0.1345,
+      "step": 2937
+    },
+    {
+      "epoch": 1.3937381404174574,
+      "grad_norm": 2.2464559078216553,
+      "learning_rate": 4.445495516710312e-06,
+      "loss": 0.1972,
+      "step": 2938
+    },
+    {
+      "epoch": 1.3942125237191652,
+      "grad_norm": 1.8251832723617554,
+      "learning_rate": 4.439108332482889e-06,
+      "loss": 0.1987,
+      "step": 2939
+    },
+    {
+      "epoch": 1.394686907020873,
+      "grad_norm": 1.6302183866500854,
+      "learning_rate": 4.432724430794786e-06,
+      "loss": 0.1673,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3951612903225805,
+      "grad_norm": 1.733634114265442,
+      "learning_rate": 4.426343815414361e-06,
+      "loss": 0.1736,
+      "step": 2941
+    },
+    {
+      "epoch": 1.3956356736242883,
+      "grad_norm": 1.841150164604187,
+      "learning_rate": 4.419966490108028e-06,
+      "loss": 0.1898,
+      "step": 2942
+    },
+    {
+      "epoch": 1.396110056925996,
+      "grad_norm": 1.559160828590393,
+      "learning_rate": 4.413592458640264e-06,
+      "loss": 0.1301,
+      "step": 2943
+    },
+    {
+      "epoch": 1.396584440227704,
+      "grad_norm": 1.6111403703689575,
+      "learning_rate": 4.407221724773587e-06,
+      "loss": 0.1824,
+      "step": 2944
+    },
+    {
+      "epoch": 1.3970588235294117,
+      "grad_norm": 1.6043332815170288,
+      "learning_rate": 4.4008542922685834e-06,
+      "loss": 0.1555,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3975332068311195,
+      "grad_norm": 1.6422481536865234,
+      "learning_rate": 4.3944901648838885e-06,
+      "loss": 0.155,
+      "step": 2946
+    },
+    {
+      "epoch": 1.3980075901328273,
+      "grad_norm": 1.5227971076965332,
+      "learning_rate": 4.388129346376177e-06,
+      "loss": 0.1495,
+      "step": 2947
+    },
+    {
+      "epoch": 1.398481973434535,
+      "grad_norm": 1.7671695947647095,
+      "learning_rate": 4.3817718405001844e-06,
+      "loss": 0.1753,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3989563567362429,
+      "grad_norm": 1.6482768058776855,
+      "learning_rate": 4.3754176510086756e-06,
+      "loss": 0.156,
+      "step": 2949
+    },
+    {
+      "epoch": 1.3994307400379506,
+      "grad_norm": 1.7106208801269531,
+      "learning_rate": 4.369066781652469e-06,
+      "loss": 0.1681,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3999051233396584,
+      "grad_norm": 1.4155083894729614,
+      "learning_rate": 4.362719236180422e-06,
+      "loss": 0.1351,
+      "step": 2951
+    },
+    {
+      "epoch": 1.4003795066413662,
+      "grad_norm": 1.725624918937683,
+      "learning_rate": 4.356375018339426e-06,
+      "loss": 0.1601,
+      "step": 2952
+    },
+    {
+      "epoch": 1.400853889943074,
+      "grad_norm": 1.8480092287063599,
+      "learning_rate": 4.350034131874414e-06,
+      "loss": 0.1942,
+      "step": 2953
+    },
+    {
+      "epoch": 1.4013282732447818,
+      "grad_norm": 1.3656977415084839,
+      "learning_rate": 4.343696580528343e-06,
+      "loss": 0.1348,
+      "step": 2954
+    },
+    {
+      "epoch": 1.4018026565464896,
+      "grad_norm": 1.9449461698532104,
+      "learning_rate": 4.33736236804221e-06,
+      "loss": 0.1573,
+      "step": 2955
+    },
+    {
+      "epoch": 1.4022770398481974,
+      "grad_norm": 1.8900002241134644,
+      "learning_rate": 4.331031498155035e-06,
+      "loss": 0.1748,
+      "step": 2956
+    },
+    {
+      "epoch": 1.4027514231499052,
+      "grad_norm": 1.8122249841690063,
+      "learning_rate": 4.324703974603873e-06,
+      "loss": 0.1863,
+      "step": 2957
+    },
+    {
+      "epoch": 1.403225806451613,
+      "grad_norm": 1.5694611072540283,
+      "learning_rate": 4.318379801123792e-06,
+      "loss": 0.1249,
+      "step": 2958
+    },
+    {
+      "epoch": 1.4037001897533208,
+      "grad_norm": 1.4751542806625366,
+      "learning_rate": 4.312058981447893e-06,
+      "loss": 0.1364,
+      "step": 2959
+    },
+    {
+      "epoch": 1.4041745730550286,
+      "grad_norm": 2.7020621299743652,
+      "learning_rate": 4.305741519307291e-06,
+      "loss": 0.1851,
+      "step": 2960
+    },
+    {
+      "epoch": 1.4046489563567364,
+      "grad_norm": 1.3586440086364746,
+      "learning_rate": 4.2994274184311245e-06,
+      "loss": 0.1246,
+      "step": 2961
+    },
+    {
+      "epoch": 1.4051233396584442,
+      "grad_norm": 1.8559433221817017,
+      "learning_rate": 4.293116682546546e-06,
+      "loss": 0.154,
+      "step": 2962
+    },
+    {
+      "epoch": 1.4055977229601517,
+      "grad_norm": 1.997769832611084,
+      "learning_rate": 4.28680931537872e-06,
+      "loss": 0.1688,
+      "step": 2963
+    },
+    {
+      "epoch": 1.4060721062618595,
+      "grad_norm": 1.8411399126052856,
+      "learning_rate": 4.280505320650814e-06,
+      "loss": 0.1979,
+      "step": 2964
+    },
+    {
+      "epoch": 1.4065464895635673,
+      "grad_norm": 1.8652249574661255,
+      "learning_rate": 4.2742047020840214e-06,
+      "loss": 0.1975,
+      "step": 2965
+    },
+    {
+      "epoch": 1.407020872865275,
+      "grad_norm": 1.6203731298446655,
+      "learning_rate": 4.2679074633975345e-06,
+      "loss": 0.149,
+      "step": 2966
+    },
+    {
+      "epoch": 1.407495256166983,
+      "grad_norm": 1.7112479209899902,
+      "learning_rate": 4.261613608308553e-06,
+      "loss": 0.1555,
+      "step": 2967
+    },
+    {
+      "epoch": 1.4079696394686907,
+      "grad_norm": 1.9960196018218994,
+      "learning_rate": 4.255323140532272e-06,
+      "loss": 0.1732,
+      "step": 2968
+    },
+    {
+      "epoch": 1.4084440227703985,
+      "grad_norm": 1.8311996459960938,
+      "learning_rate": 4.2490360637818965e-06,
+      "loss": 0.158,
+      "step": 2969
+    },
+    {
+      "epoch": 1.4089184060721063,
+      "grad_norm": 1.7324796915054321,
+      "learning_rate": 4.242752381768626e-06,
+      "loss": 0.1525,
+      "step": 2970
+    },
+    {
+      "epoch": 1.409392789373814,
+      "grad_norm": 1.6953574419021606,
+      "learning_rate": 4.236472098201651e-06,
+      "loss": 0.1494,
+      "step": 2971
+    },
+    {
+      "epoch": 1.4098671726755219,
+      "grad_norm": 1.5616384744644165,
+      "learning_rate": 4.230195216788168e-06,
+      "loss": 0.1523,
+      "step": 2972
+    },
+    {
+      "epoch": 1.4103415559772297,
+      "grad_norm": 1.4512946605682373,
+      "learning_rate": 4.223921741233349e-06,
+      "loss": 0.1483,
+      "step": 2973
+    },
+    {
+      "epoch": 1.4108159392789374,
+      "grad_norm": 1.525667667388916,
+      "learning_rate": 4.217651675240371e-06,
+      "loss": 0.1236,
+      "step": 2974
+    },
+    {
+      "epoch": 1.4112903225806452,
+      "grad_norm": 1.9730256795883179,
+      "learning_rate": 4.211385022510389e-06,
+      "loss": 0.175,
+      "step": 2975
+    },
+    {
+      "epoch": 1.4117647058823528,
+      "grad_norm": 1.4136416912078857,
+      "learning_rate": 4.205121786742552e-06,
+      "loss": 0.1287,
+      "step": 2976
+    },
+    {
+      "epoch": 1.4122390891840606,
+      "grad_norm": 1.5880212783813477,
+      "learning_rate": 4.198861971633977e-06,
+      "loss": 0.1426,
+      "step": 2977
+    },
+    {
+      "epoch": 1.4127134724857684,
+      "grad_norm": 1.5319970846176147,
+      "learning_rate": 4.1926055808797765e-06,
+      "loss": 0.1749,
+      "step": 2978
+    },
+    {
+      "epoch": 1.4131878557874762,
+      "grad_norm": 1.677323579788208,
+      "learning_rate": 4.186352618173037e-06,
+      "loss": 0.1692,
+      "step": 2979
+    },
+    {
+      "epoch": 1.413662239089184,
+      "grad_norm": 1.9051989316940308,
+      "learning_rate": 4.180103087204817e-06,
+      "loss": 0.1202,
+      "step": 2980
+    },
+    {
+      "epoch": 1.4141366223908918,
+      "grad_norm": 1.7331345081329346,
+      "learning_rate": 4.1738569916641555e-06,
+      "loss": 0.1551,
+      "step": 2981
+    },
+    {
+      "epoch": 1.4146110056925996,
+      "grad_norm": 1.4977072477340698,
+      "learning_rate": 4.167614335238058e-06,
+      "loss": 0.1707,
+      "step": 2982
+    },
+    {
+      "epoch": 1.4150853889943074,
+      "grad_norm": 2.1823627948760986,
+      "learning_rate": 4.161375121611504e-06,
+      "loss": 0.1984,
+      "step": 2983
+    },
+    {
+      "epoch": 1.4155597722960152,
+      "grad_norm": 1.9357411861419678,
+      "learning_rate": 4.155139354467439e-06,
+      "loss": 0.1967,
+      "step": 2984
+    },
+    {
+      "epoch": 1.416034155597723,
+      "grad_norm": 1.4335790872573853,
+      "learning_rate": 4.1489070374867765e-06,
+      "loss": 0.1393,
+      "step": 2985
+    },
+    {
+      "epoch": 1.4165085388994307,
+      "grad_norm": 1.459723949432373,
+      "learning_rate": 4.142678174348395e-06,
+      "loss": 0.1365,
+      "step": 2986
+    },
+    {
+      "epoch": 1.4169829222011385,
+      "grad_norm": 1.581613302230835,
+      "learning_rate": 4.136452768729126e-06,
+      "loss": 0.1507,
+      "step": 2987
+    },
+    {
+      "epoch": 1.4174573055028463,
+      "grad_norm": 1.6380817890167236,
+      "learning_rate": 4.130230824303761e-06,
+      "loss": 0.1611,
+      "step": 2988
+    },
+    {
+      "epoch": 1.4179316888045541,
+      "grad_norm": 1.8408558368682861,
+      "learning_rate": 4.1240123447450575e-06,
+      "loss": 0.1885,
+      "step": 2989
+    },
+    {
+      "epoch": 1.418406072106262,
+      "grad_norm": 1.6198720932006836,
+      "learning_rate": 4.117797333723721e-06,
+      "loss": 0.1538,
+      "step": 2990
+    },
+    {
+      "epoch": 1.4188804554079697,
+      "grad_norm": 1.727500319480896,
+      "learning_rate": 4.1115857949084145e-06,
+      "loss": 0.1834,
+      "step": 2991
+    },
+    {
+      "epoch": 1.4193548387096775,
+      "grad_norm": 2.777367353439331,
+      "learning_rate": 4.105377731965743e-06,
+      "loss": 0.1844,
+      "step": 2992
+    },
+    {
+      "epoch": 1.4198292220113853,
+      "grad_norm": 1.618923306465149,
+      "learning_rate": 4.099173148560265e-06,
+      "loss": 0.1926,
+      "step": 2993
+    },
+    {
+      "epoch": 1.420303605313093,
+      "grad_norm": 1.8688693046569824,
+      "learning_rate": 4.092972048354491e-06,
+      "loss": 0.1688,
+      "step": 2994
+    },
+    {
+      "epoch": 1.4207779886148009,
+      "grad_norm": 1.7127189636230469,
+      "learning_rate": 4.08677443500886e-06,
+      "loss": 0.1505,
+      "step": 2995
+    },
+    {
+      "epoch": 1.4212523719165087,
+      "grad_norm": 1.893202543258667,
+      "learning_rate": 4.08058031218177e-06,
+      "loss": 0.1643,
+      "step": 2996
+    },
+    {
+      "epoch": 1.4217267552182165,
+      "grad_norm": 1.7235002517700195,
+      "learning_rate": 4.074389683529542e-06,
+      "loss": 0.1644,
+      "step": 2997
+    },
+    {
+      "epoch": 1.422201138519924,
+      "grad_norm": 1.580626368522644,
+      "learning_rate": 4.0682025527064486e-06,
+      "loss": 0.1579,
+      "step": 2998
+    },
+    {
+      "epoch": 1.4226755218216318,
+      "grad_norm": 1.6993354558944702,
+      "learning_rate": 4.06201892336469e-06,
+      "loss": 0.1617,
+      "step": 2999
+    },
+    {
+      "epoch": 1.4231499051233396,
+      "grad_norm": 1.7193957567214966,
+      "learning_rate": 4.055838799154406e-06,
+      "loss": 0.1584,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4236242884250474,
+      "grad_norm": 1.6037389039993286,
+      "learning_rate": 4.049662183723655e-06,
+      "loss": 0.1507,
+      "step": 3001
+    },
+    {
+      "epoch": 1.4240986717267552,
+      "grad_norm": 1.4931392669677734,
+      "learning_rate": 4.043489080718437e-06,
+      "loss": 0.1296,
+      "step": 3002
+    },
+    {
+      "epoch": 1.424573055028463,
+      "grad_norm": 1.6914727687835693,
+      "learning_rate": 4.037319493782674e-06,
+      "loss": 0.1486,
+      "step": 3003
+    },
+    {
+      "epoch": 1.4250474383301708,
+      "grad_norm": 1.8355352878570557,
+      "learning_rate": 4.031153426558209e-06,
+      "loss": 0.1778,
+      "step": 3004
+    },
+    {
+      "epoch": 1.4255218216318786,
+      "grad_norm": 1.8768749237060547,
+      "learning_rate": 4.024990882684815e-06,
+      "loss": 0.1356,
+      "step": 3005
+    },
+    {
+      "epoch": 1.4259962049335864,
+      "grad_norm": 1.700304388999939,
+      "learning_rate": 4.018831865800174e-06,
+      "loss": 0.157,
+      "step": 3006
+    },
+    {
+      "epoch": 1.4264705882352942,
+      "grad_norm": 2.2423648834228516,
+      "learning_rate": 4.012676379539896e-06,
+      "loss": 0.1578,
+      "step": 3007
+    },
+    {
+      "epoch": 1.426944971537002,
+      "grad_norm": 1.7538981437683105,
+      "learning_rate": 4.006524427537504e-06,
+      "loss": 0.1863,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4274193548387097,
+      "grad_norm": 2.0132992267608643,
+      "learning_rate": 4.0003760134244355e-06,
+      "loss": 0.1697,
+      "step": 3009
+    },
+    {
+      "epoch": 1.4278937381404175,
+      "grad_norm": 1.686045527458191,
+      "learning_rate": 3.9942311408300395e-06,
+      "loss": 0.1412,
+      "step": 3010
+    },
+    {
+      "epoch": 1.428368121442125,
+      "grad_norm": 1.5621978044509888,
+      "learning_rate": 3.9880898133815724e-06,
+      "loss": 0.1428,
+      "step": 3011
+    },
+    {
+      "epoch": 1.428842504743833,
+      "grad_norm": 1.520124912261963,
+      "learning_rate": 3.981952034704194e-06,
+      "loss": 0.1541,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4293168880455407,
+      "grad_norm": 1.930535912513733,
+      "learning_rate": 3.975817808420978e-06,
+      "loss": 0.1873,
+      "step": 3013
+    },
+    {
+      "epoch": 1.4297912713472485,
+      "grad_norm": 1.5178654193878174,
+      "learning_rate": 3.969687138152899e-06,
+      "loss": 0.1246,
+      "step": 3014
+    },
+    {
+      "epoch": 1.4302656546489563,
+      "grad_norm": 1.252669095993042,
+      "learning_rate": 3.9635600275188335e-06,
+      "loss": 0.1106,
+      "step": 3015
+    },
+    {
+      "epoch": 1.430740037950664,
+      "grad_norm": 1.629313349723816,
+      "learning_rate": 3.957436480135547e-06,
+      "loss": 0.1768,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4312144212523719,
+      "grad_norm": 1.6184428930282593,
+      "learning_rate": 3.951316499617711e-06,
+      "loss": 0.1696,
+      "step": 3017
+    },
+    {
+      "epoch": 1.4316888045540797,
+      "grad_norm": 1.6965394020080566,
+      "learning_rate": 3.9452000895778964e-06,
+      "loss": 0.1764,
+      "step": 3018
+    },
+    {
+      "epoch": 1.4321631878557874,
+      "grad_norm": 1.5129637718200684,
+      "learning_rate": 3.93908725362655e-06,
+      "loss": 0.1325,
+      "step": 3019
+    },
+    {
+      "epoch": 1.4326375711574952,
+      "grad_norm": 1.4176459312438965,
+      "learning_rate": 3.932977995372025e-06,
+      "loss": 0.1385,
+      "step": 3020
+    },
+    {
+      "epoch": 1.433111954459203,
+      "grad_norm": 1.6068745851516724,
+      "learning_rate": 3.926872318420551e-06,
+      "loss": 0.1298,
+      "step": 3021
+    },
+    {
+      "epoch": 1.4335863377609108,
+      "grad_norm": 1.753598928451538,
+      "learning_rate": 3.920770226376251e-06,
+      "loss": 0.1772,
+      "step": 3022
+    },
+    {
+      "epoch": 1.4340607210626186,
+      "grad_norm": 1.7061421871185303,
+      "learning_rate": 3.9146717228411305e-06,
+      "loss": 0.1549,
+      "step": 3023
+    },
+    {
+      "epoch": 1.4345351043643264,
+      "grad_norm": 1.6681634187698364,
+      "learning_rate": 3.908576811415078e-06,
+      "loss": 0.1601,
+      "step": 3024
+    },
+    {
+      "epoch": 1.4350094876660342,
+      "grad_norm": 1.7460962533950806,
+      "learning_rate": 3.902485495695853e-06,
+      "loss": 0.1574,
+      "step": 3025
+    },
+    {
+      "epoch": 1.435483870967742,
+      "grad_norm": 1.7344130277633667,
+      "learning_rate": 3.896397779279102e-06,
+      "loss": 0.1805,
+      "step": 3026
+    },
+    {
+      "epoch": 1.4359582542694498,
+      "grad_norm": 1.7097632884979248,
+      "learning_rate": 3.890313665758348e-06,
+      "loss": 0.1652,
+      "step": 3027
+    },
+    {
+      "epoch": 1.4364326375711576,
+      "grad_norm": 1.7743875980377197,
+      "learning_rate": 3.884233158724976e-06,
+      "loss": 0.1655,
+      "step": 3028
+    },
+    {
+      "epoch": 1.4369070208728654,
+      "grad_norm": 1.5984655618667603,
+      "learning_rate": 3.8781562617682555e-06,
+      "loss": 0.1579,
+      "step": 3029
+    },
+    {
+      "epoch": 1.4373814041745732,
+      "grad_norm": 1.6356760263442993,
+      "learning_rate": 3.872082978475312e-06,
+      "loss": 0.1618,
+      "step": 3030
+    },
+    {
+      "epoch": 1.437855787476281,
+      "grad_norm": 1.7311320304870605,
+      "learning_rate": 3.866013312431148e-06,
+      "loss": 0.1827,
+      "step": 3031
+    },
+    {
+      "epoch": 1.4383301707779887,
+      "grad_norm": 1.62131929397583,
+      "learning_rate": 3.859947267218627e-06,
+      "loss": 0.1412,
+      "step": 3032
+    },
+    {
+      "epoch": 1.4388045540796963,
+      "grad_norm": 1.5495834350585938,
+      "learning_rate": 3.8538848464184766e-06,
+      "loss": 0.1318,
+      "step": 3033
+    },
+    {
+      "epoch": 1.439278937381404,
+      "grad_norm": 2.08760404586792,
+      "learning_rate": 3.847826053609286e-06,
+      "loss": 0.1718,
+      "step": 3034
+    },
+    {
+      "epoch": 1.439753320683112,
+      "grad_norm": 1.9067436456680298,
+      "learning_rate": 3.841770892367497e-06,
+      "loss": 0.1776,
+      "step": 3035
+    },
+    {
+      "epoch": 1.4402277039848197,
+      "grad_norm": 1.3851908445358276,
+      "learning_rate": 3.83571936626741e-06,
+      "loss": 0.1305,
+      "step": 3036
+    },
+    {
+      "epoch": 1.4407020872865275,
+      "grad_norm": 1.6880590915679932,
+      "learning_rate": 3.82967147888118e-06,
+      "loss": 0.157,
+      "step": 3037
+    },
+    {
+      "epoch": 1.4411764705882353,
+      "grad_norm": 1.8479284048080444,
+      "learning_rate": 3.823627233778824e-06,
+      "loss": 0.1465,
+      "step": 3038
+    },
+    {
+      "epoch": 1.441650853889943,
+      "grad_norm": 1.678410291671753,
+      "learning_rate": 3.8175866345281895e-06,
+      "loss": 0.1558,
+      "step": 3039
+    },
+    {
+      "epoch": 1.4421252371916509,
+      "grad_norm": 1.8292676210403442,
+      "learning_rate": 3.8115496846949885e-06,
+      "loss": 0.1587,
+      "step": 3040
+    },
+    {
+      "epoch": 1.4425996204933587,
+      "grad_norm": 1.7389070987701416,
+      "learning_rate": 3.8055163878427703e-06,
+      "loss": 0.1605,
+      "step": 3041
+    },
+    {
+      "epoch": 1.4430740037950665,
+      "grad_norm": 1.9918736219406128,
+      "learning_rate": 3.7994867475329346e-06,
+      "loss": 0.1665,
+      "step": 3042
+    },
+    {
+      "epoch": 1.4435483870967742,
+      "grad_norm": 1.6637475490570068,
+      "learning_rate": 3.7934607673247116e-06,
+      "loss": 0.1636,
+      "step": 3043
+    },
+    {
+      "epoch": 1.444022770398482,
+      "grad_norm": 1.8318939208984375,
+      "learning_rate": 3.787438450775185e-06,
+      "loss": 0.1563,
+      "step": 3044
+    },
+    {
+      "epoch": 1.4444971537001898,
+      "grad_norm": 1.7426759004592896,
+      "learning_rate": 3.781419801439261e-06,
+      "loss": 0.1709,
+      "step": 3045
+    },
+    {
+      "epoch": 1.4449715370018974,
+      "grad_norm": 1.7280033826828003,
+      "learning_rate": 3.775404822869694e-06,
+      "loss": 0.1878,
+      "step": 3046
+    },
+    {
+      "epoch": 1.4454459203036052,
+      "grad_norm": 1.8315585851669312,
+      "learning_rate": 3.7693935186170638e-06,
+      "loss": 0.1904,
+      "step": 3047
+    },
+    {
+      "epoch": 1.445920303605313,
+      "grad_norm": 1.3828792572021484,
+      "learning_rate": 3.7633858922297885e-06,
+      "loss": 0.1232,
+      "step": 3048
+    },
+    {
+      "epoch": 1.4463946869070208,
+      "grad_norm": 1.6559396982192993,
+      "learning_rate": 3.757381947254104e-06,
+      "loss": 0.1733,
+      "step": 3049
+    },
+    {
+      "epoch": 1.4468690702087286,
+      "grad_norm": 1.8938626050949097,
+      "learning_rate": 3.7513816872340826e-06,
+      "loss": 0.1362,
+      "step": 3050
+    },
+    {
+      "epoch": 1.4473434535104364,
+      "grad_norm": 2.071521282196045,
+      "learning_rate": 3.745385115711623e-06,
+      "loss": 0.1934,
+      "step": 3051
+    },
+    {
+      "epoch": 1.4478178368121442,
+      "grad_norm": 1.8912732601165771,
+      "learning_rate": 3.739392236226432e-06,
+      "loss": 0.1588,
+      "step": 3052
+    },
+    {
+      "epoch": 1.448292220113852,
+      "grad_norm": 1.723662257194519,
+      "learning_rate": 3.7334030523160582e-06,
+      "loss": 0.1656,
+      "step": 3053
+    },
+    {
+      "epoch": 1.4487666034155597,
+      "grad_norm": 1.6426823139190674,
+      "learning_rate": 3.7274175675158477e-06,
+      "loss": 0.1391,
+      "step": 3054
+    },
+    {
+      "epoch": 1.4492409867172675,
+      "grad_norm": 1.779732346534729,
+      "learning_rate": 3.7214357853589765e-06,
+      "loss": 0.1712,
+      "step": 3055
+    },
+    {
+      "epoch": 1.4497153700189753,
+      "grad_norm": 1.4211159944534302,
+      "learning_rate": 3.7154577093764334e-06,
+      "loss": 0.1523,
+      "step": 3056
+    },
+    {
+      "epoch": 1.4501897533206831,
+      "grad_norm": 2.087869167327881,
+      "learning_rate": 3.7094833430970188e-06,
+      "loss": 0.2148,
+      "step": 3057
+    },
+    {
+      "epoch": 1.450664136622391,
+      "grad_norm": 1.6574203968048096,
+      "learning_rate": 3.7035126900473363e-06,
+      "loss": 0.1594,
+      "step": 3058
+    },
+    {
+      "epoch": 1.4511385199240987,
+      "grad_norm": 1.7676061391830444,
+      "learning_rate": 3.69754575375181e-06,
+      "loss": 0.161,
+      "step": 3059
+    },
+    {
+      "epoch": 1.4516129032258065,
+      "grad_norm": 1.7256070375442505,
+      "learning_rate": 3.691582537732655e-06,
+      "loss": 0.1791,
+      "step": 3060
+    },
+    {
+      "epoch": 1.4520872865275143,
+      "grad_norm": 1.9738215208053589,
+      "learning_rate": 3.6856230455099053e-06,
+      "loss": 0.1804,
+      "step": 3061
+    },
+    {
+      "epoch": 1.452561669829222,
+      "grad_norm": 1.850088357925415,
+      "learning_rate": 3.6796672806013913e-06,
+      "loss": 0.173,
+      "step": 3062
+    },
+    {
+      "epoch": 1.4530360531309299,
+      "grad_norm": 1.7364604473114014,
+      "learning_rate": 3.6737152465227355e-06,
+      "loss": 0.1691,
+      "step": 3063
+    },
+    {
+      "epoch": 1.4535104364326377,
+      "grad_norm": 1.4408142566680908,
+      "learning_rate": 3.667766946787369e-06,
+      "loss": 0.1409,
+      "step": 3064
+    },
+    {
+      "epoch": 1.4539848197343455,
+      "grad_norm": 1.6814855337142944,
+      "learning_rate": 3.6618223849065126e-06,
+      "loss": 0.1652,
+      "step": 3065
+    },
+    {
+      "epoch": 1.4544592030360532,
+      "grad_norm": 1.748642921447754,
+      "learning_rate": 3.655881564389184e-06,
+      "loss": 0.1565,
+      "step": 3066
+    },
+    {
+      "epoch": 1.454933586337761,
+      "grad_norm": 1.5492746829986572,
+      "learning_rate": 3.649944488742194e-06,
+      "loss": 0.1557,
+      "step": 3067
+    },
+    {
+      "epoch": 1.4554079696394686,
+      "grad_norm": 1.3638722896575928,
+      "learning_rate": 3.644011161470136e-06,
+      "loss": 0.1359,
+      "step": 3068
+    },
+    {
+      "epoch": 1.4558823529411764,
+      "grad_norm": 1.5125813484191895,
+      "learning_rate": 3.6380815860753904e-06,
+      "loss": 0.1492,
+      "step": 3069
+    },
+    {
+      "epoch": 1.4563567362428842,
+      "grad_norm": 1.7100872993469238,
+      "learning_rate": 3.632155766058131e-06,
+      "loss": 0.1598,
+      "step": 3070
+    },
+    {
+      "epoch": 1.456831119544592,
+      "grad_norm": 1.9752591848373413,
+      "learning_rate": 3.6262337049163088e-06,
+      "loss": 0.1696,
+      "step": 3071
+    },
+    {
+      "epoch": 1.4573055028462998,
+      "grad_norm": 1.8669570684432983,
+      "learning_rate": 3.6203154061456648e-06,
+      "loss": 0.1718,
+      "step": 3072
+    },
+    {
+      "epoch": 1.4577798861480076,
+      "grad_norm": 1.8326796293258667,
+      "learning_rate": 3.614400873239703e-06,
+      "loss": 0.203,
+      "step": 3073
+    },
+    {
+      "epoch": 1.4582542694497154,
+      "grad_norm": 1.6635417938232422,
+      "learning_rate": 3.6084901096897163e-06,
+      "loss": 0.1458,
+      "step": 3074
+    },
+    {
+      "epoch": 1.4587286527514232,
+      "grad_norm": 1.6087132692337036,
+      "learning_rate": 3.602583118984776e-06,
+      "loss": 0.1431,
+      "step": 3075
+    },
+    {
+      "epoch": 1.459203036053131,
+      "grad_norm": 1.7828035354614258,
+      "learning_rate": 3.596679904611715e-06,
+      "loss": 0.1735,
+      "step": 3076
+    },
+    {
+      "epoch": 1.4596774193548387,
+      "grad_norm": 1.9378790855407715,
+      "learning_rate": 3.5907804700551385e-06,
+      "loss": 0.1872,
+      "step": 3077
+    },
+    {
+      "epoch": 1.4601518026565465,
+      "grad_norm": 1.23013174533844,
+      "learning_rate": 3.5848848187974294e-06,
+      "loss": 0.114,
+      "step": 3078
+    },
+    {
+      "epoch": 1.4606261859582543,
+      "grad_norm": 1.618354320526123,
+      "learning_rate": 3.5789929543187317e-06,
+      "loss": 0.147,
+      "step": 3079
+    },
+    {
+      "epoch": 1.4611005692599621,
+      "grad_norm": 2.332402229309082,
+      "learning_rate": 3.5731048800969536e-06,
+      "loss": 0.1879,
+      "step": 3080
+    },
+    {
+      "epoch": 1.4615749525616697,
+      "grad_norm": 1.8381061553955078,
+      "learning_rate": 3.5672205996077726e-06,
+      "loss": 0.1654,
+      "step": 3081
+    },
+    {
+      "epoch": 1.4620493358633775,
+      "grad_norm": 1.7428001165390015,
+      "learning_rate": 3.5613401163246118e-06,
+      "loss": 0.1625,
+      "step": 3082
+    },
+    {
+      "epoch": 1.4625237191650853,
+      "grad_norm": 1.8108954429626465,
+      "learning_rate": 3.555463433718671e-06,
+      "loss": 0.1954,
+      "step": 3083
+    },
+    {
+      "epoch": 1.462998102466793,
+      "grad_norm": 1.7431238889694214,
+      "learning_rate": 3.549590555258892e-06,
+      "loss": 0.1849,
+      "step": 3084
+    },
+    {
+      "epoch": 1.4634724857685009,
+      "grad_norm": 1.3837710618972778,
+      "learning_rate": 3.543721484411976e-06,
+      "loss": 0.1412,
+      "step": 3085
+    },
+    {
+      "epoch": 1.4639468690702087,
+      "grad_norm": 1.585083246231079,
+      "learning_rate": 3.537856224642385e-06,
+      "loss": 0.1452,
+      "step": 3086
+    },
+    {
+      "epoch": 1.4644212523719164,
+      "grad_norm": 1.6242449283599854,
+      "learning_rate": 3.5319947794123153e-06,
+      "loss": 0.157,
+      "step": 3087
+    },
+    {
+      "epoch": 1.4648956356736242,
+      "grad_norm": 1.61245596408844,
+      "learning_rate": 3.5261371521817247e-06,
+      "loss": 0.1665,
+      "step": 3088
+    },
+    {
+      "epoch": 1.465370018975332,
+      "grad_norm": 1.7214654684066772,
+      "learning_rate": 3.5202833464083096e-06,
+      "loss": 0.1626,
+      "step": 3089
+    },
+    {
+      "epoch": 1.4658444022770398,
+      "grad_norm": 1.7213759422302246,
+      "learning_rate": 3.514433365547517e-06,
+      "loss": 0.1907,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4663187855787476,
+      "grad_norm": 1.6483070850372314,
+      "learning_rate": 3.5085872130525345e-06,
+      "loss": 0.1634,
+      "step": 3091
+    },
+    {
+      "epoch": 1.4667931688804554,
+      "grad_norm": 1.7148456573486328,
+      "learning_rate": 3.5027448923742845e-06,
+      "loss": 0.1782,
+      "step": 3092
+    },
+    {
+      "epoch": 1.4672675521821632,
+      "grad_norm": 1.4626739025115967,
+      "learning_rate": 3.496906406961428e-06,
+      "loss": 0.1386,
+      "step": 3093
+    },
+    {
+      "epoch": 1.467741935483871,
+      "grad_norm": 1.7289766073226929,
+      "learning_rate": 3.491071760260368e-06,
+      "loss": 0.1474,
+      "step": 3094
+    },
+    {
+      "epoch": 1.4682163187855788,
+      "grad_norm": 1.4813730716705322,
+      "learning_rate": 3.4852409557152432e-06,
+      "loss": 0.1572,
+      "step": 3095
+    },
+    {
+      "epoch": 1.4686907020872866,
+      "grad_norm": 1.7407090663909912,
+      "learning_rate": 3.479413996767913e-06,
+      "loss": 0.1711,
+      "step": 3096
+    },
+    {
+      "epoch": 1.4691650853889944,
+      "grad_norm": 1.5457485914230347,
+      "learning_rate": 3.473590886857977e-06,
+      "loss": 0.1327,
+      "step": 3097
+    },
+    {
+      "epoch": 1.4696394686907022,
+      "grad_norm": 1.932664394378662,
+      "learning_rate": 3.4677716294227583e-06,
+      "loss": 0.1631,
+      "step": 3098
+    },
+    {
+      "epoch": 1.47011385199241,
+      "grad_norm": 1.4631233215332031,
+      "learning_rate": 3.4619562278973105e-06,
+      "loss": 0.1312,
+      "step": 3099
+    },
+    {
+      "epoch": 1.4705882352941178,
+      "grad_norm": 1.3594050407409668,
+      "learning_rate": 3.4561446857144054e-06,
+      "loss": 0.1175,
+      "step": 3100
+    },
+    {
+      "epoch": 1.4710626185958255,
+      "grad_norm": 1.4718384742736816,
+      "learning_rate": 3.4503370063045338e-06,
+      "loss": 0.1438,
+      "step": 3101
+    },
+    {
+      "epoch": 1.4715370018975333,
+      "grad_norm": 1.6673498153686523,
+      "learning_rate": 3.444533193095917e-06,
+      "loss": 0.1628,
+      "step": 3102
+    },
+    {
+      "epoch": 1.472011385199241,
+      "grad_norm": 1.9447062015533447,
+      "learning_rate": 3.4387332495144866e-06,
+      "loss": 0.2165,
+      "step": 3103
+    },
+    {
+      "epoch": 1.4724857685009487,
+      "grad_norm": 1.9136441946029663,
+      "learning_rate": 3.4329371789838916e-06,
+      "loss": 0.1972,
+      "step": 3104
+    },
+    {
+      "epoch": 1.4729601518026565,
+      "grad_norm": 2.037196636199951,
+      "learning_rate": 3.4271449849255003e-06,
+      "loss": 0.1881,
+      "step": 3105
+    },
+    {
+      "epoch": 1.4734345351043643,
+      "grad_norm": 1.9500868320465088,
+      "learning_rate": 3.42135667075838e-06,
+      "loss": 0.1469,
+      "step": 3106
+    },
+    {
+      "epoch": 1.473908918406072,
+      "grad_norm": 1.8555630445480347,
+      "learning_rate": 3.4155722398993175e-06,
+      "loss": 0.1753,
+      "step": 3107
+    },
+    {
+      "epoch": 1.4743833017077799,
+      "grad_norm": 1.6183701753616333,
+      "learning_rate": 3.4097916957628108e-06,
+      "loss": 0.1391,
+      "step": 3108
+    },
+    {
+      "epoch": 1.4748576850094877,
+      "grad_norm": 1.8170623779296875,
+      "learning_rate": 3.4040150417610483e-06,
+      "loss": 0.1789,
+      "step": 3109
+    },
+    {
+      "epoch": 1.4753320683111955,
+      "grad_norm": 1.6441960334777832,
+      "learning_rate": 3.3982422813039407e-06,
+      "loss": 0.1682,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4758064516129032,
+      "grad_norm": 1.511757254600525,
+      "learning_rate": 3.3924734177990847e-06,
+      "loss": 0.1337,
+      "step": 3111
+    },
+    {
+      "epoch": 1.476280834914611,
+      "grad_norm": 1.9087467193603516,
+      "learning_rate": 3.3867084546517847e-06,
+      "loss": 0.1792,
+      "step": 3112
+    },
+    {
+      "epoch": 1.4767552182163188,
+      "grad_norm": 1.632246732711792,
+      "learning_rate": 3.3809473952650427e-06,
+      "loss": 0.1387,
+      "step": 3113
+    },
+    {
+      "epoch": 1.4772296015180266,
+      "grad_norm": 1.7560805082321167,
+      "learning_rate": 3.3751902430395558e-06,
+      "loss": 0.1753,
+      "step": 3114
+    },
+    {
+      "epoch": 1.4777039848197344,
+      "grad_norm": 1.700164556503296,
+      "learning_rate": 3.3694370013737153e-06,
+      "loss": 0.1838,
+      "step": 3115
+    },
+    {
+      "epoch": 1.478178368121442,
+      "grad_norm": 1.6236701011657715,
+      "learning_rate": 3.3636876736636013e-06,
+      "loss": 0.1711,
+      "step": 3116
+    },
+    {
+      "epoch": 1.4786527514231498,
+      "grad_norm": 1.9994237422943115,
+      "learning_rate": 3.3579422633029813e-06,
+      "loss": 0.1642,
+      "step": 3117
+    },
+    {
+      "epoch": 1.4791271347248576,
+      "grad_norm": 1.9831045866012573,
+      "learning_rate": 3.352200773683317e-06,
+      "loss": 0.1701,
+      "step": 3118
+    },
+    {
+      "epoch": 1.4796015180265654,
+      "grad_norm": 1.4581242799758911,
+      "learning_rate": 3.3464632081937567e-06,
+      "loss": 0.1479,
+      "step": 3119
+    },
+    {
+      "epoch": 1.4800759013282732,
+      "grad_norm": 1.944639801979065,
+      "learning_rate": 3.3407295702211217e-06,
+      "loss": 0.1544,
+      "step": 3120
+    },
+    {
+      "epoch": 1.480550284629981,
+      "grad_norm": 1.6618354320526123,
+      "learning_rate": 3.3349998631499247e-06,
+      "loss": 0.1325,
+      "step": 3121
+    },
+    {
+      "epoch": 1.4810246679316887,
+      "grad_norm": 2.091897964477539,
+      "learning_rate": 3.3292740903623567e-06,
+      "loss": 0.1895,
+      "step": 3122
+    },
+    {
+      "epoch": 1.4814990512333965,
+      "grad_norm": 1.445960283279419,
+      "learning_rate": 3.323552255238286e-06,
+      "loss": 0.1448,
+      "step": 3123
+    },
+    {
+      "epoch": 1.4819734345351043,
+      "grad_norm": 1.5881474018096924,
+      "learning_rate": 3.317834361155252e-06,
+      "loss": 0.136,
+      "step": 3124
+    },
+    {
+      "epoch": 1.4824478178368121,
+      "grad_norm": 1.9209330081939697,
+      "learning_rate": 3.3121204114884696e-06,
+      "loss": 0.1705,
+      "step": 3125
+    },
+    {
+      "epoch": 1.48292220113852,
+      "grad_norm": 1.59219491481781,
+      "learning_rate": 3.3064104096108287e-06,
+      "loss": 0.1403,
+      "step": 3126
+    },
+    {
+      "epoch": 1.4833965844402277,
+      "grad_norm": 1.7049628496170044,
+      "learning_rate": 3.3007043588928866e-06,
+      "loss": 0.1602,
+      "step": 3127
+    },
+    {
+      "epoch": 1.4838709677419355,
+      "grad_norm": 1.7211378812789917,
+      "learning_rate": 3.295002262702869e-06,
+      "loss": 0.1711,
+      "step": 3128
+    },
+    {
+      "epoch": 1.4843453510436433,
+      "grad_norm": 1.8670153617858887,
+      "learning_rate": 3.2893041244066704e-06,
+      "loss": 0.1732,
+      "step": 3129
+    },
+    {
+      "epoch": 1.484819734345351,
+      "grad_norm": 1.462051510810852,
+      "learning_rate": 3.2836099473678384e-06,
+      "loss": 0.1409,
+      "step": 3130
+    },
+    {
+      "epoch": 1.4852941176470589,
+      "grad_norm": 1.5482720136642456,
+      "learning_rate": 3.2779197349475933e-06,
+      "loss": 0.1419,
+      "step": 3131
+    },
+    {
+      "epoch": 1.4857685009487667,
+      "grad_norm": 2.3937394618988037,
+      "learning_rate": 3.2722334905048146e-06,
+      "loss": 0.1878,
+      "step": 3132
+    },
+    {
+      "epoch": 1.4862428842504745,
+      "grad_norm": 1.6191223859786987,
+      "learning_rate": 3.266551217396029e-06,
+      "loss": 0.1706,
+      "step": 3133
+    },
+    {
+      "epoch": 1.4867172675521823,
+      "grad_norm": 1.8945509195327759,
+      "learning_rate": 3.260872918975432e-06,
+      "loss": 0.1589,
+      "step": 3134
+    },
+    {
+      "epoch": 1.48719165085389,
+      "grad_norm": 1.668412685394287,
+      "learning_rate": 3.255198598594862e-06,
+      "loss": 0.1757,
+      "step": 3135
+    },
+    {
+      "epoch": 1.4876660341555978,
+      "grad_norm": 1.9364664554595947,
+      "learning_rate": 3.2495282596038156e-06,
+      "loss": 0.1189,
+      "step": 3136
+    },
+    {
+      "epoch": 1.4881404174573056,
+      "grad_norm": 1.442252278327942,
+      "learning_rate": 3.243861905349439e-06,
+      "loss": 0.1294,
+      "step": 3137
+    },
+    {
+      "epoch": 1.4886148007590132,
+      "grad_norm": 1.5097479820251465,
+      "learning_rate": 3.2381995391765288e-06,
+      "loss": 0.1502,
+      "step": 3138
+    },
+    {
+      "epoch": 1.489089184060721,
+      "grad_norm": 1.8824775218963623,
+      "learning_rate": 3.2325411644275164e-06,
+      "loss": 0.1816,
+      "step": 3139
+    },
+    {
+      "epoch": 1.4895635673624288,
+      "grad_norm": 1.5267640352249146,
+      "learning_rate": 3.22688678444249e-06,
+      "loss": 0.1391,
+      "step": 3140
+    },
+    {
+      "epoch": 1.4900379506641366,
+      "grad_norm": 1.582728385925293,
+      "learning_rate": 3.221236402559169e-06,
+      "loss": 0.168,
+      "step": 3141
+    },
+    {
+      "epoch": 1.4905123339658444,
+      "grad_norm": 1.3848261833190918,
+      "learning_rate": 3.215590022112921e-06,
+      "loss": 0.1533,
+      "step": 3142
+    },
+    {
+      "epoch": 1.4909867172675522,
+      "grad_norm": 1.7733631134033203,
+      "learning_rate": 3.209947646436752e-06,
+      "loss": 0.1689,
+      "step": 3143
+    },
+    {
+      "epoch": 1.49146110056926,
+      "grad_norm": 1.8000030517578125,
+      "learning_rate": 3.204309278861294e-06,
+      "loss": 0.1392,
+      "step": 3144
+    },
+    {
+      "epoch": 1.4919354838709677,
+      "grad_norm": 1.5423595905303955,
+      "learning_rate": 3.1986749227148215e-06,
+      "loss": 0.1553,
+      "step": 3145
+    },
+    {
+      "epoch": 1.4924098671726755,
+      "grad_norm": 1.7993470430374146,
+      "learning_rate": 3.19304458132324e-06,
+      "loss": 0.1598,
+      "step": 3146
+    },
+    {
+      "epoch": 1.4928842504743833,
+      "grad_norm": 1.7650487422943115,
+      "learning_rate": 3.1874182580100874e-06,
+      "loss": 0.1995,
+      "step": 3147
+    },
+    {
+      "epoch": 1.4933586337760911,
+      "grad_norm": 2.178863525390625,
+      "learning_rate": 3.181795956096522e-06,
+      "loss": 0.1757,
+      "step": 3148
+    },
+    {
+      "epoch": 1.493833017077799,
+      "grad_norm": 1.823729395866394,
+      "learning_rate": 3.1761776789013365e-06,
+      "loss": 0.1441,
+      "step": 3149
+    },
+    {
+      "epoch": 1.4943074003795067,
+      "grad_norm": 1.6294647455215454,
+      "learning_rate": 3.1705634297409404e-06,
+      "loss": 0.1846,
+      "step": 3150
+    },
+    {
+      "epoch": 1.4947817836812145,
+      "grad_norm": 1.7616844177246094,
+      "learning_rate": 3.1649532119293713e-06,
+      "loss": 0.169,
+      "step": 3151
+    },
+    {
+      "epoch": 1.495256166982922,
+      "grad_norm": 1.9334604740142822,
+      "learning_rate": 3.1593470287782847e-06,
+      "loss": 0.1989,
+      "step": 3152
+    },
+    {
+      "epoch": 1.4957305502846299,
+      "grad_norm": 1.5212819576263428,
+      "learning_rate": 3.15374488359696e-06,
+      "loss": 0.1652,
+      "step": 3153
+    },
+    {
+      "epoch": 1.4962049335863377,
+      "grad_norm": 1.8870230913162231,
+      "learning_rate": 3.1481467796922804e-06,
+      "loss": 0.1775,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4966793168880455,
+      "grad_norm": 1.5474931001663208,
+      "learning_rate": 3.1425527203687543e-06,
+      "loss": 0.152,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4971537001897532,
+      "grad_norm": 1.711269497871399,
+      "learning_rate": 3.1369627089285036e-06,
+      "loss": 0.1527,
+      "step": 3156
+    },
+    {
+      "epoch": 1.497628083491461,
+      "grad_norm": 1.6961112022399902,
+      "learning_rate": 3.131376748671253e-06,
+      "loss": 0.1761,
+      "step": 3157
+    },
+    {
+      "epoch": 1.4981024667931688,
+      "grad_norm": 1.4259470701217651,
+      "learning_rate": 3.1257948428943375e-06,
+      "loss": 0.1497,
+      "step": 3158
+    },
+    {
+      "epoch": 1.4985768500948766,
+      "grad_norm": 1.8194340467453003,
+      "learning_rate": 3.120216994892702e-06,
+      "loss": 0.1785,
+      "step": 3159
+    },
+    {
+      "epoch": 1.4990512333965844,
+      "grad_norm": 1.514940619468689,
+      "learning_rate": 3.1146432079588963e-06,
+      "loss": 0.1585,
+      "step": 3160
+    },
+    {
+      "epoch": 1.4995256166982922,
+      "grad_norm": 1.442490577697754,
+      "learning_rate": 3.1090734853830718e-06,
+      "loss": 0.1337,
+      "step": 3161
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7581517696380615,
+      "learning_rate": 3.103507830452982e-06,
+      "loss": 0.1771,
+      "step": 3162
+    },
+    {
+      "epoch": 1.5004743833017078,
+      "grad_norm": 2.3333330154418945,
+      "learning_rate": 3.0979462464539744e-06,
+      "loss": 0.1982,
+      "step": 3163
+    },
+    {
+      "epoch": 1.5009487666034156,
+      "grad_norm": 1.7055302858352661,
+      "learning_rate": 3.092388736669002e-06,
+      "loss": 0.1643,
+      "step": 3164
+    },
+    {
+      "epoch": 1.5014231499051234,
+      "grad_norm": 1.49795401096344,
+      "learning_rate": 3.0868353043786004e-06,
+      "loss": 0.1416,
+      "step": 3165
+    },
+    {
+      "epoch": 1.5018975332068312,
+      "grad_norm": 1.4336013793945312,
+      "learning_rate": 3.0812859528609106e-06,
+      "loss": 0.1412,
+      "step": 3166
+    },
+    {
+      "epoch": 1.502371916508539,
+      "grad_norm": 1.6154780387878418,
+      "learning_rate": 3.0757406853916627e-06,
+      "loss": 0.1667,
+      "step": 3167
+    },
+    {
+      "epoch": 1.5028462998102468,
+      "grad_norm": 1.670455813407898,
+      "learning_rate": 3.0701995052441658e-06,
+      "loss": 0.1467,
+      "step": 3168
+    },
+    {
+      "epoch": 1.5033206831119545,
+      "grad_norm": 1.86698317527771,
+      "learning_rate": 3.064662415689328e-06,
+      "loss": 0.2076,
+      "step": 3169
+    },
+    {
+      "epoch": 1.5037950664136623,
+      "grad_norm": 1.5714526176452637,
+      "learning_rate": 3.059129419995638e-06,
+      "loss": 0.155,
+      "step": 3170
+    },
+    {
+      "epoch": 1.5042694497153701,
+      "grad_norm": 1.567219853401184,
+      "learning_rate": 3.053600521429172e-06,
+      "loss": 0.1538,
+      "step": 3171
+    },
+    {
+      "epoch": 1.504743833017078,
+      "grad_norm": 1.410211205482483,
+      "learning_rate": 3.0480757232535773e-06,
+      "loss": 0.1173,
+      "step": 3172
+    },
+    {
+      "epoch": 1.5052182163187857,
+      "grad_norm": 1.7881250381469727,
+      "learning_rate": 3.0425550287300943e-06,
+      "loss": 0.1827,
+      "step": 3173
+    },
+    {
+      "epoch": 1.5056925996204935,
+      "grad_norm": 1.595481514930725,
+      "learning_rate": 3.037038441117528e-06,
+      "loss": 0.1373,
+      "step": 3174
+    },
+    {
+      "epoch": 1.5061669829222013,
+      "grad_norm": 1.6253283023834229,
+      "learning_rate": 3.031525963672267e-06,
+      "loss": 0.1438,
+      "step": 3175
+    },
+    {
+      "epoch": 1.5066413662239089,
+      "grad_norm": 1.9808628559112549,
+      "learning_rate": 3.0260175996482787e-06,
+      "loss": 0.1479,
+      "step": 3176
+    },
+    {
+      "epoch": 1.5071157495256167,
+      "grad_norm": 2.177093744277954,
+      "learning_rate": 3.0205133522970865e-06,
+      "loss": 0.2158,
+      "step": 3177
+    },
+    {
+      "epoch": 1.5075901328273245,
+      "grad_norm": 1.7438995838165283,
+      "learning_rate": 3.0150132248677976e-06,
+      "loss": 0.1905,
+      "step": 3178
+    },
+    {
+      "epoch": 1.5080645161290323,
+      "grad_norm": 1.5714632272720337,
+      "learning_rate": 3.0095172206070833e-06,
+      "loss": 0.1403,
+      "step": 3179
+    },
+    {
+      "epoch": 1.50853889943074,
+      "grad_norm": 1.7036011219024658,
+      "learning_rate": 3.0040253427591827e-06,
+      "loss": 0.139,
+      "step": 3180
+    },
+    {
+      "epoch": 1.5090132827324478,
+      "grad_norm": 1.967517375946045,
+      "learning_rate": 2.9985375945658934e-06,
+      "loss": 0.1712,
+      "step": 3181
+    },
+    {
+      "epoch": 1.5094876660341556,
+      "grad_norm": 1.4856235980987549,
+      "learning_rate": 2.9930539792665767e-06,
+      "loss": 0.1472,
+      "step": 3182
+    },
+    {
+      "epoch": 1.5099620493358634,
+      "grad_norm": 1.9669393301010132,
+      "learning_rate": 2.9875745000981603e-06,
+      "loss": 0.1836,
+      "step": 3183
+    },
+    {
+      "epoch": 1.510436432637571,
+      "grad_norm": 1.9737735986709595,
+      "learning_rate": 2.9820991602951255e-06,
+      "loss": 0.1955,
+      "step": 3184
+    },
+    {
+      "epoch": 1.5109108159392788,
+      "grad_norm": 1.8585445880889893,
+      "learning_rate": 2.97662796308951e-06,
+      "loss": 0.1881,
+      "step": 3185
+    },
+    {
+      "epoch": 1.5113851992409866,
+      "grad_norm": 1.8154902458190918,
+      "learning_rate": 2.971160911710913e-06,
+      "loss": 0.1274,
+      "step": 3186
+    },
+    {
+      "epoch": 1.5118595825426944,
+      "grad_norm": 1.5582470893859863,
+      "learning_rate": 2.965698009386473e-06,
+      "loss": 0.1438,
+      "step": 3187
+    },
+    {
+      "epoch": 1.5123339658444022,
+      "grad_norm": 2.245789051055908,
+      "learning_rate": 2.9602392593408933e-06,
+      "loss": 0.2352,
+      "step": 3188
+    },
+    {
+      "epoch": 1.51280834914611,
+      "grad_norm": 1.8019410371780396,
+      "learning_rate": 2.954784664796414e-06,
+      "loss": 0.181,
+      "step": 3189
+    },
+    {
+      "epoch": 1.5132827324478177,
+      "grad_norm": 1.7983249425888062,
+      "learning_rate": 2.9493342289728334e-06,
+      "loss": 0.1187,
+      "step": 3190
+    },
+    {
+      "epoch": 1.5137571157495255,
+      "grad_norm": 1.7953065633773804,
+      "learning_rate": 2.94388795508749e-06,
+      "loss": 0.1749,
+      "step": 3191
+    },
+    {
+      "epoch": 1.5142314990512333,
+      "grad_norm": 1.9157280921936035,
+      "learning_rate": 2.93844584635526e-06,
+      "loss": 0.2026,
+      "step": 3192
+    },
+    {
+      "epoch": 1.5147058823529411,
+      "grad_norm": 1.7939811944961548,
+      "learning_rate": 2.9330079059885708e-06,
+      "loss": 0.1734,
+      "step": 3193
+    },
+    {
+      "epoch": 1.515180265654649,
+      "grad_norm": 1.3667829036712646,
+      "learning_rate": 2.927574137197383e-06,
+      "loss": 0.1348,
+      "step": 3194
+    },
+    {
+      "epoch": 1.5156546489563567,
+      "grad_norm": 1.706744909286499,
+      "learning_rate": 2.9221445431892003e-06,
+      "loss": 0.1574,
+      "step": 3195
+    },
+    {
+      "epoch": 1.5161290322580645,
+      "grad_norm": 1.9317013025283813,
+      "learning_rate": 2.916719127169053e-06,
+      "loss": 0.1782,
+      "step": 3196
+    },
+    {
+      "epoch": 1.5166034155597723,
+      "grad_norm": 1.5074424743652344,
+      "learning_rate": 2.911297892339516e-06,
+      "loss": 0.1355,
+      "step": 3197
+    },
+    {
+      "epoch": 1.51707779886148,
+      "grad_norm": 1.852922797203064,
+      "learning_rate": 2.9058808419006834e-06,
+      "loss": 0.1614,
+      "step": 3198
+    },
+    {
+      "epoch": 1.5175521821631879,
+      "grad_norm": 1.6857560873031616,
+      "learning_rate": 2.9004679790501922e-06,
+      "loss": 0.1859,
+      "step": 3199
+    },
+    {
+      "epoch": 1.5180265654648957,
+      "grad_norm": 1.6432700157165527,
+      "learning_rate": 2.895059306983201e-06,
+      "loss": 0.1501,
+      "step": 3200
+    },
+    {
+      "epoch": 1.5185009487666035,
+      "grad_norm": 2.039566993713379,
+      "learning_rate": 2.889654828892393e-06,
+      "loss": 0.1988,
+      "step": 3201
+    },
+    {
+      "epoch": 1.5189753320683113,
+      "grad_norm": 1.716520071029663,
+      "learning_rate": 2.8842545479679796e-06,
+      "loss": 0.1636,
+      "step": 3202
+    },
+    {
+      "epoch": 1.519449715370019,
+      "grad_norm": 1.249931812286377,
+      "learning_rate": 2.878858467397693e-06,
+      "loss": 0.1183,
+      "step": 3203
+    },
+    {
+      "epoch": 1.5199240986717268,
+      "grad_norm": 1.6915541887283325,
+      "learning_rate": 2.8734665903667892e-06,
+      "loss": 0.155,
+      "step": 3204
+    },
+    {
+      "epoch": 1.5203984819734346,
+      "grad_norm": 1.4063469171524048,
+      "learning_rate": 2.8680789200580373e-06,
+      "loss": 0.1344,
+      "step": 3205
+    },
+    {
+      "epoch": 1.5208728652751424,
+      "grad_norm": 1.345933198928833,
+      "learning_rate": 2.862695459651722e-06,
+      "loss": 0.1286,
+      "step": 3206
+    },
+    {
+      "epoch": 1.5213472485768502,
+      "grad_norm": 1.5287388563156128,
+      "learning_rate": 2.8573162123256504e-06,
+      "loss": 0.1617,
+      "step": 3207
+    },
+    {
+      "epoch": 1.521821631878558,
+      "grad_norm": 1.7381535768508911,
+      "learning_rate": 2.851941181255139e-06,
+      "loss": 0.1772,
+      "step": 3208
+    },
+    {
+      "epoch": 1.5222960151802658,
+      "grad_norm": 1.8976391553878784,
+      "learning_rate": 2.8465703696130142e-06,
+      "loss": 0.1806,
+      "step": 3209
+    },
+    {
+      "epoch": 1.5227703984819736,
+      "grad_norm": 1.5034171342849731,
+      "learning_rate": 2.841203780569618e-06,
+      "loss": 0.1625,
+      "step": 3210
+    },
+    {
+      "epoch": 1.5232447817836812,
+      "grad_norm": 1.4791839122772217,
+      "learning_rate": 2.835841417292788e-06,
+      "loss": 0.1411,
+      "step": 3211
+    },
+    {
+      "epoch": 1.523719165085389,
+      "grad_norm": 1.7350883483886719,
+      "learning_rate": 2.8304832829478802e-06,
+      "loss": 0.1606,
+      "step": 3212
+    },
+    {
+      "epoch": 1.5241935483870968,
+      "grad_norm": 1.4422385692596436,
+      "learning_rate": 2.825129380697741e-06,
+      "loss": 0.1368,
+      "step": 3213
+    },
+    {
+      "epoch": 1.5246679316888045,
+      "grad_norm": 1.8611582517623901,
+      "learning_rate": 2.8197797137027338e-06,
+      "loss": 0.1495,
+      "step": 3214
+    },
+    {
+      "epoch": 1.5251423149905123,
+      "grad_norm": 2.3095667362213135,
+      "learning_rate": 2.8144342851207076e-06,
+      "loss": 0.2357,
+      "step": 3215
+    },
+    {
+      "epoch": 1.5256166982922201,
+      "grad_norm": 1.809204339981079,
+      "learning_rate": 2.8090930981070176e-06,
+      "loss": 0.1656,
+      "step": 3216
+    },
+    {
+      "epoch": 1.526091081593928,
+      "grad_norm": 1.5061122179031372,
+      "learning_rate": 2.8037561558145154e-06,
+      "loss": 0.1525,
+      "step": 3217
+    },
+    {
+      "epoch": 1.5265654648956357,
+      "grad_norm": 1.6025121212005615,
+      "learning_rate": 2.7984234613935434e-06,
+      "loss": 0.137,
+      "step": 3218
+    },
+    {
+      "epoch": 1.5270398481973435,
+      "grad_norm": 1.6906027793884277,
+      "learning_rate": 2.7930950179919438e-06,
+      "loss": 0.1781,
+      "step": 3219
+    },
+    {
+      "epoch": 1.527514231499051,
+      "grad_norm": 1.538794994354248,
+      "learning_rate": 2.7877708287550366e-06,
+      "loss": 0.145,
+      "step": 3220
+    },
+    {
+      "epoch": 1.5279886148007589,
+      "grad_norm": 1.517501711845398,
+      "learning_rate": 2.7824508968256435e-06,
+      "loss": 0.1444,
+      "step": 3221
+    },
+    {
+      "epoch": 1.5284629981024667,
+      "grad_norm": 1.737318515777588,
+      "learning_rate": 2.777135225344063e-06,
+      "loss": 0.1274,
+      "step": 3222
+    },
+    {
+      "epoch": 1.5289373814041745,
+      "grad_norm": 1.742997407913208,
+      "learning_rate": 2.771823817448085e-06,
+      "loss": 0.1682,
+      "step": 3223
+    },
+    {
+      "epoch": 1.5294117647058822,
+      "grad_norm": 1.806666612625122,
+      "learning_rate": 2.7665166762729856e-06,
+      "loss": 0.1467,
+      "step": 3224
+    },
+    {
+      "epoch": 1.52988614800759,
+      "grad_norm": 2.0361621379852295,
+      "learning_rate": 2.7612138049515102e-06,
+      "loss": 0.1591,
+      "step": 3225
+    },
+    {
+      "epoch": 1.5303605313092978,
+      "grad_norm": 1.6584888696670532,
+      "learning_rate": 2.755915206613895e-06,
+      "loss": 0.1443,
+      "step": 3226
+    },
+    {
+      "epoch": 1.5308349146110056,
+      "grad_norm": 1.9782226085662842,
+      "learning_rate": 2.75062088438785e-06,
+      "loss": 0.1732,
+      "step": 3227
+    },
+    {
+      "epoch": 1.5313092979127134,
+      "grad_norm": 1.7193796634674072,
+      "learning_rate": 2.7453308413985635e-06,
+      "loss": 0.1448,
+      "step": 3228
+    },
+    {
+      "epoch": 1.5317836812144212,
+      "grad_norm": 1.6462289094924927,
+      "learning_rate": 2.740045080768694e-06,
+      "loss": 0.1444,
+      "step": 3229
+    },
+    {
+      "epoch": 1.532258064516129,
+      "grad_norm": 1.397916316986084,
+      "learning_rate": 2.73476360561837e-06,
+      "loss": 0.1503,
+      "step": 3230
+    },
+    {
+      "epoch": 1.5327324478178368,
+      "grad_norm": 2.2624073028564453,
+      "learning_rate": 2.7294864190651972e-06,
+      "loss": 0.2172,
+      "step": 3231
+    },
+    {
+      "epoch": 1.5332068311195446,
+      "grad_norm": 1.890386939048767,
+      "learning_rate": 2.724213524224246e-06,
+      "loss": 0.1538,
+      "step": 3232
+    },
+    {
+      "epoch": 1.5336812144212524,
+      "grad_norm": 2.1407437324523926,
+      "learning_rate": 2.7189449242080557e-06,
+      "loss": 0.1784,
+      "step": 3233
+    },
+    {
+      "epoch": 1.5341555977229602,
+      "grad_norm": 1.7359429597854614,
+      "learning_rate": 2.7136806221266286e-06,
+      "loss": 0.1851,
+      "step": 3234
+    },
+    {
+      "epoch": 1.534629981024668,
+      "grad_norm": 1.662856101989746,
+      "learning_rate": 2.7084206210874277e-06,
+      "loss": 0.1352,
+      "step": 3235
+    },
+    {
+      "epoch": 1.5351043643263758,
+      "grad_norm": 1.425294280052185,
+      "learning_rate": 2.7031649241953826e-06,
+      "loss": 0.134,
+      "step": 3236
+    },
+    {
+      "epoch": 1.5355787476280836,
+      "grad_norm": 1.6383399963378906,
+      "learning_rate": 2.697913534552875e-06,
+      "loss": 0.1595,
+      "step": 3237
+    },
+    {
+      "epoch": 1.5360531309297913,
+      "grad_norm": 1.8296263217926025,
+      "learning_rate": 2.6926664552597537e-06,
+      "loss": 0.1928,
+      "step": 3238
+    },
+    {
+      "epoch": 1.5365275142314991,
+      "grad_norm": 1.6858452558517456,
+      "learning_rate": 2.687423689413312e-06,
+      "loss": 0.1304,
+      "step": 3239
+    },
+    {
+      "epoch": 1.537001897533207,
+      "grad_norm": 1.7636797428131104,
+      "learning_rate": 2.6821852401083048e-06,
+      "loss": 0.167,
+      "step": 3240
+    },
+    {
+      "epoch": 1.5374762808349147,
+      "grad_norm": 1.6068103313446045,
+      "learning_rate": 2.6769511104369384e-06,
+      "loss": 0.1536,
+      "step": 3241
+    },
+    {
+      "epoch": 1.5379506641366225,
+      "grad_norm": 1.8157026767730713,
+      "learning_rate": 2.6717213034888656e-06,
+      "loss": 0.1601,
+      "step": 3242
+    },
+    {
+      "epoch": 1.5384250474383303,
+      "grad_norm": 1.6704944372177124,
+      "learning_rate": 2.6664958223511948e-06,
+      "loss": 0.165,
+      "step": 3243
+    },
+    {
+      "epoch": 1.538899430740038,
+      "grad_norm": 1.5348975658416748,
+      "learning_rate": 2.661274670108469e-06,
+      "loss": 0.1695,
+      "step": 3244
+    },
+    {
+      "epoch": 1.539373814041746,
+      "grad_norm": 1.7109907865524292,
+      "learning_rate": 2.6560578498426883e-06,
+      "loss": 0.151,
+      "step": 3245
+    },
+    {
+      "epoch": 1.5398481973434535,
+      "grad_norm": 1.4413361549377441,
+      "learning_rate": 2.6508453646332845e-06,
+      "loss": 0.1458,
+      "step": 3246
+    },
+    {
+      "epoch": 1.5403225806451613,
+      "grad_norm": 1.3164259195327759,
+      "learning_rate": 2.645637217557139e-06,
+      "loss": 0.1393,
+      "step": 3247
+    },
+    {
+      "epoch": 1.540796963946869,
+      "grad_norm": 1.6673977375030518,
+      "learning_rate": 2.640433411688572e-06,
+      "loss": 0.1613,
+      "step": 3248
+    },
+    {
+      "epoch": 1.5412713472485768,
+      "grad_norm": 1.5010228157043457,
+      "learning_rate": 2.635233950099334e-06,
+      "loss": 0.1433,
+      "step": 3249
+    },
+    {
+      "epoch": 1.5417457305502846,
+      "grad_norm": 1.5055230855941772,
+      "learning_rate": 2.630038835858617e-06,
+      "loss": 0.146,
+      "step": 3250
+    },
+    {
+      "epoch": 1.5422201138519924,
+      "grad_norm": 1.621572494506836,
+      "learning_rate": 2.624848072033046e-06,
+      "loss": 0.1707,
+      "step": 3251
+    },
+    {
+      "epoch": 1.5426944971537002,
+      "grad_norm": 2.2330636978149414,
+      "learning_rate": 2.6196616616866822e-06,
+      "loss": 0.1665,
+      "step": 3252
+    },
+    {
+      "epoch": 1.543168880455408,
+      "grad_norm": 1.9100672006607056,
+      "learning_rate": 2.6144796078810065e-06,
+      "loss": 0.1728,
+      "step": 3253
+    },
+    {
+      "epoch": 1.5436432637571158,
+      "grad_norm": 1.8044493198394775,
+      "learning_rate": 2.609301913674933e-06,
+      "loss": 0.1741,
+      "step": 3254
+    },
+    {
+      "epoch": 1.5441176470588234,
+      "grad_norm": 1.4403407573699951,
+      "learning_rate": 2.6041285821248064e-06,
+      "loss": 0.1417,
+      "step": 3255
+    },
+    {
+      "epoch": 1.5445920303605312,
+      "grad_norm": 1.5496286153793335,
+      "learning_rate": 2.598959616284391e-06,
+      "loss": 0.1356,
+      "step": 3256
+    },
+    {
+      "epoch": 1.545066413662239,
+      "grad_norm": 2.311690092086792,
+      "learning_rate": 2.5937950192048823e-06,
+      "loss": 0.1931,
+      "step": 3257
+    },
+    {
+      "epoch": 1.5455407969639468,
+      "grad_norm": 2.143083095550537,
+      "learning_rate": 2.588634793934882e-06,
+      "loss": 0.1902,
+      "step": 3258
+    },
+    {
+      "epoch": 1.5460151802656545,
+      "grad_norm": 2.0627899169921875,
+      "learning_rate": 2.5834789435204245e-06,
+      "loss": 0.1883,
+      "step": 3259
+    },
+    {
+      "epoch": 1.5464895635673623,
+      "grad_norm": 2.2060842514038086,
+      "learning_rate": 2.57832747100496e-06,
+      "loss": 0.2299,
+      "step": 3260
+    },
+    {
+      "epoch": 1.5469639468690701,
+      "grad_norm": 1.7852225303649902,
+      "learning_rate": 2.5731803794293465e-06,
+      "loss": 0.1833,
+      "step": 3261
+    },
+    {
+      "epoch": 1.547438330170778,
+      "grad_norm": 1.6334515810012817,
+      "learning_rate": 2.5680376718318657e-06,
+      "loss": 0.1617,
+      "step": 3262
+    },
+    {
+      "epoch": 1.5479127134724857,
+      "grad_norm": 1.4657764434814453,
+      "learning_rate": 2.5628993512482013e-06,
+      "loss": 0.15,
+      "step": 3263
+    },
+    {
+      "epoch": 1.5483870967741935,
+      "grad_norm": 1.775926947593689,
+      "learning_rate": 2.557765420711458e-06,
+      "loss": 0.1747,
+      "step": 3264
+    },
+    {
+      "epoch": 1.5488614800759013,
+      "grad_norm": 1.7715890407562256,
+      "learning_rate": 2.5526358832521424e-06,
+      "loss": 0.1488,
+      "step": 3265
+    },
+    {
+      "epoch": 1.549335863377609,
+      "grad_norm": 1.91201651096344,
+      "learning_rate": 2.5475107418981692e-06,
+      "loss": 0.1736,
+      "step": 3266
+    },
+    {
+      "epoch": 1.5498102466793169,
+      "grad_norm": 1.8647923469543457,
+      "learning_rate": 2.5423899996748636e-06,
+      "loss": 0.1741,
+      "step": 3267
+    },
+    {
+      "epoch": 1.5502846299810247,
+      "grad_norm": 1.4479807615280151,
+      "learning_rate": 2.5372736596049417e-06,
+      "loss": 0.1228,
+      "step": 3268
+    },
+    {
+      "epoch": 1.5507590132827325,
+      "grad_norm": 1.4096297025680542,
+      "learning_rate": 2.532161724708534e-06,
+      "loss": 0.1368,
+      "step": 3269
+    },
+    {
+      "epoch": 1.5512333965844403,
+      "grad_norm": 1.8053456544876099,
+      "learning_rate": 2.5270541980031603e-06,
+      "loss": 0.1674,
+      "step": 3270
+    },
+    {
+      "epoch": 1.551707779886148,
+      "grad_norm": 1.699059009552002,
+      "learning_rate": 2.521951082503746e-06,
+      "loss": 0.1305,
+      "step": 3271
+    },
+    {
+      "epoch": 1.5521821631878558,
+      "grad_norm": 1.85928213596344,
+      "learning_rate": 2.516852381222612e-06,
+      "loss": 0.1696,
+      "step": 3272
+    },
+    {
+      "epoch": 1.5526565464895636,
+      "grad_norm": 1.6062089204788208,
+      "learning_rate": 2.5117580971694644e-06,
+      "loss": 0.1237,
+      "step": 3273
+    },
+    {
+      "epoch": 1.5531309297912714,
+      "grad_norm": 1.8484961986541748,
+      "learning_rate": 2.5066682333514136e-06,
+      "loss": 0.1724,
+      "step": 3274
+    },
+    {
+      "epoch": 1.5536053130929792,
+      "grad_norm": 1.7327786684036255,
+      "learning_rate": 2.5015827927729554e-06,
+      "loss": 0.1314,
+      "step": 3275
+    },
+    {
+      "epoch": 1.554079696394687,
+      "grad_norm": 1.7604756355285645,
+      "learning_rate": 2.496501778435977e-06,
+      "loss": 0.1538,
+      "step": 3276
+    },
+    {
+      "epoch": 1.5545540796963948,
+      "grad_norm": 1.8276807069778442,
+      "learning_rate": 2.491425193339748e-06,
+      "loss": 0.1595,
+      "step": 3277
+    },
+    {
+      "epoch": 1.5550284629981026,
+      "grad_norm": 1.5220667123794556,
+      "learning_rate": 2.4863530404809253e-06,
+      "loss": 0.1231,
+      "step": 3278
+    },
+    {
+      "epoch": 1.5555028462998104,
+      "grad_norm": 2.1537914276123047,
+      "learning_rate": 2.4812853228535515e-06,
+      "loss": 0.1959,
+      "step": 3279
+    },
+    {
+      "epoch": 1.5559772296015182,
+      "grad_norm": 1.3721140623092651,
+      "learning_rate": 2.4762220434490504e-06,
+      "loss": 0.1232,
+      "step": 3280
+    },
+    {
+      "epoch": 1.5564516129032258,
+      "grad_norm": 1.8437730073928833,
+      "learning_rate": 2.4711632052562283e-06,
+      "loss": 0.1598,
+      "step": 3281
+    },
+    {
+      "epoch": 1.5569259962049335,
+      "grad_norm": 1.4485067129135132,
+      "learning_rate": 2.466108811261263e-06,
+      "loss": 0.161,
+      "step": 3282
+    },
+    {
+      "epoch": 1.5574003795066413,
+      "grad_norm": 1.9926724433898926,
+      "learning_rate": 2.461058864447716e-06,
+      "loss": 0.1368,
+      "step": 3283
+    },
+    {
+      "epoch": 1.5578747628083491,
+      "grad_norm": 3.5024831295013428,
+      "learning_rate": 2.456013367796519e-06,
+      "loss": 0.1822,
+      "step": 3284
+    },
+    {
+      "epoch": 1.558349146110057,
+      "grad_norm": 1.5812478065490723,
+      "learning_rate": 2.450972324285984e-06,
+      "loss": 0.1446,
+      "step": 3285
+    },
+    {
+      "epoch": 1.5588235294117647,
+      "grad_norm": 1.8127206563949585,
+      "learning_rate": 2.445935736891785e-06,
+      "loss": 0.1643,
+      "step": 3286
+    },
+    {
+      "epoch": 1.5592979127134725,
+      "grad_norm": 1.8025965690612793,
+      "learning_rate": 2.4409036085869665e-06,
+      "loss": 0.1455,
+      "step": 3287
+    },
+    {
+      "epoch": 1.5597722960151803,
+      "grad_norm": 1.6995849609375,
+      "learning_rate": 2.4358759423419476e-06,
+      "loss": 0.1461,
+      "step": 3288
+    },
+    {
+      "epoch": 1.560246679316888,
+      "grad_norm": 1.7094231843948364,
+      "learning_rate": 2.43085274112451e-06,
+      "loss": 0.1504,
+      "step": 3289
+    },
+    {
+      "epoch": 1.5607210626185957,
+      "grad_norm": 1.381062626838684,
+      "learning_rate": 2.425834007899799e-06,
+      "loss": 0.1279,
+      "step": 3290
+    },
+    {
+      "epoch": 1.5611954459203035,
+      "grad_norm": 1.921175241470337,
+      "learning_rate": 2.420819745630326e-06,
+      "loss": 0.1932,
+      "step": 3291
+    },
+    {
+      "epoch": 1.5616698292220113,
+      "grad_norm": 1.5312777757644653,
+      "learning_rate": 2.4158099572759564e-06,
+      "loss": 0.1378,
+      "step": 3292
+    },
+    {
+      "epoch": 1.562144212523719,
+      "grad_norm": 2.1276259422302246,
+      "learning_rate": 2.4108046457939215e-06,
+      "loss": 0.1814,
+      "step": 3293
+    },
+    {
+      "epoch": 1.5626185958254268,
+      "grad_norm": 1.3188360929489136,
+      "learning_rate": 2.405803814138804e-06,
+      "loss": 0.1254,
+      "step": 3294
+    },
+    {
+      "epoch": 1.5630929791271346,
+      "grad_norm": 1.5824406147003174,
+      "learning_rate": 2.40080746526255e-06,
+      "loss": 0.1318,
+      "step": 3295
+    },
+    {
+      "epoch": 1.5635673624288424,
+      "grad_norm": 1.6440813541412354,
+      "learning_rate": 2.3958156021144495e-06,
+      "loss": 0.1395,
+      "step": 3296
+    },
+    {
+      "epoch": 1.5640417457305502,
+      "grad_norm": 1.85137140750885,
+      "learning_rate": 2.390828227641152e-06,
+      "loss": 0.171,
+      "step": 3297
+    },
+    {
+      "epoch": 1.564516129032258,
+      "grad_norm": 1.7361198663711548,
+      "learning_rate": 2.385845344786656e-06,
+      "loss": 0.1485,
+      "step": 3298
+    },
+    {
+      "epoch": 1.5649905123339658,
+      "grad_norm": 1.779136061668396,
+      "learning_rate": 2.380866956492307e-06,
+      "loss": 0.1833,
+      "step": 3299
+    },
+    {
+      "epoch": 1.5654648956356736,
+      "grad_norm": 1.568875789642334,
+      "learning_rate": 2.3758930656968025e-06,
+      "loss": 0.144,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5659392789373814,
+      "grad_norm": 1.6992019414901733,
+      "learning_rate": 2.3709236753361777e-06,
+      "loss": 0.1576,
+      "step": 3301
+    },
+    {
+      "epoch": 1.5664136622390892,
+      "grad_norm": 1.7569388151168823,
+      "learning_rate": 2.3659587883438106e-06,
+      "loss": 0.1876,
+      "step": 3302
+    },
+    {
+      "epoch": 1.566888045540797,
+      "grad_norm": 1.6685066223144531,
+      "learning_rate": 2.36099840765043e-06,
+      "loss": 0.1738,
+      "step": 3303
+    },
+    {
+      "epoch": 1.5673624288425048,
+      "grad_norm": 1.763706922531128,
+      "learning_rate": 2.3560425361840976e-06,
+      "loss": 0.1597,
+      "step": 3304
+    },
+    {
+      "epoch": 1.5678368121442126,
+      "grad_norm": 1.954229712486267,
+      "learning_rate": 2.3510911768702184e-06,
+      "loss": 0.1906,
+      "step": 3305
+    },
+    {
+      "epoch": 1.5683111954459203,
+      "grad_norm": 1.7269539833068848,
+      "learning_rate": 2.346144332631526e-06,
+      "loss": 0.1543,
+      "step": 3306
+    },
+    {
+      "epoch": 1.5687855787476281,
+      "grad_norm": 1.9077547788619995,
+      "learning_rate": 2.3412020063880957e-06,
+      "loss": 0.1752,
+      "step": 3307
+    },
+    {
+      "epoch": 1.569259962049336,
+      "grad_norm": 1.3996086120605469,
+      "learning_rate": 2.336264201057333e-06,
+      "loss": 0.1241,
+      "step": 3308
+    },
+    {
+      "epoch": 1.5697343453510437,
+      "grad_norm": 1.5195519924163818,
+      "learning_rate": 2.331330919553981e-06,
+      "loss": 0.1282,
+      "step": 3309
+    },
+    {
+      "epoch": 1.5702087286527515,
+      "grad_norm": 1.725934624671936,
+      "learning_rate": 2.3264021647901014e-06,
+      "loss": 0.1716,
+      "step": 3310
+    },
+    {
+      "epoch": 1.5706831119544593,
+      "grad_norm": 1.739709734916687,
+      "learning_rate": 2.3214779396750885e-06,
+      "loss": 0.149,
+      "step": 3311
+    },
+    {
+      "epoch": 1.571157495256167,
+      "grad_norm": 1.6437108516693115,
+      "learning_rate": 2.3165582471156643e-06,
+      "loss": 0.1608,
+      "step": 3312
+    },
+    {
+      "epoch": 1.571631878557875,
+      "grad_norm": 2.1839261054992676,
+      "learning_rate": 2.311643090015877e-06,
+      "loss": 0.1525,
+      "step": 3313
+    },
+    {
+      "epoch": 1.5721062618595827,
+      "grad_norm": 1.8370964527130127,
+      "learning_rate": 2.3067324712770967e-06,
+      "loss": 0.1484,
+      "step": 3314
+    },
+    {
+      "epoch": 1.5725806451612905,
+      "grad_norm": 1.8021410703659058,
+      "learning_rate": 2.301826393798008e-06,
+      "loss": 0.1618,
+      "step": 3315
+    },
+    {
+      "epoch": 1.573055028462998,
+      "grad_norm": 1.7016773223876953,
+      "learning_rate": 2.296924860474621e-06,
+      "loss": 0.137,
+      "step": 3316
+    },
+    {
+      "epoch": 1.5735294117647058,
+      "grad_norm": 1.573151707649231,
+      "learning_rate": 2.2920278742002677e-06,
+      "loss": 0.1449,
+      "step": 3317
+    },
+    {
+      "epoch": 1.5740037950664136,
+      "grad_norm": 1.4972699880599976,
+      "learning_rate": 2.287135437865583e-06,
+      "loss": 0.1369,
+      "step": 3318
+    },
+    {
+      "epoch": 1.5744781783681214,
+      "grad_norm": 1.9173699617385864,
+      "learning_rate": 2.282247554358531e-06,
+      "loss": 0.1329,
+      "step": 3319
+    },
+    {
+      "epoch": 1.5749525616698292,
+      "grad_norm": 1.7459148168563843,
+      "learning_rate": 2.2773642265643734e-06,
+      "loss": 0.1782,
+      "step": 3320
+    },
+    {
+      "epoch": 1.575426944971537,
+      "grad_norm": 1.479753851890564,
+      "learning_rate": 2.272485457365695e-06,
+      "loss": 0.1241,
+      "step": 3321
+    },
+    {
+      "epoch": 1.5759013282732448,
+      "grad_norm": 1.529350996017456,
+      "learning_rate": 2.267611249642383e-06,
+      "loss": 0.1352,
+      "step": 3322
+    },
+    {
+      "epoch": 1.5763757115749526,
+      "grad_norm": 2.032287120819092,
+      "learning_rate": 2.2627416062716366e-06,
+      "loss": 0.1704,
+      "step": 3323
+    },
+    {
+      "epoch": 1.5768500948766604,
+      "grad_norm": 1.399459719657898,
+      "learning_rate": 2.257876530127958e-06,
+      "loss": 0.1306,
+      "step": 3324
+    },
+    {
+      "epoch": 1.577324478178368,
+      "grad_norm": 1.9230834245681763,
+      "learning_rate": 2.25301602408315e-06,
+      "loss": 0.2053,
+      "step": 3325
+    },
+    {
+      "epoch": 1.5777988614800758,
+      "grad_norm": 1.5427366495132446,
+      "learning_rate": 2.248160091006326e-06,
+      "loss": 0.137,
+      "step": 3326
+    },
+    {
+      "epoch": 1.5782732447817835,
+      "grad_norm": 1.7453058958053589,
+      "learning_rate": 2.243308733763889e-06,
+      "loss": 0.1422,
+      "step": 3327
+    },
+    {
+      "epoch": 1.5787476280834913,
+      "grad_norm": 1.7460287809371948,
+      "learning_rate": 2.2384619552195518e-06,
+      "loss": 0.1742,
+      "step": 3328
+    },
+    {
+      "epoch": 1.5792220113851991,
+      "grad_norm": 1.7111773490905762,
+      "learning_rate": 2.233619758234321e-06,
+      "loss": 0.1552,
+      "step": 3329
+    },
+    {
+      "epoch": 1.579696394686907,
+      "grad_norm": 1.5118764638900757,
+      "learning_rate": 2.2287821456664926e-06,
+      "loss": 0.1266,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5801707779886147,
+      "grad_norm": 1.605214238166809,
+      "learning_rate": 2.2239491203716644e-06,
+      "loss": 0.1548,
+      "step": 3331
+    },
+    {
+      "epoch": 1.5806451612903225,
+      "grad_norm": 1.5546764135360718,
+      "learning_rate": 2.2191206852027225e-06,
+      "loss": 0.1369,
+      "step": 3332
+    },
+    {
+      "epoch": 1.5811195445920303,
+      "grad_norm": 1.529728889465332,
+      "learning_rate": 2.214296843009848e-06,
+      "loss": 0.1205,
+      "step": 3333
+    },
+    {
+      "epoch": 1.581593927893738,
+      "grad_norm": 1.7725614309310913,
+      "learning_rate": 2.2094775966405045e-06,
+      "loss": 0.1509,
+      "step": 3334
+    },
+    {
+      "epoch": 1.5820683111954459,
+      "grad_norm": 1.8418389558792114,
+      "learning_rate": 2.2046629489394422e-06,
+      "loss": 0.1339,
+      "step": 3335
+    },
+    {
+      "epoch": 1.5825426944971537,
+      "grad_norm": 1.8874013423919678,
+      "learning_rate": 2.199852902748704e-06,
+      "loss": 0.1873,
+      "step": 3336
+    },
+    {
+      "epoch": 1.5830170777988615,
+      "grad_norm": 1.5304529666900635,
+      "learning_rate": 2.19504746090761e-06,
+      "loss": 0.1311,
+      "step": 3337
+    },
+    {
+      "epoch": 1.5834914611005693,
+      "grad_norm": 2.1148695945739746,
+      "learning_rate": 2.19024662625277e-06,
+      "loss": 0.1458,
+      "step": 3338
+    },
+    {
+      "epoch": 1.583965844402277,
+      "grad_norm": 1.3242015838623047,
+      "learning_rate": 2.185450401618062e-06,
+      "loss": 0.1238,
+      "step": 3339
+    },
+    {
+      "epoch": 1.5844402277039848,
+      "grad_norm": 2.047175645828247,
+      "learning_rate": 2.1806587898346553e-06,
+      "loss": 0.1712,
+      "step": 3340
+    },
+    {
+      "epoch": 1.5849146110056926,
+      "grad_norm": 1.6021418571472168,
+      "learning_rate": 2.17587179373099e-06,
+      "loss": 0.1615,
+      "step": 3341
+    },
+    {
+      "epoch": 1.5853889943074004,
+      "grad_norm": 1.975977897644043,
+      "learning_rate": 2.1710894161327813e-06,
+      "loss": 0.1399,
+      "step": 3342
+    },
+    {
+      "epoch": 1.5858633776091082,
+      "grad_norm": 1.3944038152694702,
+      "learning_rate": 2.1663116598630207e-06,
+      "loss": 0.1451,
+      "step": 3343
+    },
+    {
+      "epoch": 1.586337760910816,
+      "grad_norm": 1.7378476858139038,
+      "learning_rate": 2.1615385277419687e-06,
+      "loss": 0.1514,
+      "step": 3344
+    },
+    {
+      "epoch": 1.5868121442125238,
+      "grad_norm": 1.55166494846344,
+      "learning_rate": 2.156770022587157e-06,
+      "loss": 0.1545,
+      "step": 3345
+    },
+    {
+      "epoch": 1.5872865275142316,
+      "grad_norm": 2.1811869144439697,
+      "learning_rate": 2.1520061472133903e-06,
+      "loss": 0.1508,
+      "step": 3346
+    },
+    {
+      "epoch": 1.5877609108159394,
+      "grad_norm": 1.4423187971115112,
+      "learning_rate": 2.147246904432735e-06,
+      "loss": 0.1401,
+      "step": 3347
+    },
+    {
+      "epoch": 1.5882352941176472,
+      "grad_norm": 1.590010166168213,
+      "learning_rate": 2.1424922970545283e-06,
+      "loss": 0.1466,
+      "step": 3348
+    },
+    {
+      "epoch": 1.588709677419355,
+      "grad_norm": 1.5334136486053467,
+      "learning_rate": 2.1377423278853627e-06,
+      "loss": 0.1193,
+      "step": 3349
+    },
+    {
+      "epoch": 1.5891840607210628,
+      "grad_norm": 1.605485200881958,
+      "learning_rate": 2.1329969997291035e-06,
+      "loss": 0.1445,
+      "step": 3350
+    },
+    {
+      "epoch": 1.5896584440227703,
+      "grad_norm": 1.5612473487854004,
+      "learning_rate": 2.128256315386865e-06,
+      "loss": 0.1313,
+      "step": 3351
+    },
+    {
+      "epoch": 1.5901328273244781,
+      "grad_norm": 1.702793002128601,
+      "learning_rate": 2.1235202776570297e-06,
+      "loss": 0.158,
+      "step": 3352
+    },
+    {
+      "epoch": 1.590607210626186,
+      "grad_norm": 1.6103143692016602,
+      "learning_rate": 2.118788889335236e-06,
+      "loss": 0.1707,
+      "step": 3353
+    },
+    {
+      "epoch": 1.5910815939278937,
+      "grad_norm": 1.7173638343811035,
+      "learning_rate": 2.11406215321437e-06,
+      "loss": 0.1728,
+      "step": 3354
+    },
+    {
+      "epoch": 1.5915559772296015,
+      "grad_norm": 1.4429662227630615,
+      "learning_rate": 2.1093400720845813e-06,
+      "loss": 0.1154,
+      "step": 3355
+    },
+    {
+      "epoch": 1.5920303605313093,
+      "grad_norm": 1.5256520509719849,
+      "learning_rate": 2.1046226487332655e-06,
+      "loss": 0.1471,
+      "step": 3356
+    },
+    {
+      "epoch": 1.592504743833017,
+      "grad_norm": 1.6516417264938354,
+      "learning_rate": 2.099909885945075e-06,
+      "loss": 0.1527,
+      "step": 3357
+    },
+    {
+      "epoch": 1.592979127134725,
+      "grad_norm": 1.8592281341552734,
+      "learning_rate": 2.0952017865019036e-06,
+      "loss": 0.1506,
+      "step": 3358
+    },
+    {
+      "epoch": 1.5934535104364327,
+      "grad_norm": 1.5824545621871948,
+      "learning_rate": 2.0904983531828947e-06,
+      "loss": 0.1468,
+      "step": 3359
+    },
+    {
+      "epoch": 1.5939278937381403,
+      "grad_norm": 1.5531175136566162,
+      "learning_rate": 2.085799588764439e-06,
+      "loss": 0.1386,
+      "step": 3360
+    },
+    {
+      "epoch": 1.594402277039848,
+      "grad_norm": 1.6820399761199951,
+      "learning_rate": 2.081105496020173e-06,
+      "loss": 0.157,
+      "step": 3361
+    },
+    {
+      "epoch": 1.5948766603415558,
+      "grad_norm": 1.3576366901397705,
+      "learning_rate": 2.076416077720973e-06,
+      "loss": 0.1285,
+      "step": 3362
+    },
+    {
+      "epoch": 1.5953510436432636,
+      "grad_norm": 1.8949618339538574,
+      "learning_rate": 2.0717313366349534e-06,
+      "loss": 0.1585,
+      "step": 3363
+    },
+    {
+      "epoch": 1.5958254269449714,
+      "grad_norm": 1.572543978691101,
+      "learning_rate": 2.067051275527472e-06,
+      "loss": 0.1509,
+      "step": 3364
+    },
+    {
+      "epoch": 1.5962998102466792,
+      "grad_norm": 1.6558380126953125,
+      "learning_rate": 2.0623758971611252e-06,
+      "loss": 0.1409,
+      "step": 3365
+    },
+    {
+      "epoch": 1.596774193548387,
+      "grad_norm": 1.892311930656433,
+      "learning_rate": 2.0577052042957378e-06,
+      "loss": 0.189,
+      "step": 3366
+    },
+    {
+      "epoch": 1.5972485768500948,
+      "grad_norm": 1.8912239074707031,
+      "learning_rate": 2.0530391996883782e-06,
+      "loss": 0.15,
+      "step": 3367
+    },
+    {
+      "epoch": 1.5977229601518026,
+      "grad_norm": 1.8770484924316406,
+      "learning_rate": 2.0483778860933377e-06,
+      "loss": 0.1802,
+      "step": 3368
+    },
+    {
+      "epoch": 1.5981973434535104,
+      "grad_norm": 1.7015761137008667,
+      "learning_rate": 2.0437212662621477e-06,
+      "loss": 0.1437,
+      "step": 3369
+    },
+    {
+      "epoch": 1.5986717267552182,
+      "grad_norm": 1.6395004987716675,
+      "learning_rate": 2.0390693429435626e-06,
+      "loss": 0.1361,
+      "step": 3370
+    },
+    {
+      "epoch": 1.599146110056926,
+      "grad_norm": 1.7334997653961182,
+      "learning_rate": 2.0344221188835667e-06,
+      "loss": 0.1372,
+      "step": 3371
+    },
+    {
+      "epoch": 1.5996204933586338,
+      "grad_norm": 1.7484663724899292,
+      "learning_rate": 2.0297795968253753e-06,
+      "loss": 0.173,
+      "step": 3372
+    },
+    {
+      "epoch": 1.6000948766603416,
+      "grad_norm": 1.5254170894622803,
+      "learning_rate": 2.0251417795094166e-06,
+      "loss": 0.1516,
+      "step": 3373
+    },
+    {
+      "epoch": 1.6005692599620494,
+      "grad_norm": 1.8212133646011353,
+      "learning_rate": 2.020508669673352e-06,
+      "loss": 0.151,
+      "step": 3374
+    },
+    {
+      "epoch": 1.6010436432637571,
+      "grad_norm": 1.369913101196289,
+      "learning_rate": 2.0158802700520576e-06,
+      "loss": 0.1338,
+      "step": 3375
+    },
+    {
+      "epoch": 1.601518026565465,
+      "grad_norm": 1.2662460803985596,
+      "learning_rate": 2.0112565833776364e-06,
+      "loss": 0.1228,
+      "step": 3376
+    },
+    {
+      "epoch": 1.6019924098671727,
+      "grad_norm": 1.7709726095199585,
+      "learning_rate": 2.0066376123793984e-06,
+      "loss": 0.1726,
+      "step": 3377
+    },
+    {
+      "epoch": 1.6024667931688805,
+      "grad_norm": 1.5085747241973877,
+      "learning_rate": 2.0020233597838813e-06,
+      "loss": 0.1034,
+      "step": 3378
+    },
+    {
+      "epoch": 1.6029411764705883,
+      "grad_norm": 1.3893810510635376,
+      "learning_rate": 1.99741382831483e-06,
+      "loss": 0.131,
+      "step": 3379
+    },
+    {
+      "epoch": 1.603415559772296,
+      "grad_norm": 2.0741636753082275,
+      "learning_rate": 1.9928090206932083e-06,
+      "loss": 0.2064,
+      "step": 3380
+    },
+    {
+      "epoch": 1.603889943074004,
+      "grad_norm": 1.5764415264129639,
+      "learning_rate": 1.9882089396371896e-06,
+      "loss": 0.1558,
+      "step": 3381
+    },
+    {
+      "epoch": 1.6043643263757117,
+      "grad_norm": 1.7845758199691772,
+      "learning_rate": 1.983613587862153e-06,
+      "loss": 0.1599,
+      "step": 3382
+    },
+    {
+      "epoch": 1.6048387096774195,
+      "grad_norm": 1.610253095626831,
+      "learning_rate": 1.9790229680806883e-06,
+      "loss": 0.1682,
+      "step": 3383
+    },
+    {
+      "epoch": 1.6053130929791273,
+      "grad_norm": 1.7443877458572388,
+      "learning_rate": 1.9744370830025937e-06,
+      "loss": 0.155,
+      "step": 3384
+    },
+    {
+      "epoch": 1.605787476280835,
+      "grad_norm": 1.7004719972610474,
+      "learning_rate": 1.9698559353348735e-06,
+      "loss": 0.1436,
+      "step": 3385
+    },
+    {
+      "epoch": 1.6062618595825426,
+      "grad_norm": 1.878106713294983,
+      "learning_rate": 1.9652795277817348e-06,
+      "loss": 0.181,
+      "step": 3386
+    },
+    {
+      "epoch": 1.6067362428842504,
+      "grad_norm": 1.6228288412094116,
+      "learning_rate": 1.960707863044582e-06,
+      "loss": 0.1511,
+      "step": 3387
+    },
+    {
+      "epoch": 1.6072106261859582,
+      "grad_norm": 1.7127445936203003,
+      "learning_rate": 1.9561409438220245e-06,
+      "loss": 0.1445,
+      "step": 3388
+    },
+    {
+      "epoch": 1.607685009487666,
+      "grad_norm": 1.5435030460357666,
+      "learning_rate": 1.9515787728098733e-06,
+      "loss": 0.1298,
+      "step": 3389
+    },
+    {
+      "epoch": 1.6081593927893738,
+      "grad_norm": 1.4933137893676758,
+      "learning_rate": 1.9470213527011282e-06,
+      "loss": 0.1415,
+      "step": 3390
+    },
+    {
+      "epoch": 1.6086337760910816,
+      "grad_norm": 1.9454411268234253,
+      "learning_rate": 1.9424686861859933e-06,
+      "loss": 0.174,
+      "step": 3391
+    },
+    {
+      "epoch": 1.6091081593927894,
+      "grad_norm": 1.5008386373519897,
+      "learning_rate": 1.937920775951857e-06,
+      "loss": 0.1465,
+      "step": 3392
+    },
+    {
+      "epoch": 1.6095825426944972,
+      "grad_norm": 1.5946168899536133,
+      "learning_rate": 1.9333776246833092e-06,
+      "loss": 0.1379,
+      "step": 3393
+    },
+    {
+      "epoch": 1.610056925996205,
+      "grad_norm": 1.4043335914611816,
+      "learning_rate": 1.9288392350621275e-06,
+      "loss": 0.1265,
+      "step": 3394
+    },
+    {
+      "epoch": 1.6105313092979125,
+      "grad_norm": 1.7210297584533691,
+      "learning_rate": 1.9243056097672796e-06,
+      "loss": 0.1668,
+      "step": 3395
+    },
+    {
+      "epoch": 1.6110056925996203,
+      "grad_norm": 1.7458949089050293,
+      "learning_rate": 1.9197767514749156e-06,
+      "loss": 0.16,
+      "step": 3396
+    },
+    {
+      "epoch": 1.6114800759013281,
+      "grad_norm": 1.8597474098205566,
+      "learning_rate": 1.915252662858378e-06,
+      "loss": 0.1811,
+      "step": 3397
+    },
+    {
+      "epoch": 1.611954459203036,
+      "grad_norm": 1.412137508392334,
+      "learning_rate": 1.910733346588194e-06,
+      "loss": 0.1204,
+      "step": 3398
+    },
+    {
+      "epoch": 1.6124288425047437,
+      "grad_norm": 1.6285221576690674,
+      "learning_rate": 1.9062188053320663e-06,
+      "loss": 0.1567,
+      "step": 3399
+    },
+    {
+      "epoch": 1.6129032258064515,
+      "grad_norm": 2.3471789360046387,
+      "learning_rate": 1.901709041754889e-06,
+      "loss": 0.1562,
+      "step": 3400
+    },
+    {
+      "epoch": 1.6133776091081593,
+      "grad_norm": 1.5829018354415894,
+      "learning_rate": 1.8972040585187256e-06,
+      "loss": 0.1545,
+      "step": 3401
+    },
+    {
+      "epoch": 1.613851992409867,
+      "grad_norm": 1.5271830558776855,
+      "learning_rate": 1.8927038582828261e-06,
+      "loss": 0.1294,
+      "step": 3402
+    },
+    {
+      "epoch": 1.614326375711575,
+      "grad_norm": 1.8951029777526855,
+      "learning_rate": 1.8882084437036142e-06,
+      "loss": 0.1618,
+      "step": 3403
+    },
+    {
+      "epoch": 1.6148007590132827,
+      "grad_norm": 1.707849383354187,
+      "learning_rate": 1.8837178174346882e-06,
+      "loss": 0.1544,
+      "step": 3404
+    },
+    {
+      "epoch": 1.6152751423149905,
+      "grad_norm": 1.8298168182373047,
+      "learning_rate": 1.8792319821268223e-06,
+      "loss": 0.1755,
+      "step": 3405
+    },
+    {
+      "epoch": 1.6157495256166983,
+      "grad_norm": 1.6558297872543335,
+      "learning_rate": 1.8747509404279595e-06,
+      "loss": 0.1473,
+      "step": 3406
+    },
+    {
+      "epoch": 1.616223908918406,
+      "grad_norm": 1.5755760669708252,
+      "learning_rate": 1.8702746949832117e-06,
+      "loss": 0.1447,
+      "step": 3407
+    },
+    {
+      "epoch": 1.6166982922201139,
+      "grad_norm": 1.9705390930175781,
+      "learning_rate": 1.8658032484348632e-06,
+      "loss": 0.1738,
+      "step": 3408
+    },
+    {
+      "epoch": 1.6171726755218216,
+      "grad_norm": 2.1337902545928955,
+      "learning_rate": 1.8613366034223668e-06,
+      "loss": 0.2303,
+      "step": 3409
+    },
+    {
+      "epoch": 1.6176470588235294,
+      "grad_norm": 1.5434858798980713,
+      "learning_rate": 1.8568747625823403e-06,
+      "loss": 0.1602,
+      "step": 3410
+    },
+    {
+      "epoch": 1.6181214421252372,
+      "grad_norm": 2.1457488536834717,
+      "learning_rate": 1.8524177285485588e-06,
+      "loss": 0.1905,
+      "step": 3411
+    },
+    {
+      "epoch": 1.618595825426945,
+      "grad_norm": 1.500415563583374,
+      "learning_rate": 1.8479655039519683e-06,
+      "loss": 0.1368,
+      "step": 3412
+    },
+    {
+      "epoch": 1.6190702087286528,
+      "grad_norm": 1.6900932788848877,
+      "learning_rate": 1.8435180914206763e-06,
+      "loss": 0.1541,
+      "step": 3413
+    },
+    {
+      "epoch": 1.6195445920303606,
+      "grad_norm": 1.5415065288543701,
+      "learning_rate": 1.8390754935799404e-06,
+      "loss": 0.1665,
+      "step": 3414
+    },
+    {
+      "epoch": 1.6200189753320684,
+      "grad_norm": 1.7073814868927002,
+      "learning_rate": 1.8346377130521864e-06,
+      "loss": 0.1549,
+      "step": 3415
+    },
+    {
+      "epoch": 1.6204933586337762,
+      "grad_norm": 1.7535865306854248,
+      "learning_rate": 1.8302047524569888e-06,
+      "loss": 0.1497,
+      "step": 3416
+    },
+    {
+      "epoch": 1.620967741935484,
+      "grad_norm": 1.7559723854064941,
+      "learning_rate": 1.8257766144110823e-06,
+      "loss": 0.1579,
+      "step": 3417
+    },
+    {
+      "epoch": 1.6214421252371918,
+      "grad_norm": 1.74555242061615,
+      "learning_rate": 1.8213533015283524e-06,
+      "loss": 0.1379,
+      "step": 3418
+    },
+    {
+      "epoch": 1.6219165085388996,
+      "grad_norm": 2.0303471088409424,
+      "learning_rate": 1.8169348164198408e-06,
+      "loss": 0.1545,
+      "step": 3419
+    },
+    {
+      "epoch": 1.6223908918406074,
+      "grad_norm": 1.6806904077529907,
+      "learning_rate": 1.8125211616937289e-06,
+      "loss": 0.1171,
+      "step": 3420
+    },
+    {
+      "epoch": 1.6228652751423152,
+      "grad_norm": 1.7993685007095337,
+      "learning_rate": 1.8081123399553569e-06,
+      "loss": 0.1789,
+      "step": 3421
+    },
+    {
+      "epoch": 1.6233396584440227,
+      "grad_norm": 1.8345065116882324,
+      "learning_rate": 1.8037083538072109e-06,
+      "loss": 0.16,
+      "step": 3422
+    },
+    {
+      "epoch": 1.6238140417457305,
+      "grad_norm": 1.875479817390442,
+      "learning_rate": 1.7993092058489158e-06,
+      "loss": 0.1429,
+      "step": 3423
+    },
+    {
+      "epoch": 1.6242884250474383,
+      "grad_norm": 1.7342585325241089,
+      "learning_rate": 1.794914898677249e-06,
+      "loss": 0.1624,
+      "step": 3424
+    },
+    {
+      "epoch": 1.624762808349146,
+      "grad_norm": 1.4515552520751953,
+      "learning_rate": 1.7905254348861235e-06,
+      "loss": 0.1347,
+      "step": 3425
+    },
+    {
+      "epoch": 1.625237191650854,
+      "grad_norm": 1.9419041872024536,
+      "learning_rate": 1.7861408170665961e-06,
+      "loss": 0.1561,
+      "step": 3426
+    },
+    {
+      "epoch": 1.6257115749525617,
+      "grad_norm": 2.1055121421813965,
+      "learning_rate": 1.7817610478068659e-06,
+      "loss": 0.1867,
+      "step": 3427
+    },
+    {
+      "epoch": 1.6261859582542695,
+      "grad_norm": 1.6690064668655396,
+      "learning_rate": 1.7773861296922657e-06,
+      "loss": 0.1776,
+      "step": 3428
+    },
+    {
+      "epoch": 1.6266603415559773,
+      "grad_norm": 2.282166004180908,
+      "learning_rate": 1.7730160653052685e-06,
+      "loss": 0.205,
+      "step": 3429
+    },
+    {
+      "epoch": 1.6271347248576848,
+      "grad_norm": 1.7570334672927856,
+      "learning_rate": 1.7686508572254802e-06,
+      "loss": 0.1546,
+      "step": 3430
+    },
+    {
+      "epoch": 1.6276091081593926,
+      "grad_norm": 1.8625390529632568,
+      "learning_rate": 1.7642905080296346e-06,
+      "loss": 0.1668,
+      "step": 3431
+    },
+    {
+      "epoch": 1.6280834914611004,
+      "grad_norm": 1.4064631462097168,
+      "learning_rate": 1.7599350202916066e-06,
+      "loss": 0.1402,
+      "step": 3432
+    },
+    {
+      "epoch": 1.6285578747628082,
+      "grad_norm": 1.3597134351730347,
+      "learning_rate": 1.7555843965823992e-06,
+      "loss": 0.1314,
+      "step": 3433
+    },
+    {
+      "epoch": 1.629032258064516,
+      "grad_norm": 1.76325523853302,
+      "learning_rate": 1.7512386394701386e-06,
+      "loss": 0.1729,
+      "step": 3434
+    },
+    {
+      "epoch": 1.6295066413662238,
+      "grad_norm": 1.6423853635787964,
+      "learning_rate": 1.7468977515200835e-06,
+      "loss": 0.1287,
+      "step": 3435
+    },
+    {
+      "epoch": 1.6299810246679316,
+      "grad_norm": 1.6672308444976807,
+      "learning_rate": 1.7425617352946178e-06,
+      "loss": 0.1701,
+      "step": 3436
+    },
+    {
+      "epoch": 1.6304554079696394,
+      "grad_norm": 1.6599950790405273,
+      "learning_rate": 1.7382305933532494e-06,
+      "loss": 0.1618,
+      "step": 3437
+    },
+    {
+      "epoch": 1.6309297912713472,
+      "grad_norm": 1.5585594177246094,
+      "learning_rate": 1.7339043282526103e-06,
+      "loss": 0.149,
+      "step": 3438
+    },
+    {
+      "epoch": 1.631404174573055,
+      "grad_norm": 1.8305972814559937,
+      "learning_rate": 1.7295829425464494e-06,
+      "loss": 0.1789,
+      "step": 3439
+    },
+    {
+      "epoch": 1.6318785578747628,
+      "grad_norm": 1.763118028640747,
+      "learning_rate": 1.7252664387856367e-06,
+      "loss": 0.1734,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6323529411764706,
+      "grad_norm": 1.743260383605957,
+      "learning_rate": 1.7209548195181625e-06,
+      "loss": 0.1609,
+      "step": 3441
+    },
+    {
+      "epoch": 1.6328273244781784,
+      "grad_norm": 1.892848014831543,
+      "learning_rate": 1.7166480872891333e-06,
+      "loss": 0.1606,
+      "step": 3442
+    },
+    {
+      "epoch": 1.6333017077798861,
+      "grad_norm": 1.9057481288909912,
+      "learning_rate": 1.7123462446407746e-06,
+      "loss": 0.1659,
+      "step": 3443
+    },
+    {
+      "epoch": 1.633776091081594,
+      "grad_norm": 1.7405368089675903,
+      "learning_rate": 1.7080492941124139e-06,
+      "loss": 0.1593,
+      "step": 3444
+    },
+    {
+      "epoch": 1.6342504743833017,
+      "grad_norm": 1.713452696800232,
+      "learning_rate": 1.7037572382405031e-06,
+      "loss": 0.1283,
+      "step": 3445
+    },
+    {
+      "epoch": 1.6347248576850095,
+      "grad_norm": 1.595138430595398,
+      "learning_rate": 1.6994700795586027e-06,
+      "loss": 0.1319,
+      "step": 3446
+    },
+    {
+      "epoch": 1.6351992409867173,
+      "grad_norm": 1.756667137145996,
+      "learning_rate": 1.6951878205973738e-06,
+      "loss": 0.1677,
+      "step": 3447
+    },
+    {
+      "epoch": 1.635673624288425,
+      "grad_norm": 1.6868460178375244,
+      "learning_rate": 1.6909104638845986e-06,
+      "loss": 0.1546,
+      "step": 3448
+    },
+    {
+      "epoch": 1.636148007590133,
+      "grad_norm": 1.5798834562301636,
+      "learning_rate": 1.686638011945151e-06,
+      "loss": 0.1525,
+      "step": 3449
+    },
+    {
+      "epoch": 1.6366223908918407,
+      "grad_norm": 1.3165028095245361,
+      "learning_rate": 1.682370467301021e-06,
+      "loss": 0.1139,
+      "step": 3450
+    },
+    {
+      "epoch": 1.6370967741935485,
+      "grad_norm": 1.6068346500396729,
+      "learning_rate": 1.6781078324712973e-06,
+      "loss": 0.1436,
+      "step": 3451
+    },
+    {
+      "epoch": 1.6375711574952563,
+      "grad_norm": 1.4109083414077759,
+      "learning_rate": 1.6738501099721737e-06,
+      "loss": 0.1366,
+      "step": 3452
+    },
+    {
+      "epoch": 1.638045540796964,
+      "grad_norm": 1.7669192552566528,
+      "learning_rate": 1.6695973023169375e-06,
+      "loss": 0.1751,
+      "step": 3453
+    },
+    {
+      "epoch": 1.6385199240986719,
+      "grad_norm": 1.5503531694412231,
+      "learning_rate": 1.6653494120159842e-06,
+      "loss": 0.1462,
+      "step": 3454
+    },
+    {
+      "epoch": 1.6389943074003797,
+      "grad_norm": 1.9344103336334229,
+      "learning_rate": 1.6611064415767941e-06,
+      "loss": 0.2062,
+      "step": 3455
+    },
+    {
+      "epoch": 1.6394686907020875,
+      "grad_norm": 1.5301547050476074,
+      "learning_rate": 1.6568683935039554e-06,
+      "loss": 0.1375,
+      "step": 3456
+    },
+    {
+      "epoch": 1.639943074003795,
+      "grad_norm": 1.9043666124343872,
+      "learning_rate": 1.6526352702991478e-06,
+      "loss": 0.199,
+      "step": 3457
+    },
+    {
+      "epoch": 1.6404174573055028,
+      "grad_norm": 1.684772253036499,
+      "learning_rate": 1.6484070744611358e-06,
+      "loss": 0.1337,
+      "step": 3458
+    },
+    {
+      "epoch": 1.6408918406072106,
+      "grad_norm": 1.598937749862671,
+      "learning_rate": 1.6441838084857863e-06,
+      "loss": 0.1523,
+      "step": 3459
+    },
+    {
+      "epoch": 1.6413662239089184,
+      "grad_norm": 1.815584421157837,
+      "learning_rate": 1.6399654748660498e-06,
+      "loss": 0.1498,
+      "step": 3460
+    },
+    {
+      "epoch": 1.6418406072106262,
+      "grad_norm": 1.7786113023757935,
+      "learning_rate": 1.6357520760919675e-06,
+      "loss": 0.1683,
+      "step": 3461
+    },
+    {
+      "epoch": 1.642314990512334,
+      "grad_norm": 1.3854840993881226,
+      "learning_rate": 1.6315436146506702e-06,
+      "loss": 0.1122,
+      "step": 3462
+    },
+    {
+      "epoch": 1.6427893738140418,
+      "grad_norm": 1.40729558467865,
+      "learning_rate": 1.6273400930263672e-06,
+      "loss": 0.1333,
+      "step": 3463
+    },
+    {
+      "epoch": 1.6432637571157496,
+      "grad_norm": 1.844829797744751,
+      "learning_rate": 1.6231415137003536e-06,
+      "loss": 0.145,
+      "step": 3464
+    },
+    {
+      "epoch": 1.6437381404174574,
+      "grad_norm": 2.1833529472351074,
+      "learning_rate": 1.6189478791510116e-06,
+      "loss": 0.189,
+      "step": 3465
+    },
+    {
+      "epoch": 1.644212523719165,
+      "grad_norm": 1.6294044256210327,
+      "learning_rate": 1.614759191853803e-06,
+      "loss": 0.1264,
+      "step": 3466
+    },
+    {
+      "epoch": 1.6446869070208727,
+      "grad_norm": 1.6471272706985474,
+      "learning_rate": 1.6105754542812702e-06,
+      "loss": 0.1396,
+      "step": 3467
+    },
+    {
+      "epoch": 1.6451612903225805,
+      "grad_norm": 1.920088768005371,
+      "learning_rate": 1.6063966689030275e-06,
+      "loss": 0.1704,
+      "step": 3468
+    },
+    {
+      "epoch": 1.6456356736242883,
+      "grad_norm": 1.8987756967544556,
+      "learning_rate": 1.6022228381857729e-06,
+      "loss": 0.1184,
+      "step": 3469
+    },
+    {
+      "epoch": 1.646110056925996,
+      "grad_norm": 1.6714553833007812,
+      "learning_rate": 1.5980539645932802e-06,
+      "loss": 0.1568,
+      "step": 3470
+    },
+    {
+      "epoch": 1.646584440227704,
+      "grad_norm": 1.6025445461273193,
+      "learning_rate": 1.5938900505863886e-06,
+      "loss": 0.143,
+      "step": 3471
+    },
+    {
+      "epoch": 1.6470588235294117,
+      "grad_norm": 1.8814153671264648,
+      "learning_rate": 1.5897310986230196e-06,
+      "loss": 0.1493,
+      "step": 3472
+    },
+    {
+      "epoch": 1.6475332068311195,
+      "grad_norm": 1.7132648229599,
+      "learning_rate": 1.5855771111581586e-06,
+      "loss": 0.1695,
+      "step": 3473
+    },
+    {
+      "epoch": 1.6480075901328273,
+      "grad_norm": 1.6808983087539673,
+      "learning_rate": 1.5814280906438639e-06,
+      "loss": 0.1711,
+      "step": 3474
+    },
+    {
+      "epoch": 1.648481973434535,
+      "grad_norm": 2.184779405593872,
+      "learning_rate": 1.5772840395292632e-06,
+      "loss": 0.1722,
+      "step": 3475
+    },
+    {
+      "epoch": 1.6489563567362429,
+      "grad_norm": 1.5792295932769775,
+      "learning_rate": 1.5731449602605487e-06,
+      "loss": 0.152,
+      "step": 3476
+    },
+    {
+      "epoch": 1.6494307400379506,
+      "grad_norm": 1.4124369621276855,
+      "learning_rate": 1.5690108552809746e-06,
+      "loss": 0.1496,
+      "step": 3477
+    },
+    {
+      "epoch": 1.6499051233396584,
+      "grad_norm": 1.8069303035736084,
+      "learning_rate": 1.5648817270308648e-06,
+      "loss": 0.1382,
+      "step": 3478
+    },
+    {
+      "epoch": 1.6503795066413662,
+      "grad_norm": 1.832484245300293,
+      "learning_rate": 1.5607575779476047e-06,
+      "loss": 0.148,
+      "step": 3479
+    },
+    {
+      "epoch": 1.650853889943074,
+      "grad_norm": 2.0292952060699463,
+      "learning_rate": 1.556638410465635e-06,
+      "loss": 0.1711,
+      "step": 3480
+    },
+    {
+      "epoch": 1.6513282732447818,
+      "grad_norm": 1.6780633926391602,
+      "learning_rate": 1.552524227016462e-06,
+      "loss": 0.1237,
+      "step": 3481
+    },
+    {
+      "epoch": 1.6518026565464896,
+      "grad_norm": 1.6623648405075073,
+      "learning_rate": 1.548415030028645e-06,
+      "loss": 0.1589,
+      "step": 3482
+    },
+    {
+      "epoch": 1.6522770398481974,
+      "grad_norm": 1.652055263519287,
+      "learning_rate": 1.5443108219278036e-06,
+      "loss": 0.1615,
+      "step": 3483
+    },
+    {
+      "epoch": 1.6527514231499052,
+      "grad_norm": 1.6934820413589478,
+      "learning_rate": 1.5402116051366111e-06,
+      "loss": 0.1585,
+      "step": 3484
+    },
+    {
+      "epoch": 1.653225806451613,
+      "grad_norm": 1.5433070659637451,
+      "learning_rate": 1.5361173820747942e-06,
+      "loss": 0.1395,
+      "step": 3485
+    },
+    {
+      "epoch": 1.6537001897533208,
+      "grad_norm": 1.812530755996704,
+      "learning_rate": 1.5320281551591366e-06,
+      "loss": 0.1592,
+      "step": 3486
+    },
+    {
+      "epoch": 1.6541745730550286,
+      "grad_norm": 2.096843957901001,
+      "learning_rate": 1.5279439268034634e-06,
+      "loss": 0.1543,
+      "step": 3487
+    },
+    {
+      "epoch": 1.6546489563567364,
+      "grad_norm": 1.825169563293457,
+      "learning_rate": 1.5238646994186546e-06,
+      "loss": 0.153,
+      "step": 3488
+    },
+    {
+      "epoch": 1.6551233396584442,
+      "grad_norm": 1.7507545948028564,
+      "learning_rate": 1.519790475412638e-06,
+      "loss": 0.1509,
+      "step": 3489
+    },
+    {
+      "epoch": 1.655597722960152,
+      "grad_norm": 1.6845128536224365,
+      "learning_rate": 1.515721257190389e-06,
+      "loss": 0.1752,
+      "step": 3490
+    },
+    {
+      "epoch": 1.6560721062618597,
+      "grad_norm": 1.793799638748169,
+      "learning_rate": 1.5116570471539294e-06,
+      "loss": 0.1689,
+      "step": 3491
+    },
+    {
+      "epoch": 1.6565464895635673,
+      "grad_norm": 1.5899111032485962,
+      "learning_rate": 1.5075978477023156e-06,
+      "loss": 0.1291,
+      "step": 3492
+    },
+    {
+      "epoch": 1.657020872865275,
+      "grad_norm": 1.5500413179397583,
+      "learning_rate": 1.5035436612316567e-06,
+      "loss": 0.1548,
+      "step": 3493
+    },
+    {
+      "epoch": 1.657495256166983,
+      "grad_norm": 1.3642367124557495,
+      "learning_rate": 1.4994944901351006e-06,
+      "loss": 0.1452,
+      "step": 3494
+    },
+    {
+      "epoch": 1.6579696394686907,
+      "grad_norm": 1.7189817428588867,
+      "learning_rate": 1.4954503368028305e-06,
+      "loss": 0.1784,
+      "step": 3495
+    },
+    {
+      "epoch": 1.6584440227703985,
+      "grad_norm": 1.5453788042068481,
+      "learning_rate": 1.4914112036220696e-06,
+      "loss": 0.1572,
+      "step": 3496
+    },
+    {
+      "epoch": 1.6589184060721063,
+      "grad_norm": 1.3796250820159912,
+      "learning_rate": 1.4873770929770782e-06,
+      "loss": 0.1385,
+      "step": 3497
+    },
+    {
+      "epoch": 1.659392789373814,
+      "grad_norm": 1.3950660228729248,
+      "learning_rate": 1.4833480072491524e-06,
+      "loss": 0.1264,
+      "step": 3498
+    },
+    {
+      "epoch": 1.6598671726755219,
+      "grad_norm": 1.7917977571487427,
+      "learning_rate": 1.4793239488166222e-06,
+      "loss": 0.1483,
+      "step": 3499
+    },
+    {
+      "epoch": 1.6603415559772297,
+      "grad_norm": 1.921996831893921,
+      "learning_rate": 1.4753049200548519e-06,
+      "loss": 0.2267,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6608159392789372,
+      "grad_norm": 1.6829748153686523,
+      "learning_rate": 1.4712909233362304e-06,
+      "loss": 0.1554,
+      "step": 3501
+    },
+    {
+      "epoch": 1.661290322580645,
+      "grad_norm": 1.645816445350647,
+      "learning_rate": 1.4672819610301802e-06,
+      "loss": 0.1642,
+      "step": 3502
+    },
+    {
+      "epoch": 1.6617647058823528,
+      "grad_norm": 1.2264525890350342,
+      "learning_rate": 1.4632780355031573e-06,
+      "loss": 0.1073,
+      "step": 3503
+    },
+    {
+      "epoch": 1.6622390891840606,
+      "grad_norm": 1.5682717561721802,
+      "learning_rate": 1.459279149118632e-06,
+      "loss": 0.1596,
+      "step": 3504
+    },
+    {
+      "epoch": 1.6627134724857684,
+      "grad_norm": 1.7140616178512573,
+      "learning_rate": 1.455285304237114e-06,
+      "loss": 0.1499,
+      "step": 3505
+    },
+    {
+      "epoch": 1.6631878557874762,
+      "grad_norm": 2.096163749694824,
+      "learning_rate": 1.4512965032161242e-06,
+      "loss": 0.1381,
+      "step": 3506
+    },
+    {
+      "epoch": 1.663662239089184,
+      "grad_norm": 1.4052388668060303,
+      "learning_rate": 1.4473127484102157e-06,
+      "loss": 0.1269,
+      "step": 3507
+    },
+    {
+      "epoch": 1.6641366223908918,
+      "grad_norm": 1.667314052581787,
+      "learning_rate": 1.4433340421709597e-06,
+      "loss": 0.1574,
+      "step": 3508
+    },
+    {
+      "epoch": 1.6646110056925996,
+      "grad_norm": 1.8331180810928345,
+      "learning_rate": 1.4393603868469464e-06,
+      "loss": 0.1621,
+      "step": 3509
+    },
+    {
+      "epoch": 1.6650853889943074,
+      "grad_norm": 1.3572697639465332,
+      "learning_rate": 1.4353917847837883e-06,
+      "loss": 0.1127,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6655597722960152,
+      "grad_norm": 1.5188183784484863,
+      "learning_rate": 1.4314282383241097e-06,
+      "loss": 0.1292,
+      "step": 3511
+    },
+    {
+      "epoch": 1.666034155597723,
+      "grad_norm": 1.5787996053695679,
+      "learning_rate": 1.4274697498075495e-06,
+      "loss": 0.1352,
+      "step": 3512
+    },
+    {
+      "epoch": 1.6665085388994307,
+      "grad_norm": 1.8923653364181519,
+      "learning_rate": 1.423516321570767e-06,
+      "loss": 0.1811,
+      "step": 3513
+    },
+    {
+      "epoch": 1.6669829222011385,
+      "grad_norm": 1.3783034086227417,
+      "learning_rate": 1.419567955947434e-06,
+      "loss": 0.1151,
+      "step": 3514
+    },
+    {
+      "epoch": 1.6674573055028463,
+      "grad_norm": 1.7358620166778564,
+      "learning_rate": 1.4156246552682274e-06,
+      "loss": 0.178,
+      "step": 3515
+    },
+    {
+      "epoch": 1.6679316888045541,
+      "grad_norm": 1.7342180013656616,
+      "learning_rate": 1.4116864218608416e-06,
+      "loss": 0.1696,
+      "step": 3516
+    },
+    {
+      "epoch": 1.668406072106262,
+      "grad_norm": 1.8265262842178345,
+      "learning_rate": 1.4077532580499753e-06,
+      "loss": 0.1312,
+      "step": 3517
+    },
+    {
+      "epoch": 1.6688804554079697,
+      "grad_norm": 1.7043464183807373,
+      "learning_rate": 1.4038251661573387e-06,
+      "loss": 0.1454,
+      "step": 3518
+    },
+    {
+      "epoch": 1.6693548387096775,
+      "grad_norm": 1.6283822059631348,
+      "learning_rate": 1.3999021485016429e-06,
+      "loss": 0.1431,
+      "step": 3519
+    },
+    {
+      "epoch": 1.6698292220113853,
+      "grad_norm": 1.8391207456588745,
+      "learning_rate": 1.3959842073986085e-06,
+      "loss": 0.1677,
+      "step": 3520
+    },
+    {
+      "epoch": 1.670303605313093,
+      "grad_norm": 1.350468635559082,
+      "learning_rate": 1.3920713451609535e-06,
+      "loss": 0.1179,
+      "step": 3521
+    },
+    {
+      "epoch": 1.6707779886148009,
+      "grad_norm": 2.141695976257324,
+      "learning_rate": 1.3881635640984048e-06,
+      "loss": 0.1947,
+      "step": 3522
+    },
+    {
+      "epoch": 1.6712523719165087,
+      "grad_norm": 1.6811950206756592,
+      "learning_rate": 1.384260866517686e-06,
+      "loss": 0.1595,
+      "step": 3523
+    },
+    {
+      "epoch": 1.6717267552182165,
+      "grad_norm": 1.6273670196533203,
+      "learning_rate": 1.3803632547225242e-06,
+      "loss": 0.1674,
+      "step": 3524
+    },
+    {
+      "epoch": 1.6722011385199242,
+      "grad_norm": 1.6808335781097412,
+      "learning_rate": 1.376470731013636e-06,
+      "loss": 0.1413,
+      "step": 3525
+    },
+    {
+      "epoch": 1.672675521821632,
+      "grad_norm": 1.9226058721542358,
+      "learning_rate": 1.372583297688741e-06,
+      "loss": 0.1372,
+      "step": 3526
+    },
+    {
+      "epoch": 1.6731499051233396,
+      "grad_norm": 1.719659686088562,
+      "learning_rate": 1.368700957042557e-06,
+      "loss": 0.1393,
+      "step": 3527
+    },
+    {
+      "epoch": 1.6736242884250474,
+      "grad_norm": 1.4148011207580566,
+      "learning_rate": 1.3648237113667839e-06,
+      "loss": 0.144,
+      "step": 3528
+    },
+    {
+      "epoch": 1.6740986717267552,
+      "grad_norm": 1.5885848999023438,
+      "learning_rate": 1.3609515629501279e-06,
+      "loss": 0.1342,
+      "step": 3529
+    },
+    {
+      "epoch": 1.674573055028463,
+      "grad_norm": 1.828921914100647,
+      "learning_rate": 1.3570845140782752e-06,
+      "loss": 0.1664,
+      "step": 3530
+    },
+    {
+      "epoch": 1.6750474383301708,
+      "grad_norm": 1.8899797201156616,
+      "learning_rate": 1.3532225670339095e-06,
+      "loss": 0.1795,
+      "step": 3531
+    },
+    {
+      "epoch": 1.6755218216318786,
+      "grad_norm": 1.7065174579620361,
+      "learning_rate": 1.3493657240966974e-06,
+      "loss": 0.1541,
+      "step": 3532
+    },
+    {
+      "epoch": 1.6759962049335864,
+      "grad_norm": 1.6612739562988281,
+      "learning_rate": 1.3455139875433e-06,
+      "loss": 0.1717,
+      "step": 3533
+    },
+    {
+      "epoch": 1.6764705882352942,
+      "grad_norm": 2.0222203731536865,
+      "learning_rate": 1.3416673596473528e-06,
+      "loss": 0.1909,
+      "step": 3534
+    },
+    {
+      "epoch": 1.676944971537002,
+      "grad_norm": 1.3110690116882324,
+      "learning_rate": 1.3378258426794888e-06,
+      "loss": 0.1174,
+      "step": 3535
+    },
+    {
+      "epoch": 1.6774193548387095,
+      "grad_norm": 1.338375449180603,
+      "learning_rate": 1.3339894389073104e-06,
+      "loss": 0.1189,
+      "step": 3536
+    },
+    {
+      "epoch": 1.6778937381404173,
+      "grad_norm": 1.7583532333374023,
+      "learning_rate": 1.3301581505954131e-06,
+      "loss": 0.174,
+      "step": 3537
+    },
+    {
+      "epoch": 1.678368121442125,
+      "grad_norm": 1.7339435815811157,
+      "learning_rate": 1.3263319800053698e-06,
+      "loss": 0.1568,
+      "step": 3538
+    },
+    {
+      "epoch": 1.678842504743833,
+      "grad_norm": 1.5173304080963135,
+      "learning_rate": 1.3225109293957272e-06,
+      "loss": 0.1379,
+      "step": 3539
+    },
+    {
+      "epoch": 1.6793168880455407,
+      "grad_norm": 1.5127456188201904,
+      "learning_rate": 1.3186950010220156e-06,
+      "loss": 0.1355,
+      "step": 3540
+    },
+    {
+      "epoch": 1.6797912713472485,
+      "grad_norm": 1.6670554876327515,
+      "learning_rate": 1.3148841971367387e-06,
+      "loss": 0.1832,
+      "step": 3541
+    },
+    {
+      "epoch": 1.6802656546489563,
+      "grad_norm": 1.6429007053375244,
+      "learning_rate": 1.3110785199893806e-06,
+      "loss": 0.1324,
+      "step": 3542
+    },
+    {
+      "epoch": 1.680740037950664,
+      "grad_norm": 1.5100871324539185,
+      "learning_rate": 1.3072779718263884e-06,
+      "loss": 0.1413,
+      "step": 3543
+    },
+    {
+      "epoch": 1.6812144212523719,
+      "grad_norm": 1.6783661842346191,
+      "learning_rate": 1.3034825548911944e-06,
+      "loss": 0.1182,
+      "step": 3544
+    },
+    {
+      "epoch": 1.6816888045540797,
+      "grad_norm": 1.708293080329895,
+      "learning_rate": 1.29969227142419e-06,
+      "loss": 0.1654,
+      "step": 3545
+    },
+    {
+      "epoch": 1.6821631878557874,
+      "grad_norm": 1.9683902263641357,
+      "learning_rate": 1.295907123662744e-06,
+      "loss": 0.1813,
+      "step": 3546
+    },
+    {
+      "epoch": 1.6826375711574952,
+      "grad_norm": 1.7525516748428345,
+      "learning_rate": 1.2921271138411927e-06,
+      "loss": 0.1453,
+      "step": 3547
+    },
+    {
+      "epoch": 1.683111954459203,
+      "grad_norm": 1.5966471433639526,
+      "learning_rate": 1.2883522441908403e-06,
+      "loss": 0.123,
+      "step": 3548
+    },
+    {
+      "epoch": 1.6835863377609108,
+      "grad_norm": 1.6996803283691406,
+      "learning_rate": 1.2845825169399506e-06,
+      "loss": 0.1626,
+      "step": 3549
+    },
+    {
+      "epoch": 1.6840607210626186,
+      "grad_norm": 1.50482177734375,
+      "learning_rate": 1.2808179343137583e-06,
+      "loss": 0.1239,
+      "step": 3550
+    },
+    {
+      "epoch": 1.6845351043643264,
+      "grad_norm": 1.7944713830947876,
+      "learning_rate": 1.2770584985344613e-06,
+      "loss": 0.1743,
+      "step": 3551
+    },
+    {
+      "epoch": 1.6850094876660342,
+      "grad_norm": 1.602755069732666,
+      "learning_rate": 1.2733042118212157e-06,
+      "loss": 0.1602,
+      "step": 3552
+    },
+    {
+      "epoch": 1.685483870967742,
+      "grad_norm": 1.5536739826202393,
+      "learning_rate": 1.2695550763901376e-06,
+      "loss": 0.1252,
+      "step": 3553
+    },
+    {
+      "epoch": 1.6859582542694498,
+      "grad_norm": 1.7096506357192993,
+      "learning_rate": 1.2658110944543055e-06,
+      "loss": 0.1481,
+      "step": 3554
+    },
+    {
+      "epoch": 1.6864326375711576,
+      "grad_norm": 1.5002373456954956,
+      "learning_rate": 1.2620722682237575e-06,
+      "loss": 0.1237,
+      "step": 3555
+    },
+    {
+      "epoch": 1.6869070208728654,
+      "grad_norm": 1.692552924156189,
+      "learning_rate": 1.258338599905482e-06,
+      "loss": 0.176,
+      "step": 3556
+    },
+    {
+      "epoch": 1.6873814041745732,
+      "grad_norm": 1.6304970979690552,
+      "learning_rate": 1.2546100917034322e-06,
+      "loss": 0.1477,
+      "step": 3557
+    },
+    {
+      "epoch": 1.687855787476281,
+      "grad_norm": 1.4163672924041748,
+      "learning_rate": 1.2508867458185037e-06,
+      "loss": 0.111,
+      "step": 3558
+    },
+    {
+      "epoch": 1.6883301707779887,
+      "grad_norm": 1.4730753898620605,
+      "learning_rate": 1.2471685644485543e-06,
+      "loss": 0.1274,
+      "step": 3559
+    },
+    {
+      "epoch": 1.6888045540796965,
+      "grad_norm": 1.3261274099349976,
+      "learning_rate": 1.2434555497883872e-06,
+      "loss": 0.1217,
+      "step": 3560
+    },
+    {
+      "epoch": 1.6892789373814043,
+      "grad_norm": 1.5221738815307617,
+      "learning_rate": 1.239747704029758e-06,
+      "loss": 0.1276,
+      "step": 3561
+    },
+    {
+      "epoch": 1.689753320683112,
+      "grad_norm": 1.6323662996292114,
+      "learning_rate": 1.2360450293613757e-06,
+      "loss": 0.1552,
+      "step": 3562
+    },
+    {
+      "epoch": 1.6902277039848197,
+      "grad_norm": 1.506781816482544,
+      "learning_rate": 1.2323475279688869e-06,
+      "loss": 0.1304,
+      "step": 3563
+    },
+    {
+      "epoch": 1.6907020872865275,
+      "grad_norm": 1.473595380783081,
+      "learning_rate": 1.228655202034893e-06,
+      "loss": 0.1362,
+      "step": 3564
+    },
+    {
+      "epoch": 1.6911764705882353,
+      "grad_norm": 2.067612409591675,
+      "learning_rate": 1.2249680537389375e-06,
+      "loss": 0.1776,
+      "step": 3565
+    },
+    {
+      "epoch": 1.691650853889943,
+      "grad_norm": 2.5935635566711426,
+      "learning_rate": 1.2212860852575093e-06,
+      "loss": 0.1547,
+      "step": 3566
+    },
+    {
+      "epoch": 1.6921252371916509,
+      "grad_norm": 1.8970015048980713,
+      "learning_rate": 1.217609298764033e-06,
+      "loss": 0.1734,
+      "step": 3567
+    },
+    {
+      "epoch": 1.6925996204933587,
+      "grad_norm": 1.6494086980819702,
+      "learning_rate": 1.213937696428885e-06,
+      "loss": 0.1458,
+      "step": 3568
+    },
+    {
+      "epoch": 1.6930740037950665,
+      "grad_norm": 2.007638454437256,
+      "learning_rate": 1.2102712804193705e-06,
+      "loss": 0.2278,
+      "step": 3569
+    },
+    {
+      "epoch": 1.6935483870967742,
+      "grad_norm": 1.6985396146774292,
+      "learning_rate": 1.2066100528997415e-06,
+      "loss": 0.169,
+      "step": 3570
+    },
+    {
+      "epoch": 1.6940227703984818,
+      "grad_norm": 2.0554983615875244,
+      "learning_rate": 1.2029540160311859e-06,
+      "loss": 0.1328,
+      "step": 3571
+    },
+    {
+      "epoch": 1.6944971537001896,
+      "grad_norm": 1.8407206535339355,
+      "learning_rate": 1.1993031719718217e-06,
+      "loss": 0.1437,
+      "step": 3572
+    },
+    {
+      "epoch": 1.6949715370018974,
+      "grad_norm": 1.4634854793548584,
+      "learning_rate": 1.1956575228767087e-06,
+      "loss": 0.1477,
+      "step": 3573
+    },
+    {
+      "epoch": 1.6954459203036052,
+      "grad_norm": 1.489542007446289,
+      "learning_rate": 1.1920170708978374e-06,
+      "loss": 0.1353,
+      "step": 3574
+    },
+    {
+      "epoch": 1.695920303605313,
+      "grad_norm": 1.7846124172210693,
+      "learning_rate": 1.1883818181841323e-06,
+      "loss": 0.1611,
+      "step": 3575
+    },
+    {
+      "epoch": 1.6963946869070208,
+      "grad_norm": 1.4704738855361938,
+      "learning_rate": 1.1847517668814456e-06,
+      "loss": 0.1405,
+      "step": 3576
+    },
+    {
+      "epoch": 1.6968690702087286,
+      "grad_norm": 1.5131449699401855,
+      "learning_rate": 1.181126919132557e-06,
+      "loss": 0.1282,
+      "step": 3577
+    },
+    {
+      "epoch": 1.6973434535104364,
+      "grad_norm": 1.477185845375061,
+      "learning_rate": 1.1775072770771833e-06,
+      "loss": 0.1477,
+      "step": 3578
+    },
+    {
+      "epoch": 1.6978178368121442,
+      "grad_norm": 1.6830008029937744,
+      "learning_rate": 1.1738928428519603e-06,
+      "loss": 0.1419,
+      "step": 3579
+    },
+    {
+      "epoch": 1.698292220113852,
+      "grad_norm": 1.5827606916427612,
+      "learning_rate": 1.1702836185904543e-06,
+      "loss": 0.1486,
+      "step": 3580
+    },
+    {
+      "epoch": 1.6987666034155597,
+      "grad_norm": 1.622440218925476,
+      "learning_rate": 1.1666796064231566e-06,
+      "loss": 0.1392,
+      "step": 3581
+    },
+    {
+      "epoch": 1.6992409867172675,
+      "grad_norm": 1.690509557723999,
+      "learning_rate": 1.1630808084774758e-06,
+      "loss": 0.1346,
+      "step": 3582
+    },
+    {
+      "epoch": 1.6997153700189753,
+      "grad_norm": 2.058816432952881,
+      "learning_rate": 1.1594872268777513e-06,
+      "loss": 0.2435,
+      "step": 3583
+    },
+    {
+      "epoch": 1.7001897533206831,
+      "grad_norm": 1.905711054801941,
+      "learning_rate": 1.155898863745234e-06,
+      "loss": 0.1724,
+      "step": 3584
+    },
+    {
+      "epoch": 1.700664136622391,
+      "grad_norm": 1.4442384243011475,
+      "learning_rate": 1.1523157211981006e-06,
+      "loss": 0.1424,
+      "step": 3585
+    },
+    {
+      "epoch": 1.7011385199240987,
+      "grad_norm": 1.9293603897094727,
+      "learning_rate": 1.1487378013514483e-06,
+      "loss": 0.1443,
+      "step": 3586
+    },
+    {
+      "epoch": 1.7016129032258065,
+      "grad_norm": 1.8342009782791138,
+      "learning_rate": 1.145165106317282e-06,
+      "loss": 0.195,
+      "step": 3587
+    },
+    {
+      "epoch": 1.7020872865275143,
+      "grad_norm": 1.8110581636428833,
+      "learning_rate": 1.1415976382045313e-06,
+      "loss": 0.1736,
+      "step": 3588
+    },
+    {
+      "epoch": 1.702561669829222,
+      "grad_norm": 1.952837347984314,
+      "learning_rate": 1.1380353991190373e-06,
+      "loss": 0.1521,
+      "step": 3589
+    },
+    {
+      "epoch": 1.7030360531309299,
+      "grad_norm": 2.2737579345703125,
+      "learning_rate": 1.1344783911635538e-06,
+      "loss": 0.1737,
+      "step": 3590
+    },
+    {
+      "epoch": 1.7035104364326377,
+      "grad_norm": 1.83051335811615,
+      "learning_rate": 1.130926616437751e-06,
+      "loss": 0.1311,
+      "step": 3591
+    },
+    {
+      "epoch": 1.7039848197343455,
+      "grad_norm": 1.8009090423583984,
+      "learning_rate": 1.1273800770382027e-06,
+      "loss": 0.1798,
+      "step": 3592
+    },
+    {
+      "epoch": 1.7044592030360532,
+      "grad_norm": 1.824048399925232,
+      "learning_rate": 1.1238387750583945e-06,
+      "loss": 0.1588,
+      "step": 3593
+    },
+    {
+      "epoch": 1.704933586337761,
+      "grad_norm": 1.5769997835159302,
+      "learning_rate": 1.1203027125887235e-06,
+      "loss": 0.1491,
+      "step": 3594
+    },
+    {
+      "epoch": 1.7054079696394688,
+      "grad_norm": 1.6240620613098145,
+      "learning_rate": 1.1167718917164961e-06,
+      "loss": 0.1457,
+      "step": 3595
+    },
+    {
+      "epoch": 1.7058823529411766,
+      "grad_norm": 1.4127658605575562,
+      "learning_rate": 1.1132463145259144e-06,
+      "loss": 0.124,
+      "step": 3596
+    },
+    {
+      "epoch": 1.7063567362428842,
+      "grad_norm": 1.866065502166748,
+      "learning_rate": 1.1097259830980956e-06,
+      "loss": 0.1289,
+      "step": 3597
+    },
+    {
+      "epoch": 1.706831119544592,
+      "grad_norm": 2.1026611328125,
+      "learning_rate": 1.1062108995110566e-06,
+      "loss": 0.182,
+      "step": 3598
+    },
+    {
+      "epoch": 1.7073055028462998,
+      "grad_norm": 1.580645203590393,
+      "learning_rate": 1.1027010658397175e-06,
+      "loss": 0.1433,
+      "step": 3599
+    },
+    {
+      "epoch": 1.7077798861480076,
+      "grad_norm": 1.5717014074325562,
+      "learning_rate": 1.0991964841558955e-06,
+      "loss": 0.151,
+      "step": 3600
+    },
+    {
+      "epoch": 1.7082542694497154,
+      "grad_norm": 1.5343966484069824,
+      "learning_rate": 1.0956971565283114e-06,
+      "loss": 0.1533,
+      "step": 3601
+    },
+    {
+      "epoch": 1.7087286527514232,
+      "grad_norm": 1.4576107263565063,
+      "learning_rate": 1.092203085022583e-06,
+      "loss": 0.1411,
+      "step": 3602
+    },
+    {
+      "epoch": 1.709203036053131,
+      "grad_norm": 1.7388747930526733,
+      "learning_rate": 1.0887142717012266e-06,
+      "loss": 0.1765,
+      "step": 3603
+    },
+    {
+      "epoch": 1.7096774193548387,
+      "grad_norm": 1.668588638305664,
+      "learning_rate": 1.0852307186236554e-06,
+      "loss": 0.1562,
+      "step": 3604
+    },
+    {
+      "epoch": 1.7101518026565465,
+      "grad_norm": 1.5516544580459595,
+      "learning_rate": 1.0817524278461777e-06,
+      "loss": 0.1373,
+      "step": 3605
+    },
+    {
+      "epoch": 1.710626185958254,
+      "grad_norm": 1.614884614944458,
+      "learning_rate": 1.07827940142199e-06,
+      "loss": 0.1378,
+      "step": 3606
+    },
+    {
+      "epoch": 1.711100569259962,
+      "grad_norm": 1.7359789609909058,
+      "learning_rate": 1.074811641401189e-06,
+      "loss": 0.1462,
+      "step": 3607
+    },
+    {
+      "epoch": 1.7115749525616697,
+      "grad_norm": 1.587064504623413,
+      "learning_rate": 1.071349149830756e-06,
+      "loss": 0.1173,
+      "step": 3608
+    },
+    {
+      "epoch": 1.7120493358633775,
+      "grad_norm": 1.8435542583465576,
+      "learning_rate": 1.0678919287545663e-06,
+      "loss": 0.1849,
+      "step": 3609
+    },
+    {
+      "epoch": 1.7125237191650853,
+      "grad_norm": 1.614603042602539,
+      "learning_rate": 1.0644399802133876e-06,
+      "loss": 0.1441,
+      "step": 3610
+    },
+    {
+      "epoch": 1.712998102466793,
+      "grad_norm": 1.7255760431289673,
+      "learning_rate": 1.060993306244864e-06,
+      "loss": 0.1506,
+      "step": 3611
+    },
+    {
+      "epoch": 1.7134724857685009,
+      "grad_norm": 2.0267391204833984,
+      "learning_rate": 1.0575519088835374e-06,
+      "loss": 0.1878,
+      "step": 3612
+    },
+    {
+      "epoch": 1.7139468690702087,
+      "grad_norm": 1.741645336151123,
+      "learning_rate": 1.054115790160829e-06,
+      "loss": 0.1536,
+      "step": 3613
+    },
+    {
+      "epoch": 1.7144212523719164,
+      "grad_norm": 1.8468014001846313,
+      "learning_rate": 1.0506849521050478e-06,
+      "loss": 0.1775,
+      "step": 3614
+    },
+    {
+      "epoch": 1.7148956356736242,
+      "grad_norm": 2.1939375400543213,
+      "learning_rate": 1.0472593967413813e-06,
+      "loss": 0.2245,
+      "step": 3615
+    },
+    {
+      "epoch": 1.715370018975332,
+      "grad_norm": 1.7620271444320679,
+      "learning_rate": 1.0438391260919034e-06,
+      "loss": 0.1369,
+      "step": 3616
+    },
+    {
+      "epoch": 1.7158444022770398,
+      "grad_norm": 1.5060416460037231,
+      "learning_rate": 1.0404241421755623e-06,
+      "loss": 0.1334,
+      "step": 3617
+    },
+    {
+      "epoch": 1.7163187855787476,
+      "grad_norm": 1.5071996450424194,
+      "learning_rate": 1.037014447008191e-06,
+      "loss": 0.1385,
+      "step": 3618
+    },
+    {
+      "epoch": 1.7167931688804554,
+      "grad_norm": 1.741341471672058,
+      "learning_rate": 1.0336100426025008e-06,
+      "loss": 0.1588,
+      "step": 3619
+    },
+    {
+      "epoch": 1.7172675521821632,
+      "grad_norm": 1.5700851678848267,
+      "learning_rate": 1.0302109309680752e-06,
+      "loss": 0.1558,
+      "step": 3620
+    },
+    {
+      "epoch": 1.717741935483871,
+      "grad_norm": 1.5904638767242432,
+      "learning_rate": 1.0268171141113769e-06,
+      "loss": 0.1772,
+      "step": 3621
+    },
+    {
+      "epoch": 1.7182163187855788,
+      "grad_norm": 1.7748823165893555,
+      "learning_rate": 1.0234285940357424e-06,
+      "loss": 0.1702,
+      "step": 3622
+    },
+    {
+      "epoch": 1.7186907020872866,
+      "grad_norm": 1.6879323720932007,
+      "learning_rate": 1.020045372741384e-06,
+      "loss": 0.1319,
+      "step": 3623
+    },
+    {
+      "epoch": 1.7191650853889944,
+      "grad_norm": 1.3660778999328613,
+      "learning_rate": 1.0166674522253817e-06,
+      "loss": 0.1377,
+      "step": 3624
+    },
+    {
+      "epoch": 1.7196394686907022,
+      "grad_norm": 1.907179594039917,
+      "learning_rate": 1.0132948344816863e-06,
+      "loss": 0.1655,
+      "step": 3625
+    },
+    {
+      "epoch": 1.72011385199241,
+      "grad_norm": 1.814266562461853,
+      "learning_rate": 1.0099275215011227e-06,
+      "loss": 0.1645,
+      "step": 3626
+    },
+    {
+      "epoch": 1.7205882352941178,
+      "grad_norm": 1.505477786064148,
+      "learning_rate": 1.0065655152713828e-06,
+      "loss": 0.1253,
+      "step": 3627
+    },
+    {
+      "epoch": 1.7210626185958255,
+      "grad_norm": 2.314615249633789,
+      "learning_rate": 1.003208817777025e-06,
+      "loss": 0.1994,
+      "step": 3628
+    },
+    {
+      "epoch": 1.7215370018975333,
+      "grad_norm": 1.3844680786132812,
+      "learning_rate": 9.99857430999478e-07,
+      "loss": 0.1154,
+      "step": 3629
+    },
+    {
+      "epoch": 1.7220113851992411,
+      "grad_norm": 1.7867406606674194,
+      "learning_rate": 9.965113569170258e-07,
+      "loss": 0.1278,
+      "step": 3630
+    },
+    {
+      "epoch": 1.722485768500949,
+      "grad_norm": 1.7978324890136719,
+      "learning_rate": 9.931705975048279e-07,
+      "loss": 0.1922,
+      "step": 3631
+    },
+    {
+      "epoch": 1.7229601518026565,
+      "grad_norm": 1.2707836627960205,
+      "learning_rate": 9.898351547349005e-07,
+      "loss": 0.1291,
+      "step": 3632
+    },
+    {
+      "epoch": 1.7234345351043643,
+      "grad_norm": 1.5729351043701172,
+      "learning_rate": 9.865050305761226e-07,
+      "loss": 0.1654,
+      "step": 3633
+    },
+    {
+      "epoch": 1.723908918406072,
+      "grad_norm": 1.5746902227401733,
+      "learning_rate": 9.831802269942304e-07,
+      "loss": 0.1423,
+      "step": 3634
+    },
+    {
+      "epoch": 1.7243833017077799,
+      "grad_norm": 1.864458680152893,
+      "learning_rate": 9.79860745951825e-07,
+      "loss": 0.1778,
+      "step": 3635
+    },
+    {
+      "epoch": 1.7248576850094877,
+      "grad_norm": 1.7163097858428955,
+      "learning_rate": 9.765465894083637e-07,
+      "loss": 0.1507,
+      "step": 3636
+    },
+    {
+      "epoch": 1.7253320683111955,
+      "grad_norm": 1.3414804935455322,
+      "learning_rate": 9.73237759320159e-07,
+      "loss": 0.1321,
+      "step": 3637
+    },
+    {
+      "epoch": 1.7258064516129032,
+      "grad_norm": 2.13075852394104,
+      "learning_rate": 9.699342576403847e-07,
+      "loss": 0.1707,
+      "step": 3638
+    },
+    {
+      "epoch": 1.726280834914611,
+      "grad_norm": 1.949949860572815,
+      "learning_rate": 9.666360863190583e-07,
+      "loss": 0.1592,
+      "step": 3639
+    },
+    {
+      "epoch": 1.7267552182163188,
+      "grad_norm": 9.197535514831543,
+      "learning_rate": 9.633432473030635e-07,
+      "loss": 0.1525,
+      "step": 3640
+    },
+    {
+      "epoch": 1.7272296015180264,
+      "grad_norm": 1.6272482872009277,
+      "learning_rate": 9.600557425361269e-07,
+      "loss": 0.1586,
+      "step": 3641
+    },
+    {
+      "epoch": 1.7277039848197342,
+      "grad_norm": 1.5397499799728394,
+      "learning_rate": 9.567735739588314e-07,
+      "loss": 0.1292,
+      "step": 3642
+    },
+    {
+      "epoch": 1.728178368121442,
+      "grad_norm": 1.5450812578201294,
+      "learning_rate": 9.534967435086095e-07,
+      "loss": 0.1174,
+      "step": 3643
+    },
+    {
+      "epoch": 1.7286527514231498,
+      "grad_norm": 1.3916116952896118,
+      "learning_rate": 9.502252531197398e-07,
+      "loss": 0.1326,
+      "step": 3644
+    },
+    {
+      "epoch": 1.7291271347248576,
+      "grad_norm": 1.4877862930297852,
+      "learning_rate": 9.469591047233517e-07,
+      "loss": 0.1328,
+      "step": 3645
+    },
+    {
+      "epoch": 1.7296015180265654,
+      "grad_norm": 1.4843299388885498,
+      "learning_rate": 9.436983002474209e-07,
+      "loss": 0.1508,
+      "step": 3646
+    },
+    {
+      "epoch": 1.7300759013282732,
+      "grad_norm": 1.532395601272583,
+      "learning_rate": 9.404428416167688e-07,
+      "loss": 0.1403,
+      "step": 3647
+    },
+    {
+      "epoch": 1.730550284629981,
+      "grad_norm": 2.052851915359497,
+      "learning_rate": 9.371927307530593e-07,
+      "loss": 0.1483,
+      "step": 3648
+    },
+    {
+      "epoch": 1.7310246679316887,
+      "grad_norm": 1.7346041202545166,
+      "learning_rate": 9.339479695747988e-07,
+      "loss": 0.1573,
+      "step": 3649
+    },
+    {
+      "epoch": 1.7314990512333965,
+      "grad_norm": 1.8105442523956299,
+      "learning_rate": 9.307085599973387e-07,
+      "loss": 0.1597,
+      "step": 3650
+    },
+    {
+      "epoch": 1.7319734345351043,
+      "grad_norm": 2.083927631378174,
+      "learning_rate": 9.274745039328725e-07,
+      "loss": 0.164,
+      "step": 3651
+    },
+    {
+      "epoch": 1.7324478178368121,
+      "grad_norm": 1.740408182144165,
+      "learning_rate": 9.242458032904311e-07,
+      "loss": 0.1351,
+      "step": 3652
+    },
+    {
+      "epoch": 1.73292220113852,
+      "grad_norm": 1.4160617589950562,
+      "learning_rate": 9.210224599758811e-07,
+      "loss": 0.1228,
+      "step": 3653
+    },
+    {
+      "epoch": 1.7333965844402277,
+      "grad_norm": 1.3447182178497314,
+      "learning_rate": 9.178044758919336e-07,
+      "loss": 0.1153,
+      "step": 3654
+    },
+    {
+      "epoch": 1.7338709677419355,
+      "grad_norm": 2.0321731567382812,
+      "learning_rate": 9.145918529381314e-07,
+      "loss": 0.1511,
+      "step": 3655
+    },
+    {
+      "epoch": 1.7343453510436433,
+      "grad_norm": 1.9544363021850586,
+      "learning_rate": 9.113845930108567e-07,
+      "loss": 0.1599,
+      "step": 3656
+    },
+    {
+      "epoch": 1.734819734345351,
+      "grad_norm": 1.5864086151123047,
+      "learning_rate": 9.081826980033215e-07,
+      "loss": 0.1562,
+      "step": 3657
+    },
+    {
+      "epoch": 1.7352941176470589,
+      "grad_norm": 1.6874994039535522,
+      "learning_rate": 9.049861698055696e-07,
+      "loss": 0.1619,
+      "step": 3658
+    },
+    {
+      "epoch": 1.7357685009487667,
+      "grad_norm": 1.6411479711532593,
+      "learning_rate": 9.017950103044826e-07,
+      "loss": 0.1461,
+      "step": 3659
+    },
+    {
+      "epoch": 1.7362428842504745,
+      "grad_norm": 1.355292797088623,
+      "learning_rate": 8.986092213837705e-07,
+      "loss": 0.0838,
+      "step": 3660
+    },
+    {
+      "epoch": 1.7367172675521823,
+      "grad_norm": 2.7029168605804443,
+      "learning_rate": 8.954288049239734e-07,
+      "loss": 0.2356,
+      "step": 3661
+    },
+    {
+      "epoch": 1.73719165085389,
+      "grad_norm": 1.5269354581832886,
+      "learning_rate": 8.922537628024608e-07,
+      "loss": 0.1512,
+      "step": 3662
+    },
+    {
+      "epoch": 1.7376660341555978,
+      "grad_norm": 1.495399832725525,
+      "learning_rate": 8.890840968934244e-07,
+      "loss": 0.1317,
+      "step": 3663
+    },
+    {
+      "epoch": 1.7381404174573056,
+      "grad_norm": 1.3357588052749634,
+      "learning_rate": 8.859198090678923e-07,
+      "loss": 0.1108,
+      "step": 3664
+    },
+    {
+      "epoch": 1.7386148007590134,
+      "grad_norm": 1.5755971670150757,
+      "learning_rate": 8.827609011937066e-07,
+      "loss": 0.1393,
+      "step": 3665
+    },
+    {
+      "epoch": 1.7390891840607212,
+      "grad_norm": 1.437064528465271,
+      "learning_rate": 8.796073751355417e-07,
+      "loss": 0.117,
+      "step": 3666
+    },
+    {
+      "epoch": 1.739563567362429,
+      "grad_norm": 1.7386492490768433,
+      "learning_rate": 8.764592327548948e-07,
+      "loss": 0.16,
+      "step": 3667
+    },
+    {
+      "epoch": 1.7400379506641366,
+      "grad_norm": 1.099024772644043,
+      "learning_rate": 8.733164759100809e-07,
+      "loss": 0.0981,
+      "step": 3668
+    },
+    {
+      "epoch": 1.7405123339658444,
+      "grad_norm": 1.9566211700439453,
+      "learning_rate": 8.701791064562382e-07,
+      "loss": 0.157,
+      "step": 3669
+    },
+    {
+      "epoch": 1.7409867172675522,
+      "grad_norm": 1.458828091621399,
+      "learning_rate": 8.670471262453251e-07,
+      "loss": 0.1359,
+      "step": 3670
+    },
+    {
+      "epoch": 1.74146110056926,
+      "grad_norm": 1.4478492736816406,
+      "learning_rate": 8.639205371261217e-07,
+      "loss": 0.1664,
+      "step": 3671
+    },
+    {
+      "epoch": 1.7419354838709677,
+      "grad_norm": 1.6490832567214966,
+      "learning_rate": 8.607993409442173e-07,
+      "loss": 0.1427,
+      "step": 3672
+    },
+    {
+      "epoch": 1.7424098671726755,
+      "grad_norm": 3.0351619720458984,
+      "learning_rate": 8.57683539542028e-07,
+      "loss": 0.1973,
+      "step": 3673
+    },
+    {
+      "epoch": 1.7428842504743833,
+      "grad_norm": 1.4072052240371704,
+      "learning_rate": 8.54573134758776e-07,
+      "loss": 0.1351,
+      "step": 3674
+    },
+    {
+      "epoch": 1.7433586337760911,
+      "grad_norm": 1.6349687576293945,
+      "learning_rate": 8.514681284305048e-07,
+      "loss": 0.1405,
+      "step": 3675
+    },
+    {
+      "epoch": 1.7438330170777987,
+      "grad_norm": 1.8339117765426636,
+      "learning_rate": 8.483685223900706e-07,
+      "loss": 0.1787,
+      "step": 3676
+    },
+    {
+      "epoch": 1.7443074003795065,
+      "grad_norm": 1.4211890697479248,
+      "learning_rate": 8.452743184671363e-07,
+      "loss": 0.1482,
+      "step": 3677
+    },
+    {
+      "epoch": 1.7447817836812143,
+      "grad_norm": 1.6451423168182373,
+      "learning_rate": 8.421855184881822e-07,
+      "loss": 0.1287,
+      "step": 3678
+    },
+    {
+      "epoch": 1.745256166982922,
+      "grad_norm": 1.9902081489562988,
+      "learning_rate": 8.391021242764962e-07,
+      "loss": 0.1923,
+      "step": 3679
+    },
+    {
+      "epoch": 1.7457305502846299,
+      "grad_norm": 1.723975658416748,
+      "learning_rate": 8.360241376521772e-07,
+      "loss": 0.1539,
+      "step": 3680
+    },
+    {
+      "epoch": 1.7462049335863377,
+      "grad_norm": 1.7645349502563477,
+      "learning_rate": 8.329515604321281e-07,
+      "loss": 0.15,
+      "step": 3681
+    },
+    {
+      "epoch": 1.7466793168880455,
+      "grad_norm": 1.3427985906600952,
+      "learning_rate": 8.298843944300583e-07,
+      "loss": 0.113,
+      "step": 3682
+    },
+    {
+      "epoch": 1.7471537001897532,
+      "grad_norm": 1.576675295829773,
+      "learning_rate": 8.268226414564895e-07,
+      "loss": 0.1492,
+      "step": 3683
+    },
+    {
+      "epoch": 1.747628083491461,
+      "grad_norm": 1.534294605255127,
+      "learning_rate": 8.237663033187426e-07,
+      "loss": 0.135,
+      "step": 3684
+    },
+    {
+      "epoch": 1.7481024667931688,
+      "grad_norm": 1.7654008865356445,
+      "learning_rate": 8.207153818209446e-07,
+      "loss": 0.1498,
+      "step": 3685
+    },
+    {
+      "epoch": 1.7485768500948766,
+      "grad_norm": 1.6576253175735474,
+      "learning_rate": 8.176698787640247e-07,
+      "loss": 0.1322,
+      "step": 3686
+    },
+    {
+      "epoch": 1.7490512333965844,
+      "grad_norm": 1.7810031175613403,
+      "learning_rate": 8.146297959457116e-07,
+      "loss": 0.1576,
+      "step": 3687
+    },
+    {
+      "epoch": 1.7495256166982922,
+      "grad_norm": 1.609955906867981,
+      "learning_rate": 8.115951351605378e-07,
+      "loss": 0.1367,
+      "step": 3688
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7572312355041504,
+      "learning_rate": 8.085658981998312e-07,
+      "loss": 0.1378,
+      "step": 3689
+    },
+    {
+      "epoch": 1.7504743833017078,
+      "grad_norm": 1.3481042385101318,
+      "learning_rate": 8.055420868517227e-07,
+      "loss": 0.1182,
+      "step": 3690
+    },
+    {
+      "epoch": 1.7509487666034156,
+      "grad_norm": 2.354755401611328,
+      "learning_rate": 8.025237029011368e-07,
+      "loss": 0.201,
+      "step": 3691
+    },
+    {
+      "epoch": 1.7514231499051234,
+      "grad_norm": 1.7035918235778809,
+      "learning_rate": 7.995107481297948e-07,
+      "loss": 0.1286,
+      "step": 3692
+    },
+    {
+      "epoch": 1.7518975332068312,
+      "grad_norm": 1.6000009775161743,
+      "learning_rate": 7.965032243162163e-07,
+      "loss": 0.1307,
+      "step": 3693
+    },
+    {
+      "epoch": 1.752371916508539,
+      "grad_norm": 1.5351053476333618,
+      "learning_rate": 7.935011332357113e-07,
+      "loss": 0.1476,
+      "step": 3694
+    },
+    {
+      "epoch": 1.7528462998102468,
+      "grad_norm": 1.4815095663070679,
+      "learning_rate": 7.905044766603876e-07,
+      "loss": 0.1445,
+      "step": 3695
+    },
+    {
+      "epoch": 1.7533206831119545,
+      "grad_norm": 1.7617803812026978,
+      "learning_rate": 7.875132563591382e-07,
+      "loss": 0.153,
+      "step": 3696
+    },
+    {
+      "epoch": 1.7537950664136623,
+      "grad_norm": 1.4343780279159546,
+      "learning_rate": 7.845274740976527e-07,
+      "loss": 0.135,
+      "step": 3697
+    },
+    {
+      "epoch": 1.7542694497153701,
+      "grad_norm": 1.5575170516967773,
+      "learning_rate": 7.815471316384071e-07,
+      "loss": 0.1198,
+      "step": 3698
+    },
+    {
+      "epoch": 1.754743833017078,
+      "grad_norm": 1.9399551153182983,
+      "learning_rate": 7.785722307406685e-07,
+      "loss": 0.1777,
+      "step": 3699
+    },
+    {
+      "epoch": 1.7552182163187857,
+      "grad_norm": 1.8243434429168701,
+      "learning_rate": 7.756027731604943e-07,
+      "loss": 0.1456,
+      "step": 3700
+    },
+    {
+      "epoch": 1.7556925996204935,
+      "grad_norm": 1.7485800981521606,
+      "learning_rate": 7.726387606507224e-07,
+      "loss": 0.1182,
+      "step": 3701
+    },
+    {
+      "epoch": 1.7561669829222013,
+      "grad_norm": 1.8186720609664917,
+      "learning_rate": 7.696801949609811e-07,
+      "loss": 0.1679,
+      "step": 3702
+    },
+    {
+      "epoch": 1.7566413662239089,
+      "grad_norm": 2.1172068119049072,
+      "learning_rate": 7.667270778376834e-07,
+      "loss": 0.2089,
+      "step": 3703
+    },
+    {
+      "epoch": 1.7571157495256167,
+      "grad_norm": 1.8532425165176392,
+      "learning_rate": 7.637794110240259e-07,
+      "loss": 0.1866,
+      "step": 3704
+    },
+    {
+      "epoch": 1.7575901328273245,
+      "grad_norm": 1.3793625831604004,
+      "learning_rate": 7.608371962599847e-07,
+      "loss": 0.1241,
+      "step": 3705
+    },
+    {
+      "epoch": 1.7580645161290323,
+      "grad_norm": 1.5093717575073242,
+      "learning_rate": 7.579004352823205e-07,
+      "loss": 0.1404,
+      "step": 3706
+    },
+    {
+      "epoch": 1.75853889943074,
+      "grad_norm": 2.2254135608673096,
+      "learning_rate": 7.549691298245754e-07,
+      "loss": 0.1306,
+      "step": 3707
+    },
+    {
+      "epoch": 1.7590132827324478,
+      "grad_norm": 1.539953351020813,
+      "learning_rate": 7.520432816170686e-07,
+      "loss": 0.1244,
+      "step": 3708
+    },
+    {
+      "epoch": 1.7594876660341556,
+      "grad_norm": 1.3199529647827148,
+      "learning_rate": 7.491228923868999e-07,
+      "loss": 0.1135,
+      "step": 3709
+    },
+    {
+      "epoch": 1.7599620493358634,
+      "grad_norm": 1.6439963579177856,
+      "learning_rate": 7.462079638579489e-07,
+      "loss": 0.124,
+      "step": 3710
+    },
+    {
+      "epoch": 1.760436432637571,
+      "grad_norm": 1.6355787515640259,
+      "learning_rate": 7.432984977508639e-07,
+      "loss": 0.1471,
+      "step": 3711
+    },
+    {
+      "epoch": 1.7609108159392788,
+      "grad_norm": 1.7077136039733887,
+      "learning_rate": 7.4039449578308e-07,
+      "loss": 0.1542,
+      "step": 3712
+    },
+    {
+      "epoch": 1.7613851992409866,
+      "grad_norm": 1.7139383554458618,
+      "learning_rate": 7.374959596687948e-07,
+      "loss": 0.1534,
+      "step": 3713
+    },
+    {
+      "epoch": 1.7618595825426944,
+      "grad_norm": 1.5415153503417969,
+      "learning_rate": 7.346028911189895e-07,
+      "loss": 0.1537,
+      "step": 3714
+    },
+    {
+      "epoch": 1.7623339658444022,
+      "grad_norm": 1.2754297256469727,
+      "learning_rate": 7.317152918414116e-07,
+      "loss": 0.1296,
+      "step": 3715
+    },
+    {
+      "epoch": 1.76280834914611,
+      "grad_norm": 1.7097855806350708,
+      "learning_rate": 7.288331635405832e-07,
+      "loss": 0.1495,
+      "step": 3716
+    },
+    {
+      "epoch": 1.7632827324478177,
+      "grad_norm": 1.5699681043624878,
+      "learning_rate": 7.259565079177966e-07,
+      "loss": 0.1435,
+      "step": 3717
+    },
+    {
+      "epoch": 1.7637571157495255,
+      "grad_norm": 1.5883891582489014,
+      "learning_rate": 7.230853266711124e-07,
+      "loss": 0.142,
+      "step": 3718
+    },
+    {
+      "epoch": 1.7642314990512333,
+      "grad_norm": 1.6266002655029297,
+      "learning_rate": 7.202196214953616e-07,
+      "loss": 0.1283,
+      "step": 3719
+    },
+    {
+      "epoch": 1.7647058823529411,
+      "grad_norm": 1.50413978099823,
+      "learning_rate": 7.173593940821411e-07,
+      "loss": 0.1272,
+      "step": 3720
+    },
+    {
+      "epoch": 1.765180265654649,
+      "grad_norm": 1.3868768215179443,
+      "learning_rate": 7.145046461198146e-07,
+      "loss": 0.1253,
+      "step": 3721
+    },
+    {
+      "epoch": 1.7656546489563567,
+      "grad_norm": 1.3816770315170288,
+      "learning_rate": 7.11655379293511e-07,
+      "loss": 0.1432,
+      "step": 3722
+    },
+    {
+      "epoch": 1.7661290322580645,
+      "grad_norm": 1.8432587385177612,
+      "learning_rate": 7.088115952851238e-07,
+      "loss": 0.1617,
+      "step": 3723
+    },
+    {
+      "epoch": 1.7666034155597723,
+      "grad_norm": 1.6249598264694214,
+      "learning_rate": 7.059732957733145e-07,
+      "loss": 0.1462,
+      "step": 3724
+    },
+    {
+      "epoch": 1.76707779886148,
+      "grad_norm": 1.4073657989501953,
+      "learning_rate": 7.031404824334986e-07,
+      "loss": 0.126,
+      "step": 3725
+    },
+    {
+      "epoch": 1.7675521821631879,
+      "grad_norm": 1.3608930110931396,
+      "learning_rate": 7.003131569378586e-07,
+      "loss": 0.1174,
+      "step": 3726
+    },
+    {
+      "epoch": 1.7680265654648957,
+      "grad_norm": 1.6498671770095825,
+      "learning_rate": 6.97491320955338e-07,
+      "loss": 0.1563,
+      "step": 3727
+    },
+    {
+      "epoch": 1.7685009487666035,
+      "grad_norm": 1.5158390998840332,
+      "learning_rate": 6.946749761516402e-07,
+      "loss": 0.1455,
+      "step": 3728
+    },
+    {
+      "epoch": 1.7689753320683113,
+      "grad_norm": 1.5832974910736084,
+      "learning_rate": 6.918641241892243e-07,
+      "loss": 0.1263,
+      "step": 3729
+    },
+    {
+      "epoch": 1.769449715370019,
+      "grad_norm": 1.628330945968628,
+      "learning_rate": 6.890587667273064e-07,
+      "loss": 0.1484,
+      "step": 3730
+    },
+    {
+      "epoch": 1.7699240986717268,
+      "grad_norm": 1.4703549146652222,
+      "learning_rate": 6.862589054218616e-07,
+      "loss": 0.1334,
+      "step": 3731
+    },
+    {
+      "epoch": 1.7703984819734346,
+      "grad_norm": 1.8686482906341553,
+      "learning_rate": 6.834645419256225e-07,
+      "loss": 0.1633,
+      "step": 3732
+    },
+    {
+      "epoch": 1.7708728652751424,
+      "grad_norm": 1.9761866331100464,
+      "learning_rate": 6.806756778880752e-07,
+      "loss": 0.1436,
+      "step": 3733
+    },
+    {
+      "epoch": 1.7713472485768502,
+      "grad_norm": 1.775627851486206,
+      "learning_rate": 6.77892314955454e-07,
+      "loss": 0.1542,
+      "step": 3734
+    },
+    {
+      "epoch": 1.771821631878558,
+      "grad_norm": 1.536481261253357,
+      "learning_rate": 6.751144547707533e-07,
+      "loss": 0.1278,
+      "step": 3735
+    },
+    {
+      "epoch": 1.7722960151802658,
+      "grad_norm": 2.131673812866211,
+      "learning_rate": 6.723420989737184e-07,
+      "loss": 0.1947,
+      "step": 3736
+    },
+    {
+      "epoch": 1.7727703984819736,
+      "grad_norm": 1.6633423566818237,
+      "learning_rate": 6.695752492008389e-07,
+      "loss": 0.1661,
+      "step": 3737
+    },
+    {
+      "epoch": 1.7732447817836812,
+      "grad_norm": 1.718865990638733,
+      "learning_rate": 6.66813907085363e-07,
+      "loss": 0.1533,
+      "step": 3738
+    },
+    {
+      "epoch": 1.773719165085389,
+      "grad_norm": 1.7374719381332397,
+      "learning_rate": 6.64058074257281e-07,
+      "loss": 0.1821,
+      "step": 3739
+    },
+    {
+      "epoch": 1.7741935483870968,
+      "grad_norm": 1.5633835792541504,
+      "learning_rate": 6.613077523433342e-07,
+      "loss": 0.1299,
+      "step": 3740
+    },
+    {
+      "epoch": 1.7746679316888045,
+      "grad_norm": 1.8587796688079834,
+      "learning_rate": 6.585629429670115e-07,
+      "loss": 0.1701,
+      "step": 3741
+    },
+    {
+      "epoch": 1.7751423149905123,
+      "grad_norm": 1.9442617893218994,
+      "learning_rate": 6.558236477485458e-07,
+      "loss": 0.1211,
+      "step": 3742
+    },
+    {
+      "epoch": 1.7756166982922201,
+      "grad_norm": 1.8159979581832886,
+      "learning_rate": 6.53089868304917e-07,
+      "loss": 0.1365,
+      "step": 3743
+    },
+    {
+      "epoch": 1.776091081593928,
+      "grad_norm": 1.4900180101394653,
+      "learning_rate": 6.503616062498464e-07,
+      "loss": 0.1381,
+      "step": 3744
+    },
+    {
+      "epoch": 1.7765654648956357,
+      "grad_norm": 1.4936360120773315,
+      "learning_rate": 6.476388631938024e-07,
+      "loss": 0.1397,
+      "step": 3745
+    },
+    {
+      "epoch": 1.7770398481973435,
+      "grad_norm": 1.368524432182312,
+      "learning_rate": 6.449216407439906e-07,
+      "loss": 0.1033,
+      "step": 3746
+    },
+    {
+      "epoch": 1.777514231499051,
+      "grad_norm": 1.771744728088379,
+      "learning_rate": 6.422099405043613e-07,
+      "loss": 0.1609,
+      "step": 3747
+    },
+    {
+      "epoch": 1.7779886148007589,
+      "grad_norm": 1.7894208431243896,
+      "learning_rate": 6.395037640756074e-07,
+      "loss": 0.1453,
+      "step": 3748
+    },
+    {
+      "epoch": 1.7784629981024667,
+      "grad_norm": 1.6382660865783691,
+      "learning_rate": 6.368031130551533e-07,
+      "loss": 0.1232,
+      "step": 3749
+    },
+    {
+      "epoch": 1.7789373814041745,
+      "grad_norm": 1.9367256164550781,
+      "learning_rate": 6.341079890371682e-07,
+      "loss": 0.183,
+      "step": 3750
+    },
+    {
+      "epoch": 1.7794117647058822,
+      "grad_norm": 1.570556640625,
+      "learning_rate": 6.314183936125584e-07,
+      "loss": 0.1278,
+      "step": 3751
+    },
+    {
+      "epoch": 1.77988614800759,
+      "grad_norm": 1.5493059158325195,
+      "learning_rate": 6.287343283689662e-07,
+      "loss": 0.1329,
+      "step": 3752
+    },
+    {
+      "epoch": 1.7803605313092978,
+      "grad_norm": 1.3096294403076172,
+      "learning_rate": 6.260557948907664e-07,
+      "loss": 0.1195,
+      "step": 3753
+    },
+    {
+      "epoch": 1.7808349146110056,
+      "grad_norm": 2.0561397075653076,
+      "learning_rate": 6.233827947590709e-07,
+      "loss": 0.1499,
+      "step": 3754
+    },
+    {
+      "epoch": 1.7813092979127134,
+      "grad_norm": 1.4813035726547241,
+      "learning_rate": 6.207153295517265e-07,
+      "loss": 0.1271,
+      "step": 3755
+    },
+    {
+      "epoch": 1.7817836812144212,
+      "grad_norm": 1.5294699668884277,
+      "learning_rate": 6.180534008433114e-07,
+      "loss": 0.1312,
+      "step": 3756
+    },
+    {
+      "epoch": 1.782258064516129,
+      "grad_norm": 1.5780595541000366,
+      "learning_rate": 6.153970102051354e-07,
+      "loss": 0.1269,
+      "step": 3757
+    },
+    {
+      "epoch": 1.7827324478178368,
+      "grad_norm": 1.5694869756698608,
+      "learning_rate": 6.127461592052397e-07,
+      "loss": 0.1167,
+      "step": 3758
+    },
+    {
+      "epoch": 1.7832068311195446,
+      "grad_norm": 1.5196350812911987,
+      "learning_rate": 6.101008494083948e-07,
+      "loss": 0.1269,
+      "step": 3759
+    },
+    {
+      "epoch": 1.7836812144212524,
+      "grad_norm": 1.7971737384796143,
+      "learning_rate": 6.074610823761029e-07,
+      "loss": 0.1443,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7841555977229602,
+      "grad_norm": 1.4723081588745117,
+      "learning_rate": 6.048268596665896e-07,
+      "loss": 0.1194,
+      "step": 3761
+    },
+    {
+      "epoch": 1.784629981024668,
+      "grad_norm": 1.2657811641693115,
+      "learning_rate": 6.021981828348123e-07,
+      "loss": 0.1006,
+      "step": 3762
+    },
+    {
+      "epoch": 1.7851043643263758,
+      "grad_norm": 1.2274057865142822,
+      "learning_rate": 5.995750534324518e-07,
+      "loss": 0.1081,
+      "step": 3763
+    },
+    {
+      "epoch": 1.7855787476280836,
+      "grad_norm": 1.5427498817443848,
+      "learning_rate": 5.969574730079164e-07,
+      "loss": 0.1491,
+      "step": 3764
+    },
+    {
+      "epoch": 1.7860531309297913,
+      "grad_norm": 1.7528079748153687,
+      "learning_rate": 5.943454431063367e-07,
+      "loss": 0.1405,
+      "step": 3765
+    },
+    {
+      "epoch": 1.7865275142314991,
+      "grad_norm": 1.619002342224121,
+      "learning_rate": 5.917389652695693e-07,
+      "loss": 0.1516,
+      "step": 3766
+    },
+    {
+      "epoch": 1.787001897533207,
+      "grad_norm": 2.0210466384887695,
+      "learning_rate": 5.891380410361947e-07,
+      "loss": 0.1638,
+      "step": 3767
+    },
+    {
+      "epoch": 1.7874762808349147,
+      "grad_norm": 1.4195687770843506,
+      "learning_rate": 5.865426719415068e-07,
+      "loss": 0.1199,
+      "step": 3768
+    },
+    {
+      "epoch": 1.7879506641366225,
+      "grad_norm": 1.5800368785858154,
+      "learning_rate": 5.839528595175314e-07,
+      "loss": 0.1301,
+      "step": 3769
+    },
+    {
+      "epoch": 1.7884250474383303,
+      "grad_norm": 1.5409759283065796,
+      "learning_rate": 5.813686052930068e-07,
+      "loss": 0.1405,
+      "step": 3770
+    },
+    {
+      "epoch": 1.788899430740038,
+      "grad_norm": 1.754284143447876,
+      "learning_rate": 5.787899107933936e-07,
+      "loss": 0.1502,
+      "step": 3771
+    },
+    {
+      "epoch": 1.789373814041746,
+      "grad_norm": 1.586987853050232,
+      "learning_rate": 5.762167775408678e-07,
+      "loss": 0.1244,
+      "step": 3772
+    },
+    {
+      "epoch": 1.7898481973434535,
+      "grad_norm": 1.611585021018982,
+      "learning_rate": 5.736492070543265e-07,
+      "loss": 0.1423,
+      "step": 3773
+    },
+    {
+      "epoch": 1.7903225806451613,
+      "grad_norm": 1.8729734420776367,
+      "learning_rate": 5.710872008493795e-07,
+      "loss": 0.1478,
+      "step": 3774
+    },
+    {
+      "epoch": 1.790796963946869,
+      "grad_norm": 1.65080988407135,
+      "learning_rate": 5.685307604383561e-07,
+      "loss": 0.1352,
+      "step": 3775
+    },
+    {
+      "epoch": 1.7912713472485768,
+      "grad_norm": 1.5004171133041382,
+      "learning_rate": 5.659798873302968e-07,
+      "loss": 0.1132,
+      "step": 3776
+    },
+    {
+      "epoch": 1.7917457305502846,
+      "grad_norm": 1.4567018747329712,
+      "learning_rate": 5.634345830309563e-07,
+      "loss": 0.1126,
+      "step": 3777
+    },
+    {
+      "epoch": 1.7922201138519924,
+      "grad_norm": 1.566493034362793,
+      "learning_rate": 5.608948490428023e-07,
+      "loss": 0.1527,
+      "step": 3778
+    },
+    {
+      "epoch": 1.7926944971537002,
+      "grad_norm": 1.4656519889831543,
+      "learning_rate": 5.583606868650138e-07,
+      "loss": 0.1614,
+      "step": 3779
+    },
+    {
+      "epoch": 1.793168880455408,
+      "grad_norm": 1.7221026420593262,
+      "learning_rate": 5.558320979934839e-07,
+      "loss": 0.1515,
+      "step": 3780
+    },
+    {
+      "epoch": 1.7936432637571158,
+      "grad_norm": 1.484522819519043,
+      "learning_rate": 5.533090839208133e-07,
+      "loss": 0.1394,
+      "step": 3781
+    },
+    {
+      "epoch": 1.7941176470588234,
+      "grad_norm": 1.717825174331665,
+      "learning_rate": 5.507916461363094e-07,
+      "loss": 0.1472,
+      "step": 3782
+    },
+    {
+      "epoch": 1.7945920303605312,
+      "grad_norm": 1.826720952987671,
+      "learning_rate": 5.482797861259937e-07,
+      "loss": 0.1465,
+      "step": 3783
+    },
+    {
+      "epoch": 1.795066413662239,
+      "grad_norm": 1.4461770057678223,
+      "learning_rate": 5.45773505372591e-07,
+      "loss": 0.1204,
+      "step": 3784
+    },
+    {
+      "epoch": 1.7955407969639468,
+      "grad_norm": 1.7469446659088135,
+      "learning_rate": 5.43272805355537e-07,
+      "loss": 0.1501,
+      "step": 3785
+    },
+    {
+      "epoch": 1.7960151802656545,
+      "grad_norm": 1.5683156251907349,
+      "learning_rate": 5.407776875509663e-07,
+      "loss": 0.1272,
+      "step": 3786
+    },
+    {
+      "epoch": 1.7964895635673623,
+      "grad_norm": 1.7831333875656128,
+      "learning_rate": 5.382881534317231e-07,
+      "loss": 0.1567,
+      "step": 3787
+    },
+    {
+      "epoch": 1.7969639468690701,
+      "grad_norm": 1.6136335134506226,
+      "learning_rate": 5.35804204467355e-07,
+      "loss": 0.1244,
+      "step": 3788
+    },
+    {
+      "epoch": 1.797438330170778,
+      "grad_norm": 2.210510015487671,
+      "learning_rate": 5.333258421241127e-07,
+      "loss": 0.1729,
+      "step": 3789
+    },
+    {
+      "epoch": 1.7979127134724857,
+      "grad_norm": 1.5324962139129639,
+      "learning_rate": 5.308530678649504e-07,
+      "loss": 0.1455,
+      "step": 3790
+    },
+    {
+      "epoch": 1.7983870967741935,
+      "grad_norm": 2.263535737991333,
+      "learning_rate": 5.283858831495192e-07,
+      "loss": 0.2427,
+      "step": 3791
+    },
+    {
+      "epoch": 1.7988614800759013,
+      "grad_norm": 1.561944603919983,
+      "learning_rate": 5.259242894341765e-07,
+      "loss": 0.1292,
+      "step": 3792
+    },
+    {
+      "epoch": 1.799335863377609,
+      "grad_norm": 1.5755988359451294,
+      "learning_rate": 5.234682881719766e-07,
+      "loss": 0.1342,
+      "step": 3793
+    },
+    {
+      "epoch": 1.7998102466793169,
+      "grad_norm": 1.7627415657043457,
+      "learning_rate": 5.210178808126698e-07,
+      "loss": 0.1818,
+      "step": 3794
+    },
+    {
+      "epoch": 1.8002846299810247,
+      "grad_norm": 1.6085638999938965,
+      "learning_rate": 5.185730688027124e-07,
+      "loss": 0.166,
+      "step": 3795
+    },
+    {
+      "epoch": 1.8007590132827325,
+      "grad_norm": 1.764938473701477,
+      "learning_rate": 5.161338535852467e-07,
+      "loss": 0.204,
+      "step": 3796
+    },
+    {
+      "epoch": 1.8012333965844403,
+      "grad_norm": 1.4336494207382202,
+      "learning_rate": 5.137002366001209e-07,
+      "loss": 0.1188,
+      "step": 3797
+    },
+    {
+      "epoch": 1.801707779886148,
+      "grad_norm": 1.5212832689285278,
+      "learning_rate": 5.112722192838748e-07,
+      "loss": 0.1357,
+      "step": 3798
+    },
+    {
+      "epoch": 1.8021821631878558,
+      "grad_norm": 1.5787197351455688,
+      "learning_rate": 5.08849803069743e-07,
+      "loss": 0.1273,
+      "step": 3799
+    },
+    {
+      "epoch": 1.8026565464895636,
+      "grad_norm": 1.7871791124343872,
+      "learning_rate": 5.064329893876541e-07,
+      "loss": 0.1423,
+      "step": 3800
+    },
+    {
+      "epoch": 1.8031309297912714,
+      "grad_norm": 1.4555021524429321,
+      "learning_rate": 5.0402177966423e-07,
+      "loss": 0.1356,
+      "step": 3801
+    },
+    {
+      "epoch": 1.8036053130929792,
+      "grad_norm": 1.40778648853302,
+      "learning_rate": 5.016161753227799e-07,
+      "loss": 0.1366,
+      "step": 3802
+    },
+    {
+      "epoch": 1.804079696394687,
+      "grad_norm": 1.7542588710784912,
+      "learning_rate": 4.992161777833116e-07,
+      "loss": 0.1631,
+      "step": 3803
+    },
+    {
+      "epoch": 1.8045540796963948,
+      "grad_norm": 1.4451662302017212,
+      "learning_rate": 4.968217884625182e-07,
+      "loss": 0.1096,
+      "step": 3804
+    },
+    {
+      "epoch": 1.8050284629981026,
+      "grad_norm": 1.571986198425293,
+      "learning_rate": 4.944330087737881e-07,
+      "loss": 0.1449,
+      "step": 3805
+    },
+    {
+      "epoch": 1.8055028462998104,
+      "grad_norm": 1.793839693069458,
+      "learning_rate": 4.92049840127189e-07,
+      "loss": 0.1494,
+      "step": 3806
+    },
+    {
+      "epoch": 1.8059772296015182,
+      "grad_norm": 1.783685564994812,
+      "learning_rate": 4.896722839294843e-07,
+      "loss": 0.1737,
+      "step": 3807
+    },
+    {
+      "epoch": 1.8064516129032258,
+      "grad_norm": 1.5244357585906982,
+      "learning_rate": 4.873003415841215e-07,
+      "loss": 0.1348,
+      "step": 3808
+    },
+    {
+      "epoch": 1.8069259962049335,
+      "grad_norm": 1.776548147201538,
+      "learning_rate": 4.849340144912363e-07,
+      "loss": 0.1773,
+      "step": 3809
+    },
+    {
+      "epoch": 1.8074003795066413,
+      "grad_norm": 1.6930649280548096,
+      "learning_rate": 4.825733040476465e-07,
+      "loss": 0.1131,
+      "step": 3810
+    },
+    {
+      "epoch": 1.8078747628083491,
+      "grad_norm": 1.6838048696517944,
+      "learning_rate": 4.802182116468556e-07,
+      "loss": 0.1474,
+      "step": 3811
+    },
+    {
+      "epoch": 1.808349146110057,
+      "grad_norm": 1.8660080432891846,
+      "learning_rate": 4.778687386790515e-07,
+      "loss": 0.1772,
+      "step": 3812
+    },
+    {
+      "epoch": 1.8088235294117647,
+      "grad_norm": 1.328919768333435,
+      "learning_rate": 4.7552488653110575e-07,
+      "loss": 0.1042,
+      "step": 3813
+    },
+    {
+      "epoch": 1.8092979127134725,
+      "grad_norm": 1.7662233114242554,
+      "learning_rate": 4.731866565865717e-07,
+      "loss": 0.1594,
+      "step": 3814
+    },
+    {
+      "epoch": 1.8097722960151803,
+      "grad_norm": 1.4320173263549805,
+      "learning_rate": 4.7085405022568196e-07,
+      "loss": 0.1277,
+      "step": 3815
+    },
+    {
+      "epoch": 1.810246679316888,
+      "grad_norm": 1.8004564046859741,
+      "learning_rate": 4.685270688253507e-07,
+      "loss": 0.1586,
+      "step": 3816
+    },
+    {
+      "epoch": 1.8107210626185957,
+      "grad_norm": 1.5289167165756226,
+      "learning_rate": 4.6620571375917356e-07,
+      "loss": 0.1296,
+      "step": 3817
+    },
+    {
+      "epoch": 1.8111954459203035,
+      "grad_norm": 1.3791871070861816,
+      "learning_rate": 4.638899863974222e-07,
+      "loss": 0.1142,
+      "step": 3818
+    },
+    {
+      "epoch": 1.8116698292220113,
+      "grad_norm": 1.5164226293563843,
+      "learning_rate": 4.615798881070499e-07,
+      "loss": 0.157,
+      "step": 3819
+    },
+    {
+      "epoch": 1.812144212523719,
+      "grad_norm": 1.3034436702728271,
+      "learning_rate": 4.5927542025168025e-07,
+      "loss": 0.1208,
+      "step": 3820
+    },
+    {
+      "epoch": 1.8126185958254268,
+      "grad_norm": 1.3577231168746948,
+      "learning_rate": 4.5697658419162183e-07,
+      "loss": 0.1204,
+      "step": 3821
+    },
+    {
+      "epoch": 1.8130929791271346,
+      "grad_norm": 1.599459171295166,
+      "learning_rate": 4.5468338128385247e-07,
+      "loss": 0.1329,
+      "step": 3822
+    },
+    {
+      "epoch": 1.8135673624288424,
+      "grad_norm": 1.78299081325531,
+      "learning_rate": 4.523958128820283e-07,
+      "loss": 0.1555,
+      "step": 3823
+    },
+    {
+      "epoch": 1.8140417457305502,
+      "grad_norm": 2.62516450881958,
+      "learning_rate": 4.501138803364802e-07,
+      "loss": 0.2332,
+      "step": 3824
+    },
+    {
+      "epoch": 1.814516129032258,
+      "grad_norm": 1.7529395818710327,
+      "learning_rate": 4.478375849942063e-07,
+      "loss": 0.1568,
+      "step": 3825
+    },
+    {
+      "epoch": 1.8149905123339658,
+      "grad_norm": 1.5792466402053833,
+      "learning_rate": 4.4556692819888504e-07,
+      "loss": 0.1675,
+      "step": 3826
+    },
+    {
+      "epoch": 1.8154648956356736,
+      "grad_norm": 1.464310646057129,
+      "learning_rate": 4.4330191129085873e-07,
+      "loss": 0.1283,
+      "step": 3827
+    },
+    {
+      "epoch": 1.8159392789373814,
+      "grad_norm": 1.2922571897506714,
+      "learning_rate": 4.4104253560714794e-07,
+      "loss": 0.1136,
+      "step": 3828
+    },
+    {
+      "epoch": 1.8164136622390892,
+      "grad_norm": 1.4315319061279297,
+      "learning_rate": 4.3878880248143904e-07,
+      "loss": 0.1296,
+      "step": 3829
+    },
+    {
+      "epoch": 1.816888045540797,
+      "grad_norm": 1.5491856336593628,
+      "learning_rate": 4.3654071324408685e-07,
+      "loss": 0.1488,
+      "step": 3830
+    },
+    {
+      "epoch": 1.8173624288425048,
+      "grad_norm": 1.3317991495132446,
+      "learning_rate": 4.342982692221165e-07,
+      "loss": 0.1164,
+      "step": 3831
+    },
+    {
+      "epoch": 1.8178368121442126,
+      "grad_norm": 1.8910882472991943,
+      "learning_rate": 4.3206147173922133e-07,
+      "loss": 0.176,
+      "step": 3832
+    },
+    {
+      "epoch": 1.8183111954459203,
+      "grad_norm": 1.317647933959961,
+      "learning_rate": 4.298303221157618e-07,
+      "loss": 0.1262,
+      "step": 3833
+    },
+    {
+      "epoch": 1.8187855787476281,
+      "grad_norm": 1.5027046203613281,
+      "learning_rate": 4.276048216687634e-07,
+      "loss": 0.1289,
+      "step": 3834
+    },
+    {
+      "epoch": 1.819259962049336,
+      "grad_norm": 1.5207945108413696,
+      "learning_rate": 4.25384971711913e-07,
+      "loss": 0.1304,
+      "step": 3835
+    },
+    {
+      "epoch": 1.8197343453510437,
+      "grad_norm": 2.8607802391052246,
+      "learning_rate": 4.231707735555701e-07,
+      "loss": 0.1902,
+      "step": 3836
+    },
+    {
+      "epoch": 1.8202087286527515,
+      "grad_norm": 1.8455995321273804,
+      "learning_rate": 4.209622285067516e-07,
+      "loss": 0.1518,
+      "step": 3837
+    },
+    {
+      "epoch": 1.8206831119544593,
+      "grad_norm": 1.3427563905715942,
+      "learning_rate": 4.187593378691435e-07,
+      "loss": 0.1137,
+      "step": 3838
+    },
+    {
+      "epoch": 1.821157495256167,
+      "grad_norm": 1.9709796905517578,
+      "learning_rate": 4.165621029430855e-07,
+      "loss": 0.1531,
+      "step": 3839
+    },
+    {
+      "epoch": 1.821631878557875,
+      "grad_norm": 2.085360527038574,
+      "learning_rate": 4.1437052502558693e-07,
+      "loss": 0.1924,
+      "step": 3840
+    },
+    {
+      "epoch": 1.8221062618595827,
+      "grad_norm": 1.684356927871704,
+      "learning_rate": 4.1218460541031404e-07,
+      "loss": 0.1503,
+      "step": 3841
+    },
+    {
+      "epoch": 1.8225806451612905,
+      "grad_norm": 1.6647909879684448,
+      "learning_rate": 4.1000434538759235e-07,
+      "loss": 0.1498,
+      "step": 3842
+    },
+    {
+      "epoch": 1.823055028462998,
+      "grad_norm": 1.9033278226852417,
+      "learning_rate": 4.078297462444092e-07,
+      "loss": 0.1553,
+      "step": 3843
+    },
+    {
+      "epoch": 1.8235294117647058,
+      "grad_norm": 1.4398269653320312,
+      "learning_rate": 4.0566080926440765e-07,
+      "loss": 0.128,
+      "step": 3844
+    },
+    {
+      "epoch": 1.8240037950664136,
+      "grad_norm": 1.634743571281433,
+      "learning_rate": 4.034975357278903e-07,
+      "loss": 0.1454,
+      "step": 3845
+    },
+    {
+      "epoch": 1.8244781783681214,
+      "grad_norm": 1.7548015117645264,
+      "learning_rate": 4.013399269118157e-07,
+      "loss": 0.1535,
+      "step": 3846
+    },
+    {
+      "epoch": 1.8249525616698292,
+      "grad_norm": 1.925663948059082,
+      "learning_rate": 3.991879840897994e-07,
+      "loss": 0.1483,
+      "step": 3847
+    },
+    {
+      "epoch": 1.825426944971537,
+      "grad_norm": 1.2624423503875732,
+      "learning_rate": 3.970417085321143e-07,
+      "loss": 0.1018,
+      "step": 3848
+    },
+    {
+      "epoch": 1.8259013282732448,
+      "grad_norm": 1.4971909523010254,
+      "learning_rate": 3.949011015056803e-07,
+      "loss": 0.1474,
+      "step": 3849
+    },
+    {
+      "epoch": 1.8263757115749526,
+      "grad_norm": 1.7054779529571533,
+      "learning_rate": 3.9276616427408207e-07,
+      "loss": 0.1422,
+      "step": 3850
+    },
+    {
+      "epoch": 1.8268500948766604,
+      "grad_norm": 1.506289005279541,
+      "learning_rate": 3.9063689809754837e-07,
+      "loss": 0.1337,
+      "step": 3851
+    },
+    {
+      "epoch": 1.827324478178368,
+      "grad_norm": 1.4159045219421387,
+      "learning_rate": 3.8851330423296476e-07,
+      "loss": 0.1259,
+      "step": 3852
+    },
+    {
+      "epoch": 1.8277988614800758,
+      "grad_norm": 1.5308058261871338,
+      "learning_rate": 3.8639538393386854e-07,
+      "loss": 0.1109,
+      "step": 3853
+    },
+    {
+      "epoch": 1.8282732447817835,
+      "grad_norm": 1.6169154644012451,
+      "learning_rate": 3.842831384504453e-07,
+      "loss": 0.1385,
+      "step": 3854
+    },
+    {
+      "epoch": 1.8287476280834913,
+      "grad_norm": 1.4959263801574707,
+      "learning_rate": 3.821765690295343e-07,
+      "loss": 0.1364,
+      "step": 3855
+    },
+    {
+      "epoch": 1.8292220113851991,
+      "grad_norm": 1.550232172012329,
+      "learning_rate": 3.8007567691462187e-07,
+      "loss": 0.1415,
+      "step": 3856
+    },
+    {
+      "epoch": 1.829696394686907,
+      "grad_norm": 1.8426754474639893,
+      "learning_rate": 3.77980463345845e-07,
+      "loss": 0.1443,
+      "step": 3857
+    },
+    {
+      "epoch": 1.8301707779886147,
+      "grad_norm": 1.846116065979004,
+      "learning_rate": 3.758909295599877e-07,
+      "loss": 0.1743,
+      "step": 3858
+    },
+    {
+      "epoch": 1.8306451612903225,
+      "grad_norm": 2.2521839141845703,
+      "learning_rate": 3.738070767904778e-07,
+      "loss": 0.1485,
+      "step": 3859
+    },
+    {
+      "epoch": 1.8311195445920303,
+      "grad_norm": 1.8919899463653564,
+      "learning_rate": 3.7172890626739566e-07,
+      "loss": 0.1651,
+      "step": 3860
+    },
+    {
+      "epoch": 1.831593927893738,
+      "grad_norm": 1.477987289428711,
+      "learning_rate": 3.696564192174645e-07,
+      "loss": 0.1399,
+      "step": 3861
+    },
+    {
+      "epoch": 1.8320683111954459,
+      "grad_norm": 1.4697281122207642,
+      "learning_rate": 3.675896168640536e-07,
+      "loss": 0.114,
+      "step": 3862
+    },
+    {
+      "epoch": 1.8325426944971537,
+      "grad_norm": 1.2176628112792969,
+      "learning_rate": 3.655285004271747e-07,
+      "loss": 0.1069,
+      "step": 3863
+    },
+    {
+      "epoch": 1.8330170777988615,
+      "grad_norm": 1.5733720064163208,
+      "learning_rate": 3.634730711234835e-07,
+      "loss": 0.1506,
+      "step": 3864
+    },
+    {
+      "epoch": 1.8334914611005693,
+      "grad_norm": 2.247699737548828,
+      "learning_rate": 3.6142333016628286e-07,
+      "loss": 0.1617,
+      "step": 3865
+    },
+    {
+      "epoch": 1.833965844402277,
+      "grad_norm": 1.3761510848999023,
+      "learning_rate": 3.593792787655115e-07,
+      "loss": 0.1083,
+      "step": 3866
+    },
+    {
+      "epoch": 1.8344402277039848,
+      "grad_norm": 1.6789445877075195,
+      "learning_rate": 3.573409181277554e-07,
+      "loss": 0.1403,
+      "step": 3867
+    },
+    {
+      "epoch": 1.8349146110056926,
+      "grad_norm": 1.8942904472351074,
+      "learning_rate": 3.553082494562354e-07,
+      "loss": 0.1702,
+      "step": 3868
+    },
+    {
+      "epoch": 1.8353889943074004,
+      "grad_norm": 1.7401469945907593,
+      "learning_rate": 3.5328127395081736e-07,
+      "loss": 0.1493,
+      "step": 3869
+    },
+    {
+      "epoch": 1.8358633776091082,
+      "grad_norm": 1.4459508657455444,
+      "learning_rate": 3.5125999280800517e-07,
+      "loss": 0.1332,
+      "step": 3870
+    },
+    {
+      "epoch": 1.836337760910816,
+      "grad_norm": 1.4247829914093018,
+      "learning_rate": 3.492444072209411e-07,
+      "loss": 0.1384,
+      "step": 3871
+    },
+    {
+      "epoch": 1.8368121442125238,
+      "grad_norm": 1.431828260421753,
+      "learning_rate": 3.472345183794046e-07,
+      "loss": 0.1207,
+      "step": 3872
+    },
+    {
+      "epoch": 1.8372865275142316,
+      "grad_norm": 1.6826831102371216,
+      "learning_rate": 3.4523032746981434e-07,
+      "loss": 0.1252,
+      "step": 3873
+    },
+    {
+      "epoch": 1.8377609108159394,
+      "grad_norm": 1.5859261751174927,
+      "learning_rate": 3.43231835675224e-07,
+      "loss": 0.1273,
+      "step": 3874
+    },
+    {
+      "epoch": 1.8382352941176472,
+      "grad_norm": 1.906586766242981,
+      "learning_rate": 3.4123904417532325e-07,
+      "loss": 0.1509,
+      "step": 3875
+    },
+    {
+      "epoch": 1.838709677419355,
+      "grad_norm": 1.7677314281463623,
+      "learning_rate": 3.3925195414643677e-07,
+      "loss": 0.1258,
+      "step": 3876
+    },
+    {
+      "epoch": 1.8391840607210628,
+      "grad_norm": 1.9575536251068115,
+      "learning_rate": 3.37270566761525e-07,
+      "loss": 0.1903,
+      "step": 3877
+    },
+    {
+      "epoch": 1.8396584440227703,
+      "grad_norm": 1.5801265239715576,
+      "learning_rate": 3.3529488319017924e-07,
+      "loss": 0.123,
+      "step": 3878
+    },
+    {
+      "epoch": 1.8401328273244781,
+      "grad_norm": 1.6547151803970337,
+      "learning_rate": 3.3332490459862865e-07,
+      "loss": 0.1577,
+      "step": 3879
+    },
+    {
+      "epoch": 1.840607210626186,
+      "grad_norm": 1.688603401184082,
+      "learning_rate": 3.313606321497309e-07,
+      "loss": 0.1416,
+      "step": 3880
+    },
+    {
+      "epoch": 1.8410815939278937,
+      "grad_norm": 1.4524204730987549,
+      "learning_rate": 3.294020670029785e-07,
+      "loss": 0.1121,
+      "step": 3881
+    },
+    {
+      "epoch": 1.8415559772296015,
+      "grad_norm": 1.886781930923462,
+      "learning_rate": 3.2744921031448997e-07,
+      "loss": 0.1375,
+      "step": 3882
+    },
+    {
+      "epoch": 1.8420303605313093,
+      "grad_norm": 1.9067200422286987,
+      "learning_rate": 3.255020632370176e-07,
+      "loss": 0.182,
+      "step": 3883
+    },
+    {
+      "epoch": 1.842504743833017,
+      "grad_norm": 1.91090989112854,
+      "learning_rate": 3.235606269199454e-07,
+      "loss": 0.1711,
+      "step": 3884
+    },
+    {
+      "epoch": 1.842979127134725,
+      "grad_norm": 2.2755870819091797,
+      "learning_rate": 3.2162490250928103e-07,
+      "loss": 0.1389,
+      "step": 3885
+    },
+    {
+      "epoch": 1.8434535104364327,
+      "grad_norm": 1.7041386365890503,
+      "learning_rate": 3.1969489114766715e-07,
+      "loss": 0.1558,
+      "step": 3886
+    },
+    {
+      "epoch": 1.8439278937381403,
+      "grad_norm": 1.5372099876403809,
+      "learning_rate": 3.1777059397436693e-07,
+      "loss": 0.1196,
+      "step": 3887
+    },
+    {
+      "epoch": 1.844402277039848,
+      "grad_norm": 1.4192637205123901,
+      "learning_rate": 3.1585201212527507e-07,
+      "loss": 0.1458,
+      "step": 3888
+    },
+    {
+      "epoch": 1.8448766603415558,
+      "grad_norm": 2.2096190452575684,
+      "learning_rate": 3.1393914673291335e-07,
+      "loss": 0.1621,
+      "step": 3889
+    },
+    {
+      "epoch": 1.8453510436432636,
+      "grad_norm": 1.54441499710083,
+      "learning_rate": 3.120319989264242e-07,
+      "loss": 0.1255,
+      "step": 3890
+    },
+    {
+      "epoch": 1.8458254269449714,
+      "grad_norm": 1.6018781661987305,
+      "learning_rate": 3.101305698315815e-07,
+      "loss": 0.1354,
+      "step": 3891
+    },
+    {
+      "epoch": 1.8462998102466792,
+      "grad_norm": 2.1039795875549316,
+      "learning_rate": 3.082348605707752e-07,
+      "loss": 0.1501,
+      "step": 3892
+    },
+    {
+      "epoch": 1.846774193548387,
+      "grad_norm": 1.5671883821487427,
+      "learning_rate": 3.06344872263028e-07,
+      "loss": 0.1483,
+      "step": 3893
+    },
+    {
+      "epoch": 1.8472485768500948,
+      "grad_norm": 1.5464131832122803,
+      "learning_rate": 3.0446060602397965e-07,
+      "loss": 0.1455,
+      "step": 3894
+    },
+    {
+      "epoch": 1.8477229601518026,
+      "grad_norm": 1.674782156944275,
+      "learning_rate": 3.0258206296589487e-07,
+      "loss": 0.158,
+      "step": 3895
+    },
+    {
+      "epoch": 1.8481973434535104,
+      "grad_norm": 1.4829376935958862,
+      "learning_rate": 3.007092441976567e-07,
+      "loss": 0.1428,
+      "step": 3896
+    },
+    {
+      "epoch": 1.8486717267552182,
+      "grad_norm": 1.6832263469696045,
+      "learning_rate": 2.988421508247741e-07,
+      "loss": 0.1594,
+      "step": 3897
+    },
+    {
+      "epoch": 1.849146110056926,
+      "grad_norm": 1.5657440423965454,
+      "learning_rate": 2.9698078394937325e-07,
+      "loss": 0.1129,
+      "step": 3898
+    },
+    {
+      "epoch": 1.8496204933586338,
+      "grad_norm": 1.9195455312728882,
+      "learning_rate": 2.951251446701997e-07,
+      "loss": 0.1411,
+      "step": 3899
+    },
+    {
+      "epoch": 1.8500948766603416,
+      "grad_norm": 1.7796403169631958,
+      "learning_rate": 2.932752340826195e-07,
+      "loss": 0.1521,
+      "step": 3900
+    },
+    {
+      "epoch": 1.8505692599620494,
+      "grad_norm": 1.448384404182434,
+      "learning_rate": 2.914310532786158e-07,
+      "loss": 0.1057,
+      "step": 3901
+    },
+    {
+      "epoch": 1.8510436432637571,
+      "grad_norm": 1.8354464769363403,
+      "learning_rate": 2.8959260334679107e-07,
+      "loss": 0.1109,
+      "step": 3902
+    },
+    {
+      "epoch": 1.851518026565465,
+      "grad_norm": 1.9993449449539185,
+      "learning_rate": 2.87759885372364e-07,
+      "loss": 0.194,
+      "step": 3903
+    },
+    {
+      "epoch": 1.8519924098671727,
+      "grad_norm": 1.7258479595184326,
+      "learning_rate": 2.859329004371703e-07,
+      "loss": 0.1546,
+      "step": 3904
+    },
+    {
+      "epoch": 1.8524667931688805,
+      "grad_norm": 1.6797391176223755,
+      "learning_rate": 2.8411164961966164e-07,
+      "loss": 0.1433,
+      "step": 3905
+    },
+    {
+      "epoch": 1.8529411764705883,
+      "grad_norm": 1.4090880155563354,
+      "learning_rate": 2.8229613399490265e-07,
+      "loss": 0.1238,
+      "step": 3906
+    },
+    {
+      "epoch": 1.853415559772296,
+      "grad_norm": 1.9143239259719849,
+      "learning_rate": 2.8048635463457485e-07,
+      "loss": 0.1373,
+      "step": 3907
+    },
+    {
+      "epoch": 1.853889943074004,
+      "grad_norm": 1.5854467153549194,
+      "learning_rate": 2.7868231260697267e-07,
+      "loss": 0.1538,
+      "step": 3908
+    },
+    {
+      "epoch": 1.8543643263757117,
+      "grad_norm": 1.3684797286987305,
+      "learning_rate": 2.768840089770053e-07,
+      "loss": 0.0993,
+      "step": 3909
+    },
+    {
+      "epoch": 1.8548387096774195,
+      "grad_norm": 1.3780193328857422,
+      "learning_rate": 2.750914448061925e-07,
+      "loss": 0.0877,
+      "step": 3910
+    },
+    {
+      "epoch": 1.8553130929791273,
+      "grad_norm": 1.7516002655029297,
+      "learning_rate": 2.73304621152668e-07,
+      "loss": 0.1584,
+      "step": 3911
+    },
+    {
+      "epoch": 1.855787476280835,
+      "grad_norm": 2.2483954429626465,
+      "learning_rate": 2.7152353907117566e-07,
+      "loss": 0.2224,
+      "step": 3912
+    },
+    {
+      "epoch": 1.8562618595825426,
+      "grad_norm": 1.5955204963684082,
+      "learning_rate": 2.697481996130713e-07,
+      "loss": 0.127,
+      "step": 3913
+    },
+    {
+      "epoch": 1.8567362428842504,
+      "grad_norm": 1.3224315643310547,
+      "learning_rate": 2.6797860382631993e-07,
+      "loss": 0.1119,
+      "step": 3914
+    },
+    {
+      "epoch": 1.8572106261859582,
+      "grad_norm": 1.8325411081314087,
+      "learning_rate": 2.6621475275549593e-07,
+      "loss": 0.1764,
+      "step": 3915
+    },
+    {
+      "epoch": 1.857685009487666,
+      "grad_norm": 1.426345705986023,
+      "learning_rate": 2.644566474417831e-07,
+      "loss": 0.1312,
+      "step": 3916
+    },
+    {
+      "epoch": 1.8581593927893738,
+      "grad_norm": 1.5736240148544312,
+      "learning_rate": 2.627042889229736e-07,
+      "loss": 0.1433,
+      "step": 3917
+    },
+    {
+      "epoch": 1.8586337760910816,
+      "grad_norm": 1.5317918062210083,
+      "learning_rate": 2.609576782334688e-07,
+      "loss": 0.1323,
+      "step": 3918
+    },
+    {
+      "epoch": 1.8591081593927894,
+      "grad_norm": 1.8912248611450195,
+      "learning_rate": 2.592168164042741e-07,
+      "loss": 0.14,
+      "step": 3919
+    },
+    {
+      "epoch": 1.8595825426944972,
+      "grad_norm": 1.658992052078247,
+      "learning_rate": 2.57481704463004e-07,
+      "loss": 0.1498,
+      "step": 3920
+    },
+    {
+      "epoch": 1.860056925996205,
+      "grad_norm": 1.9872993230819702,
+      "learning_rate": 2.5575234343387603e-07,
+      "loss": 0.1534,
+      "step": 3921
+    },
+    {
+      "epoch": 1.8605313092979125,
+      "grad_norm": 1.5195602178573608,
+      "learning_rate": 2.5402873433771793e-07,
+      "loss": 0.1353,
+      "step": 3922
+    },
+    {
+      "epoch": 1.8610056925996203,
+      "grad_norm": 1.7144728899002075,
+      "learning_rate": 2.52310878191957e-07,
+      "loss": 0.1466,
+      "step": 3923
+    },
+    {
+      "epoch": 1.8614800759013281,
+      "grad_norm": 1.6180931329727173,
+      "learning_rate": 2.5059877601062655e-07,
+      "loss": 0.1519,
+      "step": 3924
+    },
+    {
+      "epoch": 1.861954459203036,
+      "grad_norm": 1.5871514081954956,
+      "learning_rate": 2.488924288043648e-07,
+      "loss": 0.1451,
+      "step": 3925
+    },
+    {
+      "epoch": 1.8624288425047437,
+      "grad_norm": 1.9041343927383423,
+      "learning_rate": 2.4719183758041056e-07,
+      "loss": 0.1313,
+      "step": 3926
+    },
+    {
+      "epoch": 1.8629032258064515,
+      "grad_norm": 1.3394379615783691,
+      "learning_rate": 2.454970033426052e-07,
+      "loss": 0.1156,
+      "step": 3927
+    },
+    {
+      "epoch": 1.8633776091081593,
+      "grad_norm": 1.8027927875518799,
+      "learning_rate": 2.4380792709139513e-07,
+      "loss": 0.1709,
+      "step": 3928
+    },
+    {
+      "epoch": 1.863851992409867,
+      "grad_norm": 1.628211498260498,
+      "learning_rate": 2.4212460982382503e-07,
+      "loss": 0.1529,
+      "step": 3929
+    },
+    {
+      "epoch": 1.864326375711575,
+      "grad_norm": 1.6816056966781616,
+      "learning_rate": 2.4044705253353897e-07,
+      "loss": 0.171,
+      "step": 3930
+    },
+    {
+      "epoch": 1.8648007590132827,
+      "grad_norm": 1.5056520700454712,
+      "learning_rate": 2.387752562107826e-07,
+      "loss": 0.1233,
+      "step": 3931
+    },
+    {
+      "epoch": 1.8652751423149905,
+      "grad_norm": 1.4784669876098633,
+      "learning_rate": 2.3710922184239983e-07,
+      "loss": 0.103,
+      "step": 3932
+    },
+    {
+      "epoch": 1.8657495256166983,
+      "grad_norm": 1.5319194793701172,
+      "learning_rate": 2.3544895041183736e-07,
+      "loss": 0.131,
+      "step": 3933
+    },
+    {
+      "epoch": 1.866223908918406,
+      "grad_norm": 1.5678507089614868,
+      "learning_rate": 2.3379444289913344e-07,
+      "loss": 0.1255,
+      "step": 3934
+    },
+    {
+      "epoch": 1.8666982922201139,
+      "grad_norm": 1.686842441558838,
+      "learning_rate": 2.321457002809302e-07,
+      "loss": 0.1496,
+      "step": 3935
+    },
+    {
+      "epoch": 1.8671726755218216,
+      "grad_norm": 1.3267841339111328,
+      "learning_rate": 2.3050272353046244e-07,
+      "loss": 0.0997,
+      "step": 3936
+    },
+    {
+      "epoch": 1.8676470588235294,
+      "grad_norm": 1.3671629428863525,
+      "learning_rate": 2.2886551361756326e-07,
+      "loss": 0.1393,
+      "step": 3937
+    },
+    {
+      "epoch": 1.8681214421252372,
+      "grad_norm": 1.4104747772216797,
+      "learning_rate": 2.2723407150866295e-07,
+      "loss": 0.117,
+      "step": 3938
+    },
+    {
+      "epoch": 1.868595825426945,
+      "grad_norm": 1.6997783184051514,
+      "learning_rate": 2.2560839816678447e-07,
+      "loss": 0.1638,
+      "step": 3939
+    },
+    {
+      "epoch": 1.8690702087286528,
+      "grad_norm": 1.3959341049194336,
+      "learning_rate": 2.2398849455154693e-07,
+      "loss": 0.1404,
+      "step": 3940
+    },
+    {
+      "epoch": 1.8695445920303606,
+      "grad_norm": 1.8219751119613647,
+      "learning_rate": 2.2237436161916204e-07,
+      "loss": 0.163,
+      "step": 3941
+    },
+    {
+      "epoch": 1.8700189753320684,
+      "grad_norm": 1.9857854843139648,
+      "learning_rate": 2.2076600032243766e-07,
+      "loss": 0.1881,
+      "step": 3942
+    },
+    {
+      "epoch": 1.8704933586337762,
+      "grad_norm": 1.7564761638641357,
+      "learning_rate": 2.1916341161077547e-07,
+      "loss": 0.1525,
+      "step": 3943
+    },
+    {
+      "epoch": 1.870967741935484,
+      "grad_norm": 1.8521637916564941,
+      "learning_rate": 2.175665964301643e-07,
+      "loss": 0.16,
+      "step": 3944
+    },
+    {
+      "epoch": 1.8714421252371918,
+      "grad_norm": 1.7699942588806152,
+      "learning_rate": 2.1597555572319017e-07,
+      "loss": 0.1384,
+      "step": 3945
+    },
+    {
+      "epoch": 1.8719165085388996,
+      "grad_norm": 1.621044635772705,
+      "learning_rate": 2.143902904290296e-07,
+      "loss": 0.1489,
+      "step": 3946
+    },
+    {
+      "epoch": 1.8723908918406074,
+      "grad_norm": 1.7144032716751099,
+      "learning_rate": 2.1281080148344734e-07,
+      "loss": 0.1488,
+      "step": 3947
+    },
+    {
+      "epoch": 1.8728652751423152,
+      "grad_norm": 1.5608563423156738,
+      "learning_rate": 2.1123708981880097e-07,
+      "loss": 0.133,
+      "step": 3948
+    },
+    {
+      "epoch": 1.8733396584440227,
+      "grad_norm": 1.6698534488677979,
+      "learning_rate": 2.0966915636403518e-07,
+      "loss": 0.1631,
+      "step": 3949
+    },
+    {
+      "epoch": 1.8738140417457305,
+      "grad_norm": 1.595363974571228,
+      "learning_rate": 2.0810700204468737e-07,
+      "loss": 0.1502,
+      "step": 3950
+    },
+    {
+      "epoch": 1.8742884250474383,
+      "grad_norm": 1.279978632926941,
+      "learning_rate": 2.0655062778288103e-07,
+      "loss": 0.1074,
+      "step": 3951
+    },
+    {
+      "epoch": 1.874762808349146,
+      "grad_norm": 1.4724317789077759,
+      "learning_rate": 2.050000344973302e-07,
+      "loss": 0.1373,
+      "step": 3952
+    },
+    {
+      "epoch": 1.875237191650854,
+      "grad_norm": 1.8594779968261719,
+      "learning_rate": 2.0345522310333154e-07,
+      "loss": 0.1505,
+      "step": 3953
+    },
+    {
+      "epoch": 1.8757115749525617,
+      "grad_norm": 1.8657863140106201,
+      "learning_rate": 2.0191619451277568e-07,
+      "loss": 0.1591,
+      "step": 3954
+    },
+    {
+      "epoch": 1.8761859582542695,
+      "grad_norm": 1.5500346422195435,
+      "learning_rate": 2.0038294963413251e-07,
+      "loss": 0.1185,
+      "step": 3955
+    },
+    {
+      "epoch": 1.8766603415559773,
+      "grad_norm": 2.1276473999023438,
+      "learning_rate": 1.9885548937246259e-07,
+      "loss": 0.1951,
+      "step": 3956
+    },
+    {
+      "epoch": 1.8771347248576848,
+      "grad_norm": 1.9857920408248901,
+      "learning_rate": 1.9733381462941237e-07,
+      "loss": 0.1628,
+      "step": 3957
+    },
+    {
+      "epoch": 1.8776091081593926,
+      "grad_norm": 1.5573140382766724,
+      "learning_rate": 1.9581792630320784e-07,
+      "loss": 0.142,
+      "step": 3958
+    },
+    {
+      "epoch": 1.8780834914611004,
+      "grad_norm": 1.2747137546539307,
+      "learning_rate": 1.9430782528866655e-07,
+      "loss": 0.0852,
+      "step": 3959
+    },
+    {
+      "epoch": 1.8785578747628082,
+      "grad_norm": 1.549103856086731,
+      "learning_rate": 1.928035124771832e-07,
+      "loss": 0.1298,
+      "step": 3960
+    },
+    {
+      "epoch": 1.879032258064516,
+      "grad_norm": 1.432792067527771,
+      "learning_rate": 1.9130498875673975e-07,
+      "loss": 0.1194,
+      "step": 3961
+    },
+    {
+      "epoch": 1.8795066413662238,
+      "grad_norm": 1.63291335105896,
+      "learning_rate": 1.8981225501190193e-07,
+      "loss": 0.1418,
+      "step": 3962
+    },
+    {
+      "epoch": 1.8799810246679316,
+      "grad_norm": 1.7714699506759644,
+      "learning_rate": 1.8832531212381378e-07,
+      "loss": 0.1696,
+      "step": 3963
+    },
+    {
+      "epoch": 1.8804554079696394,
+      "grad_norm": 1.4121677875518799,
+      "learning_rate": 1.8684416097020318e-07,
+      "loss": 0.1029,
+      "step": 3964
+    },
+    {
+      "epoch": 1.8809297912713472,
+      "grad_norm": 2.2035553455352783,
+      "learning_rate": 1.853688024253786e-07,
+      "loss": 0.1971,
+      "step": 3965
+    },
+    {
+      "epoch": 1.881404174573055,
+      "grad_norm": 1.4989498853683472,
+      "learning_rate": 1.8389923736022886e-07,
+      "loss": 0.1317,
+      "step": 3966
+    },
+    {
+      "epoch": 1.8818785578747628,
+      "grad_norm": 1.5162180662155151,
+      "learning_rate": 1.824354666422268e-07,
+      "loss": 0.1308,
+      "step": 3967
+    },
+    {
+      "epoch": 1.8823529411764706,
+      "grad_norm": 1.9501628875732422,
+      "learning_rate": 1.8097749113541897e-07,
+      "loss": 0.1492,
+      "step": 3968
+    },
+    {
+      "epoch": 1.8828273244781784,
+      "grad_norm": 1.5917279720306396,
+      "learning_rate": 1.7952531170043474e-07,
+      "loss": 0.1331,
+      "step": 3969
+    },
+    {
+      "epoch": 1.8833017077798861,
+      "grad_norm": 1.2856277227401733,
+      "learning_rate": 1.7807892919448178e-07,
+      "loss": 0.1212,
+      "step": 3970
+    },
+    {
+      "epoch": 1.883776091081594,
+      "grad_norm": 1.876271367073059,
+      "learning_rate": 1.7663834447134488e-07,
+      "loss": 0.184,
+      "step": 3971
+    },
+    {
+      "epoch": 1.8842504743833017,
+      "grad_norm": 2.1068382263183594,
+      "learning_rate": 1.7520355838138603e-07,
+      "loss": 0.1388,
+      "step": 3972
+    },
+    {
+      "epoch": 1.8847248576850095,
+      "grad_norm": 1.8254120349884033,
+      "learning_rate": 1.7377457177154554e-07,
+      "loss": 0.1451,
+      "step": 3973
+    },
+    {
+      "epoch": 1.8851992409867173,
+      "grad_norm": 2.037038564682007,
+      "learning_rate": 1.7235138548534202e-07,
+      "loss": 0.1183,
+      "step": 3974
+    },
+    {
+      "epoch": 1.885673624288425,
+      "grad_norm": 1.3412225246429443,
+      "learning_rate": 1.7093400036286567e-07,
+      "loss": 0.1199,
+      "step": 3975
+    },
+    {
+      "epoch": 1.886148007590133,
+      "grad_norm": 1.9562183618545532,
+      "learning_rate": 1.6952241724078723e-07,
+      "loss": 0.1583,
+      "step": 3976
+    },
+    {
+      "epoch": 1.8866223908918407,
+      "grad_norm": 1.518302321434021,
+      "learning_rate": 1.6811663695234903e-07,
+      "loss": 0.1554,
+      "step": 3977
+    },
+    {
+      "epoch": 1.8870967741935485,
+      "grad_norm": 1.9183396100997925,
+      "learning_rate": 1.6671666032736845e-07,
+      "loss": 0.1279,
+      "step": 3978
+    },
+    {
+      "epoch": 1.8875711574952563,
+      "grad_norm": 2.229943037033081,
+      "learning_rate": 1.6532248819223995e-07,
+      "loss": 0.1979,
+      "step": 3979
+    },
+    {
+      "epoch": 1.888045540796964,
+      "grad_norm": 1.8653277158737183,
+      "learning_rate": 1.639341213699286e-07,
+      "loss": 0.1692,
+      "step": 3980
+    },
+    {
+      "epoch": 1.8885199240986719,
+      "grad_norm": 1.746902346611023,
+      "learning_rate": 1.6255156067997325e-07,
+      "loss": 0.1305,
+      "step": 3981
+    },
+    {
+      "epoch": 1.8889943074003797,
+      "grad_norm": 1.417380928993225,
+      "learning_rate": 1.6117480693848442e-07,
+      "loss": 0.1378,
+      "step": 3982
+    },
+    {
+      "epoch": 1.8894686907020875,
+      "grad_norm": 1.7626910209655762,
+      "learning_rate": 1.598038609581487e-07,
+      "loss": 0.1556,
+      "step": 3983
+    },
+    {
+      "epoch": 1.889943074003795,
+      "grad_norm": 1.6275900602340698,
+      "learning_rate": 1.5843872354822099e-07,
+      "loss": 0.1346,
+      "step": 3984
+    },
+    {
+      "epoch": 1.8904174573055028,
+      "grad_norm": 2.031325340270996,
+      "learning_rate": 1.5707939551452778e-07,
+      "loss": 0.1854,
+      "step": 3985
+    },
+    {
+      "epoch": 1.8908918406072106,
+      "grad_norm": 1.9907748699188232,
+      "learning_rate": 1.5572587765946833e-07,
+      "loss": 0.1645,
+      "step": 3986
+    },
+    {
+      "epoch": 1.8913662239089184,
+      "grad_norm": 1.80055570602417,
+      "learning_rate": 1.5437817078201024e-07,
+      "loss": 0.1441,
+      "step": 3987
+    },
+    {
+      "epoch": 1.8918406072106262,
+      "grad_norm": 1.7448289394378662,
+      "learning_rate": 1.5303627567769043e-07,
+      "loss": 0.1543,
+      "step": 3988
+    },
+    {
+      "epoch": 1.892314990512334,
+      "grad_norm": 1.6348083019256592,
+      "learning_rate": 1.517001931386175e-07,
+      "loss": 0.1384,
+      "step": 3989
+    },
+    {
+      "epoch": 1.8927893738140418,
+      "grad_norm": 1.6425923109054565,
+      "learning_rate": 1.5036992395346838e-07,
+      "loss": 0.1273,
+      "step": 3990
+    },
+    {
+      "epoch": 1.8932637571157496,
+      "grad_norm": 1.3642024993896484,
+      "learning_rate": 1.4904546890748606e-07,
+      "loss": 0.123,
+      "step": 3991
+    },
+    {
+      "epoch": 1.8937381404174574,
+      "grad_norm": 2.2443578243255615,
+      "learning_rate": 1.477268287824829e-07,
+      "loss": 0.1664,
+      "step": 3992
+    },
+    {
+      "epoch": 1.894212523719165,
+      "grad_norm": 1.5792819261550903,
+      "learning_rate": 1.4641400435684184e-07,
+      "loss": 0.1412,
+      "step": 3993
+    },
+    {
+      "epoch": 1.8946869070208727,
+      "grad_norm": 1.4068413972854614,
+      "learning_rate": 1.4510699640550852e-07,
+      "loss": 0.1264,
+      "step": 3994
+    },
+    {
+      "epoch": 1.8951612903225805,
+      "grad_norm": 1.7615771293640137,
+      "learning_rate": 1.438058056999969e-07,
+      "loss": 0.1467,
+      "step": 3995
+    },
+    {
+      "epoch": 1.8956356736242883,
+      "grad_norm": 1.797843098640442,
+      "learning_rate": 1.4251043300838706e-07,
+      "loss": 0.1566,
+      "step": 3996
+    },
+    {
+      "epoch": 1.896110056925996,
+      "grad_norm": 1.5007370710372925,
+      "learning_rate": 1.412208790953229e-07,
+      "loss": 0.1244,
+      "step": 3997
+    },
+    {
+      "epoch": 1.896584440227704,
+      "grad_norm": 1.8236267566680908,
+      "learning_rate": 1.399371447220188e-07,
+      "loss": 0.1535,
+      "step": 3998
+    },
+    {
+      "epoch": 1.8970588235294117,
+      "grad_norm": 1.9256770610809326,
+      "learning_rate": 1.3865923064624753e-07,
+      "loss": 0.1759,
+      "step": 3999
+    },
+    {
+      "epoch": 1.8975332068311195,
+      "grad_norm": 1.9802489280700684,
+      "learning_rate": 1.3738713762235124e-07,
+      "loss": 0.1589,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8980075901328273,
+      "grad_norm": 1.7271060943603516,
+      "learning_rate": 1.3612086640123257e-07,
+      "loss": 0.1405,
+      "step": 4001
+    },
+    {
+      "epoch": 1.898481973434535,
+      "grad_norm": 1.7396737337112427,
+      "learning_rate": 1.348604177303592e-07,
+      "loss": 0.1848,
+      "step": 4002
+    },
+    {
+      "epoch": 1.8989563567362429,
+      "grad_norm": 1.5223190784454346,
+      "learning_rate": 1.3360579235376148e-07,
+      "loss": 0.1409,
+      "step": 4003
+    },
+    {
+      "epoch": 1.8994307400379506,
+      "grad_norm": 1.359562635421753,
+      "learning_rate": 1.323569910120326e-07,
+      "loss": 0.1323,
+      "step": 4004
+    },
+    {
+      "epoch": 1.8999051233396584,
+      "grad_norm": 1.640760064125061,
+      "learning_rate": 1.311140144423273e-07,
+      "loss": 0.1185,
+      "step": 4005
+    },
+    {
+      "epoch": 1.9003795066413662,
+      "grad_norm": 1.469146966934204,
+      "learning_rate": 1.2987686337836202e-07,
+      "loss": 0.1387,
+      "step": 4006
+    },
+    {
+      "epoch": 1.900853889943074,
+      "grad_norm": 1.6434742212295532,
+      "learning_rate": 1.2864553855041484e-07,
+      "loss": 0.159,
+      "step": 4007
+    },
+    {
+      "epoch": 1.9013282732447818,
+      "grad_norm": 1.5248254537582397,
+      "learning_rate": 1.2742004068532544e-07,
+      "loss": 0.1352,
+      "step": 4008
+    },
+    {
+      "epoch": 1.9018026565464896,
+      "grad_norm": 1.5884958505630493,
+      "learning_rate": 1.2620037050649404e-07,
+      "loss": 0.1388,
+      "step": 4009
+    },
+    {
+      "epoch": 1.9022770398481974,
+      "grad_norm": 1.4101170301437378,
+      "learning_rate": 1.2498652873387696e-07,
+      "loss": 0.1303,
+      "step": 4010
+    },
+    {
+      "epoch": 1.9027514231499052,
+      "grad_norm": 1.5898159742355347,
+      "learning_rate": 1.237785160839955e-07,
+      "loss": 0.1243,
+      "step": 4011
+    },
+    {
+      "epoch": 1.903225806451613,
+      "grad_norm": 1.8250586986541748,
+      "learning_rate": 1.225763332699259e-07,
+      "loss": 0.1687,
+      "step": 4012
+    },
+    {
+      "epoch": 1.9037001897533208,
+      "grad_norm": 1.5457115173339844,
+      "learning_rate": 1.21379981001305e-07,
+      "loss": 0.137,
+      "step": 4013
+    },
+    {
+      "epoch": 1.9041745730550286,
+      "grad_norm": 2.082979917526245,
+      "learning_rate": 1.2018945998433007e-07,
+      "loss": 0.1687,
+      "step": 4014
+    },
+    {
+      "epoch": 1.9046489563567364,
+      "grad_norm": 1.4070978164672852,
+      "learning_rate": 1.190047709217501e-07,
+      "loss": 0.0983,
+      "step": 4015
+    },
+    {
+      "epoch": 1.9051233396584442,
+      "grad_norm": 1.6972084045410156,
+      "learning_rate": 1.1782591451287795e-07,
+      "loss": 0.1345,
+      "step": 4016
+    },
+    {
+      "epoch": 1.905597722960152,
+      "grad_norm": 1.9089219570159912,
+      "learning_rate": 1.1665289145357916e-07,
+      "loss": 0.1535,
+      "step": 4017
+    },
+    {
+      "epoch": 1.9060721062618597,
+      "grad_norm": 1.5392357110977173,
+      "learning_rate": 1.1548570243627988e-07,
+      "loss": 0.1201,
+      "step": 4018
+    },
+    {
+      "epoch": 1.9065464895635673,
+      "grad_norm": 1.7641630172729492,
+      "learning_rate": 1.1432434814995897e-07,
+      "loss": 0.1616,
+      "step": 4019
+    },
+    {
+      "epoch": 1.907020872865275,
+      "grad_norm": 1.4591128826141357,
+      "learning_rate": 1.131688292801525e-07,
+      "loss": 0.125,
+      "step": 4020
+    },
+    {
+      "epoch": 1.907495256166983,
+      "grad_norm": 1.7952362298965454,
+      "learning_rate": 1.1201914650895152e-07,
+      "loss": 0.1476,
+      "step": 4021
+    },
+    {
+      "epoch": 1.9079696394686907,
+      "grad_norm": 1.4461826086044312,
+      "learning_rate": 1.1087530051500206e-07,
+      "loss": 0.1369,
+      "step": 4022
+    },
+    {
+      "epoch": 1.9084440227703985,
+      "grad_norm": 1.7202457189559937,
+      "learning_rate": 1.0973729197350514e-07,
+      "loss": 0.1535,
+      "step": 4023
+    },
+    {
+      "epoch": 1.9089184060721063,
+      "grad_norm": 1.419438362121582,
+      "learning_rate": 1.0860512155621783e-07,
+      "loss": 0.1482,
+      "step": 4024
+    },
+    {
+      "epoch": 1.909392789373814,
+      "grad_norm": 1.293198585510254,
+      "learning_rate": 1.0747878993144667e-07,
+      "loss": 0.1156,
+      "step": 4025
+    },
+    {
+      "epoch": 1.9098671726755219,
+      "grad_norm": 1.2855961322784424,
+      "learning_rate": 1.0635829776405537e-07,
+      "loss": 0.1263,
+      "step": 4026
+    },
+    {
+      "epoch": 1.9103415559772297,
+      "grad_norm": 1.5536890029907227,
+      "learning_rate": 1.0524364571546042e-07,
+      "loss": 0.1277,
+      "step": 4027
+    },
+    {
+      "epoch": 1.9108159392789372,
+      "grad_norm": 1.8725106716156006,
+      "learning_rate": 1.041348344436277e-07,
+      "loss": 0.1618,
+      "step": 4028
+    },
+    {
+      "epoch": 1.911290322580645,
+      "grad_norm": 1.922491431236267,
+      "learning_rate": 1.0303186460307813e-07,
+      "loss": 0.1475,
+      "step": 4029
+    },
+    {
+      "epoch": 1.9117647058823528,
+      "grad_norm": 1.6635925769805908,
+      "learning_rate": 1.019347368448842e-07,
+      "loss": 0.1678,
+      "step": 4030
+    },
+    {
+      "epoch": 1.9122390891840606,
+      "grad_norm": 1.6013275384902954,
+      "learning_rate": 1.0084345181666899e-07,
+      "loss": 0.1227,
+      "step": 4031
+    },
+    {
+      "epoch": 1.9127134724857684,
+      "grad_norm": 1.4993422031402588,
+      "learning_rate": 9.975801016260834e-08,
+      "loss": 0.1227,
+      "step": 4032
+    },
+    {
+      "epoch": 1.9131878557874762,
+      "grad_norm": 1.8088953495025635,
+      "learning_rate": 9.867841252342747e-08,
+      "loss": 0.1511,
+      "step": 4033
+    },
+    {
+      "epoch": 1.913662239089184,
+      "grad_norm": 1.366092562675476,
+      "learning_rate": 9.76046595364022e-08,
+      "loss": 0.1118,
+      "step": 4034
+    },
+    {
+      "epoch": 1.9141366223908918,
+      "grad_norm": 1.6480141878128052,
+      "learning_rate": 9.65367518353577e-08,
+      "loss": 0.1707,
+      "step": 4035
+    },
+    {
+      "epoch": 1.9146110056925996,
+      "grad_norm": 1.374222755432129,
+      "learning_rate": 9.547469005066979e-08,
+      "loss": 0.1033,
+      "step": 4036
+    },
+    {
+      "epoch": 1.9150853889943074,
+      "grad_norm": 1.432600736618042,
+      "learning_rate": 9.441847480926247e-08,
+      "loss": 0.1277,
+      "step": 4037
+    },
+    {
+      "epoch": 1.9155597722960152,
+      "grad_norm": 1.8599258661270142,
+      "learning_rate": 9.336810673460928e-08,
+      "loss": 0.1714,
+      "step": 4038
+    },
+    {
+      "epoch": 1.916034155597723,
+      "grad_norm": 1.7460694313049316,
+      "learning_rate": 9.232358644673311e-08,
+      "loss": 0.141,
+      "step": 4039
+    },
+    {
+      "epoch": 1.9165085388994307,
+      "grad_norm": 1.5455800294876099,
+      "learning_rate": 9.12849145622019e-08,
+      "loss": 0.1159,
+      "step": 4040
+    },
+    {
+      "epoch": 1.9169829222011385,
+      "grad_norm": 1.9061877727508545,
+      "learning_rate": 9.025209169413629e-08,
+      "loss": 0.1449,
+      "step": 4041
+    },
+    {
+      "epoch": 1.9174573055028463,
+      "grad_norm": 1.8176791667938232,
+      "learning_rate": 8.922511845219972e-08,
+      "loss": 0.1377,
+      "step": 4042
+    },
+    {
+      "epoch": 1.9179316888045541,
+      "grad_norm": 1.5344781875610352,
+      "learning_rate": 8.820399544260283e-08,
+      "loss": 0.1394,
+      "step": 4043
+    },
+    {
+      "epoch": 1.918406072106262,
+      "grad_norm": 1.6512079238891602,
+      "learning_rate": 8.71887232681079e-08,
+      "loss": 0.1129,
+      "step": 4044
+    },
+    {
+      "epoch": 1.9188804554079697,
+      "grad_norm": 2.159686803817749,
+      "learning_rate": 8.617930252801665e-08,
+      "loss": 0.1642,
+      "step": 4045
+    },
+    {
+      "epoch": 1.9193548387096775,
+      "grad_norm": 1.8301466703414917,
+      "learning_rate": 8.517573381818245e-08,
+      "loss": 0.1814,
+      "step": 4046
+    },
+    {
+      "epoch": 1.9198292220113853,
+      "grad_norm": 1.6383099555969238,
+      "learning_rate": 8.417801773100032e-08,
+      "loss": 0.1308,
+      "step": 4047
+    },
+    {
+      "epoch": 1.920303605313093,
+      "grad_norm": 1.4430229663848877,
+      "learning_rate": 8.31861548554147e-08,
+      "loss": 0.1266,
+      "step": 4048
+    },
+    {
+      "epoch": 1.9207779886148009,
+      "grad_norm": 1.6547898054122925,
+      "learning_rate": 8.220014577690949e-08,
+      "loss": 0.1496,
+      "step": 4049
+    },
+    {
+      "epoch": 1.9212523719165087,
+      "grad_norm": 1.4231595993041992,
+      "learning_rate": 8.1219991077518e-08,
+      "loss": 0.1162,
+      "step": 4050
+    },
+    {
+      "epoch": 1.9217267552182165,
+      "grad_norm": 1.9546478986740112,
+      "learning_rate": 8.024569133581517e-08,
+      "loss": 0.1404,
+      "step": 4051
+    },
+    {
+      "epoch": 1.9222011385199242,
+      "grad_norm": 1.541332483291626,
+      "learning_rate": 7.927724712692098e-08,
+      "loss": 0.1339,
+      "step": 4052
+    },
+    {
+      "epoch": 1.922675521821632,
+      "grad_norm": 2.084918260574341,
+      "learning_rate": 7.831465902249701e-08,
+      "loss": 0.1533,
+      "step": 4053
+    },
+    {
+      "epoch": 1.9231499051233396,
+      "grad_norm": 1.7567437887191772,
+      "learning_rate": 7.7357927590751e-08,
+      "loss": 0.1356,
+      "step": 4054
+    },
+    {
+      "epoch": 1.9236242884250474,
+      "grad_norm": 1.616256833076477,
+      "learning_rate": 7.640705339643118e-08,
+      "loss": 0.1404,
+      "step": 4055
+    },
+    {
+      "epoch": 1.9240986717267552,
+      "grad_norm": 1.6991674900054932,
+      "learning_rate": 7.546203700082966e-08,
+      "loss": 0.1553,
+      "step": 4056
+    },
+    {
+      "epoch": 1.924573055028463,
+      "grad_norm": 1.4188369512557983,
+      "learning_rate": 7.452287896178134e-08,
+      "loss": 0.1166,
+      "step": 4057
+    },
+    {
+      "epoch": 1.9250474383301708,
+      "grad_norm": 1.779510498046875,
+      "learning_rate": 7.358957983365944e-08,
+      "loss": 0.1478,
+      "step": 4058
+    },
+    {
+      "epoch": 1.9255218216318786,
+      "grad_norm": 1.4632707834243774,
+      "learning_rate": 7.266214016738326e-08,
+      "loss": 0.1639,
+      "step": 4059
+    },
+    {
+      "epoch": 1.9259962049335864,
+      "grad_norm": 2.640392303466797,
+      "learning_rate": 7.174056051041045e-08,
+      "loss": 0.1873,
+      "step": 4060
+    },
+    {
+      "epoch": 1.9264705882352942,
+      "grad_norm": 2.0957512855529785,
+      "learning_rate": 7.082484140674029e-08,
+      "loss": 0.1609,
+      "step": 4061
+    },
+    {
+      "epoch": 1.926944971537002,
+      "grad_norm": 1.6926616430282593,
+      "learning_rate": 6.99149833969126e-08,
+      "loss": 0.1288,
+      "step": 4062
+    },
+    {
+      "epoch": 1.9274193548387095,
+      "grad_norm": 1.697201132774353,
+      "learning_rate": 6.901098701800779e-08,
+      "loss": 0.1569,
+      "step": 4063
+    },
+    {
+      "epoch": 1.9278937381404173,
+      "grad_norm": 1.5255944728851318,
+      "learning_rate": 6.811285280364677e-08,
+      "loss": 0.1255,
+      "step": 4064
+    },
+    {
+      "epoch": 1.928368121442125,
+      "grad_norm": 1.38019859790802,
+      "learning_rate": 6.722058128398768e-08,
+      "loss": 0.1255,
+      "step": 4065
+    },
+    {
+      "epoch": 1.928842504743833,
+      "grad_norm": 1.9084728956222534,
+      "learning_rate": 6.633417298573142e-08,
+      "loss": 0.1617,
+      "step": 4066
+    },
+    {
+      "epoch": 1.9293168880455407,
+      "grad_norm": 1.2430599927902222,
+      "learning_rate": 6.54536284321139e-08,
+      "loss": 0.1083,
+      "step": 4067
+    },
+    {
+      "epoch": 1.9297912713472485,
+      "grad_norm": 1.3795536756515503,
+      "learning_rate": 6.457894814291376e-08,
+      "loss": 0.1128,
+      "step": 4068
+    },
+    {
+      "epoch": 1.9302656546489563,
+      "grad_norm": 1.5916107892990112,
+      "learning_rate": 6.371013263444469e-08,
+      "loss": 0.1367,
+      "step": 4069
+    },
+    {
+      "epoch": 1.930740037950664,
+      "grad_norm": 1.4485095739364624,
+      "learning_rate": 6.284718241956089e-08,
+      "loss": 0.1246,
+      "step": 4070
+    },
+    {
+      "epoch": 1.9312144212523719,
+      "grad_norm": 1.8069593906402588,
+      "learning_rate": 6.199009800765265e-08,
+      "loss": 0.1763,
+      "step": 4071
+    },
+    {
+      "epoch": 1.9316888045540797,
+      "grad_norm": 1.608431339263916,
+      "learning_rate": 6.113887990464862e-08,
+      "loss": 0.1233,
+      "step": 4072
+    },
+    {
+      "epoch": 1.9321631878557874,
+      "grad_norm": 1.4641162157058716,
+      "learning_rate": 6.029352861301462e-08,
+      "loss": 0.127,
+      "step": 4073
+    },
+    {
+      "epoch": 1.9326375711574952,
+      "grad_norm": 1.5734367370605469,
+      "learning_rate": 5.945404463175375e-08,
+      "loss": 0.1413,
+      "step": 4074
+    },
+    {
+      "epoch": 1.933111954459203,
+      "grad_norm": 1.354025959968567,
+      "learning_rate": 5.862042845640403e-08,
+      "loss": 0.1094,
+      "step": 4075
+    },
+    {
+      "epoch": 1.9335863377609108,
+      "grad_norm": 1.6101434230804443,
+      "learning_rate": 5.779268057904186e-08,
+      "loss": 0.1551,
+      "step": 4076
+    },
+    {
+      "epoch": 1.9340607210626186,
+      "grad_norm": 1.3197808265686035,
+      "learning_rate": 5.6970801488276385e-08,
+      "loss": 0.1253,
+      "step": 4077
+    },
+    {
+      "epoch": 1.9345351043643264,
+      "grad_norm": 1.7028498649597168,
+      "learning_rate": 5.61547916692573e-08,
+      "loss": 0.1485,
+      "step": 4078
+    },
+    {
+      "epoch": 1.9350094876660342,
+      "grad_norm": 1.6578190326690674,
+      "learning_rate": 5.534465160366598e-08,
+      "loss": 0.1275,
+      "step": 4079
+    },
+    {
+      "epoch": 1.935483870967742,
+      "grad_norm": 1.70431649684906,
+      "learning_rate": 5.454038176971987e-08,
+      "loss": 0.1387,
+      "step": 4080
+    },
+    {
+      "epoch": 1.9359582542694498,
+      "grad_norm": 1.3666795492172241,
+      "learning_rate": 5.3741982642173675e-08,
+      "loss": 0.1215,
+      "step": 4081
+    },
+    {
+      "epoch": 1.9364326375711576,
+      "grad_norm": 1.2814855575561523,
+      "learning_rate": 5.294945469231039e-08,
+      "loss": 0.1162,
+      "step": 4082
+    },
+    {
+      "epoch": 1.9369070208728654,
+      "grad_norm": 1.7028858661651611,
+      "learning_rate": 5.2162798387954686e-08,
+      "loss": 0.1455,
+      "step": 4083
+    },
+    {
+      "epoch": 1.9373814041745732,
+      "grad_norm": 1.8575924634933472,
+      "learning_rate": 5.1382014193461783e-08,
+      "loss": 0.1607,
+      "step": 4084
+    },
+    {
+      "epoch": 1.937855787476281,
+      "grad_norm": 1.5589218139648438,
+      "learning_rate": 5.0607102569718566e-08,
+      "loss": 0.1597,
+      "step": 4085
+    },
+    {
+      "epoch": 1.9383301707779887,
+      "grad_norm": 1.469462275505066,
+      "learning_rate": 4.9838063974150255e-08,
+      "loss": 0.1403,
+      "step": 4086
+    },
+    {
+      "epoch": 1.9388045540796965,
+      "grad_norm": 1.900517225265503,
+      "learning_rate": 4.9074898860711485e-08,
+      "loss": 0.1851,
+      "step": 4087
+    },
+    {
+      "epoch": 1.9392789373814043,
+      "grad_norm": 1.6202280521392822,
+      "learning_rate": 4.83176076798908e-08,
+      "loss": 0.1457,
+      "step": 4088
+    },
+    {
+      "epoch": 1.939753320683112,
+      "grad_norm": 1.608129620552063,
+      "learning_rate": 4.7566190878710615e-08,
+      "loss": 0.1075,
+      "step": 4089
+    },
+    {
+      "epoch": 1.9402277039848197,
+      "grad_norm": 1.531738519668579,
+      "learning_rate": 4.6820648900725016e-08,
+      "loss": 0.1195,
+      "step": 4090
+    },
+    {
+      "epoch": 1.9407020872865275,
+      "grad_norm": 1.6448765993118286,
+      "learning_rate": 4.608098218601864e-08,
+      "loss": 0.1508,
+      "step": 4091
+    },
+    {
+      "epoch": 1.9411764705882353,
+      "grad_norm": 1.4417756795883179,
+      "learning_rate": 4.5347191171211114e-08,
+      "loss": 0.1439,
+      "step": 4092
+    },
+    {
+      "epoch": 1.941650853889943,
+      "grad_norm": 1.6759575605392456,
+      "learning_rate": 4.4619276289450394e-08,
+      "loss": 0.1487,
+      "step": 4093
+    },
+    {
+      "epoch": 1.9421252371916509,
+      "grad_norm": 1.6265183687210083,
+      "learning_rate": 4.3897237970418336e-08,
+      "loss": 0.1338,
+      "step": 4094
+    },
+    {
+      "epoch": 1.9425996204933587,
+      "grad_norm": 1.5453293323516846,
+      "learning_rate": 4.318107664032622e-08,
+      "loss": 0.1366,
+      "step": 4095
+    },
+    {
+      "epoch": 1.9430740037950665,
+      "grad_norm": 1.4695607423782349,
+      "learning_rate": 4.247079272191812e-08,
+      "loss": 0.1329,
+      "step": 4096
+    },
+    {
+      "epoch": 1.9435483870967742,
+      "grad_norm": 1.945720911026001,
+      "learning_rate": 4.1766386634467523e-08,
+      "loss": 0.169,
+      "step": 4097
+    },
+    {
+      "epoch": 1.9440227703984818,
+      "grad_norm": 1.592432975769043,
+      "learning_rate": 4.1067858793777394e-08,
+      "loss": 0.1179,
+      "step": 4098
+    },
+    {
+      "epoch": 1.9444971537001896,
+      "grad_norm": 1.6521458625793457,
+      "learning_rate": 4.037520961218233e-08,
+      "loss": 0.1522,
+      "step": 4099
+    },
+    {
+      "epoch": 1.9449715370018974,
+      "grad_norm": 1.886146903038025,
+      "learning_rate": 3.96884394985475e-08,
+      "loss": 0.1636,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9454459203036052,
+      "grad_norm": 1.8324472904205322,
+      "learning_rate": 3.900754885826419e-08,
+      "loss": 0.1596,
+      "step": 4101
+    },
+    {
+      "epoch": 1.945920303605313,
+      "grad_norm": 1.935429573059082,
+      "learning_rate": 3.833253809325643e-08,
+      "loss": 0.171,
+      "step": 4102
+    },
+    {
+      "epoch": 1.9463946869070208,
+      "grad_norm": 1.5601022243499756,
+      "learning_rate": 3.766340760197662e-08,
+      "loss": 0.1629,
+      "step": 4103
+    },
+    {
+      "epoch": 1.9468690702087286,
+      "grad_norm": 1.2988762855529785,
+      "learning_rate": 3.700015777940547e-08,
+      "loss": 0.1098,
+      "step": 4104
+    },
+    {
+      "epoch": 1.9473434535104364,
+      "grad_norm": 1.6322436332702637,
+      "learning_rate": 3.634278901705424e-08,
+      "loss": 0.167,
+      "step": 4105
+    },
+    {
+      "epoch": 1.9478178368121442,
+      "grad_norm": 1.707542061805725,
+      "learning_rate": 3.56913017029592e-08,
+      "loss": 0.1659,
+      "step": 4106
+    },
+    {
+      "epoch": 1.948292220113852,
+      "grad_norm": 1.5153127908706665,
+      "learning_rate": 3.50456962216883e-08,
+      "loss": 0.1431,
+      "step": 4107
+    },
+    {
+      "epoch": 1.9487666034155597,
+      "grad_norm": 1.735724687576294,
+      "learning_rate": 3.4405972954334454e-08,
+      "loss": 0.1478,
+      "step": 4108
+    },
+    {
+      "epoch": 1.9492409867172675,
+      "grad_norm": 1.4995523691177368,
+      "learning_rate": 3.3772132278522276e-08,
+      "loss": 0.1218,
+      "step": 4109
+    },
+    {
+      "epoch": 1.9497153700189753,
+      "grad_norm": 1.6257517337799072,
+      "learning_rate": 3.3144174568399135e-08,
+      "loss": 0.1306,
+      "step": 4110
+    },
+    {
+      "epoch": 1.9501897533206831,
+      "grad_norm": 1.814677357673645,
+      "learning_rate": 3.252210019464408e-08,
+      "loss": 0.1423,
+      "step": 4111
+    },
+    {
+      "epoch": 1.950664136622391,
+      "grad_norm": 1.7298712730407715,
+      "learning_rate": 3.190590952446115e-08,
+      "loss": 0.1366,
+      "step": 4112
+    },
+    {
+      "epoch": 1.9511385199240987,
+      "grad_norm": 1.5716830492019653,
+      "learning_rate": 3.129560292158051e-08,
+      "loss": 0.1091,
+      "step": 4113
+    },
+    {
+      "epoch": 1.9516129032258065,
+      "grad_norm": 1.3613415956497192,
+      "learning_rate": 3.069118074626176e-08,
+      "loss": 0.1238,
+      "step": 4114
+    },
+    {
+      "epoch": 1.9520872865275143,
+      "grad_norm": 1.6825555562973022,
+      "learning_rate": 3.0092643355287274e-08,
+      "loss": 0.1441,
+      "step": 4115
+    },
+    {
+      "epoch": 1.952561669829222,
+      "grad_norm": 1.650367021560669,
+      "learning_rate": 2.9499991101969995e-08,
+      "loss": 0.156,
+      "step": 4116
+    },
+    {
+      "epoch": 1.9530360531309299,
+      "grad_norm": 1.4215662479400635,
+      "learning_rate": 2.8913224336145628e-08,
+      "loss": 0.1157,
+      "step": 4117
+    },
+    {
+      "epoch": 1.9535104364326377,
+      "grad_norm": 1.7951689958572388,
+      "learning_rate": 2.8332343404177122e-08,
+      "loss": 0.1527,
+      "step": 4118
+    },
+    {
+      "epoch": 1.9539848197343455,
+      "grad_norm": 1.3944616317749023,
+      "learning_rate": 2.7757348648951298e-08,
+      "loss": 0.1123,
+      "step": 4119
+    },
+    {
+      "epoch": 1.9544592030360532,
+      "grad_norm": 1.5503982305526733,
+      "learning_rate": 2.7188240409883325e-08,
+      "loss": 0.1412,
+      "step": 4120
+    },
+    {
+      "epoch": 1.954933586337761,
+      "grad_norm": 1.899158000946045,
+      "learning_rate": 2.6625019022912256e-08,
+      "loss": 0.1543,
+      "step": 4121
+    },
+    {
+      "epoch": 1.9554079696394688,
+      "grad_norm": 1.7116118669509888,
+      "learning_rate": 2.606768482050215e-08,
+      "loss": 0.1359,
+      "step": 4122
+    },
+    {
+      "epoch": 1.9558823529411766,
+      "grad_norm": 1.7955915927886963,
+      "learning_rate": 2.5516238131640945e-08,
+      "loss": 0.1608,
+      "step": 4123
+    },
+    {
+      "epoch": 1.9563567362428842,
+      "grad_norm": 1.3384437561035156,
+      "learning_rate": 2.4970679281842715e-08,
+      "loss": 0.1349,
+      "step": 4124
+    },
+    {
+      "epoch": 1.956831119544592,
+      "grad_norm": 1.5795972347259521,
+      "learning_rate": 2.443100859314429e-08,
+      "loss": 0.131,
+      "step": 4125
+    },
+    {
+      "epoch": 1.9573055028462998,
+      "grad_norm": 1.428765892982483,
+      "learning_rate": 2.389722638410974e-08,
+      "loss": 0.1195,
+      "step": 4126
+    },
+    {
+      "epoch": 1.9577798861480076,
+      "grad_norm": 1.754357933998108,
+      "learning_rate": 2.3369332969824798e-08,
+      "loss": 0.1282,
+      "step": 4127
+    },
+    {
+      "epoch": 1.9582542694497154,
+      "grad_norm": 1.8691462278366089,
+      "learning_rate": 2.2847328661900203e-08,
+      "loss": 0.1735,
+      "step": 4128
+    },
+    {
+      "epoch": 1.9587286527514232,
+      "grad_norm": 1.630744218826294,
+      "learning_rate": 2.2331213768468363e-08,
+      "loss": 0.119,
+      "step": 4129
+    },
+    {
+      "epoch": 1.959203036053131,
+      "grad_norm": 1.8132566213607788,
+      "learning_rate": 2.1820988594187796e-08,
+      "loss": 0.1588,
+      "step": 4130
+    },
+    {
+      "epoch": 1.9596774193548387,
+      "grad_norm": 1.4721542596817017,
+      "learning_rate": 2.131665344023981e-08,
+      "loss": 0.1301,
+      "step": 4131
+    },
+    {
+      "epoch": 1.9601518026565465,
+      "grad_norm": 1.846588373184204,
+      "learning_rate": 2.0818208604328482e-08,
+      "loss": 0.1776,
+      "step": 4132
+    },
+    {
+      "epoch": 1.960626185958254,
+      "grad_norm": 1.6939235925674438,
+      "learning_rate": 2.032565438067957e-08,
+      "loss": 0.163,
+      "step": 4133
+    },
+    {
+      "epoch": 1.961100569259962,
+      "grad_norm": 1.7470312118530273,
+      "learning_rate": 1.9838991060043833e-08,
+      "loss": 0.1481,
+      "step": 4134
+    },
+    {
+      "epoch": 1.9615749525616697,
+      "grad_norm": 1.7479218244552612,
+      "learning_rate": 1.9358218929693695e-08,
+      "loss": 0.1545,
+      "step": 4135
+    },
+    {
+      "epoch": 1.9620493358633775,
+      "grad_norm": 1.5178728103637695,
+      "learning_rate": 1.8883338273425478e-08,
+      "loss": 0.134,
+      "step": 4136
+    },
+    {
+      "epoch": 1.9625237191650853,
+      "grad_norm": 1.7611669301986694,
+      "learning_rate": 1.8414349371553842e-08,
+      "loss": 0.1345,
+      "step": 4137
+    },
+    {
+      "epoch": 1.962998102466793,
+      "grad_norm": 1.5717504024505615,
+      "learning_rate": 1.7951252500920668e-08,
+      "loss": 0.1209,
+      "step": 4138
+    },
+    {
+      "epoch": 1.9634724857685009,
+      "grad_norm": 1.52717924118042,
+      "learning_rate": 1.7494047934885073e-08,
+      "loss": 0.1201,
+      "step": 4139
+    },
+    {
+      "epoch": 1.9639468690702087,
+      "grad_norm": 1.8689799308776855,
+      "learning_rate": 1.7042735943333388e-08,
+      "loss": 0.1422,
+      "step": 4140
+    },
+    {
+      "epoch": 1.9644212523719164,
+      "grad_norm": 1.4372001886367798,
+      "learning_rate": 1.659731679266807e-08,
+      "loss": 0.123,
+      "step": 4141
+    },
+    {
+      "epoch": 1.9648956356736242,
+      "grad_norm": 1.98843514919281,
+      "learning_rate": 1.6157790745817692e-08,
+      "loss": 0.1896,
+      "step": 4142
+    },
+    {
+      "epoch": 1.965370018975332,
+      "grad_norm": 1.4792766571044922,
+      "learning_rate": 1.5724158062228046e-08,
+      "loss": 0.1424,
+      "step": 4143
+    },
+    {
+      "epoch": 1.9658444022770398,
+      "grad_norm": 1.795911192893982,
+      "learning_rate": 1.5296418997869932e-08,
+      "loss": 0.1412,
+      "step": 4144
+    },
+    {
+      "epoch": 1.9663187855787476,
+      "grad_norm": 1.8548011779785156,
+      "learning_rate": 1.4874573805232495e-08,
+      "loss": 0.1624,
+      "step": 4145
+    },
+    {
+      "epoch": 1.9667931688804554,
+      "grad_norm": 2.306521415710449,
+      "learning_rate": 1.4458622733327654e-08,
+      "loss": 0.1668,
+      "step": 4146
+    },
+    {
+      "epoch": 1.9672675521821632,
+      "grad_norm": 2.005544900894165,
+      "learning_rate": 1.4048566027685673e-08,
+      "loss": 0.1667,
+      "step": 4147
+    },
+    {
+      "epoch": 1.967741935483871,
+      "grad_norm": 2.2230448722839355,
+      "learning_rate": 1.3644403930360706e-08,
+      "loss": 0.1654,
+      "step": 4148
+    },
+    {
+      "epoch": 1.9682163187855788,
+      "grad_norm": 1.799142837524414,
+      "learning_rate": 1.3246136679925249e-08,
+      "loss": 0.138,
+      "step": 4149
+    },
+    {
+      "epoch": 1.9686907020872866,
+      "grad_norm": 1.5519241094589233,
+      "learning_rate": 1.2853764511471245e-08,
+      "loss": 0.114,
+      "step": 4150
+    },
+    {
+      "epoch": 1.9691650853889944,
+      "grad_norm": 1.872344732284546,
+      "learning_rate": 1.2467287656613425e-08,
+      "loss": 0.1696,
+      "step": 4151
+    },
+    {
+      "epoch": 1.9696394686907022,
+      "grad_norm": 2.0996015071868896,
+      "learning_rate": 1.2086706343484855e-08,
+      "loss": 0.1841,
+      "step": 4152
+    },
+    {
+      "epoch": 1.97011385199241,
+      "grad_norm": 1.880236268043518,
+      "learning_rate": 1.1712020796738056e-08,
+      "loss": 0.1657,
+      "step": 4153
+    },
+    {
+      "epoch": 1.9705882352941178,
+      "grad_norm": 1.773377537727356,
+      "learning_rate": 1.1343231237548324e-08,
+      "loss": 0.1686,
+      "step": 4154
+    },
+    {
+      "epoch": 1.9710626185958255,
+      "grad_norm": 2.0221047401428223,
+      "learning_rate": 1.0980337883605973e-08,
+      "loss": 0.1789,
+      "step": 4155
+    },
+    {
+      "epoch": 1.9715370018975333,
+      "grad_norm": 1.4797416925430298,
+      "learning_rate": 1.0623340949125204e-08,
+      "loss": 0.1164,
+      "step": 4156
+    },
+    {
+      "epoch": 1.9720113851992411,
+      "grad_norm": 1.4807063341140747,
+      "learning_rate": 1.027224064483745e-08,
+      "loss": 0.1294,
+      "step": 4157
+    },
+    {
+      "epoch": 1.972485768500949,
+      "grad_norm": 1.8948291540145874,
+      "learning_rate": 9.927037177993593e-09,
+      "loss": 0.1705,
+      "step": 4158
+    },
+    {
+      "epoch": 1.9729601518026565,
+      "grad_norm": 1.3457918167114258,
+      "learning_rate": 9.587730752362855e-09,
+      "loss": 0.1067,
+      "step": 4159
+    },
+    {
+      "epoch": 1.9734345351043643,
+      "grad_norm": 1.3780224323272705,
+      "learning_rate": 9.254321568236135e-09,
+      "loss": 0.1239,
+      "step": 4160
+    },
+    {
+      "epoch": 1.973908918406072,
+      "grad_norm": 1.7696219682693481,
+      "learning_rate": 8.926809822420446e-09,
+      "loss": 0.1581,
+      "step": 4161
+    },
+    {
+      "epoch": 1.9743833017077799,
+      "grad_norm": 1.508495807647705,
+      "learning_rate": 8.605195708242254e-09,
+      "loss": 0.1172,
+      "step": 4162
+    },
+    {
+      "epoch": 1.9748576850094877,
+      "grad_norm": 1.7179173231124878,
+      "learning_rate": 8.289479415548585e-09,
+      "loss": 0.1668,
+      "step": 4163
+    },
+    {
+      "epoch": 1.9753320683111955,
+      "grad_norm": 1.8635634183883667,
+      "learning_rate": 7.979661130703697e-09,
+      "loss": 0.148,
+      "step": 4164
+    },
+    {
+      "epoch": 1.9758064516129032,
+      "grad_norm": 1.8939173221588135,
+      "learning_rate": 7.67574103658797e-09,
+      "loss": 0.1754,
+      "step": 4165
+    },
+    {
+      "epoch": 1.976280834914611,
+      "grad_norm": 1.3412449359893799,
+      "learning_rate": 7.377719312605669e-09,
+      "loss": 0.1145,
+      "step": 4166
+    },
+    {
+      "epoch": 1.9767552182163188,
+      "grad_norm": 1.482282280921936,
+      "learning_rate": 7.085596134673855e-09,
+      "loss": 0.1171,
+      "step": 4167
+    },
+    {
+      "epoch": 1.9772296015180264,
+      "grad_norm": 1.7333260774612427,
+      "learning_rate": 6.799371675230149e-09,
+      "loss": 0.154,
+      "step": 4168
+    },
+    {
+      "epoch": 1.9777039848197342,
+      "grad_norm": 2.1031572818756104,
+      "learning_rate": 6.5190461032305085e-09,
+      "loss": 0.1674,
+      "step": 4169
+    },
+    {
+      "epoch": 1.978178368121442,
+      "grad_norm": 2.2297379970550537,
+      "learning_rate": 6.244619584148126e-09,
+      "loss": 0.1709,
+      "step": 4170
+    },
+    {
+      "epoch": 1.9786527514231498,
+      "grad_norm": 1.7986024618148804,
+      "learning_rate": 5.976092279974533e-09,
+      "loss": 0.1359,
+      "step": 4171
+    },
+    {
+      "epoch": 1.9791271347248576,
+      "grad_norm": 1.8482614755630493,
+      "learning_rate": 5.713464349218489e-09,
+      "loss": 0.1645,
+      "step": 4172
+    },
+    {
+      "epoch": 1.9796015180265654,
+      "grad_norm": 1.9954028129577637,
+      "learning_rate": 5.456735946907099e-09,
+      "loss": 0.15,
+      "step": 4173
+    },
+    {
+      "epoch": 1.9800759013282732,
+      "grad_norm": 1.6475908756256104,
+      "learning_rate": 5.205907224583584e-09,
+      "loss": 0.1412,
+      "step": 4174
+    },
+    {
+      "epoch": 1.980550284629981,
+      "grad_norm": 1.6490331888198853,
+      "learning_rate": 4.960978330310618e-09,
+      "loss": 0.1739,
+      "step": 4175
+    },
+    {
+      "epoch": 1.9810246679316887,
+      "grad_norm": 1.4961657524108887,
+      "learning_rate": 4.721949408666993e-09,
+      "loss": 0.1371,
+      "step": 4176
+    },
+    {
+      "epoch": 1.9814990512333965,
+      "grad_norm": 2.557554006576538,
+      "learning_rate": 4.488820600749844e-09,
+      "loss": 0.2002,
+      "step": 4177
+    },
+    {
+      "epoch": 1.9819734345351043,
+      "grad_norm": 1.2743830680847168,
+      "learning_rate": 4.261592044171314e-09,
+      "loss": 0.1035,
+      "step": 4178
+    },
+    {
+      "epoch": 1.9824478178368121,
+      "grad_norm": 2.122196674346924,
+      "learning_rate": 4.040263873063e-09,
+      "loss": 0.1985,
+      "step": 4179
+    },
+    {
+      "epoch": 1.98292220113852,
+      "grad_norm": 2.3841426372528076,
+      "learning_rate": 3.824836218072614e-09,
+      "loss": 0.1923,
+      "step": 4180
+    },
+    {
+      "epoch": 1.9833965844402277,
+      "grad_norm": 1.5701267719268799,
+      "learning_rate": 3.615309206365103e-09,
+      "loss": 0.12,
+      "step": 4181
+    },
+    {
+      "epoch": 1.9838709677419355,
+      "grad_norm": 1.7213834524154663,
+      "learning_rate": 3.411682961621532e-09,
+      "loss": 0.1528,
+      "step": 4182
+    },
+    {
+      "epoch": 1.9843453510436433,
+      "grad_norm": 1.8331454992294312,
+      "learning_rate": 3.2139576040413067e-09,
+      "loss": 0.1536,
+      "step": 4183
+    },
+    {
+      "epoch": 1.984819734345351,
+      "grad_norm": 1.744813084602356,
+      "learning_rate": 3.0221332503399534e-09,
+      "loss": 0.1609,
+      "step": 4184
+    },
+    {
+      "epoch": 1.9852941176470589,
+      "grad_norm": 1.6270570755004883,
+      "learning_rate": 2.8362100137491187e-09,
+      "loss": 0.1473,
+      "step": 4185
+    },
+    {
+      "epoch": 1.9857685009487667,
+      "grad_norm": 1.7102569341659546,
+      "learning_rate": 2.656188004016569e-09,
+      "loss": 0.1371,
+      "step": 4186
+    },
+    {
+      "epoch": 1.9862428842504745,
+      "grad_norm": 1.3755574226379395,
+      "learning_rate": 2.4820673274095207e-09,
+      "loss": 0.1176,
+      "step": 4187
+    },
+    {
+      "epoch": 1.9867172675521823,
+      "grad_norm": 1.6135125160217285,
+      "learning_rate": 2.3138480867079814e-09,
+      "loss": 0.1593,
+      "step": 4188
+    },
+    {
+      "epoch": 1.98719165085389,
+      "grad_norm": 1.431528925895691,
+      "learning_rate": 2.1515303812091883e-09,
+      "loss": 0.1384,
+      "step": 4189
+    },
+    {
+      "epoch": 1.9876660341555978,
+      "grad_norm": 1.3943769931793213,
+      "learning_rate": 1.9951143067309385e-09,
+      "loss": 0.1049,
+      "step": 4190
+    },
+    {
+      "epoch": 1.9881404174573056,
+      "grad_norm": 1.5564050674438477,
+      "learning_rate": 1.8445999556016003e-09,
+      "loss": 0.1207,
+      "step": 4191
+    },
+    {
+      "epoch": 1.9886148007590134,
+      "grad_norm": 1.4701181650161743,
+      "learning_rate": 1.6999874166678809e-09,
+      "loss": 0.1306,
+      "step": 4192
+    },
+    {
+      "epoch": 1.9890891840607212,
+      "grad_norm": 2.03861141204834,
+      "learning_rate": 1.561276775295939e-09,
+      "loss": 0.1344,
+      "step": 4193
+    },
+    {
+      "epoch": 1.989563567362429,
+      "grad_norm": 1.9190205335617065,
+      "learning_rate": 1.4284681133625022e-09,
+      "loss": 0.1536,
+      "step": 4194
+    },
+    {
+      "epoch": 1.9900379506641366,
+      "grad_norm": 1.4114086627960205,
+      "learning_rate": 1.301561509263749e-09,
+      "loss": 0.1194,
+      "step": 4195
+    },
+    {
+      "epoch": 1.9905123339658444,
+      "grad_norm": 1.5192102193832397,
+      "learning_rate": 1.1805570379130882e-09,
+      "loss": 0.1374,
+      "step": 4196
+    },
+    {
+      "epoch": 1.9909867172675522,
+      "grad_norm": 1.6114188432693481,
+      "learning_rate": 1.0654547707367179e-09,
+      "loss": 0.1747,
+      "step": 4197
+    },
+    {
+      "epoch": 1.99146110056926,
+      "grad_norm": 1.4619154930114746,
+      "learning_rate": 9.562547756780672e-10,
+      "loss": 0.1304,
+      "step": 4198
+    },
+    {
+      "epoch": 1.9919354838709677,
+      "grad_norm": 1.5712562799453735,
+      "learning_rate": 8.529571171977946e-10,
+      "loss": 0.1691,
+      "step": 4199
+    },
+    {
+      "epoch": 1.9924098671726755,
+      "grad_norm": 1.8165565729141235,
+      "learning_rate": 7.555618562715695e-10,
+      "loss": 0.1524,
+      "step": 4200
+    },
+    {
+      "epoch": 1.9928842504743833,
+      "grad_norm": 1.9051848649978638,
+      "learning_rate": 6.640690503889601e-10,
+      "loss": 0.1687,
+      "step": 4201
+    },
+    {
+      "epoch": 1.9933586337760911,
+      "grad_norm": 1.6580159664154053,
+      "learning_rate": 5.784787535600966e-10,
+      "loss": 0.1748,
+      "step": 4202
+    },
+    {
+      "epoch": 1.9938330170777987,
+      "grad_norm": 1.8379294872283936,
+      "learning_rate": 4.987910163067878e-10,
+      "loss": 0.127,
+      "step": 4203
+    },
+    {
+      "epoch": 1.9943074003795065,
+      "grad_norm": 1.5118616819381714,
+      "learning_rate": 4.2500588566696254e-10,
+      "loss": 0.1409,
+      "step": 4204
+    },
+    {
+      "epoch": 1.9947817836812143,
+      "grad_norm": 1.5703513622283936,
+      "learning_rate": 3.5712340519689083e-10,
+      "loss": 0.1385,
+      "step": 4205
+    },
+    {
+      "epoch": 1.995256166982922,
+      "grad_norm": 1.7153023481369019,
+      "learning_rate": 2.9514361496563173e-10,
+      "loss": 0.1148,
+      "step": 4206
+    },
+    {
+      "epoch": 1.9957305502846299,
+      "grad_norm": 1.4218897819519043,
+      "learning_rate": 2.39066551560585e-10,
+      "loss": 0.1276,
+      "step": 4207
+    },
+    {
+      "epoch": 1.9962049335863377,
+      "grad_norm": 1.6402268409729004,
+      "learning_rate": 1.8889224808193995e-10,
+      "loss": 0.1318,
+      "step": 4208
+    },
+    {
+      "epoch": 1.9966793168880455,
+      "grad_norm": 1.7963615655899048,
+      "learning_rate": 1.446207341482264e-10,
+      "loss": 0.1481,
+      "step": 4209
+    },
+    {
+      "epoch": 1.9971537001897532,
+      "grad_norm": 1.9880362749099731,
+      "learning_rate": 1.0625203589187394e-10,
+      "loss": 0.1913,
+      "step": 4210
+    },
+    {
+      "epoch": 1.997628083491461,
+      "grad_norm": 1.616383671760559,
+      "learning_rate": 7.378617596143223e-11,
+      "loss": 0.1222,
+      "step": 4211
+    },
+    {
+      "epoch": 1.9981024667931688,
+      "grad_norm": 1.5262385606765747,
+      "learning_rate": 4.722317352157113e-11,
+      "loss": 0.1253,
+      "step": 4212
+    },
+    {
+      "epoch": 1.9985768500948766,
+      "grad_norm": 1.7264385223388672,
+      "learning_rate": 2.656304425308065e-11,
+      "loss": 0.1395,
+      "step": 4213
+    },
+    {
+      "epoch": 1.9990512333965844,
+      "grad_norm": 1.826885461807251,
+      "learning_rate": 1.1805800349540264e-11,
+      "loss": 0.1372,
+      "step": 4214
+    },
+    {
+      "epoch": 1.9995256166982922,
+      "grad_norm": 1.5402823686599731,
+      "learning_rate": 2.9514505228700473e-12,
+      "loss": 0.1458,
+      "step": 4215
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4708012342453003,
+      "learning_rate": 0.0,
+      "loss": 0.1223,
+      "step": 4216
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4216,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 8.236846139047936e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}