diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25263 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9991675915649276,
+  "eval_steps": 500,
+  "global_step": 3603,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000832408435072142,
+      "grad_norm": 5.852675437927246,
+      "learning_rate": 2.770083102493075e-08,
+      "loss": 0.8812,
+      "step": 1
+    },
+    {
+      "epoch": 0.001664816870144284,
+      "grad_norm": 5.757553577423096,
+      "learning_rate": 5.54016620498615e-08,
+      "loss": 0.8425,
+      "step": 2
+    },
+    {
+      "epoch": 0.0024972253052164264,
+      "grad_norm": 5.973855972290039,
+      "learning_rate": 8.310249307479226e-08,
+      "loss": 0.8658,
+      "step": 3
+    },
+    {
+      "epoch": 0.003329633740288568,
+      "grad_norm": 6.0041680335998535,
+      "learning_rate": 1.10803324099723e-07,
+      "loss": 0.8471,
+      "step": 4
+    },
+    {
+      "epoch": 0.004162042175360711,
+      "grad_norm": 5.798811912536621,
+      "learning_rate": 1.3850415512465375e-07,
+      "loss": 0.8696,
+      "step": 5
+    },
+    {
+      "epoch": 0.004994450610432853,
+      "grad_norm": 5.87969970703125,
+      "learning_rate": 1.662049861495845e-07,
+      "loss": 0.8461,
+      "step": 6
+    },
+    {
+      "epoch": 0.005826859045504994,
+      "grad_norm": 5.775433540344238,
+      "learning_rate": 1.9390581717451524e-07,
+      "loss": 0.859,
+      "step": 7
+    },
+    {
+      "epoch": 0.006659267480577136,
+      "grad_norm": 6.0282087326049805,
+      "learning_rate": 2.21606648199446e-07,
+      "loss": 0.8566,
+      "step": 8
+    },
+    {
+      "epoch": 0.007491675915649278,
+      "grad_norm": 5.845994472503662,
+      "learning_rate": 2.4930747922437677e-07,
+      "loss": 0.8533,
+      "step": 9
+    },
+    {
+      "epoch": 0.008324084350721421,
+      "grad_norm": 5.8350605964660645,
+      "learning_rate": 2.770083102493075e-07,
+      "loss": 0.8669,
+      "step": 10
+    },
+    {
+      "epoch": 0.009156492785793563,
+      "grad_norm": 5.791550636291504,
+      "learning_rate": 3.0470914127423823e-07,
+      "loss": 0.8701,
+      "step": 11
+    },
+    {
+      "epoch": 0.009988901220865706,
+      "grad_norm": 5.940352439880371,
+      "learning_rate": 3.32409972299169e-07,
+      "loss": 0.8909,
+      "step": 12
+    },
+    {
+      "epoch": 0.010821309655937847,
+      "grad_norm": 5.590038299560547,
+      "learning_rate": 3.601108033240998e-07,
+      "loss": 0.8443,
+      "step": 13
+    },
+    {
+      "epoch": 0.011653718091009988,
+      "grad_norm": 5.473329067230225,
+      "learning_rate": 3.878116343490305e-07,
+      "loss": 0.8345,
+      "step": 14
+    },
+    {
+      "epoch": 0.012486126526082131,
+      "grad_norm": 5.688844680786133,
+      "learning_rate": 4.155124653739612e-07,
+      "loss": 0.8813,
+      "step": 15
+    },
+    {
+      "epoch": 0.013318534961154272,
+      "grad_norm": 5.5658440589904785,
+      "learning_rate": 4.43213296398892e-07,
+      "loss": 0.8678,
+      "step": 16
+    },
+    {
+      "epoch": 0.014150943396226415,
+      "grad_norm": 5.2306599617004395,
+      "learning_rate": 4.7091412742382274e-07,
+      "loss": 0.8587,
+      "step": 17
+    },
+    {
+      "epoch": 0.014983351831298557,
+      "grad_norm": 5.296006202697754,
+      "learning_rate": 4.986149584487535e-07,
+      "loss": 0.8373,
+      "step": 18
+    },
+    {
+      "epoch": 0.015815760266370698,
+      "grad_norm": 4.465157508850098,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 0.7817,
+      "step": 19
+    },
+    {
+      "epoch": 0.016648168701442843,
+      "grad_norm": 4.338566780090332,
+      "learning_rate": 5.54016620498615e-07,
+      "loss": 0.7938,
+      "step": 20
+    },
+    {
+      "epoch": 0.017480577136514984,
+      "grad_norm": 4.399113655090332,
+      "learning_rate": 5.817174515235457e-07,
+      "loss": 0.8391,
+      "step": 21
+    },
+    {
+      "epoch": 0.018312985571587125,
+      "grad_norm": 4.371567726135254,
+      "learning_rate": 6.094182825484765e-07,
+      "loss": 0.8577,
+      "step": 22
+    },
+    {
+      "epoch": 0.019145394006659266,
+      "grad_norm": 3.9851796627044678,
+      "learning_rate": 6.371191135734073e-07,
+      "loss": 0.7962,
+      "step": 23
+    },
+    {
+      "epoch": 0.01997780244173141,
+      "grad_norm": 4.0287652015686035,
+      "learning_rate": 6.64819944598338e-07,
+      "loss": 0.7759,
+      "step": 24
+    },
+    {
+      "epoch": 0.020810210876803552,
+      "grad_norm": 3.2178382873535156,
+      "learning_rate": 6.925207756232688e-07,
+      "loss": 0.7951,
+      "step": 25
+    },
+    {
+      "epoch": 0.021642619311875694,
+      "grad_norm": 2.373326301574707,
+      "learning_rate": 7.202216066481996e-07,
+      "loss": 0.7714,
+      "step": 26
+    },
+    {
+      "epoch": 0.022475027746947835,
+      "grad_norm": 2.38152813911438,
+      "learning_rate": 7.479224376731302e-07,
+      "loss": 0.7554,
+      "step": 27
+    },
+    {
+      "epoch": 0.023307436182019976,
+      "grad_norm": 2.2056140899658203,
+      "learning_rate": 7.75623268698061e-07,
+      "loss": 0.7232,
+      "step": 28
+    },
+    {
+      "epoch": 0.02413984461709212,
+      "grad_norm": 2.3134679794311523,
+      "learning_rate": 8.033240997229917e-07,
+      "loss": 0.7865,
+      "step": 29
+    },
+    {
+      "epoch": 0.024972253052164262,
+      "grad_norm": 2.2102293968200684,
+      "learning_rate": 8.310249307479224e-07,
+      "loss": 0.7796,
+      "step": 30
+    },
+    {
+      "epoch": 0.025804661487236404,
+      "grad_norm": 1.9953457117080688,
+      "learning_rate": 8.587257617728533e-07,
+      "loss": 0.7473,
+      "step": 31
+    },
+    {
+      "epoch": 0.026637069922308545,
+      "grad_norm": 1.9345048666000366,
+      "learning_rate": 8.86426592797784e-07,
+      "loss": 0.7113,
+      "step": 32
+    },
+    {
+      "epoch": 0.02746947835738069,
+      "grad_norm": 1.8699673414230347,
+      "learning_rate": 9.141274238227148e-07,
+      "loss": 0.7649,
+      "step": 33
+    },
+    {
+      "epoch": 0.02830188679245283,
+      "grad_norm": 1.523240566253662,
+      "learning_rate": 9.418282548476455e-07,
+      "loss": 0.7485,
+      "step": 34
+    },
+    {
+      "epoch": 0.029134295227524972,
+      "grad_norm": 1.7488731145858765,
+      "learning_rate": 9.695290858725762e-07,
+      "loss": 0.7514,
+      "step": 35
+    },
+    {
+      "epoch": 0.029966703662597113,
+      "grad_norm": 1.9269617795944214,
+      "learning_rate": 9.97229916897507e-07,
+      "loss": 0.7005,
+      "step": 36
+    },
+    {
+      "epoch": 0.030799112097669258,
+      "grad_norm": 2.1249892711639404,
+      "learning_rate": 1.024930747922438e-06,
+      "loss": 0.7309,
+      "step": 37
+    },
+    {
+      "epoch": 0.031631520532741396,
+      "grad_norm": 2.0195670127868652,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.6937,
+      "step": 38
+    },
+    {
+      "epoch": 0.03246392896781354,
+      "grad_norm": 2.0689985752105713,
+      "learning_rate": 1.0803324099722992e-06,
+      "loss": 0.7177,
+      "step": 39
+    },
+    {
+      "epoch": 0.033296337402885685,
+      "grad_norm": 2.0408408641815186,
+      "learning_rate": 1.10803324099723e-06,
+      "loss": 0.7019,
+      "step": 40
+    },
+    {
+      "epoch": 0.03412874583795782,
+      "grad_norm": 1.8314168453216553,
+      "learning_rate": 1.1357340720221608e-06,
+      "loss": 0.6972,
+      "step": 41
+    },
+    {
+      "epoch": 0.03496115427302997,
+      "grad_norm": 1.612196922302246,
+      "learning_rate": 1.1634349030470915e-06,
+      "loss": 0.6842,
+      "step": 42
+    },
+    {
+      "epoch": 0.035793562708102106,
+      "grad_norm": 1.4777947664260864,
+      "learning_rate": 1.1911357340720223e-06,
+      "loss": 0.6992,
+      "step": 43
+    },
+    {
+      "epoch": 0.03662597114317425,
+      "grad_norm": 1.207147479057312,
+      "learning_rate": 1.218836565096953e-06,
+      "loss": 0.7215,
+      "step": 44
+    },
+    {
+      "epoch": 0.037458379578246395,
+      "grad_norm": 0.8992202877998352,
+      "learning_rate": 1.2465373961218838e-06,
+      "loss": 0.663,
+      "step": 45
+    },
+    {
+      "epoch": 0.03829078801331853,
+      "grad_norm": 0.8715615272521973,
+      "learning_rate": 1.2742382271468146e-06,
+      "loss": 0.696,
+      "step": 46
+    },
+    {
+      "epoch": 0.03912319644839068,
+      "grad_norm": 0.9824283719062805,
+      "learning_rate": 1.3019390581717452e-06,
+      "loss": 0.6709,
+      "step": 47
+    },
+    {
+      "epoch": 0.03995560488346282,
+      "grad_norm": 1.060660481452942,
+      "learning_rate": 1.329639889196676e-06,
+      "loss": 0.6799,
+      "step": 48
+    },
+    {
+      "epoch": 0.04078801331853496,
+      "grad_norm": 1.121728777885437,
+      "learning_rate": 1.357340720221607e-06,
+      "loss": 0.6862,
+      "step": 49
+    },
+    {
+      "epoch": 0.041620421753607105,
+      "grad_norm": 0.9326874017715454,
+      "learning_rate": 1.3850415512465375e-06,
+      "loss": 0.6305,
+      "step": 50
+    },
+    {
+      "epoch": 0.04245283018867924,
+      "grad_norm": 0.9469634294509888,
+      "learning_rate": 1.4127423822714684e-06,
+      "loss": 0.6881,
+      "step": 51
+    },
+    {
+      "epoch": 0.04328523862375139,
+      "grad_norm": 0.790377140045166,
+      "learning_rate": 1.4404432132963992e-06,
+      "loss": 0.6712,
+      "step": 52
+    },
+    {
+      "epoch": 0.04411764705882353,
+      "grad_norm": 0.8012718558311462,
+      "learning_rate": 1.4681440443213299e-06,
+      "loss": 0.6611,
+      "step": 53
+    },
+    {
+      "epoch": 0.04495005549389567,
+      "grad_norm": 0.7442087531089783,
+      "learning_rate": 1.4958448753462605e-06,
+      "loss": 0.6363,
+      "step": 54
+    },
+    {
+      "epoch": 0.045782463928967815,
+      "grad_norm": 0.6595564484596252,
+      "learning_rate": 1.5235457063711911e-06,
+      "loss": 0.6289,
+      "step": 55
+    },
+    {
+      "epoch": 0.04661487236403995,
+      "grad_norm": 0.7440547943115234,
+      "learning_rate": 1.551246537396122e-06,
+      "loss": 0.6602,
+      "step": 56
+    },
+    {
+      "epoch": 0.0474472807991121,
+      "grad_norm": 0.6962677240371704,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.6092,
+      "step": 57
+    },
+    {
+      "epoch": 0.04827968923418424,
+      "grad_norm": 0.5988077521324158,
+      "learning_rate": 1.6066481994459834e-06,
+      "loss": 0.6156,
+      "step": 58
+    },
+    {
+      "epoch": 0.04911209766925638,
+      "grad_norm": 0.579781711101532,
+      "learning_rate": 1.6343490304709143e-06,
+      "loss": 0.5902,
+      "step": 59
+    },
+    {
+      "epoch": 0.049944506104328525,
+      "grad_norm": 0.658952534198761,
+      "learning_rate": 1.6620498614958449e-06,
+      "loss": 0.6049,
+      "step": 60
+    },
+    {
+      "epoch": 0.05077691453940067,
+      "grad_norm": 0.7643985748291016,
+      "learning_rate": 1.6897506925207757e-06,
+      "loss": 0.6195,
+      "step": 61
+    },
+    {
+      "epoch": 0.05160932297447281,
+      "grad_norm": 0.6020765900611877,
+      "learning_rate": 1.7174515235457066e-06,
+      "loss": 0.6154,
+      "step": 62
+    },
+    {
+      "epoch": 0.05244173140954495,
+      "grad_norm": 0.5499704480171204,
+      "learning_rate": 1.7451523545706372e-06,
+      "loss": 0.6152,
+      "step": 63
+    },
+    {
+      "epoch": 0.05327413984461709,
+      "grad_norm": 0.5000544190406799,
+      "learning_rate": 1.772853185595568e-06,
+      "loss": 0.6228,
+      "step": 64
+    },
+    {
+      "epoch": 0.054106548279689234,
+      "grad_norm": 0.591923713684082,
+      "learning_rate": 1.8005540166204989e-06,
+      "loss": 0.6174,
+      "step": 65
+    },
+    {
+      "epoch": 0.05493895671476138,
+      "grad_norm": 0.6395013332366943,
+      "learning_rate": 1.8282548476454295e-06,
+      "loss": 0.6169,
+      "step": 66
+    },
+    {
+      "epoch": 0.05577136514983352,
+      "grad_norm": 0.5679120421409607,
+      "learning_rate": 1.8559556786703603e-06,
+      "loss": 0.6119,
+      "step": 67
+    },
+    {
+      "epoch": 0.05660377358490566,
+      "grad_norm": 0.5052242279052734,
+      "learning_rate": 1.883656509695291e-06,
+      "loss": 0.6275,
+      "step": 68
+    },
+    {
+      "epoch": 0.0574361820199778,
+      "grad_norm": 0.5107065439224243,
+      "learning_rate": 1.911357340720222e-06,
+      "loss": 0.6044,
+      "step": 69
+    },
+    {
+      "epoch": 0.058268590455049944,
+      "grad_norm": 0.5278398990631104,
+      "learning_rate": 1.9390581717451524e-06,
+      "loss": 0.6005,
+      "step": 70
+    },
+    {
+      "epoch": 0.05910099889012209,
+      "grad_norm": 0.5229373574256897,
+      "learning_rate": 1.9667590027700835e-06,
+      "loss": 0.605,
+      "step": 71
+    },
+    {
+      "epoch": 0.05993340732519423,
+      "grad_norm": 0.45075175166130066,
+      "learning_rate": 1.994459833795014e-06,
+      "loss": 0.5785,
+      "step": 72
+    },
+    {
+      "epoch": 0.06076581576026637,
+      "grad_norm": 0.44926148653030396,
+      "learning_rate": 2.0221606648199448e-06,
+      "loss": 0.584,
+      "step": 73
+    },
+    {
+      "epoch": 0.061598224195338516,
+      "grad_norm": 0.49701979756355286,
+      "learning_rate": 2.049861495844876e-06,
+      "loss": 0.5674,
+      "step": 74
+    },
+    {
+      "epoch": 0.062430632630410654,
+      "grad_norm": 0.4435327351093292,
+      "learning_rate": 2.077562326869806e-06,
+      "loss": 0.6039,
+      "step": 75
+    },
+    {
+      "epoch": 0.06326304106548279,
+      "grad_norm": 0.5081650614738464,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.6047,
+      "step": 76
+    },
+    {
+      "epoch": 0.06409544950055494,
+      "grad_norm": 0.47305235266685486,
+      "learning_rate": 2.1329639889196677e-06,
+      "loss": 0.6117,
+      "step": 77
+    },
+    {
+      "epoch": 0.06492785793562708,
+      "grad_norm": 0.501122772693634,
+      "learning_rate": 2.1606648199445983e-06,
+      "loss": 0.5932,
+      "step": 78
+    },
+    {
+      "epoch": 0.06576026637069922,
+      "grad_norm": 0.4220408797264099,
+      "learning_rate": 2.1883656509695294e-06,
+      "loss": 0.5489,
+      "step": 79
+    },
+    {
+      "epoch": 0.06659267480577137,
+      "grad_norm": 0.45031723380088806,
+      "learning_rate": 2.21606648199446e-06,
+      "loss": 0.5997,
+      "step": 80
+    },
+    {
+      "epoch": 0.06742508324084351,
+      "grad_norm": 0.4954591691493988,
+      "learning_rate": 2.2437673130193906e-06,
+      "loss": 0.6027,
+      "step": 81
+    },
+    {
+      "epoch": 0.06825749167591565,
+      "grad_norm": 0.471713662147522,
+      "learning_rate": 2.2714681440443217e-06,
+      "loss": 0.5924,
+      "step": 82
+    },
+    {
+      "epoch": 0.0690899001109878,
+      "grad_norm": 0.41615861654281616,
+      "learning_rate": 2.2991689750692523e-06,
+      "loss": 0.5723,
+      "step": 83
+    },
+    {
+      "epoch": 0.06992230854605994,
+      "grad_norm": 0.3945973515510559,
+      "learning_rate": 2.326869806094183e-06,
+      "loss": 0.5716,
+      "step": 84
+    },
+    {
+      "epoch": 0.07075471698113207,
+      "grad_norm": 0.4024456739425659,
+      "learning_rate": 2.3545706371191136e-06,
+      "loss": 0.5668,
+      "step": 85
+    },
+    {
+      "epoch": 0.07158712541620421,
+      "grad_norm": 0.4162571132183075,
+      "learning_rate": 2.3822714681440446e-06,
+      "loss": 0.5706,
+      "step": 86
+    },
+    {
+      "epoch": 0.07241953385127636,
+      "grad_norm": 0.41097456216812134,
+      "learning_rate": 2.4099722991689752e-06,
+      "loss": 0.6016,
+      "step": 87
+    },
+    {
+      "epoch": 0.0732519422863485,
+      "grad_norm": 0.4599824845790863,
+      "learning_rate": 2.437673130193906e-06,
+      "loss": 0.6005,
+      "step": 88
+    },
+    {
+      "epoch": 0.07408435072142064,
+      "grad_norm": 0.3909335136413574,
+      "learning_rate": 2.465373961218837e-06,
+      "loss": 0.5682,
+      "step": 89
+    },
+    {
+      "epoch": 0.07491675915649279,
+      "grad_norm": 0.4187878370285034,
+      "learning_rate": 2.4930747922437675e-06,
+      "loss": 0.5648,
+      "step": 90
+    },
+    {
+      "epoch": 0.07574916759156493,
+      "grad_norm": 0.41880685091018677,
+      "learning_rate": 2.520775623268698e-06,
+      "loss": 0.5944,
+      "step": 91
+    },
+    {
+      "epoch": 0.07658157602663707,
+      "grad_norm": 0.4465304911136627,
+      "learning_rate": 2.5484764542936292e-06,
+      "loss": 0.6203,
+      "step": 92
+    },
+    {
+      "epoch": 0.07741398446170922,
+      "grad_norm": 0.43194279074668884,
+      "learning_rate": 2.5761772853185594e-06,
+      "loss": 0.5817,
+      "step": 93
+    },
+    {
+      "epoch": 0.07824639289678136,
+      "grad_norm": 0.41561976075172424,
+      "learning_rate": 2.6038781163434905e-06,
+      "loss": 0.5741,
+      "step": 94
+    },
+    {
+      "epoch": 0.0790788013318535,
+      "grad_norm": 0.4317433536052704,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.5727,
+      "step": 95
+    },
+    {
+      "epoch": 0.07991120976692564,
+      "grad_norm": 0.43518805503845215,
+      "learning_rate": 2.659279778393352e-06,
+      "loss": 0.5826,
+      "step": 96
+    },
+    {
+      "epoch": 0.08074361820199778,
+      "grad_norm": 0.42625486850738525,
+      "learning_rate": 2.686980609418283e-06,
+      "loss": 0.5984,
+      "step": 97
+    },
+    {
+      "epoch": 0.08157602663706992,
+      "grad_norm": 0.41016989946365356,
+      "learning_rate": 2.714681440443214e-06,
+      "loss": 0.5642,
+      "step": 98
+    },
+    {
+      "epoch": 0.08240843507214206,
+      "grad_norm": 0.48712158203125,
+      "learning_rate": 2.742382271468144e-06,
+      "loss": 0.5968,
+      "step": 99
+    },
+    {
+      "epoch": 0.08324084350721421,
+      "grad_norm": 0.4600513279438019,
+      "learning_rate": 2.770083102493075e-06,
+      "loss": 0.5559,
+      "step": 100
+    },
+    {
+      "epoch": 0.08407325194228635,
+      "grad_norm": 0.4251794219017029,
+      "learning_rate": 2.7977839335180057e-06,
+      "loss": 0.5698,
+      "step": 101
+    },
+    {
+      "epoch": 0.08490566037735849,
+      "grad_norm": 0.4187605082988739,
+      "learning_rate": 2.8254847645429368e-06,
+      "loss": 0.5463,
+      "step": 102
+    },
+    {
+      "epoch": 0.08573806881243064,
+      "grad_norm": 0.45978236198425293,
+      "learning_rate": 2.8531855955678674e-06,
+      "loss": 0.5605,
+      "step": 103
+    },
+    {
+      "epoch": 0.08657047724750278,
+      "grad_norm": 0.4345948398113251,
+      "learning_rate": 2.8808864265927985e-06,
+      "loss": 0.548,
+      "step": 104
+    },
+    {
+      "epoch": 0.08740288568257491,
+      "grad_norm": 0.4089731276035309,
+      "learning_rate": 2.9085872576177287e-06,
+      "loss": 0.5333,
+      "step": 105
+    },
+    {
+      "epoch": 0.08823529411764706,
+      "grad_norm": 0.4466627240180969,
+      "learning_rate": 2.9362880886426597e-06,
+      "loss": 0.5653,
+      "step": 106
+    },
+    {
+      "epoch": 0.0890677025527192,
+      "grad_norm": 0.4312840700149536,
+      "learning_rate": 2.9639889196675903e-06,
+      "loss": 0.544,
+      "step": 107
+    },
+    {
+      "epoch": 0.08990011098779134,
+      "grad_norm": 0.3958422541618347,
+      "learning_rate": 2.991689750692521e-06,
+      "loss": 0.5314,
+      "step": 108
+    },
+    {
+      "epoch": 0.09073251942286349,
+      "grad_norm": 0.4052140712738037,
+      "learning_rate": 3.0193905817174516e-06,
+      "loss": 0.5724,
+      "step": 109
+    },
+    {
+      "epoch": 0.09156492785793563,
+      "grad_norm": 0.4097498059272766,
+      "learning_rate": 3.0470914127423822e-06,
+      "loss": 0.5559,
+      "step": 110
+    },
+    {
+      "epoch": 0.09239733629300777,
+      "grad_norm": 0.43090808391571045,
+      "learning_rate": 3.0747922437673133e-06,
+      "loss": 0.5456,
+      "step": 111
+    },
+    {
+      "epoch": 0.0932297447280799,
+      "grad_norm": 0.43639659881591797,
+      "learning_rate": 3.102493074792244e-06,
+      "loss": 0.5572,
+      "step": 112
+    },
+    {
+      "epoch": 0.09406215316315206,
+      "grad_norm": 0.39143499732017517,
+      "learning_rate": 3.130193905817175e-06,
+      "loss": 0.5627,
+      "step": 113
+    },
+    {
+      "epoch": 0.0948945615982242,
+      "grad_norm": 0.4211403727531433,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.5701,
+      "step": 114
+    },
+    {
+      "epoch": 0.09572697003329633,
+      "grad_norm": 0.42813047766685486,
+      "learning_rate": 3.1855955678670362e-06,
+      "loss": 0.5594,
+      "step": 115
+    },
+    {
+      "epoch": 0.09655937846836848,
+      "grad_norm": 0.4047672152519226,
+      "learning_rate": 3.213296398891967e-06,
+      "loss": 0.5467,
+      "step": 116
+    },
+    {
+      "epoch": 0.09739178690344062,
+      "grad_norm": 0.4235449731349945,
+      "learning_rate": 3.240997229916898e-06,
+      "loss": 0.5596,
+      "step": 117
+    },
+    {
+      "epoch": 0.09822419533851276,
+      "grad_norm": 0.5001515746116638,
+      "learning_rate": 3.2686980609418285e-06,
+      "loss": 0.55,
+      "step": 118
+    },
+    {
+      "epoch": 0.09905660377358491,
+      "grad_norm": 0.4500423073768616,
+      "learning_rate": 3.2963988919667596e-06,
+      "loss": 0.5578,
+      "step": 119
+    },
+    {
+      "epoch": 0.09988901220865705,
+      "grad_norm": 0.43707484006881714,
+      "learning_rate": 3.3240997229916898e-06,
+      "loss": 0.544,
+      "step": 120
+    },
+    {
+      "epoch": 0.10072142064372919,
+      "grad_norm": 0.4369399845600128,
+      "learning_rate": 3.351800554016621e-06,
+      "loss": 0.5551,
+      "step": 121
+    },
+    {
+      "epoch": 0.10155382907880134,
+      "grad_norm": 0.4763941764831543,
+      "learning_rate": 3.3795013850415515e-06,
+      "loss": 0.5628,
+      "step": 122
+    },
+    {
+      "epoch": 0.10238623751387348,
+      "grad_norm": 0.4882802963256836,
+      "learning_rate": 3.4072022160664825e-06,
+      "loss": 0.5585,
+      "step": 123
+    },
+    {
+      "epoch": 0.10321864594894561,
+      "grad_norm": 0.441522479057312,
+      "learning_rate": 3.434903047091413e-06,
+      "loss": 0.5494,
+      "step": 124
+    },
+    {
+      "epoch": 0.10405105438401775,
+      "grad_norm": 0.4602923095226288,
+      "learning_rate": 3.462603878116344e-06,
+      "loss": 0.5152,
+      "step": 125
+    },
+    {
+      "epoch": 0.1048834628190899,
+      "grad_norm": 0.5620714426040649,
+      "learning_rate": 3.4903047091412744e-06,
+      "loss": 0.5424,
+      "step": 126
+    },
+    {
+      "epoch": 0.10571587125416204,
+      "grad_norm": 0.5059617161750793,
+      "learning_rate": 3.5180055401662054e-06,
+      "loss": 0.5758,
+      "step": 127
+    },
+    {
+      "epoch": 0.10654827968923418,
+      "grad_norm": 0.44737622141838074,
+      "learning_rate": 3.545706371191136e-06,
+      "loss": 0.53,
+      "step": 128
+    },
+    {
+      "epoch": 0.10738068812430633,
+      "grad_norm": 0.5812498331069946,
+      "learning_rate": 3.5734072022160667e-06,
+      "loss": 0.5437,
+      "step": 129
+    },
+    {
+      "epoch": 0.10821309655937847,
+      "grad_norm": 0.4357758164405823,
+      "learning_rate": 3.6011080332409978e-06,
+      "loss": 0.5482,
+      "step": 130
+    },
+    {
+      "epoch": 0.1090455049944506,
+      "grad_norm": 0.46321603655815125,
+      "learning_rate": 3.628808864265928e-06,
+      "loss": 0.5671,
+      "step": 131
+    },
+    {
+      "epoch": 0.10987791342952276,
+      "grad_norm": 0.4342377185821533,
+      "learning_rate": 3.656509695290859e-06,
+      "loss": 0.5532,
+      "step": 132
+    },
+    {
+      "epoch": 0.1107103218645949,
+      "grad_norm": 0.4561794102191925,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.5395,
+      "step": 133
+    },
+    {
+      "epoch": 0.11154273029966703,
+      "grad_norm": 0.40543898940086365,
+      "learning_rate": 3.7119113573407207e-06,
+      "loss": 0.5023,
+      "step": 134
+    },
+    {
+      "epoch": 0.11237513873473919,
+      "grad_norm": 0.4146581292152405,
+      "learning_rate": 3.739612188365651e-06,
+      "loss": 0.5455,
+      "step": 135
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.3970547914505005,
+      "learning_rate": 3.767313019390582e-06,
+      "loss": 0.5341,
+      "step": 136
+    },
+    {
+      "epoch": 0.11403995560488346,
+      "grad_norm": 0.45214366912841797,
+      "learning_rate": 3.7950138504155126e-06,
+      "loss": 0.5509,
+      "step": 137
+    },
+    {
+      "epoch": 0.1148723640399556,
+      "grad_norm": 0.4272449314594269,
+      "learning_rate": 3.822714681440444e-06,
+      "loss": 0.503,
+      "step": 138
+    },
+    {
+      "epoch": 0.11570477247502775,
+      "grad_norm": 0.4434300661087036,
+      "learning_rate": 3.850415512465374e-06,
+      "loss": 0.5212,
+      "step": 139
+    },
+    {
+      "epoch": 0.11653718091009989,
+      "grad_norm": 0.4421811103820801,
+      "learning_rate": 3.878116343490305e-06,
+      "loss": 0.5389,
+      "step": 140
+    },
+    {
+      "epoch": 0.11736958934517203,
+      "grad_norm": 0.42131364345550537,
+      "learning_rate": 3.9058171745152355e-06,
+      "loss": 0.5443,
+      "step": 141
+    },
+    {
+      "epoch": 0.11820199778024418,
+      "grad_norm": 0.47368770837783813,
+      "learning_rate": 3.933518005540167e-06,
+      "loss": 0.5534,
+      "step": 142
+    },
+    {
+      "epoch": 0.11903440621531632,
+      "grad_norm": 0.4222336709499359,
+      "learning_rate": 3.961218836565098e-06,
+      "loss": 0.5484,
+      "step": 143
+    },
+    {
+      "epoch": 0.11986681465038845,
+      "grad_norm": 0.4217516779899597,
+      "learning_rate": 3.988919667590028e-06,
+      "loss": 0.5359,
+      "step": 144
+    },
+    {
+      "epoch": 0.1206992230854606,
+      "grad_norm": 0.46646854281425476,
+      "learning_rate": 4.016620498614959e-06,
+      "loss": 0.5608,
+      "step": 145
+    },
+    {
+      "epoch": 0.12153163152053274,
+      "grad_norm": 0.4836723506450653,
+      "learning_rate": 4.0443213296398895e-06,
+      "loss": 0.5471,
+      "step": 146
+    },
+    {
+      "epoch": 0.12236403995560488,
+      "grad_norm": 0.42951828241348267,
+      "learning_rate": 4.07202216066482e-06,
+      "loss": 0.5387,
+      "step": 147
+    },
+    {
+      "epoch": 0.12319644839067703,
+      "grad_norm": 0.4030103385448456,
+      "learning_rate": 4.099722991689752e-06,
+      "loss": 0.5426,
+      "step": 148
+    },
+    {
+      "epoch": 0.12402885682574917,
+      "grad_norm": 0.4542948305606842,
+      "learning_rate": 4.127423822714681e-06,
+      "loss": 0.5331,
+      "step": 149
+    },
+    {
+      "epoch": 0.12486126526082131,
+      "grad_norm": 0.481719046831131,
+      "learning_rate": 4.155124653739612e-06,
+      "loss": 0.5384,
+      "step": 150
+    },
+    {
+      "epoch": 0.12569367369589346,
+      "grad_norm": 0.4303570091724396,
+      "learning_rate": 4.1828254847645435e-06,
+      "loss": 0.5398,
+      "step": 151
+    },
+    {
+      "epoch": 0.12652608213096558,
+      "grad_norm": 0.4159233272075653,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.5373,
+      "step": 152
+    },
+    {
+      "epoch": 0.12735849056603774,
+      "grad_norm": 0.37868037819862366,
+      "learning_rate": 4.238227146814405e-06,
+      "loss": 0.5276,
+      "step": 153
+    },
+    {
+      "epoch": 0.1281908990011099,
+      "grad_norm": 0.4570467472076416,
+      "learning_rate": 4.265927977839335e-06,
+      "loss": 0.5514,
+      "step": 154
+    },
+    {
+      "epoch": 0.129023307436182,
+      "grad_norm": 0.39249682426452637,
+      "learning_rate": 4.293628808864266e-06,
+      "loss": 0.5566,
+      "step": 155
+    },
+    {
+      "epoch": 0.12985571587125416,
+      "grad_norm": 0.37955617904663086,
+      "learning_rate": 4.321329639889197e-06,
+      "loss": 0.5411,
+      "step": 156
+    },
+    {
+      "epoch": 0.13068812430632631,
+      "grad_norm": 0.42655855417251587,
+      "learning_rate": 4.349030470914128e-06,
+      "loss": 0.538,
+      "step": 157
+    },
+    {
+      "epoch": 0.13152053274139844,
+      "grad_norm": 0.41612133383750916,
+      "learning_rate": 4.376731301939059e-06,
+      "loss": 0.5628,
+      "step": 158
+    },
+    {
+      "epoch": 0.1323529411764706,
+      "grad_norm": 0.4542618989944458,
+      "learning_rate": 4.404432132963989e-06,
+      "loss": 0.5541,
+      "step": 159
+    },
+    {
+      "epoch": 0.13318534961154274,
+      "grad_norm": 0.4357217252254486,
+      "learning_rate": 4.43213296398892e-06,
+      "loss": 0.536,
+      "step": 160
+    },
+    {
+      "epoch": 0.13401775804661487,
+      "grad_norm": 0.4199989140033722,
+      "learning_rate": 4.459833795013851e-06,
+      "loss": 0.5207,
+      "step": 161
+    },
+    {
+      "epoch": 0.13485016648168702,
+      "grad_norm": 0.43055838346481323,
+      "learning_rate": 4.487534626038781e-06,
+      "loss": 0.5649,
+      "step": 162
+    },
+    {
+      "epoch": 0.13568257491675917,
+      "grad_norm": 0.4102376103401184,
+      "learning_rate": 4.515235457063713e-06,
+      "loss": 0.5379,
+      "step": 163
+    },
+    {
+      "epoch": 0.1365149833518313,
+      "grad_norm": 0.40180259943008423,
+      "learning_rate": 4.542936288088643e-06,
+      "loss": 0.5461,
+      "step": 164
+    },
+    {
+      "epoch": 0.13734739178690344,
+      "grad_norm": 0.4794843792915344,
+      "learning_rate": 4.570637119113574e-06,
+      "loss": 0.5279,
+      "step": 165
+    },
+    {
+      "epoch": 0.1381798002219756,
+      "grad_norm": 0.5056953430175781,
+      "learning_rate": 4.598337950138505e-06,
+      "loss": 0.5348,
+      "step": 166
+    },
+    {
+      "epoch": 0.13901220865704772,
+      "grad_norm": 0.45556968450546265,
+      "learning_rate": 4.626038781163435e-06,
+      "loss": 0.4995,
+      "step": 167
+    },
+    {
+      "epoch": 0.13984461709211987,
+      "grad_norm": 0.40867018699645996,
+      "learning_rate": 4.653739612188366e-06,
+      "loss": 0.5301,
+      "step": 168
+    },
+    {
+      "epoch": 0.140677025527192,
+      "grad_norm": 0.44523248076438904,
+      "learning_rate": 4.681440443213297e-06,
+      "loss": 0.5097,
+      "step": 169
+    },
+    {
+      "epoch": 0.14150943396226415,
+      "grad_norm": 0.44204166531562805,
+      "learning_rate": 4.709141274238227e-06,
+      "loss": 0.5348,
+      "step": 170
+    },
+    {
+      "epoch": 0.1423418423973363,
+      "grad_norm": 0.39262545108795166,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.5127,
+      "step": 171
+    },
+    {
+      "epoch": 0.14317425083240842,
+      "grad_norm": 0.4378338158130646,
+      "learning_rate": 4.764542936288089e-06,
+      "loss": 0.498,
+      "step": 172
+    },
+    {
+      "epoch": 0.14400665926748057,
+      "grad_norm": 0.4008118510246277,
+      "learning_rate": 4.79224376731302e-06,
+      "loss": 0.5389,
+      "step": 173
+    },
+    {
+      "epoch": 0.14483906770255273,
+      "grad_norm": 0.41733667254447937,
+      "learning_rate": 4.8199445983379505e-06,
+      "loss": 0.5216,
+      "step": 174
+    },
+    {
+      "epoch": 0.14567147613762485,
+      "grad_norm": 0.4296148121356964,
+      "learning_rate": 4.847645429362881e-06,
+      "loss": 0.5328,
+      "step": 175
+    },
+    {
+      "epoch": 0.146503884572697,
+      "grad_norm": 0.4301013648509979,
+      "learning_rate": 4.875346260387812e-06,
+      "loss": 0.5292,
+      "step": 176
+    },
+    {
+      "epoch": 0.14733629300776915,
+      "grad_norm": 0.46143457293510437,
+      "learning_rate": 4.903047091412742e-06,
+      "loss": 0.5336,
+      "step": 177
+    },
+    {
+      "epoch": 0.14816870144284128,
+      "grad_norm": 0.4127751886844635,
+      "learning_rate": 4.930747922437674e-06,
+      "loss": 0.5322,
+      "step": 178
+    },
+    {
+      "epoch": 0.14900110987791343,
+      "grad_norm": 0.4373955726623535,
+      "learning_rate": 4.9584487534626045e-06,
+      "loss": 0.5486,
+      "step": 179
+    },
+    {
+      "epoch": 0.14983351831298558,
+      "grad_norm": 0.4892722964286804,
+      "learning_rate": 4.986149584487535e-06,
+      "loss": 0.5416,
+      "step": 180
+    },
+    {
+      "epoch": 0.1506659267480577,
+      "grad_norm": 0.40266159176826477,
+      "learning_rate": 5.013850415512466e-06,
+      "loss": 0.5149,
+      "step": 181
+    },
+    {
+      "epoch": 0.15149833518312986,
+      "grad_norm": 0.4328942894935608,
+      "learning_rate": 5.041551246537396e-06,
+      "loss": 0.5364,
+      "step": 182
+    },
+    {
+      "epoch": 0.152330743618202,
+      "grad_norm": 0.46720901131629944,
+      "learning_rate": 5.069252077562328e-06,
+      "loss": 0.5481,
+      "step": 183
+    },
+    {
+      "epoch": 0.15316315205327413,
+      "grad_norm": 0.5006843209266663,
+      "learning_rate": 5.0969529085872585e-06,
+      "loss": 0.5599,
+      "step": 184
+    },
+    {
+      "epoch": 0.15399556048834628,
+      "grad_norm": 0.4212501347064972,
+      "learning_rate": 5.124653739612189e-06,
+      "loss": 0.5316,
+      "step": 185
+    },
+    {
+      "epoch": 0.15482796892341844,
+      "grad_norm": 0.4488828778266907,
+      "learning_rate": 5.152354570637119e-06,
+      "loss": 0.5163,
+      "step": 186
+    },
+    {
+      "epoch": 0.15566037735849056,
+      "grad_norm": 0.40712884068489075,
+      "learning_rate": 5.180055401662051e-06,
+      "loss": 0.529,
+      "step": 187
+    },
+    {
+      "epoch": 0.1564927857935627,
+      "grad_norm": 0.4387264847755432,
+      "learning_rate": 5.207756232686981e-06,
+      "loss": 0.5097,
+      "step": 188
+    },
+    {
+      "epoch": 0.15732519422863486,
+      "grad_norm": 0.4043707549571991,
+      "learning_rate": 5.235457063711912e-06,
+      "loss": 0.4989,
+      "step": 189
+    },
+    {
+      "epoch": 0.158157602663707,
+      "grad_norm": 0.4202433228492737,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.5577,
+      "step": 190
+    },
+    {
+      "epoch": 0.15899001109877914,
+      "grad_norm": 0.4169784188270569,
+      "learning_rate": 5.290858725761774e-06,
+      "loss": 0.5331,
+      "step": 191
+    },
+    {
+      "epoch": 0.1598224195338513,
+      "grad_norm": 0.42586076259613037,
+      "learning_rate": 5.318559556786704e-06,
+      "loss": 0.5323,
+      "step": 192
+    },
+    {
+      "epoch": 0.1606548279689234,
+      "grad_norm": 0.4811337888240814,
+      "learning_rate": 5.346260387811635e-06,
+      "loss": 0.5305,
+      "step": 193
+    },
+    {
+      "epoch": 0.16148723640399557,
+      "grad_norm": 0.4163426458835602,
+      "learning_rate": 5.373961218836566e-06,
+      "loss": 0.5038,
+      "step": 194
+    },
+    {
+      "epoch": 0.1623196448390677,
+      "grad_norm": 0.4817521572113037,
+      "learning_rate": 5.401662049861495e-06,
+      "loss": 0.5423,
+      "step": 195
+    },
+    {
+      "epoch": 0.16315205327413984,
+      "grad_norm": 0.49575290083885193,
+      "learning_rate": 5.429362880886428e-06,
+      "loss": 0.5233,
+      "step": 196
+    },
+    {
+      "epoch": 0.163984461709212,
+      "grad_norm": 0.4486011266708374,
+      "learning_rate": 5.4570637119113575e-06,
+      "loss": 0.5004,
+      "step": 197
+    },
+    {
+      "epoch": 0.16481687014428412,
+      "grad_norm": 0.4150060713291168,
+      "learning_rate": 5.484764542936288e-06,
+      "loss": 0.52,
+      "step": 198
+    },
+    {
+      "epoch": 0.16564927857935627,
+      "grad_norm": 0.4181516468524933,
+      "learning_rate": 5.512465373961219e-06,
+      "loss": 0.498,
+      "step": 199
+    },
+    {
+      "epoch": 0.16648168701442842,
+      "grad_norm": 0.5380068421363831,
+      "learning_rate": 5.54016620498615e-06,
+      "loss": 0.5458,
+      "step": 200
+    },
+    {
+      "epoch": 0.16731409544950054,
+      "grad_norm": 0.4043119549751282,
+      "learning_rate": 5.567867036011081e-06,
+      "loss": 0.5297,
+      "step": 201
+    },
+    {
+      "epoch": 0.1681465038845727,
+      "grad_norm": 0.481564462184906,
+      "learning_rate": 5.5955678670360115e-06,
+      "loss": 0.5275,
+      "step": 202
+    },
+    {
+      "epoch": 0.16897891231964485,
+      "grad_norm": 0.4668130874633789,
+      "learning_rate": 5.623268698060942e-06,
+      "loss": 0.5195,
+      "step": 203
+    },
+    {
+      "epoch": 0.16981132075471697,
+      "grad_norm": 0.4354064166545868,
+      "learning_rate": 5.6509695290858736e-06,
+      "loss": 0.5147,
+      "step": 204
+    },
+    {
+      "epoch": 0.17064372918978912,
+      "grad_norm": 0.45728379487991333,
+      "learning_rate": 5.678670360110804e-06,
+      "loss": 0.5487,
+      "step": 205
+    },
+    {
+      "epoch": 0.17147613762486127,
+      "grad_norm": 0.4586714804172516,
+      "learning_rate": 5.706371191135735e-06,
+      "loss": 0.5398,
+      "step": 206
+    },
+    {
+      "epoch": 0.1723085460599334,
+      "grad_norm": 0.4355411231517792,
+      "learning_rate": 5.734072022160665e-06,
+      "loss": 0.5143,
+      "step": 207
+    },
+    {
+      "epoch": 0.17314095449500555,
+      "grad_norm": 0.4318200647830963,
+      "learning_rate": 5.761772853185597e-06,
+      "loss": 0.5153,
+      "step": 208
+    },
+    {
+      "epoch": 0.1739733629300777,
+      "grad_norm": 0.45262500643730164,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.5367,
+      "step": 209
+    },
+    {
+      "epoch": 0.17480577136514983,
+      "grad_norm": 0.41200655698776245,
+      "learning_rate": 5.817174515235457e-06,
+      "loss": 0.4899,
+      "step": 210
+    },
+    {
+      "epoch": 0.17563817980022198,
+      "grad_norm": 0.4550832211971283,
+      "learning_rate": 5.844875346260388e-06,
+      "loss": 0.5312,
+      "step": 211
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 0.48159360885620117,
+      "learning_rate": 5.8725761772853194e-06,
+      "loss": 0.5105,
+      "step": 212
+    },
+    {
+      "epoch": 0.17730299667036625,
+      "grad_norm": 0.4670851230621338,
+      "learning_rate": 5.90027700831025e-06,
+      "loss": 0.5193,
+      "step": 213
+    },
+    {
+      "epoch": 0.1781354051054384,
+      "grad_norm": 0.4833730161190033,
+      "learning_rate": 5.927977839335181e-06,
+      "loss": 0.5374,
+      "step": 214
+    },
+    {
+      "epoch": 0.17896781354051056,
+      "grad_norm": 0.41837170720100403,
+      "learning_rate": 5.955678670360111e-06,
+      "loss": 0.549,
+      "step": 215
+    },
+    {
+      "epoch": 0.17980022197558268,
+      "grad_norm": 0.45265597105026245,
+      "learning_rate": 5.983379501385042e-06,
+      "loss": 0.5009,
+      "step": 216
+    },
+    {
+      "epoch": 0.18063263041065483,
+      "grad_norm": 0.4560681879520416,
+      "learning_rate": 6.011080332409973e-06,
+      "loss": 0.5201,
+      "step": 217
+    },
+    {
+      "epoch": 0.18146503884572698,
+      "grad_norm": 0.41413623094558716,
+      "learning_rate": 6.038781163434903e-06,
+      "loss": 0.5098,
+      "step": 218
+    },
+    {
+      "epoch": 0.1822974472807991,
+      "grad_norm": 0.45979252457618713,
+      "learning_rate": 6.066481994459834e-06,
+      "loss": 0.5127,
+      "step": 219
+    },
+    {
+      "epoch": 0.18312985571587126,
+      "grad_norm": 0.42055100202560425,
+      "learning_rate": 6.0941828254847645e-06,
+      "loss": 0.4953,
+      "step": 220
+    },
+    {
+      "epoch": 0.18396226415094338,
+      "grad_norm": 0.4109812378883362,
+      "learning_rate": 6.121883656509696e-06,
+      "loss": 0.5282,
+      "step": 221
+    },
+    {
+      "epoch": 0.18479467258601553,
+      "grad_norm": 0.405984103679657,
+      "learning_rate": 6.1495844875346266e-06,
+      "loss": 0.5089,
+      "step": 222
+    },
+    {
+      "epoch": 0.1856270810210877,
+      "grad_norm": 0.388094037771225,
+      "learning_rate": 6.177285318559557e-06,
+      "loss": 0.5212,
+      "step": 223
+    },
+    {
+      "epoch": 0.1864594894561598,
+      "grad_norm": 0.4275857210159302,
+      "learning_rate": 6.204986149584488e-06,
+      "loss": 0.5238,
+      "step": 224
+    },
+    {
+      "epoch": 0.18729189789123196,
+      "grad_norm": 0.4158404469490051,
+      "learning_rate": 6.232686980609419e-06,
+      "loss": 0.5246,
+      "step": 225
+    },
+    {
+      "epoch": 0.1881243063263041,
+      "grad_norm": 0.44389262795448303,
+      "learning_rate": 6.26038781163435e-06,
+      "loss": 0.5089,
+      "step": 226
+    },
+    {
+      "epoch": 0.18895671476137624,
+      "grad_norm": 0.4087418019771576,
+      "learning_rate": 6.2880886426592805e-06,
+      "loss": 0.5089,
+      "step": 227
+    },
+    {
+      "epoch": 0.1897891231964484,
+      "grad_norm": 0.5286365747451782,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.5551,
+      "step": 228
+    },
+    {
+      "epoch": 0.19062153163152054,
+      "grad_norm": 0.4365133047103882,
+      "learning_rate": 6.343490304709143e-06,
+      "loss": 0.511,
+      "step": 229
+    },
+    {
+      "epoch": 0.19145394006659266,
+      "grad_norm": 0.44955843687057495,
+      "learning_rate": 6.3711911357340724e-06,
+      "loss": 0.4946,
+      "step": 230
+    },
+    {
+      "epoch": 0.19228634850166482,
+      "grad_norm": 0.41679856181144714,
+      "learning_rate": 6.398891966759003e-06,
+      "loss": 0.5041,
+      "step": 231
+    },
+    {
+      "epoch": 0.19311875693673697,
+      "grad_norm": 0.4691244959831238,
+      "learning_rate": 6.426592797783934e-06,
+      "loss": 0.4901,
+      "step": 232
+    },
+    {
+      "epoch": 0.1939511653718091,
+      "grad_norm": 0.49247369170188904,
+      "learning_rate": 6.454293628808865e-06,
+      "loss": 0.5176,
+      "step": 233
+    },
+    {
+      "epoch": 0.19478357380688124,
+      "grad_norm": 0.4147561192512512,
+      "learning_rate": 6.481994459833796e-06,
+      "loss": 0.4995,
+      "step": 234
+    },
+    {
+      "epoch": 0.1956159822419534,
+      "grad_norm": 0.4889633059501648,
+      "learning_rate": 6.509695290858726e-06,
+      "loss": 0.5404,
+      "step": 235
+    },
+    {
+      "epoch": 0.19644839067702552,
+      "grad_norm": 0.5010858774185181,
+      "learning_rate": 6.537396121883657e-06,
+      "loss": 0.5358,
+      "step": 236
+    },
+    {
+      "epoch": 0.19728079911209767,
+      "grad_norm": 0.38324692845344543,
+      "learning_rate": 6.565096952908588e-06,
+      "loss": 0.4914,
+      "step": 237
+    },
+    {
+      "epoch": 0.19811320754716982,
+      "grad_norm": 0.489378958940506,
+      "learning_rate": 6.592797783933519e-06,
+      "loss": 0.4954,
+      "step": 238
+    },
+    {
+      "epoch": 0.19894561598224195,
+      "grad_norm": 0.4189784526824951,
+      "learning_rate": 6.62049861495845e-06,
+      "loss": 0.5062,
+      "step": 239
+    },
+    {
+      "epoch": 0.1997780244173141,
+      "grad_norm": 0.42447060346603394,
+      "learning_rate": 6.6481994459833796e-06,
+      "loss": 0.5128,
+      "step": 240
+    },
+    {
+      "epoch": 0.20061043285238625,
+      "grad_norm": 0.4346916079521179,
+      "learning_rate": 6.67590027700831e-06,
+      "loss": 0.5257,
+      "step": 241
+    },
+    {
+      "epoch": 0.20144284128745837,
+      "grad_norm": 0.4251374304294586,
+      "learning_rate": 6.703601108033242e-06,
+      "loss": 0.5267,
+      "step": 242
+    },
+    {
+      "epoch": 0.20227524972253053,
+      "grad_norm": 0.40433380007743835,
+      "learning_rate": 6.731301939058172e-06,
+      "loss": 0.4952,
+      "step": 243
+    },
+    {
+      "epoch": 0.20310765815760268,
+      "grad_norm": 0.44367408752441406,
+      "learning_rate": 6.759002770083103e-06,
+      "loss": 0.514,
+      "step": 244
+    },
+    {
+      "epoch": 0.2039400665926748,
+      "grad_norm": 0.45639634132385254,
+      "learning_rate": 6.7867036011080335e-06,
+      "loss": 0.5206,
+      "step": 245
+    },
+    {
+      "epoch": 0.20477247502774695,
+      "grad_norm": 0.4419868290424347,
+      "learning_rate": 6.814404432132965e-06,
+      "loss": 0.5412,
+      "step": 246
+    },
+    {
+      "epoch": 0.20560488346281908,
+      "grad_norm": 0.4839082658290863,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.5237,
+      "step": 247
+    },
+    {
+      "epoch": 0.20643729189789123,
+      "grad_norm": 0.42248091101646423,
+      "learning_rate": 6.869806094182826e-06,
+      "loss": 0.5242,
+      "step": 248
+    },
+    {
+      "epoch": 0.20726970033296338,
+      "grad_norm": 0.46019700169563293,
+      "learning_rate": 6.897506925207756e-06,
+      "loss": 0.4951,
+      "step": 249
+    },
+    {
+      "epoch": 0.2081021087680355,
+      "grad_norm": 0.5231823325157166,
+      "learning_rate": 6.925207756232688e-06,
+      "loss": 0.5459,
+      "step": 250
+    },
+    {
+      "epoch": 0.20893451720310766,
+      "grad_norm": 0.45339393615722656,
+      "learning_rate": 6.952908587257618e-06,
+      "loss": 0.5099,
+      "step": 251
+    },
+    {
+      "epoch": 0.2097669256381798,
+      "grad_norm": 0.4783773124217987,
+      "learning_rate": 6.980609418282549e-06,
+      "loss": 0.4946,
+      "step": 252
+    },
+    {
+      "epoch": 0.21059933407325193,
+      "grad_norm": 0.46553727984428406,
+      "learning_rate": 7.008310249307479e-06,
+      "loss": 0.482,
+      "step": 253
+    },
+    {
+      "epoch": 0.21143174250832408,
+      "grad_norm": 0.4713238775730133,
+      "learning_rate": 7.036011080332411e-06,
+      "loss": 0.4964,
+      "step": 254
+    },
+    {
+      "epoch": 0.21226415094339623,
+      "grad_norm": 0.48873159289360046,
+      "learning_rate": 7.0637119113573415e-06,
+      "loss": 0.5337,
+      "step": 255
+    },
+    {
+      "epoch": 0.21309655937846836,
+      "grad_norm": 0.45275193452835083,
+      "learning_rate": 7.091412742382272e-06,
+      "loss": 0.5168,
+      "step": 256
+    },
+    {
+      "epoch": 0.2139289678135405,
+      "grad_norm": 0.45386070013046265,
+      "learning_rate": 7.119113573407203e-06,
+      "loss": 0.5093,
+      "step": 257
+    },
+    {
+      "epoch": 0.21476137624861266,
+      "grad_norm": 0.46968305110931396,
+      "learning_rate": 7.146814404432133e-06,
+      "loss": 0.5128,
+      "step": 258
+    },
+    {
+      "epoch": 0.21559378468368479,
+      "grad_norm": 0.4741690754890442,
+      "learning_rate": 7.174515235457065e-06,
+      "loss": 0.5301,
+      "step": 259
+    },
+    {
+      "epoch": 0.21642619311875694,
+      "grad_norm": 0.48019328713417053,
+      "learning_rate": 7.2022160664819955e-06,
+      "loss": 0.4837,
+      "step": 260
+    },
+    {
+      "epoch": 0.2172586015538291,
+      "grad_norm": 0.424640029668808,
+      "learning_rate": 7.229916897506925e-06,
+      "loss": 0.4917,
+      "step": 261
+    },
+    {
+      "epoch": 0.2180910099889012,
+      "grad_norm": 0.5357044339179993,
+      "learning_rate": 7.257617728531856e-06,
+      "loss": 0.5159,
+      "step": 262
+    },
+    {
+      "epoch": 0.21892341842397336,
+      "grad_norm": 0.46231383085250854,
+      "learning_rate": 7.285318559556787e-06,
+      "loss": 0.5235,
+      "step": 263
+    },
+    {
+      "epoch": 0.21975582685904552,
+      "grad_norm": 0.4579526484012604,
+      "learning_rate": 7.313019390581718e-06,
+      "loss": 0.4905,
+      "step": 264
+    },
+    {
+      "epoch": 0.22058823529411764,
+      "grad_norm": 0.459525465965271,
+      "learning_rate": 7.340720221606649e-06,
+      "loss": 0.525,
+      "step": 265
+    },
+    {
+      "epoch": 0.2214206437291898,
+      "grad_norm": 0.5286341309547424,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.5093,
+      "step": 266
+    },
+    {
+      "epoch": 0.22225305216426194,
+      "grad_norm": 0.4147150218486786,
+      "learning_rate": 7.396121883656511e-06,
+      "loss": 0.4996,
+      "step": 267
+    },
+    {
+      "epoch": 0.22308546059933407,
+      "grad_norm": 0.4209958612918854,
+      "learning_rate": 7.423822714681441e-06,
+      "loss": 0.5076,
+      "step": 268
+    },
+    {
+      "epoch": 0.22391786903440622,
+      "grad_norm": 0.41367340087890625,
+      "learning_rate": 7.451523545706372e-06,
+      "loss": 0.5261,
+      "step": 269
+    },
+    {
+      "epoch": 0.22475027746947837,
+      "grad_norm": 0.4930833578109741,
+      "learning_rate": 7.479224376731302e-06,
+      "loss": 0.4987,
+      "step": 270
+    },
+    {
+      "epoch": 0.2255826859045505,
+      "grad_norm": 0.45146897435188293,
+      "learning_rate": 7.506925207756234e-06,
+      "loss": 0.4857,
+      "step": 271
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.4591994881629944,
+      "learning_rate": 7.534626038781164e-06,
+      "loss": 0.5007,
+      "step": 272
+    },
+    {
+      "epoch": 0.22724750277469477,
+      "grad_norm": 0.5006039142608643,
+      "learning_rate": 7.5623268698060945e-06,
+      "loss": 0.5174,
+      "step": 273
+    },
+    {
+      "epoch": 0.22807991120976692,
+      "grad_norm": 0.4668283462524414,
+      "learning_rate": 7.590027700831025e-06,
+      "loss": 0.5019,
+      "step": 274
+    },
+    {
+      "epoch": 0.22891231964483907,
+      "grad_norm": 0.5253505110740662,
+      "learning_rate": 7.617728531855957e-06,
+      "loss": 0.4974,
+      "step": 275
+    },
+    {
+      "epoch": 0.2297447280799112,
+      "grad_norm": 0.5322619676589966,
+      "learning_rate": 7.645429362880887e-06,
+      "loss": 0.5147,
+      "step": 276
+    },
+    {
+      "epoch": 0.23057713651498335,
+      "grad_norm": 0.44390517473220825,
+      "learning_rate": 7.673130193905818e-06,
+      "loss": 0.5246,
+      "step": 277
+    },
+    {
+      "epoch": 0.2314095449500555,
+      "grad_norm": 0.5119628310203552,
+      "learning_rate": 7.700831024930749e-06,
+      "loss": 0.4879,
+      "step": 278
+    },
+    {
+      "epoch": 0.23224195338512763,
+      "grad_norm": 0.4466327726840973,
+      "learning_rate": 7.728531855955679e-06,
+      "loss": 0.4999,
+      "step": 279
+    },
+    {
+      "epoch": 0.23307436182019978,
+      "grad_norm": 0.4270954728126526,
+      "learning_rate": 7.75623268698061e-06,
+      "loss": 0.4955,
+      "step": 280
+    },
+    {
+      "epoch": 0.23390677025527193,
+      "grad_norm": 0.5279788970947266,
+      "learning_rate": 7.78393351800554e-06,
+      "loss": 0.5138,
+      "step": 281
+    },
+    {
+      "epoch": 0.23473917869034405,
+      "grad_norm": 0.4631377160549164,
+      "learning_rate": 7.811634349030471e-06,
+      "loss": 0.4915,
+      "step": 282
+    },
+    {
+      "epoch": 0.2355715871254162,
+      "grad_norm": 0.509636402130127,
+      "learning_rate": 7.839335180055402e-06,
+      "loss": 0.5205,
+      "step": 283
+    },
+    {
+      "epoch": 0.23640399556048836,
+      "grad_norm": 0.42661571502685547,
+      "learning_rate": 7.867036011080334e-06,
+      "loss": 0.4931,
+      "step": 284
+    },
+    {
+      "epoch": 0.23723640399556048,
+      "grad_norm": 0.4809859097003937,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.514,
+      "step": 285
+    },
+    {
+      "epoch": 0.23806881243063263,
+      "grad_norm": 0.4813212752342224,
+      "learning_rate": 7.922437673130195e-06,
+      "loss": 0.5209,
+      "step": 286
+    },
+    {
+      "epoch": 0.23890122086570478,
+      "grad_norm": 0.4047999083995819,
+      "learning_rate": 7.950138504155124e-06,
+      "loss": 0.4689,
+      "step": 287
+    },
+    {
+      "epoch": 0.2397336293007769,
+      "grad_norm": 0.48839592933654785,
+      "learning_rate": 7.977839335180056e-06,
+      "loss": 0.495,
+      "step": 288
+    },
+    {
+      "epoch": 0.24056603773584906,
+      "grad_norm": 0.4675354063510895,
+      "learning_rate": 8.005540166204987e-06,
+      "loss": 0.5214,
+      "step": 289
+    },
+    {
+      "epoch": 0.2413984461709212,
+      "grad_norm": 0.5134713053703308,
+      "learning_rate": 8.033240997229918e-06,
+      "loss": 0.4755,
+      "step": 290
+    },
+    {
+      "epoch": 0.24223085460599333,
+      "grad_norm": 0.46296653151512146,
+      "learning_rate": 8.060941828254848e-06,
+      "loss": 0.5137,
+      "step": 291
+    },
+    {
+      "epoch": 0.24306326304106549,
+      "grad_norm": 0.43857628107070923,
+      "learning_rate": 8.088642659279779e-06,
+      "loss": 0.5045,
+      "step": 292
+    },
+    {
+      "epoch": 0.24389567147613764,
+      "grad_norm": 0.4630931615829468,
+      "learning_rate": 8.11634349030471e-06,
+      "loss": 0.5113,
+      "step": 293
+    },
+    {
+      "epoch": 0.24472807991120976,
+      "grad_norm": 0.4841103255748749,
+      "learning_rate": 8.14404432132964e-06,
+      "loss": 0.519,
+      "step": 294
+    },
+    {
+      "epoch": 0.2455604883462819,
+      "grad_norm": 0.4541115462779999,
+      "learning_rate": 8.171745152354571e-06,
+      "loss": 0.5125,
+      "step": 295
+    },
+    {
+      "epoch": 0.24639289678135406,
+      "grad_norm": 0.4210924208164215,
+      "learning_rate": 8.199445983379503e-06,
+      "loss": 0.5035,
+      "step": 296
+    },
+    {
+      "epoch": 0.2472253052164262,
+      "grad_norm": 0.41758114099502563,
+      "learning_rate": 8.227146814404434e-06,
+      "loss": 0.4851,
+      "step": 297
+    },
+    {
+      "epoch": 0.24805771365149834,
+      "grad_norm": 0.5241228938102722,
+      "learning_rate": 8.254847645429363e-06,
+      "loss": 0.5046,
+      "step": 298
+    },
+    {
+      "epoch": 0.24889012208657046,
+      "grad_norm": 0.5144554972648621,
+      "learning_rate": 8.282548476454293e-06,
+      "loss": 0.54,
+      "step": 299
+    },
+    {
+      "epoch": 0.24972253052164262,
+      "grad_norm": 0.5133737921714783,
+      "learning_rate": 8.310249307479224e-06,
+      "loss": 0.499,
+      "step": 300
+    },
+    {
+      "epoch": 0.25055493895671477,
+      "grad_norm": 0.5024670362472534,
+      "learning_rate": 8.337950138504156e-06,
+      "loss": 0.5042,
+      "step": 301
+    },
+    {
+      "epoch": 0.2513873473917869,
+      "grad_norm": 0.5267788767814636,
+      "learning_rate": 8.365650969529087e-06,
+      "loss": 0.5023,
+      "step": 302
+    },
+    {
+      "epoch": 0.25221975582685907,
+      "grad_norm": 0.43696895241737366,
+      "learning_rate": 8.393351800554018e-06,
+      "loss": 0.4858,
+      "step": 303
+    },
+    {
+      "epoch": 0.25305216426193117,
+      "grad_norm": 0.5760444402694702,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.5106,
+      "step": 304
+    },
+    {
+      "epoch": 0.2538845726970033,
+      "grad_norm": 0.4616737961769104,
+      "learning_rate": 8.448753462603879e-06,
+      "loss": 0.4843,
+      "step": 305
+    },
+    {
+      "epoch": 0.25471698113207547,
+      "grad_norm": 0.5310185551643372,
+      "learning_rate": 8.47645429362881e-06,
+      "loss": 0.5263,
+      "step": 306
+    },
+    {
+      "epoch": 0.2555493895671476,
+      "grad_norm": 0.4836473762989044,
+      "learning_rate": 8.50415512465374e-06,
+      "loss": 0.4889,
+      "step": 307
+    },
+    {
+      "epoch": 0.2563817980022198,
+      "grad_norm": 0.5024266839027405,
+      "learning_rate": 8.53185595567867e-06,
+      "loss": 0.491,
+      "step": 308
+    },
+    {
+      "epoch": 0.25721420643729187,
+      "grad_norm": 0.47080838680267334,
+      "learning_rate": 8.559556786703603e-06,
+      "loss": 0.467,
+      "step": 309
+    },
+    {
+      "epoch": 0.258046614872364,
+      "grad_norm": 0.44176098704338074,
+      "learning_rate": 8.587257617728532e-06,
+      "loss": 0.4699,
+      "step": 310
+    },
+    {
+      "epoch": 0.2588790233074362,
+      "grad_norm": 0.43928641080856323,
+      "learning_rate": 8.614958448753463e-06,
+      "loss": 0.5042,
+      "step": 311
+    },
+    {
+      "epoch": 0.2597114317425083,
+      "grad_norm": 0.4728699326515198,
+      "learning_rate": 8.642659279778393e-06,
+      "loss": 0.4923,
+      "step": 312
+    },
+    {
+      "epoch": 0.2605438401775805,
+      "grad_norm": 0.45453718304634094,
+      "learning_rate": 8.670360110803326e-06,
+      "loss": 0.4983,
+      "step": 313
+    },
+    {
+      "epoch": 0.26137624861265263,
+      "grad_norm": 0.46343910694122314,
+      "learning_rate": 8.698060941828256e-06,
+      "loss": 0.4985,
+      "step": 314
+    },
+    {
+      "epoch": 0.2622086570477247,
+      "grad_norm": 0.5328369140625,
+      "learning_rate": 8.725761772853187e-06,
+      "loss": 0.5047,
+      "step": 315
+    },
+    {
+      "epoch": 0.2630410654827969,
+      "grad_norm": 0.4438888728618622,
+      "learning_rate": 8.753462603878117e-06,
+      "loss": 0.5047,
+      "step": 316
+    },
+    {
+      "epoch": 0.26387347391786903,
+      "grad_norm": 0.5838459730148315,
+      "learning_rate": 8.781163434903048e-06,
+      "loss": 0.5239,
+      "step": 317
+    },
+    {
+      "epoch": 0.2647058823529412,
+      "grad_norm": 0.48084893822669983,
+      "learning_rate": 8.808864265927979e-06,
+      "loss": 0.5101,
+      "step": 318
+    },
+    {
+      "epoch": 0.26553829078801333,
+      "grad_norm": 0.3967612385749817,
+      "learning_rate": 8.83656509695291e-06,
+      "loss": 0.474,
+      "step": 319
+    },
+    {
+      "epoch": 0.2663706992230855,
+      "grad_norm": 0.49214819073677063,
+      "learning_rate": 8.86426592797784e-06,
+      "loss": 0.5064,
+      "step": 320
+    },
+    {
+      "epoch": 0.2672031076581576,
+      "grad_norm": 0.46390125155448914,
+      "learning_rate": 8.89196675900277e-06,
+      "loss": 0.4726,
+      "step": 321
+    },
+    {
+      "epoch": 0.26803551609322973,
+      "grad_norm": 0.43828120827674866,
+      "learning_rate": 8.919667590027701e-06,
+      "loss": 0.508,
+      "step": 322
+    },
+    {
+      "epoch": 0.2688679245283019,
+      "grad_norm": 0.4651317596435547,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.4844,
+      "step": 323
+    },
+    {
+      "epoch": 0.26970033296337403,
+      "grad_norm": 0.45498353242874146,
+      "learning_rate": 8.975069252077562e-06,
+      "loss": 0.4959,
+      "step": 324
+    },
+    {
+      "epoch": 0.2705327413984462,
+      "grad_norm": 0.5337496995925903,
+      "learning_rate": 9.002770083102493e-06,
+      "loss": 0.4892,
+      "step": 325
+    },
+    {
+      "epoch": 0.27136514983351834,
+      "grad_norm": 0.4983648657798767,
+      "learning_rate": 9.030470914127425e-06,
+      "loss": 0.5105,
+      "step": 326
+    },
+    {
+      "epoch": 0.27219755826859043,
+      "grad_norm": 0.4203820824623108,
+      "learning_rate": 9.058171745152356e-06,
+      "loss": 0.4674,
+      "step": 327
+    },
+    {
+      "epoch": 0.2730299667036626,
+      "grad_norm": 0.4564470052719116,
+      "learning_rate": 9.085872576177287e-06,
+      "loss": 0.4779,
+      "step": 328
+    },
+    {
+      "epoch": 0.27386237513873474,
+      "grad_norm": 0.48127999901771545,
+      "learning_rate": 9.113573407202216e-06,
+      "loss": 0.504,
+      "step": 329
+    },
+    {
+      "epoch": 0.2746947835738069,
+      "grad_norm": 0.41335824131965637,
+      "learning_rate": 9.141274238227148e-06,
+      "loss": 0.452,
+      "step": 330
+    },
+    {
+      "epoch": 0.27552719200887904,
+      "grad_norm": 0.48663002252578735,
+      "learning_rate": 9.168975069252079e-06,
+      "loss": 0.497,
+      "step": 331
+    },
+    {
+      "epoch": 0.2763596004439512,
+      "grad_norm": 0.469848096370697,
+      "learning_rate": 9.19667590027701e-06,
+      "loss": 0.4722,
+      "step": 332
+    },
+    {
+      "epoch": 0.2771920088790233,
+      "grad_norm": 0.5372227430343628,
+      "learning_rate": 9.22437673130194e-06,
+      "loss": 0.5136,
+      "step": 333
+    },
+    {
+      "epoch": 0.27802441731409544,
+      "grad_norm": 0.4874361753463745,
+      "learning_rate": 9.25207756232687e-06,
+      "loss": 0.508,
+      "step": 334
+    },
+    {
+      "epoch": 0.2788568257491676,
+      "grad_norm": 0.42719605565071106,
+      "learning_rate": 9.279778393351801e-06,
+      "loss": 0.4939,
+      "step": 335
+    },
+    {
+      "epoch": 0.27968923418423974,
+      "grad_norm": 0.4991985261440277,
+      "learning_rate": 9.307479224376732e-06,
+      "loss": 0.5072,
+      "step": 336
+    },
+    {
+      "epoch": 0.2805216426193119,
+      "grad_norm": 0.5013337135314941,
+      "learning_rate": 9.335180055401662e-06,
+      "loss": 0.4705,
+      "step": 337
+    },
+    {
+      "epoch": 0.281354051054384,
+      "grad_norm": 0.40710508823394775,
+      "learning_rate": 9.362880886426595e-06,
+      "loss": 0.4713,
+      "step": 338
+    },
+    {
+      "epoch": 0.28218645948945614,
+      "grad_norm": 0.4505155682563782,
+      "learning_rate": 9.390581717451525e-06,
+      "loss": 0.4912,
+      "step": 339
+    },
+    {
+      "epoch": 0.2830188679245283,
+      "grad_norm": 0.4827728569507599,
+      "learning_rate": 9.418282548476454e-06,
+      "loss": 0.5223,
+      "step": 340
+    },
+    {
+      "epoch": 0.28385127635960045,
+      "grad_norm": 0.5375002026557922,
+      "learning_rate": 9.445983379501385e-06,
+      "loss": 0.5245,
+      "step": 341
+    },
+    {
+      "epoch": 0.2846836847946726,
+      "grad_norm": 0.41850459575653076,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.5029,
+      "step": 342
+    },
+    {
+      "epoch": 0.28551609322974475,
+      "grad_norm": 0.508848249912262,
+      "learning_rate": 9.501385041551248e-06,
+      "loss": 0.4864,
+      "step": 343
+    },
+    {
+      "epoch": 0.28634850166481685,
+      "grad_norm": 0.5495002865791321,
+      "learning_rate": 9.529085872576178e-06,
+      "loss": 0.5147,
+      "step": 344
+    },
+    {
+      "epoch": 0.287180910099889,
+      "grad_norm": 0.49234721064567566,
+      "learning_rate": 9.556786703601109e-06,
+      "loss": 0.5126,
+      "step": 345
+    },
+    {
+      "epoch": 0.28801331853496115,
+      "grad_norm": 0.4997720718383789,
+      "learning_rate": 9.58448753462604e-06,
+      "loss": 0.5154,
+      "step": 346
+    },
+    {
+      "epoch": 0.2888457269700333,
+      "grad_norm": 0.5731498599052429,
+      "learning_rate": 9.61218836565097e-06,
+      "loss": 0.5084,
+      "step": 347
+    },
+    {
+      "epoch": 0.28967813540510545,
+      "grad_norm": 0.5639533400535583,
+      "learning_rate": 9.639889196675901e-06,
+      "loss": 0.458,
+      "step": 348
+    },
+    {
+      "epoch": 0.2905105438401776,
+      "grad_norm": 0.5300630927085876,
+      "learning_rate": 9.667590027700832e-06,
+      "loss": 0.499,
+      "step": 349
+    },
+    {
+      "epoch": 0.2913429522752497,
+      "grad_norm": 0.5944730639457703,
+      "learning_rate": 9.695290858725762e-06,
+      "loss": 0.5156,
+      "step": 350
+    },
+    {
+      "epoch": 0.29217536071032185,
+      "grad_norm": 0.4813040494918823,
+      "learning_rate": 9.722991689750695e-06,
+      "loss": 0.492,
+      "step": 351
+    },
+    {
+      "epoch": 0.293007769145394,
+      "grad_norm": 0.4312509000301361,
+      "learning_rate": 9.750692520775623e-06,
+      "loss": 0.4801,
+      "step": 352
+    },
+    {
+      "epoch": 0.29384017758046616,
+      "grad_norm": 0.474255234003067,
+      "learning_rate": 9.778393351800554e-06,
+      "loss": 0.4755,
+      "step": 353
+    },
+    {
+      "epoch": 0.2946725860155383,
+      "grad_norm": 0.48329421877861023,
+      "learning_rate": 9.806094182825485e-06,
+      "loss": 0.4974,
+      "step": 354
+    },
+    {
+      "epoch": 0.29550499445061046,
+      "grad_norm": 0.47372984886169434,
+      "learning_rate": 9.833795013850417e-06,
+      "loss": 0.5232,
+      "step": 355
+    },
+    {
+      "epoch": 0.29633740288568255,
+      "grad_norm": 0.4526323080062866,
+      "learning_rate": 9.861495844875348e-06,
+      "loss": 0.4792,
+      "step": 356
+    },
+    {
+      "epoch": 0.2971698113207547,
+      "grad_norm": 0.5251845121383667,
+      "learning_rate": 9.889196675900278e-06,
+      "loss": 0.4866,
+      "step": 357
+    },
+    {
+      "epoch": 0.29800221975582686,
+      "grad_norm": 0.4455892741680145,
+      "learning_rate": 9.916897506925209e-06,
+      "loss": 0.4904,
+      "step": 358
+    },
+    {
+      "epoch": 0.298834628190899,
+      "grad_norm": 0.46031251549720764,
+      "learning_rate": 9.94459833795014e-06,
+      "loss": 0.4744,
+      "step": 359
+    },
+    {
+      "epoch": 0.29966703662597116,
+      "grad_norm": 0.4524519443511963,
+      "learning_rate": 9.97229916897507e-06,
+      "loss": 0.4958,
+      "step": 360
+    },
+    {
+      "epoch": 0.30049944506104326,
+      "grad_norm": 0.4535054862499237,
+      "learning_rate": 1e-05,
+      "loss": 0.4906,
+      "step": 361
+    },
+    {
+      "epoch": 0.3013318534961154,
+      "grad_norm": 0.4776564836502075,
+      "learning_rate": 9.999997652456228e-06,
+      "loss": 0.5017,
+      "step": 362
+    },
+    {
+      "epoch": 0.30216426193118756,
+      "grad_norm": 0.5730841159820557,
+      "learning_rate": 9.999990609827113e-06,
+      "loss": 0.4679,
+      "step": 363
+    },
+    {
+      "epoch": 0.3029966703662597,
+      "grad_norm": 0.4576025903224945,
+      "learning_rate": 9.999978872119267e-06,
+      "loss": 0.4968,
+      "step": 364
+    },
+    {
+      "epoch": 0.30382907880133186,
+      "grad_norm": 0.6305115818977356,
+      "learning_rate": 9.999962439343715e-06,
+      "loss": 0.4939,
+      "step": 365
+    },
+    {
+      "epoch": 0.304661487236404,
+      "grad_norm": 0.4445456564426422,
+      "learning_rate": 9.999941311515888e-06,
+      "loss": 0.4895,
+      "step": 366
+    },
+    {
+      "epoch": 0.3054938956714761,
+      "grad_norm": 0.535879909992218,
+      "learning_rate": 9.999915488655623e-06,
+      "loss": 0.4841,
+      "step": 367
+    },
+    {
+      "epoch": 0.30632630410654826,
+      "grad_norm": 0.5554977059364319,
+      "learning_rate": 9.999884970787168e-06,
+      "loss": 0.5098,
+      "step": 368
+    },
+    {
+      "epoch": 0.3071587125416204,
+      "grad_norm": 0.5069095492362976,
+      "learning_rate": 9.999849757939182e-06,
+      "loss": 0.4879,
+      "step": 369
+    },
+    {
+      "epoch": 0.30799112097669257,
+      "grad_norm": 0.48924490809440613,
+      "learning_rate": 9.99980985014473e-06,
+      "loss": 0.489,
+      "step": 370
+    },
+    {
+      "epoch": 0.3088235294117647,
+      "grad_norm": 0.5114119052886963,
+      "learning_rate": 9.999765247441285e-06,
+      "loss": 0.4927,
+      "step": 371
+    },
+    {
+      "epoch": 0.30965593784683687,
+      "grad_norm": 0.6173185706138611,
+      "learning_rate": 9.999715949870729e-06,
+      "loss": 0.5053,
+      "step": 372
+    },
+    {
+      "epoch": 0.31048834628190897,
+      "grad_norm": 0.4988461136817932,
+      "learning_rate": 9.999661957479354e-06,
+      "loss": 0.4579,
+      "step": 373
+    },
+    {
+      "epoch": 0.3113207547169811,
+      "grad_norm": 0.5641622543334961,
+      "learning_rate": 9.999603270317863e-06,
+      "loss": 0.5126,
+      "step": 374
+    },
+    {
+      "epoch": 0.31215316315205327,
+      "grad_norm": 0.5810229778289795,
+      "learning_rate": 9.99953988844136e-06,
+      "loss": 0.4952,
+      "step": 375
+    },
+    {
+      "epoch": 0.3129855715871254,
+      "grad_norm": 0.6129553914070129,
+      "learning_rate": 9.999471811909363e-06,
+      "loss": 0.5012,
+      "step": 376
+    },
+    {
+      "epoch": 0.3138179800221976,
+      "grad_norm": 0.5977503061294556,
+      "learning_rate": 9.999399040785797e-06,
+      "loss": 0.492,
+      "step": 377
+    },
+    {
+      "epoch": 0.3146503884572697,
+      "grad_norm": 0.6007142663002014,
+      "learning_rate": 9.999321575138997e-06,
+      "loss": 0.501,
+      "step": 378
+    },
+    {
+      "epoch": 0.3154827968923418,
+      "grad_norm": 0.6089385747909546,
+      "learning_rate": 9.999239415041701e-06,
+      "loss": 0.4894,
+      "step": 379
+    },
+    {
+      "epoch": 0.316315205327414,
+      "grad_norm": 0.6288129091262817,
+      "learning_rate": 9.999152560571064e-06,
+      "loss": 0.4886,
+      "step": 380
+    },
+    {
+      "epoch": 0.3171476137624861,
+      "grad_norm": 0.6314293146133423,
+      "learning_rate": 9.99906101180864e-06,
+      "loss": 0.5023,
+      "step": 381
+    },
+    {
+      "epoch": 0.3179800221975583,
+      "grad_norm": 0.6294896006584167,
+      "learning_rate": 9.998964768840393e-06,
+      "loss": 0.495,
+      "step": 382
+    },
+    {
+      "epoch": 0.31881243063263043,
+      "grad_norm": 0.5049759745597839,
+      "learning_rate": 9.998863831756702e-06,
+      "loss": 0.5074,
+      "step": 383
+    },
+    {
+      "epoch": 0.3196448390677026,
+      "grad_norm": 0.5267007350921631,
+      "learning_rate": 9.998758200652346e-06,
+      "loss": 0.4751,
+      "step": 384
+    },
+    {
+      "epoch": 0.3204772475027747,
+      "grad_norm": 0.5030413866043091,
+      "learning_rate": 9.998647875626514e-06,
+      "loss": 0.5,
+      "step": 385
+    },
+    {
+      "epoch": 0.3213096559378468,
+      "grad_norm": 0.49688029289245605,
+      "learning_rate": 9.998532856782805e-06,
+      "loss": 0.4716,
+      "step": 386
+    },
+    {
+      "epoch": 0.322142064372919,
+      "grad_norm": 0.4237273335456848,
+      "learning_rate": 9.998413144229224e-06,
+      "loss": 0.4842,
+      "step": 387
+    },
+    {
+      "epoch": 0.32297447280799113,
+      "grad_norm": 0.5279334783554077,
+      "learning_rate": 9.998288738078179e-06,
+      "loss": 0.4842,
+      "step": 388
+    },
+    {
+      "epoch": 0.3238068812430633,
+      "grad_norm": 0.43567541241645813,
+      "learning_rate": 9.998159638446495e-06,
+      "loss": 0.4772,
+      "step": 389
+    },
+    {
+      "epoch": 0.3246392896781354,
+      "grad_norm": 0.4781498610973358,
+      "learning_rate": 9.998025845455394e-06,
+      "loss": 0.4918,
+      "step": 390
+    },
+    {
+      "epoch": 0.32547169811320753,
+      "grad_norm": 0.5459445118904114,
+      "learning_rate": 9.99788735923051e-06,
+      "loss": 0.4865,
+      "step": 391
+    },
+    {
+      "epoch": 0.3263041065482797,
+      "grad_norm": 0.4947822391986847,
+      "learning_rate": 9.997744179901891e-06,
+      "loss": 0.4784,
+      "step": 392
+    },
+    {
+      "epoch": 0.32713651498335183,
+      "grad_norm": 0.5778188109397888,
+      "learning_rate": 9.997596307603979e-06,
+      "loss": 0.5046,
+      "step": 393
+    },
+    {
+      "epoch": 0.327968923418424,
+      "grad_norm": 0.46349483728408813,
+      "learning_rate": 9.997443742475628e-06,
+      "loss": 0.5002,
+      "step": 394
+    },
+    {
+      "epoch": 0.32880133185349614,
+      "grad_norm": 0.41382893919944763,
+      "learning_rate": 9.997286484660101e-06,
+      "loss": 0.5169,
+      "step": 395
+    },
+    {
+      "epoch": 0.32963374028856823,
+      "grad_norm": 0.5306146740913391,
+      "learning_rate": 9.997124534305065e-06,
+      "loss": 0.5053,
+      "step": 396
+    },
+    {
+      "epoch": 0.3304661487236404,
+      "grad_norm": 0.4569046199321747,
+      "learning_rate": 9.996957891562598e-06,
+      "loss": 0.5099,
+      "step": 397
+    },
+    {
+      "epoch": 0.33129855715871254,
+      "grad_norm": 0.440790057182312,
+      "learning_rate": 9.996786556589175e-06,
+      "loss": 0.5094,
+      "step": 398
+    },
+    {
+      "epoch": 0.3321309655937847,
+      "grad_norm": 0.5211203694343567,
+      "learning_rate": 9.996610529545685e-06,
+      "loss": 0.4936,
+      "step": 399
+    },
+    {
+      "epoch": 0.33296337402885684,
+      "grad_norm": 0.4785847067832947,
+      "learning_rate": 9.996429810597421e-06,
+      "loss": 0.4808,
+      "step": 400
+    },
+    {
+      "epoch": 0.333795782463929,
+      "grad_norm": 0.5769135355949402,
+      "learning_rate": 9.996244399914083e-06,
+      "loss": 0.5027,
+      "step": 401
+    },
+    {
+      "epoch": 0.3346281908990011,
+      "grad_norm": 0.49235013127326965,
+      "learning_rate": 9.99605429766977e-06,
+      "loss": 0.4918,
+      "step": 402
+    },
+    {
+      "epoch": 0.33546059933407324,
+      "grad_norm": 0.559281587600708,
+      "learning_rate": 9.995859504042994e-06,
+      "loss": 0.5126,
+      "step": 403
+    },
+    {
+      "epoch": 0.3362930077691454,
+      "grad_norm": 0.4233754873275757,
+      "learning_rate": 9.99566001921667e-06,
+      "loss": 0.4942,
+      "step": 404
+    },
+    {
+      "epoch": 0.33712541620421754,
+      "grad_norm": 0.5626436471939087,
+      "learning_rate": 9.995455843378118e-06,
+      "loss": 0.4992,
+      "step": 405
+    },
+    {
+      "epoch": 0.3379578246392897,
+      "grad_norm": 0.4755416810512543,
+      "learning_rate": 9.995246976719063e-06,
+      "loss": 0.497,
+      "step": 406
+    },
+    {
+      "epoch": 0.33879023307436185,
+      "grad_norm": 0.5172478556632996,
+      "learning_rate": 9.995033419435632e-06,
+      "loss": 0.4947,
+      "step": 407
+    },
+    {
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.5225632786750793,
+      "learning_rate": 9.994815171728362e-06,
+      "loss": 0.5041,
+      "step": 408
+    },
+    {
+      "epoch": 0.3404550499445061,
+      "grad_norm": 0.5508219003677368,
+      "learning_rate": 9.994592233802189e-06,
+      "loss": 0.4589,
+      "step": 409
+    },
+    {
+      "epoch": 0.34128745837957825,
+      "grad_norm": 0.5616341233253479,
+      "learning_rate": 9.994364605866455e-06,
+      "loss": 0.4893,
+      "step": 410
+    },
+    {
+      "epoch": 0.3421198668146504,
+      "grad_norm": 0.5048859119415283,
+      "learning_rate": 9.99413228813491e-06,
+      "loss": 0.4887,
+      "step": 411
+    },
+    {
+      "epoch": 0.34295227524972255,
+      "grad_norm": 0.5626086592674255,
+      "learning_rate": 9.993895280825702e-06,
+      "loss": 0.4853,
+      "step": 412
+    },
+    {
+      "epoch": 0.34378468368479465,
+      "grad_norm": 0.46970364451408386,
+      "learning_rate": 9.993653584161387e-06,
+      "loss": 0.4965,
+      "step": 413
+    },
+    {
+      "epoch": 0.3446170921198668,
+      "grad_norm": 0.6829771995544434,
+      "learning_rate": 9.993407198368918e-06,
+      "loss": 0.4942,
+      "step": 414
+    },
+    {
+      "epoch": 0.34544950055493895,
+      "grad_norm": 0.5054067373275757,
+      "learning_rate": 9.993156123679662e-06,
+      "loss": 0.4761,
+      "step": 415
+    },
+    {
+      "epoch": 0.3462819089900111,
+      "grad_norm": 0.6721978783607483,
+      "learning_rate": 9.992900360329376e-06,
+      "loss": 0.4938,
+      "step": 416
+    },
+    {
+      "epoch": 0.34711431742508325,
+      "grad_norm": 0.4733940362930298,
+      "learning_rate": 9.992639908558232e-06,
+      "loss": 0.4801,
+      "step": 417
+    },
+    {
+      "epoch": 0.3479467258601554,
+      "grad_norm": 0.5857818722724915,
+      "learning_rate": 9.992374768610795e-06,
+      "loss": 0.4631,
+      "step": 418
+    },
+    {
+      "epoch": 0.3487791342952275,
+      "grad_norm": 0.4833131432533264,
+      "learning_rate": 9.992104940736038e-06,
+      "loss": 0.469,
+      "step": 419
+    },
+    {
+      "epoch": 0.34961154273029965,
+      "grad_norm": 0.5365222692489624,
+      "learning_rate": 9.991830425187333e-06,
+      "loss": 0.4771,
+      "step": 420
+    },
+    {
+      "epoch": 0.3504439511653718,
+      "grad_norm": 0.4678042531013489,
+      "learning_rate": 9.991551222222455e-06,
+      "loss": 0.4632,
+      "step": 421
+    },
+    {
+      "epoch": 0.35127635960044395,
+      "grad_norm": 0.4540461599826813,
+      "learning_rate": 9.99126733210358e-06,
+      "loss": 0.4764,
+      "step": 422
+    },
+    {
+      "epoch": 0.3521087680355161,
+      "grad_norm": 0.47178414463996887,
+      "learning_rate": 9.990978755097287e-06,
+      "loss": 0.4711,
+      "step": 423
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 0.4962081015110016,
+      "learning_rate": 9.990685491474555e-06,
+      "loss": 0.5143,
+      "step": 424
+    },
+    {
+      "epoch": 0.35377358490566035,
+      "grad_norm": 0.42502203583717346,
+      "learning_rate": 9.990387541510761e-06,
+      "loss": 0.4839,
+      "step": 425
+    },
+    {
+      "epoch": 0.3546059933407325,
+      "grad_norm": 0.4943620264530182,
+      "learning_rate": 9.990084905485689e-06,
+      "loss": 0.4881,
+      "step": 426
+    },
+    {
+      "epoch": 0.35543840177580466,
+      "grad_norm": 0.4736219644546509,
+      "learning_rate": 9.989777583683517e-06,
+      "loss": 0.4774,
+      "step": 427
+    },
+    {
+      "epoch": 0.3562708102108768,
+      "grad_norm": 0.5000523924827576,
+      "learning_rate": 9.989465576392828e-06,
+      "loss": 0.5023,
+      "step": 428
+    },
+    {
+      "epoch": 0.35710321864594896,
+      "grad_norm": 0.5680121183395386,
+      "learning_rate": 9.989148883906599e-06,
+      "loss": 0.4816,
+      "step": 429
+    },
+    {
+      "epoch": 0.3579356270810211,
+      "grad_norm": 0.4986157715320587,
+      "learning_rate": 9.988827506522211e-06,
+      "loss": 0.4624,
+      "step": 430
+    },
+    {
+      "epoch": 0.3587680355160932,
+      "grad_norm": 0.48457589745521545,
+      "learning_rate": 9.988501444541445e-06,
+      "loss": 0.4946,
+      "step": 431
+    },
+    {
+      "epoch": 0.35960044395116536,
+      "grad_norm": 0.5403779149055481,
+      "learning_rate": 9.988170698270477e-06,
+      "loss": 0.4924,
+      "step": 432
+    },
+    {
+      "epoch": 0.3604328523862375,
+      "grad_norm": 0.388967901468277,
+      "learning_rate": 9.987835268019883e-06,
+      "loss": 0.4922,
+      "step": 433
+    },
+    {
+      "epoch": 0.36126526082130966,
+      "grad_norm": 0.459904283285141,
+      "learning_rate": 9.98749515410464e-06,
+      "loss": 0.4948,
+      "step": 434
+    },
+    {
+      "epoch": 0.3620976692563818,
+      "grad_norm": 0.46121054887771606,
+      "learning_rate": 9.987150356844118e-06,
+      "loss": 0.4833,
+      "step": 435
+    },
+    {
+      "epoch": 0.36293007769145397,
+      "grad_norm": 0.46013152599334717,
+      "learning_rate": 9.98680087656209e-06,
+      "loss": 0.4854,
+      "step": 436
+    },
+    {
+      "epoch": 0.36376248612652606,
+      "grad_norm": 0.5025128126144409,
+      "learning_rate": 9.986446713586724e-06,
+      "loss": 0.4917,
+      "step": 437
+    },
+    {
+      "epoch": 0.3645948945615982,
+      "grad_norm": 0.4966227412223816,
+      "learning_rate": 9.986087868250584e-06,
+      "loss": 0.4776,
+      "step": 438
+    },
+    {
+      "epoch": 0.36542730299667037,
+      "grad_norm": 0.45798709988594055,
+      "learning_rate": 9.985724340890633e-06,
+      "loss": 0.4892,
+      "step": 439
+    },
+    {
+      "epoch": 0.3662597114317425,
+      "grad_norm": 0.4821065068244934,
+      "learning_rate": 9.98535613184823e-06,
+      "loss": 0.4889,
+      "step": 440
+    },
+    {
+      "epoch": 0.36709211986681467,
+      "grad_norm": 0.4548647999763489,
+      "learning_rate": 9.984983241469129e-06,
+      "loss": 0.4701,
+      "step": 441
+    },
+    {
+      "epoch": 0.36792452830188677,
+      "grad_norm": 0.43688225746154785,
+      "learning_rate": 9.984605670103478e-06,
+      "loss": 0.5215,
+      "step": 442
+    },
+    {
+      "epoch": 0.3687569367369589,
+      "grad_norm": 0.549662709236145,
+      "learning_rate": 9.98422341810583e-06,
+      "loss": 0.4871,
+      "step": 443
+    },
+    {
+      "epoch": 0.36958934517203107,
+      "grad_norm": 0.40335264801979065,
+      "learning_rate": 9.98383648583512e-06,
+      "loss": 0.4678,
+      "step": 444
+    },
+    {
+      "epoch": 0.3704217536071032,
+      "grad_norm": 0.5139138102531433,
+      "learning_rate": 9.983444873654683e-06,
+      "loss": 0.4765,
+      "step": 445
+    },
+    {
+      "epoch": 0.3712541620421754,
+      "grad_norm": 0.5220227837562561,
+      "learning_rate": 9.983048581932257e-06,
+      "loss": 0.4893,
+      "step": 446
+    },
+    {
+      "epoch": 0.3720865704772475,
+      "grad_norm": 0.47806721925735474,
+      "learning_rate": 9.982647611039961e-06,
+      "loss": 0.4884,
+      "step": 447
+    },
+    {
+      "epoch": 0.3729189789123196,
+      "grad_norm": 0.49555501341819763,
+      "learning_rate": 9.982241961354317e-06,
+      "loss": 0.4835,
+      "step": 448
+    },
+    {
+      "epoch": 0.3737513873473918,
+      "grad_norm": 0.4484332203865051,
+      "learning_rate": 9.981831633256236e-06,
+      "loss": 0.4983,
+      "step": 449
+    },
+    {
+      "epoch": 0.3745837957824639,
+      "grad_norm": 0.41702842712402344,
+      "learning_rate": 9.981416627131022e-06,
+      "loss": 0.5161,
+      "step": 450
+    },
+    {
+      "epoch": 0.3754162042175361,
+      "grad_norm": 0.43207862973213196,
+      "learning_rate": 9.980996943368373e-06,
+      "loss": 0.5019,
+      "step": 451
+    },
+    {
+      "epoch": 0.3762486126526082,
+      "grad_norm": 0.460387647151947,
+      "learning_rate": 9.98057258236238e-06,
+      "loss": 0.4698,
+      "step": 452
+    },
+    {
+      "epoch": 0.3770810210876804,
+      "grad_norm": 0.4922404885292053,
+      "learning_rate": 9.980143544511527e-06,
+      "loss": 0.4838,
+      "step": 453
+    },
+    {
+      "epoch": 0.3779134295227525,
+      "grad_norm": 0.4728861153125763,
+      "learning_rate": 9.979709830218688e-06,
+      "loss": 0.4578,
+      "step": 454
+    },
+    {
+      "epoch": 0.3787458379578246,
+      "grad_norm": 0.5862690210342407,
+      "learning_rate": 9.979271439891125e-06,
+      "loss": 0.4802,
+      "step": 455
+    },
+    {
+      "epoch": 0.3795782463928968,
+      "grad_norm": 0.4514050781726837,
+      "learning_rate": 9.978828373940498e-06,
+      "loss": 0.4834,
+      "step": 456
+    },
+    {
+      "epoch": 0.38041065482796893,
+      "grad_norm": 0.5906931757926941,
+      "learning_rate": 9.97838063278285e-06,
+      "loss": 0.4621,
+      "step": 457
+    },
+    {
+      "epoch": 0.3812430632630411,
+      "grad_norm": 0.4813970625400543,
+      "learning_rate": 9.977928216838622e-06,
+      "loss": 0.4894,
+      "step": 458
+    },
+    {
+      "epoch": 0.38207547169811323,
+      "grad_norm": 0.5167549848556519,
+      "learning_rate": 9.977471126532636e-06,
+      "loss": 0.5006,
+      "step": 459
+    },
+    {
+      "epoch": 0.38290788013318533,
+      "grad_norm": 0.4821680188179016,
+      "learning_rate": 9.97700936229411e-06,
+      "loss": 0.4993,
+      "step": 460
+    },
+    {
+      "epoch": 0.3837402885682575,
+      "grad_norm": 0.4512309432029724,
+      "learning_rate": 9.976542924556652e-06,
+      "loss": 0.487,
+      "step": 461
+    },
+    {
+      "epoch": 0.38457269700332963,
+      "grad_norm": 0.4451935887336731,
+      "learning_rate": 9.976071813758249e-06,
+      "loss": 0.4546,
+      "step": 462
+    },
+    {
+      "epoch": 0.3854051054384018,
+      "grad_norm": 0.4715465307235718,
+      "learning_rate": 9.975596030341287e-06,
+      "loss": 0.4891,
+      "step": 463
+    },
+    {
+      "epoch": 0.38623751387347394,
+      "grad_norm": 0.39830783009529114,
+      "learning_rate": 9.975115574752532e-06,
+      "loss": 0.4996,
+      "step": 464
+    },
+    {
+      "epoch": 0.38706992230854603,
+      "grad_norm": 0.41202181577682495,
+      "learning_rate": 9.974630447443142e-06,
+      "loss": 0.4827,
+      "step": 465
+    },
+    {
+      "epoch": 0.3879023307436182,
+      "grad_norm": 0.42218315601348877,
+      "learning_rate": 9.974140648868659e-06,
+      "loss": 0.4635,
+      "step": 466
+    },
+    {
+      "epoch": 0.38873473917869034,
+      "grad_norm": 0.4501170516014099,
+      "learning_rate": 9.973646179489014e-06,
+      "loss": 0.4941,
+      "step": 467
+    },
+    {
+      "epoch": 0.3895671476137625,
+      "grad_norm": 0.44711339473724365,
+      "learning_rate": 9.97314703976852e-06,
+      "loss": 0.4581,
+      "step": 468
+    },
+    {
+      "epoch": 0.39039955604883464,
+      "grad_norm": 0.486054927110672,
+      "learning_rate": 9.97264323017588e-06,
+      "loss": 0.4863,
+      "step": 469
+    },
+    {
+      "epoch": 0.3912319644839068,
+      "grad_norm": 0.4822620153427124,
+      "learning_rate": 9.97213475118418e-06,
+      "loss": 0.5206,
+      "step": 470
+    },
+    {
+      "epoch": 0.3920643729189789,
+      "grad_norm": 0.40667545795440674,
+      "learning_rate": 9.971621603270887e-06,
+      "loss": 0.5174,
+      "step": 471
+    },
+    {
+      "epoch": 0.39289678135405104,
+      "grad_norm": 0.46628639101982117,
+      "learning_rate": 9.971103786917862e-06,
+      "loss": 0.4741,
+      "step": 472
+    },
+    {
+      "epoch": 0.3937291897891232,
+      "grad_norm": 0.40411314368247986,
+      "learning_rate": 9.97058130261134e-06,
+      "loss": 0.4768,
+      "step": 473
+    },
+    {
+      "epoch": 0.39456159822419534,
+      "grad_norm": 0.4988131523132324,
+      "learning_rate": 9.970054150841942e-06,
+      "loss": 0.4741,
+      "step": 474
+    },
+    {
+      "epoch": 0.3953940066592675,
+      "grad_norm": 0.42282071709632874,
+      "learning_rate": 9.969522332104675e-06,
+      "loss": 0.4761,
+      "step": 475
+    },
+    {
+      "epoch": 0.39622641509433965,
+      "grad_norm": 0.4693518877029419,
+      "learning_rate": 9.968985846898924e-06,
+      "loss": 0.4808,
+      "step": 476
+    },
+    {
+      "epoch": 0.39705882352941174,
+      "grad_norm": 0.5171592235565186,
+      "learning_rate": 9.968444695728461e-06,
+      "loss": 0.4797,
+      "step": 477
+    },
+    {
+      "epoch": 0.3978912319644839,
+      "grad_norm": 0.4162120819091797,
+      "learning_rate": 9.967898879101434e-06,
+      "loss": 0.5033,
+      "step": 478
+    },
+    {
+      "epoch": 0.39872364039955605,
+      "grad_norm": 0.5173420906066895,
+      "learning_rate": 9.967348397530373e-06,
+      "loss": 0.4704,
+      "step": 479
+    },
+    {
+      "epoch": 0.3995560488346282,
+      "grad_norm": 0.48679623007774353,
+      "learning_rate": 9.966793251532197e-06,
+      "loss": 0.4607,
+      "step": 480
+    },
+    {
+      "epoch": 0.40038845726970035,
+      "grad_norm": 0.4463588297367096,
+      "learning_rate": 9.966233441628188e-06,
+      "loss": 0.4836,
+      "step": 481
+    },
+    {
+      "epoch": 0.4012208657047725,
+      "grad_norm": 0.5191836357116699,
+      "learning_rate": 9.965668968344023e-06,
+      "loss": 0.4966,
+      "step": 482
+    },
+    {
+      "epoch": 0.4020532741398446,
+      "grad_norm": 0.4401721656322479,
+      "learning_rate": 9.965099832209753e-06,
+      "loss": 0.4864,
+      "step": 483
+    },
+    {
+      "epoch": 0.40288568257491675,
+      "grad_norm": 0.5379999876022339,
+      "learning_rate": 9.964526033759803e-06,
+      "loss": 0.4431,
+      "step": 484
+    },
+    {
+      "epoch": 0.4037180910099889,
+      "grad_norm": 0.46185803413391113,
+      "learning_rate": 9.963947573532983e-06,
+      "loss": 0.4894,
+      "step": 485
+    },
+    {
+      "epoch": 0.40455049944506105,
+      "grad_norm": 0.45974308252334595,
+      "learning_rate": 9.963364452072475e-06,
+      "loss": 0.4609,
+      "step": 486
+    },
+    {
+      "epoch": 0.4053829078801332,
+      "grad_norm": 0.46747586131095886,
+      "learning_rate": 9.962776669925842e-06,
+      "loss": 0.4949,
+      "step": 487
+    },
+    {
+      "epoch": 0.40621531631520535,
+      "grad_norm": 0.45227017998695374,
+      "learning_rate": 9.962184227645021e-06,
+      "loss": 0.5051,
+      "step": 488
+    },
+    {
+      "epoch": 0.40704772475027745,
+      "grad_norm": 0.46203479170799255,
+      "learning_rate": 9.961587125786328e-06,
+      "loss": 0.476,
+      "step": 489
+    },
+    {
+      "epoch": 0.4078801331853496,
+      "grad_norm": 0.4355171024799347,
+      "learning_rate": 9.960985364910448e-06,
+      "loss": 0.4809,
+      "step": 490
+    },
+    {
+      "epoch": 0.40871254162042175,
+      "grad_norm": 0.47708234190940857,
+      "learning_rate": 9.960378945582446e-06,
+      "loss": 0.4802,
+      "step": 491
+    },
+    {
+      "epoch": 0.4095449500554939,
+      "grad_norm": 0.46055299043655396,
+      "learning_rate": 9.959767868371761e-06,
+      "loss": 0.479,
+      "step": 492
+    },
+    {
+      "epoch": 0.41037735849056606,
+      "grad_norm": 0.502031147480011,
+      "learning_rate": 9.959152133852209e-06,
+      "loss": 0.4968,
+      "step": 493
+    },
+    {
+      "epoch": 0.41120976692563815,
+      "grad_norm": 0.48754680156707764,
+      "learning_rate": 9.958531742601968e-06,
+      "loss": 0.4791,
+      "step": 494
+    },
+    {
+      "epoch": 0.4120421753607103,
+      "grad_norm": 0.4217134416103363,
+      "learning_rate": 9.9579066952036e-06,
+      "loss": 0.4927,
+      "step": 495
+    },
+    {
+      "epoch": 0.41287458379578246,
+      "grad_norm": 0.474809467792511,
+      "learning_rate": 9.957276992244039e-06,
+      "loss": 0.491,
+      "step": 496
+    },
+    {
+      "epoch": 0.4137069922308546,
+      "grad_norm": 0.48426496982574463,
+      "learning_rate": 9.956642634314582e-06,
+      "loss": 0.4944,
+      "step": 497
+    },
+    {
+      "epoch": 0.41453940066592676,
+      "grad_norm": 0.4359665811061859,
+      "learning_rate": 9.956003622010904e-06,
+      "loss": 0.4695,
+      "step": 498
+    },
+    {
+      "epoch": 0.4153718091009989,
+      "grad_norm": 0.4425244927406311,
+      "learning_rate": 9.955359955933048e-06,
+      "loss": 0.49,
+      "step": 499
+    },
+    {
+      "epoch": 0.416204217536071,
+      "grad_norm": 0.5776862502098083,
+      "learning_rate": 9.95471163668543e-06,
+      "loss": 0.5233,
+      "step": 500
+    },
+    {
+      "epoch": 0.41703662597114316,
+      "grad_norm": 0.46279191970825195,
+      "learning_rate": 9.954058664876832e-06,
+      "loss": 0.4912,
+      "step": 501
+    },
+    {
+      "epoch": 0.4178690344062153,
+      "grad_norm": 0.5074960589408875,
+      "learning_rate": 9.953401041120403e-06,
+      "loss": 0.4979,
+      "step": 502
+    },
+    {
+      "epoch": 0.41870144284128746,
+      "grad_norm": 0.51701819896698,
+      "learning_rate": 9.952738766033668e-06,
+      "loss": 0.4649,
+      "step": 503
+    },
+    {
+      "epoch": 0.4195338512763596,
+      "grad_norm": 0.424464613199234,
+      "learning_rate": 9.952071840238511e-06,
+      "loss": 0.4885,
+      "step": 504
+    },
+    {
+      "epoch": 0.42036625971143177,
+      "grad_norm": 0.45139414072036743,
+      "learning_rate": 9.951400264361188e-06,
+      "loss": 0.4702,
+      "step": 505
+    },
+    {
+      "epoch": 0.42119866814650386,
+      "grad_norm": 0.46182501316070557,
+      "learning_rate": 9.950724039032324e-06,
+      "loss": 0.4643,
+      "step": 506
+    },
+    {
+      "epoch": 0.422031076581576,
+      "grad_norm": 0.43030858039855957,
+      "learning_rate": 9.950043164886902e-06,
+      "loss": 0.4903,
+      "step": 507
+    },
+    {
+      "epoch": 0.42286348501664817,
+      "grad_norm": 0.4888497292995453,
+      "learning_rate": 9.949357642564275e-06,
+      "loss": 0.5164,
+      "step": 508
+    },
+    {
+      "epoch": 0.4236958934517203,
+      "grad_norm": 0.48107990622520447,
+      "learning_rate": 9.948667472708163e-06,
+      "loss": 0.4811,
+      "step": 509
+    },
+    {
+      "epoch": 0.42452830188679247,
+      "grad_norm": 0.4666735827922821,
+      "learning_rate": 9.947972655966647e-06,
+      "loss": 0.4719,
+      "step": 510
+    },
+    {
+      "epoch": 0.4253607103218646,
+      "grad_norm": 0.4709594249725342,
+      "learning_rate": 9.947273192992171e-06,
+      "loss": 0.5038,
+      "step": 511
+    },
+    {
+      "epoch": 0.4261931187569367,
+      "grad_norm": 0.48332247138023376,
+      "learning_rate": 9.946569084441542e-06,
+      "loss": 0.4892,
+      "step": 512
+    },
+    {
+      "epoch": 0.42702552719200887,
+      "grad_norm": 0.507439911365509,
+      "learning_rate": 9.945860330975933e-06,
+      "loss": 0.482,
+      "step": 513
+    },
+    {
+      "epoch": 0.427857935627081,
+      "grad_norm": 0.4573519825935364,
+      "learning_rate": 9.945146933260876e-06,
+      "loss": 0.4808,
+      "step": 514
+    },
+    {
+      "epoch": 0.4286903440621532,
+      "grad_norm": 0.5414947867393494,
+      "learning_rate": 9.94442889196626e-06,
+      "loss": 0.4983,
+      "step": 515
+    },
+    {
+      "epoch": 0.4295227524972253,
+      "grad_norm": 0.43500539660453796,
+      "learning_rate": 9.94370620776634e-06,
+      "loss": 0.5014,
+      "step": 516
+    },
+    {
+      "epoch": 0.4303551609322974,
+      "grad_norm": 0.5725677609443665,
+      "learning_rate": 9.942978881339732e-06,
+      "loss": 0.487,
+      "step": 517
+    },
+    {
+      "epoch": 0.43118756936736957,
+      "grad_norm": 0.4238481819629669,
+      "learning_rate": 9.942246913369409e-06,
+      "loss": 0.4854,
+      "step": 518
+    },
+    {
+      "epoch": 0.4320199778024417,
+      "grad_norm": 0.44186079502105713,
+      "learning_rate": 9.941510304542695e-06,
+      "loss": 0.5183,
+      "step": 519
+    },
+    {
+      "epoch": 0.4328523862375139,
+      "grad_norm": 0.45779183506965637,
+      "learning_rate": 9.940769055551284e-06,
+      "loss": 0.4754,
+      "step": 520
+    },
+    {
+      "epoch": 0.433684794672586,
+      "grad_norm": 0.4756607115268707,
+      "learning_rate": 9.940023167091219e-06,
+      "loss": 0.4907,
+      "step": 521
+    },
+    {
+      "epoch": 0.4345172031076582,
+      "grad_norm": 0.47078803181648254,
+      "learning_rate": 9.939272639862905e-06,
+      "loss": 0.4845,
+      "step": 522
+    },
+    {
+      "epoch": 0.4353496115427303,
+      "grad_norm": 0.48096713423728943,
+      "learning_rate": 9.9385174745711e-06,
+      "loss": 0.4778,
+      "step": 523
+    },
+    {
+      "epoch": 0.4361820199778024,
+      "grad_norm": 0.48447316884994507,
+      "learning_rate": 9.937757671924915e-06,
+      "loss": 0.4883,
+      "step": 524
+    },
+    {
+      "epoch": 0.4370144284128746,
+      "grad_norm": 0.5201837420463562,
+      "learning_rate": 9.936993232637818e-06,
+      "loss": 0.5062,
+      "step": 525
+    },
+    {
+      "epoch": 0.43784683684794673,
+      "grad_norm": 0.5595635175704956,
+      "learning_rate": 9.936224157427635e-06,
+      "loss": 0.4843,
+      "step": 526
+    },
+    {
+      "epoch": 0.4386792452830189,
+      "grad_norm": 0.4055692255496979,
+      "learning_rate": 9.935450447016535e-06,
+      "loss": 0.4659,
+      "step": 527
+    },
+    {
+      "epoch": 0.43951165371809103,
+      "grad_norm": 0.5074310302734375,
+      "learning_rate": 9.934672102131052e-06,
+      "loss": 0.4681,
+      "step": 528
+    },
+    {
+      "epoch": 0.44034406215316313,
+      "grad_norm": 0.4834982454776764,
+      "learning_rate": 9.933889123502059e-06,
+      "loss": 0.5108,
+      "step": 529
+    },
+    {
+      "epoch": 0.4411764705882353,
+      "grad_norm": 0.48070859909057617,
+      "learning_rate": 9.933101511864793e-06,
+      "loss": 0.4833,
+      "step": 530
+    },
+    {
+      "epoch": 0.44200887902330743,
+      "grad_norm": 0.4138585031032562,
+      "learning_rate": 9.93230926795883e-06,
+      "loss": 0.4869,
+      "step": 531
+    },
+    {
+      "epoch": 0.4428412874583796,
+      "grad_norm": 0.5085632801055908,
+      "learning_rate": 9.931512392528104e-06,
+      "loss": 0.4832,
+      "step": 532
+    },
+    {
+      "epoch": 0.44367369589345174,
+      "grad_norm": 0.5004951357841492,
+      "learning_rate": 9.930710886320895e-06,
+      "loss": 0.4908,
+      "step": 533
+    },
+    {
+      "epoch": 0.4445061043285239,
+      "grad_norm": 0.498270720243454,
+      "learning_rate": 9.929904750089829e-06,
+      "loss": 0.4734,
+      "step": 534
+    },
+    {
+      "epoch": 0.445338512763596,
+      "grad_norm": 0.4868960976600647,
+      "learning_rate": 9.929093984591884e-06,
+      "loss": 0.4477,
+      "step": 535
+    },
+    {
+      "epoch": 0.44617092119866814,
+      "grad_norm": 0.5053864121437073,
+      "learning_rate": 9.928278590588382e-06,
+      "loss": 0.464,
+      "step": 536
+    },
+    {
+      "epoch": 0.4470033296337403,
+      "grad_norm": 0.47487616539001465,
+      "learning_rate": 9.927458568844994e-06,
+      "loss": 0.471,
+      "step": 537
+    },
+    {
+      "epoch": 0.44783573806881244,
+      "grad_norm": 0.4798586368560791,
+      "learning_rate": 9.926633920131732e-06,
+      "loss": 0.4628,
+      "step": 538
+    },
+    {
+      "epoch": 0.4486681465038846,
+      "grad_norm": 0.44665879011154175,
+      "learning_rate": 9.925804645222957e-06,
+      "loss": 0.464,
+      "step": 539
+    },
+    {
+      "epoch": 0.44950055493895674,
+      "grad_norm": 0.5611540675163269,
+      "learning_rate": 9.924970744897373e-06,
+      "loss": 0.4741,
+      "step": 540
+    },
+    {
+      "epoch": 0.45033296337402884,
+      "grad_norm": 0.4786832332611084,
+      "learning_rate": 9.924132219938027e-06,
+      "loss": 0.4732,
+      "step": 541
+    },
+    {
+      "epoch": 0.451165371809101,
+      "grad_norm": 0.4666735529899597,
+      "learning_rate": 9.923289071132308e-06,
+      "loss": 0.502,
+      "step": 542
+    },
+    {
+      "epoch": 0.45199778024417314,
+      "grad_norm": 0.5262482166290283,
+      "learning_rate": 9.922441299271948e-06,
+      "loss": 0.4744,
+      "step": 543
+    },
+    {
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.5326946377754211,
+      "learning_rate": 9.92158890515302e-06,
+      "loss": 0.4882,
+      "step": 544
+    },
+    {
+      "epoch": 0.45366259711431745,
+      "grad_norm": 0.4670204222202301,
+      "learning_rate": 9.920731889575935e-06,
+      "loss": 0.5037,
+      "step": 545
+    },
+    {
+      "epoch": 0.45449500554938954,
+      "grad_norm": 0.5759830474853516,
+      "learning_rate": 9.919870253345446e-06,
+      "loss": 0.4822,
+      "step": 546
+    },
+    {
+      "epoch": 0.4553274139844617,
+      "grad_norm": 0.42398568987846375,
+      "learning_rate": 9.919003997270648e-06,
+      "loss": 0.4776,
+      "step": 547
+    },
+    {
+      "epoch": 0.45615982241953384,
+      "grad_norm": 0.5048391819000244,
+      "learning_rate": 9.918133122164968e-06,
+      "loss": 0.4621,
+      "step": 548
+    },
+    {
+      "epoch": 0.456992230854606,
+      "grad_norm": 0.5019201636314392,
+      "learning_rate": 9.917257628846172e-06,
+      "loss": 0.4885,
+      "step": 549
+    },
+    {
+      "epoch": 0.45782463928967815,
+      "grad_norm": 0.43558669090270996,
+      "learning_rate": 9.916377518136367e-06,
+      "loss": 0.4843,
+      "step": 550
+    },
+    {
+      "epoch": 0.4586570477247503,
+      "grad_norm": 0.526262104511261,
+      "learning_rate": 9.915492790861986e-06,
+      "loss": 0.4854,
+      "step": 551
+    },
+    {
+      "epoch": 0.4594894561598224,
+      "grad_norm": 0.5632872581481934,
+      "learning_rate": 9.91460344785381e-06,
+      "loss": 0.4879,
+      "step": 552
+    },
+    {
+      "epoch": 0.46032186459489455,
+      "grad_norm": 0.5479772686958313,
+      "learning_rate": 9.913709489946946e-06,
+      "loss": 0.5225,
+      "step": 553
+    },
+    {
+      "epoch": 0.4611542730299667,
+      "grad_norm": 0.5169483423233032,
+      "learning_rate": 9.912810917980834e-06,
+      "loss": 0.4791,
+      "step": 554
+    },
+    {
+      "epoch": 0.46198668146503885,
+      "grad_norm": 0.5335115194320679,
+      "learning_rate": 9.911907732799251e-06,
+      "loss": 0.4974,
+      "step": 555
+    },
+    {
+      "epoch": 0.462819089900111,
+      "grad_norm": 0.49141526222229004,
+      "learning_rate": 9.910999935250302e-06,
+      "loss": 0.4835,
+      "step": 556
+    },
+    {
+      "epoch": 0.46365149833518315,
+      "grad_norm": 0.7199199795722961,
+      "learning_rate": 9.910087526186424e-06,
+      "loss": 0.502,
+      "step": 557
+    },
+    {
+      "epoch": 0.46448390677025525,
+      "grad_norm": 7.59892463684082,
+      "learning_rate": 9.909170506464389e-06,
+      "loss": 0.453,
+      "step": 558
+    },
+    {
+      "epoch": 0.4653163152053274,
+      "grad_norm": 4702.46728515625,
+      "learning_rate": 9.908248876945291e-06,
+      "loss": 1.725,
+      "step": 559
+    },
+    {
+      "epoch": 0.46614872364039955,
+      "grad_norm": 13.055675506591797,
+      "learning_rate": 9.907322638494558e-06,
+      "loss": 0.492,
+      "step": 560
+    },
+    {
+      "epoch": 0.4669811320754717,
+      "grad_norm": 2.1198976039886475,
+      "learning_rate": 9.906391791981944e-06,
+      "loss": 0.504,
+      "step": 561
+    },
+    {
+      "epoch": 0.46781354051054386,
+      "grad_norm": 0.6735315322875977,
+      "learning_rate": 9.90545633828153e-06,
+      "loss": 0.4779,
+      "step": 562
+    },
+    {
+      "epoch": 0.468645948945616,
+      "grad_norm": 0.7947428226470947,
+      "learning_rate": 9.904516278271721e-06,
+      "loss": 0.5156,
+      "step": 563
+    },
+    {
+      "epoch": 0.4694783573806881,
+      "grad_norm": 0.8350366950035095,
+      "learning_rate": 9.903571612835254e-06,
+      "loss": 0.4695,
+      "step": 564
+    },
+    {
+      "epoch": 0.47031076581576026,
+      "grad_norm": 0.6595231294631958,
+      "learning_rate": 9.902622342859183e-06,
+      "loss": 0.5045,
+      "step": 565
+    },
+    {
+      "epoch": 0.4711431742508324,
+      "grad_norm": 0.7960402369499207,
+      "learning_rate": 9.901668469234892e-06,
+      "loss": 0.4872,
+      "step": 566
+    },
+    {
+      "epoch": 0.47197558268590456,
+      "grad_norm": 0.6408143639564514,
+      "learning_rate": 9.900709992858083e-06,
+      "loss": 0.4483,
+      "step": 567
+    },
+    {
+      "epoch": 0.4728079911209767,
+      "grad_norm": 0.8016465306282043,
+      "learning_rate": 9.899746914628782e-06,
+      "loss": 0.4663,
+      "step": 568
+    },
+    {
+      "epoch": 0.4736403995560488,
+      "grad_norm": 0.5370069146156311,
+      "learning_rate": 9.898779235451337e-06,
+      "loss": 0.4861,
+      "step": 569
+    },
+    {
+      "epoch": 0.47447280799112096,
+      "grad_norm": 0.6317716836929321,
+      "learning_rate": 9.897806956234417e-06,
+      "loss": 0.4831,
+      "step": 570
+    },
+    {
+      "epoch": 0.4753052164261931,
+      "grad_norm": 0.649116575717926,
+      "learning_rate": 9.896830077891007e-06,
+      "loss": 0.484,
+      "step": 571
+    },
+    {
+      "epoch": 0.47613762486126526,
+      "grad_norm": 0.5388279557228088,
+      "learning_rate": 9.895848601338414e-06,
+      "loss": 0.5064,
+      "step": 572
+    },
+    {
+      "epoch": 0.4769700332963374,
+      "grad_norm": 0.6505693793296814,
+      "learning_rate": 9.894862527498259e-06,
+      "loss": 0.4757,
+      "step": 573
+    },
+    {
+      "epoch": 0.47780244173140957,
+      "grad_norm": 0.503017783164978,
+      "learning_rate": 9.893871857296487e-06,
+      "loss": 0.5006,
+      "step": 574
+    },
+    {
+      "epoch": 0.47863485016648166,
+      "grad_norm": 0.7261723875999451,
+      "learning_rate": 9.892876591663355e-06,
+      "loss": 0.4688,
+      "step": 575
+    },
+    {
+      "epoch": 0.4794672586015538,
+      "grad_norm": 0.49724769592285156,
+      "learning_rate": 9.891876731533429e-06,
+      "loss": 0.4621,
+      "step": 576
+    },
+    {
+      "epoch": 0.48029966703662597,
+      "grad_norm": 0.6609309315681458,
+      "learning_rate": 9.8908722778456e-06,
+      "loss": 0.4631,
+      "step": 577
+    },
+    {
+      "epoch": 0.4811320754716981,
+      "grad_norm": 0.5452806353569031,
+      "learning_rate": 9.889863231543065e-06,
+      "loss": 0.4881,
+      "step": 578
+    },
+    {
+      "epoch": 0.48196448390677027,
+      "grad_norm": 0.675403892993927,
+      "learning_rate": 9.888849593573339e-06,
+      "loss": 0.4709,
+      "step": 579
+    },
+    {
+      "epoch": 0.4827968923418424,
+      "grad_norm": 0.6482904553413391,
+      "learning_rate": 9.887831364888243e-06,
+      "loss": 0.4907,
+      "step": 580
+    },
+    {
+      "epoch": 0.4836293007769145,
+      "grad_norm": 0.6234882473945618,
+      "learning_rate": 9.886808546443914e-06,
+      "loss": 0.4801,
+      "step": 581
+    },
+    {
+      "epoch": 0.48446170921198667,
+      "grad_norm": 0.60924232006073,
+      "learning_rate": 9.885781139200794e-06,
+      "loss": 0.4668,
+      "step": 582
+    },
+    {
+      "epoch": 0.4852941176470588,
+      "grad_norm": 0.5140445828437805,
+      "learning_rate": 9.88474914412364e-06,
+      "loss": 0.471,
+      "step": 583
+    },
+    {
+      "epoch": 0.48612652608213097,
+      "grad_norm": 0.6285945773124695,
+      "learning_rate": 9.88371256218151e-06,
+      "loss": 0.469,
+      "step": 584
+    },
+    {
+      "epoch": 0.4869589345172031,
+      "grad_norm": 0.45175155997276306,
+      "learning_rate": 9.882671394347771e-06,
+      "loss": 0.4646,
+      "step": 585
+    },
+    {
+      "epoch": 0.4877913429522753,
+      "grad_norm": 0.4605112373828888,
+      "learning_rate": 9.881625641600104e-06,
+      "loss": 0.4653,
+      "step": 586
+    },
+    {
+      "epoch": 0.48862375138734737,
+      "grad_norm": 0.5589239597320557,
+      "learning_rate": 9.880575304920484e-06,
+      "loss": 0.4641,
+      "step": 587
+    },
+    {
+      "epoch": 0.4894561598224195,
+      "grad_norm": 0.38805943727493286,
+      "learning_rate": 9.879520385295197e-06,
+      "loss": 0.4584,
+      "step": 588
+    },
+    {
+      "epoch": 0.4902885682574917,
+      "grad_norm": 0.5301197171211243,
+      "learning_rate": 9.878460883714831e-06,
+      "loss": 0.5203,
+      "step": 589
+    },
+    {
+      "epoch": 0.4911209766925638,
+      "grad_norm": 0.47221046686172485,
+      "learning_rate": 9.877396801174277e-06,
+      "loss": 0.489,
+      "step": 590
+    },
+    {
+      "epoch": 0.491953385127636,
+      "grad_norm": 0.4699765145778656,
+      "learning_rate": 9.876328138672726e-06,
+      "loss": 0.4739,
+      "step": 591
+    },
+    {
+      "epoch": 0.49278579356270813,
+      "grad_norm": 0.4534553587436676,
+      "learning_rate": 9.875254897213674e-06,
+      "loss": 0.4624,
+      "step": 592
+    },
+    {
+      "epoch": 0.4936182019977802,
+      "grad_norm": 0.39965343475341797,
+      "learning_rate": 9.87417707780491e-06,
+      "loss": 0.4865,
+      "step": 593
+    },
+    {
+      "epoch": 0.4944506104328524,
+      "grad_norm": 0.43681901693344116,
+      "learning_rate": 9.873094681458525e-06,
+      "loss": 0.4793,
+      "step": 594
+    },
+    {
+      "epoch": 0.49528301886792453,
+      "grad_norm": 0.4199521839618683,
+      "learning_rate": 9.87200770919091e-06,
+      "loss": 0.4561,
+      "step": 595
+    },
+    {
+      "epoch": 0.4961154273029967,
+      "grad_norm": 0.42258721590042114,
+      "learning_rate": 9.870916162022752e-06,
+      "loss": 0.4763,
+      "step": 596
+    },
+    {
+      "epoch": 0.49694783573806883,
+      "grad_norm": 0.4213769733905792,
+      "learning_rate": 9.86982004097903e-06,
+      "loss": 0.4815,
+      "step": 597
+    },
+    {
+      "epoch": 0.49778024417314093,
+      "grad_norm": 0.41223376989364624,
+      "learning_rate": 9.868719347089024e-06,
+      "loss": 0.4688,
+      "step": 598
+    },
+    {
+      "epoch": 0.4986126526082131,
+      "grad_norm": 0.49295932054519653,
+      "learning_rate": 9.867614081386302e-06,
+      "loss": 0.5049,
+      "step": 599
+    },
+    {
+      "epoch": 0.49944506104328523,
+      "grad_norm": 0.3732836842536926,
+      "learning_rate": 9.866504244908728e-06,
+      "loss": 0.5007,
+      "step": 600
+    },
+    {
+      "epoch": 0.5002774694783574,
+      "grad_norm": 0.39977091550827026,
+      "learning_rate": 9.86538983869846e-06,
+      "loss": 0.4807,
+      "step": 601
+    },
+    {
+      "epoch": 0.5011098779134295,
+      "grad_norm": 0.4643203318119049,
+      "learning_rate": 9.864270863801944e-06,
+      "loss": 0.4807,
+      "step": 602
+    },
+    {
+      "epoch": 0.5019422863485017,
+      "grad_norm": 0.391868531703949,
+      "learning_rate": 9.863147321269918e-06,
+      "loss": 0.4536,
+      "step": 603
+    },
+    {
+      "epoch": 0.5027746947835738,
+      "grad_norm": 0.4176725149154663,
+      "learning_rate": 9.862019212157406e-06,
+      "loss": 0.4658,
+      "step": 604
+    },
+    {
+      "epoch": 0.503607103218646,
+      "grad_norm": 0.4765459895133972,
+      "learning_rate": 9.860886537523721e-06,
+      "loss": 0.4891,
+      "step": 605
+    },
+    {
+      "epoch": 0.5044395116537181,
+      "grad_norm": 0.41668006777763367,
+      "learning_rate": 9.859749298432468e-06,
+      "loss": 0.4859,
+      "step": 606
+    },
+    {
+      "epoch": 0.5052719200887902,
+      "grad_norm": 0.45202815532684326,
+      "learning_rate": 9.858607495951534e-06,
+      "loss": 0.4732,
+      "step": 607
+    },
+    {
+      "epoch": 0.5061043285238623,
+      "grad_norm": 0.41504961252212524,
+      "learning_rate": 9.857461131153089e-06,
+      "loss": 0.4529,
+      "step": 608
+    },
+    {
+      "epoch": 0.5069367369589345,
+      "grad_norm": 0.4873095452785492,
+      "learning_rate": 9.856310205113594e-06,
+      "loss": 0.4693,
+      "step": 609
+    },
+    {
+      "epoch": 0.5077691453940066,
+      "grad_norm": 0.42755189538002014,
+      "learning_rate": 9.855154718913782e-06,
+      "loss": 0.4502,
+      "step": 610
+    },
+    {
+      "epoch": 0.5086015538290788,
+      "grad_norm": 0.40592634677886963,
+      "learning_rate": 9.853994673638679e-06,
+      "loss": 0.469,
+      "step": 611
+    },
+    {
+      "epoch": 0.5094339622641509,
+      "grad_norm": 0.4807058274745941,
+      "learning_rate": 9.852830070377588e-06,
+      "loss": 0.489,
+      "step": 612
+    },
+    {
+      "epoch": 0.5102663706992231,
+      "grad_norm": 0.4219500720500946,
+      "learning_rate": 9.851660910224092e-06,
+      "loss": 0.4927,
+      "step": 613
+    },
+    {
+      "epoch": 0.5110987791342952,
+      "grad_norm": 0.45034486055374146,
+      "learning_rate": 9.85048719427605e-06,
+      "loss": 0.4877,
+      "step": 614
+    },
+    {
+      "epoch": 0.5119311875693674,
+      "grad_norm": 0.39540550112724304,
+      "learning_rate": 9.849308923635606e-06,
+      "loss": 0.4769,
+      "step": 615
+    },
+    {
+      "epoch": 0.5127635960044395,
+      "grad_norm": 0.40875479578971863,
+      "learning_rate": 9.848126099409175e-06,
+      "loss": 0.4732,
+      "step": 616
+    },
+    {
+      "epoch": 0.5135960044395117,
+      "grad_norm": 0.4183914363384247,
+      "learning_rate": 9.846938722707446e-06,
+      "loss": 0.4671,
+      "step": 617
+    },
+    {
+      "epoch": 0.5144284128745837,
+      "grad_norm": 0.3762715458869934,
+      "learning_rate": 9.845746794645393e-06,
+      "loss": 0.4749,
+      "step": 618
+    },
+    {
+      "epoch": 0.5152608213096559,
+      "grad_norm": 0.4459075927734375,
+      "learning_rate": 9.844550316342252e-06,
+      "loss": 0.5014,
+      "step": 619
+    },
+    {
+      "epoch": 0.516093229744728,
+      "grad_norm": 0.4192522466182709,
+      "learning_rate": 9.843349288921543e-06,
+      "loss": 0.4872,
+      "step": 620
+    },
+    {
+      "epoch": 0.5169256381798002,
+      "grad_norm": 0.40821659564971924,
+      "learning_rate": 9.842143713511044e-06,
+      "loss": 0.479,
+      "step": 621
+    },
+    {
+      "epoch": 0.5177580466148723,
+      "grad_norm": 0.45011478662490845,
+      "learning_rate": 9.840933591242817e-06,
+      "loss": 0.4743,
+      "step": 622
+    },
+    {
+      "epoch": 0.5185904550499445,
+      "grad_norm": 0.45597994327545166,
+      "learning_rate": 9.839718923253186e-06,
+      "loss": 0.4667,
+      "step": 623
+    },
+    {
+      "epoch": 0.5194228634850167,
+      "grad_norm": 0.4411117136478424,
+      "learning_rate": 9.838499710682745e-06,
+      "loss": 0.4849,
+      "step": 624
+    },
+    {
+      "epoch": 0.5202552719200888,
+      "grad_norm": 0.46066632866859436,
+      "learning_rate": 9.837275954676357e-06,
+      "loss": 0.4375,
+      "step": 625
+    },
+    {
+      "epoch": 0.521087680355161,
+      "grad_norm": 0.42149725556373596,
+      "learning_rate": 9.836047656383152e-06,
+      "loss": 0.4505,
+      "step": 626
+    },
+    {
+      "epoch": 0.5219200887902331,
+      "grad_norm": 0.4638105630874634,
+      "learning_rate": 9.834814816956521e-06,
+      "loss": 0.4442,
+      "step": 627
+    },
+    {
+      "epoch": 0.5227524972253053,
+      "grad_norm": 0.5468646883964539,
+      "learning_rate": 9.833577437554121e-06,
+      "loss": 0.4775,
+      "step": 628
+    },
+    {
+      "epoch": 0.5235849056603774,
+      "grad_norm": 0.44078686833381653,
+      "learning_rate": 9.832335519337877e-06,
+      "loss": 0.4493,
+      "step": 629
+    },
+    {
+      "epoch": 0.5244173140954494,
+      "grad_norm": 0.5681177973747253,
+      "learning_rate": 9.831089063473967e-06,
+      "loss": 0.4996,
+      "step": 630
+    },
+    {
+      "epoch": 0.5252497225305216,
+      "grad_norm": 0.4630327522754669,
+      "learning_rate": 9.82983807113284e-06,
+      "loss": 0.4714,
+      "step": 631
+    },
+    {
+      "epoch": 0.5260821309655938,
+      "grad_norm": 0.47248589992523193,
+      "learning_rate": 9.828582543489194e-06,
+      "loss": 0.4838,
+      "step": 632
+    },
+    {
+      "epoch": 0.5269145394006659,
+      "grad_norm": 0.5343927145004272,
+      "learning_rate": 9.827322481721998e-06,
+      "loss": 0.4678,
+      "step": 633
+    },
+    {
+      "epoch": 0.5277469478357381,
+      "grad_norm": 0.48827970027923584,
+      "learning_rate": 9.826057887014466e-06,
+      "loss": 0.4841,
+      "step": 634
+    },
+    {
+      "epoch": 0.5285793562708102,
+      "grad_norm": 0.42138999700546265,
+      "learning_rate": 9.824788760554078e-06,
+      "loss": 0.4821,
+      "step": 635
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 0.46162131428718567,
+      "learning_rate": 9.823515103532564e-06,
+      "loss": 0.463,
+      "step": 636
+    },
+    {
+      "epoch": 0.5302441731409545,
+      "grad_norm": 0.41701412200927734,
+      "learning_rate": 9.822236917145914e-06,
+      "loss": 0.5037,
+      "step": 637
+    },
+    {
+      "epoch": 0.5310765815760267,
+      "grad_norm": 0.43682774901390076,
+      "learning_rate": 9.820954202594362e-06,
+      "loss": 0.468,
+      "step": 638
+    },
+    {
+      "epoch": 0.5319089900110988,
+      "grad_norm": 0.5141327977180481,
+      "learning_rate": 9.819666961082402e-06,
+      "loss": 0.5035,
+      "step": 639
+    },
+    {
+      "epoch": 0.532741398446171,
+      "grad_norm": 0.4251098036766052,
+      "learning_rate": 9.81837519381878e-06,
+      "loss": 0.4749,
+      "step": 640
+    },
+    {
+      "epoch": 0.5335738068812431,
+      "grad_norm": 0.4675353765487671,
+      "learning_rate": 9.817078902016481e-06,
+      "loss": 0.4837,
+      "step": 641
+    },
+    {
+      "epoch": 0.5344062153163152,
+      "grad_norm": 0.44004005193710327,
+      "learning_rate": 9.81577808689275e-06,
+      "loss": 0.4729,
+      "step": 642
+    },
+    {
+      "epoch": 0.5352386237513873,
+      "grad_norm": 0.45024535059928894,
+      "learning_rate": 9.814472749669076e-06,
+      "loss": 0.4783,
+      "step": 643
+    },
+    {
+      "epoch": 0.5360710321864595,
+      "grad_norm": 0.3834257125854492,
+      "learning_rate": 9.813162891571189e-06,
+      "loss": 0.4503,
+      "step": 644
+    },
+    {
+      "epoch": 0.5369034406215316,
+      "grad_norm": 0.47850501537323,
+      "learning_rate": 9.811848513829074e-06,
+      "loss": 0.466,
+      "step": 645
+    },
+    {
+      "epoch": 0.5377358490566038,
+      "grad_norm": 0.397339403629303,
+      "learning_rate": 9.810529617676952e-06,
+      "loss": 0.4804,
+      "step": 646
+    },
+    {
+      "epoch": 0.5385682574916759,
+      "grad_norm": 0.45498281717300415,
+      "learning_rate": 9.809206204353289e-06,
+      "loss": 0.4587,
+      "step": 647
+    },
+    {
+      "epoch": 0.5394006659267481,
+      "grad_norm": 0.46825650334358215,
+      "learning_rate": 9.807878275100795e-06,
+      "loss": 0.4772,
+      "step": 648
+    },
+    {
+      "epoch": 0.5402330743618202,
+      "grad_norm": 0.40103840827941895,
+      "learning_rate": 9.806545831166417e-06,
+      "loss": 0.4725,
+      "step": 649
+    },
+    {
+      "epoch": 0.5410654827968924,
+      "grad_norm": 0.4803685247898102,
+      "learning_rate": 9.805208873801346e-06,
+      "loss": 0.4799,
+      "step": 650
+    },
+    {
+      "epoch": 0.5418978912319645,
+      "grad_norm": 0.4705343544483185,
+      "learning_rate": 9.803867404261005e-06,
+      "loss": 0.4769,
+      "step": 651
+    },
+    {
+      "epoch": 0.5427302996670367,
+      "grad_norm": 0.4226051867008209,
+      "learning_rate": 9.80252142380506e-06,
+      "loss": 0.4734,
+      "step": 652
+    },
+    {
+      "epoch": 0.5435627081021087,
+      "grad_norm": 0.42430418729782104,
+      "learning_rate": 9.80117093369741e-06,
+      "loss": 0.474,
+      "step": 653
+    },
+    {
+      "epoch": 0.5443951165371809,
+      "grad_norm": 0.4154928922653198,
+      "learning_rate": 9.799815935206187e-06,
+      "loss": 0.4819,
+      "step": 654
+    },
+    {
+      "epoch": 0.545227524972253,
+      "grad_norm": 0.359573096036911,
+      "learning_rate": 9.798456429603758e-06,
+      "loss": 0.4564,
+      "step": 655
+    },
+    {
+      "epoch": 0.5460599334073252,
+      "grad_norm": 0.5043289661407471,
+      "learning_rate": 9.797092418166725e-06,
+      "loss": 0.4784,
+      "step": 656
+    },
+    {
+      "epoch": 0.5468923418423973,
+      "grad_norm": 0.3750476539134979,
+      "learning_rate": 9.795723902175918e-06,
+      "loss": 0.4542,
+      "step": 657
+    },
+    {
+      "epoch": 0.5477247502774695,
+      "grad_norm": 0.4907470643520355,
+      "learning_rate": 9.794350882916397e-06,
+      "loss": 0.4908,
+      "step": 658
+    },
+    {
+      "epoch": 0.5485571587125416,
+      "grad_norm": 0.457126259803772,
+      "learning_rate": 9.79297336167745e-06,
+      "loss": 0.4323,
+      "step": 659
+    },
+    {
+      "epoch": 0.5493895671476138,
+      "grad_norm": 0.42247408628463745,
+      "learning_rate": 9.791591339752596e-06,
+      "loss": 0.4752,
+      "step": 660
+    },
+    {
+      "epoch": 0.5502219755826859,
+      "grad_norm": 0.49953287839889526,
+      "learning_rate": 9.790204818439576e-06,
+      "loss": 0.4852,
+      "step": 661
+    },
+    {
+      "epoch": 0.5510543840177581,
+      "grad_norm": 0.466348260641098,
+      "learning_rate": 9.788813799040358e-06,
+      "loss": 0.4654,
+      "step": 662
+    },
+    {
+      "epoch": 0.5518867924528302,
+      "grad_norm": 0.47069528698921204,
+      "learning_rate": 9.787418282861135e-06,
+      "loss": 0.4919,
+      "step": 663
+    },
+    {
+      "epoch": 0.5527192008879024,
+      "grad_norm": 0.3939701020717621,
+      "learning_rate": 9.786018271212318e-06,
+      "loss": 0.4712,
+      "step": 664
+    },
+    {
+      "epoch": 0.5535516093229744,
+      "grad_norm": 0.4161626994609833,
+      "learning_rate": 9.784613765408546e-06,
+      "loss": 0.4791,
+      "step": 665
+    },
+    {
+      "epoch": 0.5543840177580466,
+      "grad_norm": 0.39920780062675476,
+      "learning_rate": 9.783204766768672e-06,
+      "loss": 0.4669,
+      "step": 666
+    },
+    {
+      "epoch": 0.5552164261931187,
+      "grad_norm": 0.47431665658950806,
+      "learning_rate": 9.781791276615774e-06,
+      "loss": 0.4623,
+      "step": 667
+    },
+    {
+      "epoch": 0.5560488346281909,
+      "grad_norm": 0.4187391400337219,
+      "learning_rate": 9.780373296277137e-06,
+      "loss": 0.4766,
+      "step": 668
+    },
+    {
+      "epoch": 0.556881243063263,
+      "grad_norm": 0.39394712448120117,
+      "learning_rate": 9.778950827084277e-06,
+      "loss": 0.4857,
+      "step": 669
+    },
+    {
+      "epoch": 0.5577136514983352,
+      "grad_norm": 0.4316924512386322,
+      "learning_rate": 9.777523870372913e-06,
+      "loss": 0.472,
+      "step": 670
+    },
+    {
+      "epoch": 0.5585460599334073,
+      "grad_norm": 0.54462730884552,
+      "learning_rate": 9.776092427482984e-06,
+      "loss": 0.4716,
+      "step": 671
+    },
+    {
+      "epoch": 0.5593784683684795,
+      "grad_norm": 0.40116629004478455,
+      "learning_rate": 9.774656499758639e-06,
+      "loss": 0.4498,
+      "step": 672
+    },
+    {
+      "epoch": 0.5602108768035516,
+      "grad_norm": 0.5211840271949768,
+      "learning_rate": 9.77321608854824e-06,
+      "loss": 0.4702,
+      "step": 673
+    },
+    {
+      "epoch": 0.5610432852386238,
+      "grad_norm": 0.5067088603973389,
+      "learning_rate": 9.771771195204358e-06,
+      "loss": 0.4796,
+      "step": 674
+    },
+    {
+      "epoch": 0.5618756936736959,
+      "grad_norm": 0.5115644335746765,
+      "learning_rate": 9.770321821083774e-06,
+      "loss": 0.4809,
+      "step": 675
+    },
+    {
+      "epoch": 0.562708102108768,
+      "grad_norm": 0.4665434956550598,
+      "learning_rate": 9.768867967547472e-06,
+      "loss": 0.4649,
+      "step": 676
+    },
+    {
+      "epoch": 0.5635405105438401,
+      "grad_norm": 0.5498418211936951,
+      "learning_rate": 9.767409635960653e-06,
+      "loss": 0.4799,
+      "step": 677
+    },
+    {
+      "epoch": 0.5643729189789123,
+      "grad_norm": 0.41861289739608765,
+      "learning_rate": 9.76594682769271e-06,
+      "loss": 0.4655,
+      "step": 678
+    },
+    {
+      "epoch": 0.5652053274139844,
+      "grad_norm": 0.4563583731651306,
+      "learning_rate": 9.764479544117247e-06,
+      "loss": 0.4608,
+      "step": 679
+    },
+    {
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.45897558331489563,
+      "learning_rate": 9.76300778661207e-06,
+      "loss": 0.4476,
+      "step": 680
+    },
+    {
+      "epoch": 0.5668701442841287,
+      "grad_norm": 0.42618608474731445,
+      "learning_rate": 9.761531556559183e-06,
+      "loss": 0.4758,
+      "step": 681
+    },
+    {
+      "epoch": 0.5677025527192009,
+      "grad_norm": 0.3950604796409607,
+      "learning_rate": 9.760050855344795e-06,
+      "loss": 0.4566,
+      "step": 682
+    },
+    {
+      "epoch": 0.568534961154273,
+      "grad_norm": 0.46975719928741455,
+      "learning_rate": 9.758565684359307e-06,
+      "loss": 0.4612,
+      "step": 683
+    },
+    {
+      "epoch": 0.5693673695893452,
+      "grad_norm": 0.4233880639076233,
+      "learning_rate": 9.757076044997324e-06,
+      "loss": 0.4591,
+      "step": 684
+    },
+    {
+      "epoch": 0.5701997780244173,
+      "grad_norm": 0.42603635787963867,
+      "learning_rate": 9.75558193865764e-06,
+      "loss": 0.4841,
+      "step": 685
+    },
+    {
+      "epoch": 0.5710321864594895,
+      "grad_norm": 0.4705389738082886,
+      "learning_rate": 9.754083366743249e-06,
+      "loss": 0.4501,
+      "step": 686
+    },
+    {
+      "epoch": 0.5718645948945617,
+      "grad_norm": 0.4174215793609619,
+      "learning_rate": 9.752580330661336e-06,
+      "loss": 0.4804,
+      "step": 687
+    },
+    {
+      "epoch": 0.5726970033296337,
+      "grad_norm": 0.47212427854537964,
+      "learning_rate": 9.751072831823279e-06,
+      "loss": 0.4987,
+      "step": 688
+    },
+    {
+      "epoch": 0.5735294117647058,
+      "grad_norm": 0.453812837600708,
+      "learning_rate": 9.749560871644643e-06,
+      "loss": 0.4655,
+      "step": 689
+    },
+    {
+      "epoch": 0.574361820199778,
+      "grad_norm": 0.44600024819374084,
+      "learning_rate": 9.748044451545188e-06,
+      "loss": 0.4926,
+      "step": 690
+    },
+    {
+      "epoch": 0.5751942286348501,
+      "grad_norm": 0.5240420699119568,
+      "learning_rate": 9.746523572948857e-06,
+      "loss": 0.4919,
+      "step": 691
+    },
+    {
+      "epoch": 0.5760266370699223,
+      "grad_norm": 0.415798544883728,
+      "learning_rate": 9.744998237283785e-06,
+      "loss": 0.4447,
+      "step": 692
+    },
+    {
+      "epoch": 0.5768590455049944,
+      "grad_norm": 0.39677396416664124,
+      "learning_rate": 9.743468445982284e-06,
+      "loss": 0.4626,
+      "step": 693
+    },
+    {
+      "epoch": 0.5776914539400666,
+      "grad_norm": 0.4481954276561737,
+      "learning_rate": 9.741934200480857e-06,
+      "loss": 0.4845,
+      "step": 694
+    },
+    {
+      "epoch": 0.5785238623751388,
+      "grad_norm": 0.4083997309207916,
+      "learning_rate": 9.740395502220192e-06,
+      "loss": 0.4673,
+      "step": 695
+    },
+    {
+      "epoch": 0.5793562708102109,
+      "grad_norm": 0.4454471170902252,
+      "learning_rate": 9.738852352645145e-06,
+      "loss": 0.4606,
+      "step": 696
+    },
+    {
+      "epoch": 0.5801886792452831,
+      "grad_norm": 0.45555463433265686,
+      "learning_rate": 9.737304753204767e-06,
+      "loss": 0.482,
+      "step": 697
+    },
+    {
+      "epoch": 0.5810210876803552,
+      "grad_norm": 0.5052851438522339,
+      "learning_rate": 9.735752705352278e-06,
+      "loss": 0.4909,
+      "step": 698
+    },
+    {
+      "epoch": 0.5818534961154272,
+      "grad_norm": 0.40319934487342834,
+      "learning_rate": 9.734196210545079e-06,
+      "loss": 0.4804,
+      "step": 699
+    },
+    {
+      "epoch": 0.5826859045504994,
+      "grad_norm": 0.4926762878894806,
+      "learning_rate": 9.732635270244745e-06,
+      "loss": 0.455,
+      "step": 700
+    },
+    {
+      "epoch": 0.5835183129855716,
+      "grad_norm": 0.41983771324157715,
+      "learning_rate": 9.731069885917029e-06,
+      "loss": 0.4748,
+      "step": 701
+    },
+    {
+      "epoch": 0.5843507214206437,
+      "grad_norm": 0.4828875958919525,
+      "learning_rate": 9.729500059031851e-06,
+      "loss": 0.4664,
+      "step": 702
+    },
+    {
+      "epoch": 0.5851831298557159,
+      "grad_norm": 0.4084136188030243,
+      "learning_rate": 9.727925791063306e-06,
+      "loss": 0.4576,
+      "step": 703
+    },
+    {
+      "epoch": 0.586015538290788,
+      "grad_norm": 0.5071380138397217,
+      "learning_rate": 9.726347083489661e-06,
+      "loss": 0.4702,
+      "step": 704
+    },
+    {
+      "epoch": 0.5868479467258602,
+      "grad_norm": 0.4290992021560669,
+      "learning_rate": 9.724763937793352e-06,
+      "loss": 0.4669,
+      "step": 705
+    },
+    {
+      "epoch": 0.5876803551609323,
+      "grad_norm": 0.46184346079826355,
+      "learning_rate": 9.723176355460978e-06,
+      "loss": 0.4622,
+      "step": 706
+    },
+    {
+      "epoch": 0.5885127635960045,
+      "grad_norm": 0.5367664694786072,
+      "learning_rate": 9.721584337983303e-06,
+      "loss": 0.5039,
+      "step": 707
+    },
+    {
+      "epoch": 0.5893451720310766,
+      "grad_norm": 0.47268781065940857,
+      "learning_rate": 9.719987886855264e-06,
+      "loss": 0.4887,
+      "step": 708
+    },
+    {
+      "epoch": 0.5901775804661488,
+      "grad_norm": 0.5064681768417358,
+      "learning_rate": 9.718387003575957e-06,
+      "loss": 0.4529,
+      "step": 709
+    },
+    {
+      "epoch": 0.5910099889012209,
+      "grad_norm": 0.4127980172634125,
+      "learning_rate": 9.716781689648638e-06,
+      "loss": 0.4661,
+      "step": 710
+    },
+    {
+      "epoch": 0.591842397336293,
+      "grad_norm": 0.525039553642273,
+      "learning_rate": 9.715171946580724e-06,
+      "loss": 0.4509,
+      "step": 711
+    },
+    {
+      "epoch": 0.5926748057713651,
+      "grad_norm": 0.5468900799751282,
+      "learning_rate": 9.713557775883793e-06,
+      "loss": 0.4854,
+      "step": 712
+    },
+    {
+      "epoch": 0.5935072142064373,
+      "grad_norm": 0.4610214829444885,
+      "learning_rate": 9.71193917907358e-06,
+      "loss": 0.4783,
+      "step": 713
+    },
+    {
+      "epoch": 0.5943396226415094,
+      "grad_norm": 0.5212530493736267,
+      "learning_rate": 9.710316157669972e-06,
+      "loss": 0.4461,
+      "step": 714
+    },
+    {
+      "epoch": 0.5951720310765816,
+      "grad_norm": 0.4037787616252899,
+      "learning_rate": 9.708688713197021e-06,
+      "loss": 0.4931,
+      "step": 715
+    },
+    {
+      "epoch": 0.5960044395116537,
+      "grad_norm": 0.4619676172733307,
+      "learning_rate": 9.707056847182921e-06,
+      "loss": 0.4709,
+      "step": 716
+    },
+    {
+      "epoch": 0.5968368479467259,
+      "grad_norm": 0.4411291778087616,
+      "learning_rate": 9.705420561160024e-06,
+      "loss": 0.4923,
+      "step": 717
+    },
+    {
+      "epoch": 0.597669256381798,
+      "grad_norm": 0.45065465569496155,
+      "learning_rate": 9.703779856664833e-06,
+      "loss": 0.457,
+      "step": 718
+    },
+    {
+      "epoch": 0.5985016648168702,
+      "grad_norm": 0.4192379415035248,
+      "learning_rate": 9.702134735237994e-06,
+      "loss": 0.4815,
+      "step": 719
+    },
+    {
+      "epoch": 0.5993340732519423,
+      "grad_norm": 0.5066999793052673,
+      "learning_rate": 9.700485198424307e-06,
+      "loss": 0.4763,
+      "step": 720
+    },
+    {
+      "epoch": 0.6001664816870145,
+      "grad_norm": 0.4297652840614319,
+      "learning_rate": 9.69883124777272e-06,
+      "loss": 0.4545,
+      "step": 721
+    },
+    {
+      "epoch": 0.6009988901220865,
+      "grad_norm": 0.4559658169746399,
+      "learning_rate": 9.697172884836315e-06,
+      "loss": 0.4581,
+      "step": 722
+    },
+    {
+      "epoch": 0.6018312985571587,
+      "grad_norm": 0.41913849115371704,
+      "learning_rate": 9.695510111172329e-06,
+      "loss": 0.443,
+      "step": 723
+    },
+    {
+      "epoch": 0.6026637069922308,
+      "grad_norm": 0.40670254826545715,
+      "learning_rate": 9.693842928342132e-06,
+      "loss": 0.4689,
+      "step": 724
+    },
+    {
+      "epoch": 0.603496115427303,
+      "grad_norm": 0.43783038854599,
+      "learning_rate": 9.69217133791124e-06,
+      "loss": 0.4633,
+      "step": 725
+    },
+    {
+      "epoch": 0.6043285238623751,
+      "grad_norm": 0.4118957221508026,
+      "learning_rate": 9.690495341449304e-06,
+      "loss": 0.4812,
+      "step": 726
+    },
+    {
+      "epoch": 0.6051609322974473,
+      "grad_norm": 0.41395294666290283,
+      "learning_rate": 9.688814940530115e-06,
+      "loss": 0.4975,
+      "step": 727
+    },
+    {
+      "epoch": 0.6059933407325194,
+      "grad_norm": 0.485917866230011,
+      "learning_rate": 9.6871301367316e-06,
+      "loss": 0.5011,
+      "step": 728
+    },
+    {
+      "epoch": 0.6068257491675916,
+      "grad_norm": 0.39293450117111206,
+      "learning_rate": 9.68544093163582e-06,
+      "loss": 0.4541,
+      "step": 729
+    },
+    {
+      "epoch": 0.6076581576026637,
+      "grad_norm": 0.47208675742149353,
+      "learning_rate": 9.683747326828962e-06,
+      "loss": 0.4707,
+      "step": 730
+    },
+    {
+      "epoch": 0.6084905660377359,
+      "grad_norm": 0.41672036051750183,
+      "learning_rate": 9.682049323901358e-06,
+      "loss": 0.4689,
+      "step": 731
+    },
+    {
+      "epoch": 0.609322974472808,
+      "grad_norm": 0.45332202315330505,
+      "learning_rate": 9.680346924447458e-06,
+      "loss": 0.5,
+      "step": 732
+    },
+    {
+      "epoch": 0.6101553829078802,
+      "grad_norm": 0.4699179530143738,
+      "learning_rate": 9.678640130065846e-06,
+      "loss": 0.4914,
+      "step": 733
+    },
+    {
+      "epoch": 0.6109877913429522,
+      "grad_norm": 0.42764753103256226,
+      "learning_rate": 9.676928942359233e-06,
+      "loss": 0.4713,
+      "step": 734
+    },
+    {
+      "epoch": 0.6118201997780244,
+      "grad_norm": 0.4510103464126587,
+      "learning_rate": 9.675213362934454e-06,
+      "loss": 0.4422,
+      "step": 735
+    },
+    {
+      "epoch": 0.6126526082130965,
+      "grad_norm": 0.44094187021255493,
+      "learning_rate": 9.673493393402466e-06,
+      "loss": 0.478,
+      "step": 736
+    },
+    {
+      "epoch": 0.6134850166481687,
+      "grad_norm": 0.4860018491744995,
+      "learning_rate": 9.671769035378352e-06,
+      "loss": 0.4597,
+      "step": 737
+    },
+    {
+      "epoch": 0.6143174250832408,
+      "grad_norm": 0.4602179229259491,
+      "learning_rate": 9.670040290481315e-06,
+      "loss": 0.477,
+      "step": 738
+    },
+    {
+      "epoch": 0.615149833518313,
+      "grad_norm": 0.4050056040287018,
+      "learning_rate": 9.668307160334676e-06,
+      "loss": 0.4669,
+      "step": 739
+    },
+    {
+      "epoch": 0.6159822419533851,
+      "grad_norm": 0.49610546231269836,
+      "learning_rate": 9.666569646565875e-06,
+      "loss": 0.4783,
+      "step": 740
+    },
+    {
+      "epoch": 0.6168146503884573,
+      "grad_norm": 0.37179750204086304,
+      "learning_rate": 9.664827750806465e-06,
+      "loss": 0.4661,
+      "step": 741
+    },
+    {
+      "epoch": 0.6176470588235294,
+      "grad_norm": 0.42984944581985474,
+      "learning_rate": 9.663081474692123e-06,
+      "loss": 0.4516,
+      "step": 742
+    },
+    {
+      "epoch": 0.6184794672586016,
+      "grad_norm": 0.4149248003959656,
+      "learning_rate": 9.661330819862626e-06,
+      "loss": 0.4644,
+      "step": 743
+    },
+    {
+      "epoch": 0.6193118756936737,
+      "grad_norm": 0.4002021551132202,
+      "learning_rate": 9.659575787961872e-06,
+      "loss": 0.4571,
+      "step": 744
+    },
+    {
+      "epoch": 0.6201442841287459,
+      "grad_norm": 0.3851153552532196,
+      "learning_rate": 9.657816380637868e-06,
+      "loss": 0.4761,
+      "step": 745
+    },
+    {
+      "epoch": 0.6209766925638179,
+      "grad_norm": 0.4163179397583008,
+      "learning_rate": 9.656052599542728e-06,
+      "loss": 0.4712,
+      "step": 746
+    },
+    {
+      "epoch": 0.6218091009988901,
+      "grad_norm": 0.41883882880210876,
+      "learning_rate": 9.654284446332673e-06,
+      "loss": 0.4481,
+      "step": 747
+    },
+    {
+      "epoch": 0.6226415094339622,
+      "grad_norm": 0.35763731598854065,
+      "learning_rate": 9.652511922668029e-06,
+      "loss": 0.4512,
+      "step": 748
+    },
+    {
+      "epoch": 0.6234739178690344,
+      "grad_norm": 0.4117819368839264,
+      "learning_rate": 9.650735030213228e-06,
+      "loss": 0.483,
+      "step": 749
+    },
+    {
+      "epoch": 0.6243063263041065,
+      "grad_norm": 0.4240819811820984,
+      "learning_rate": 9.648953770636801e-06,
+      "loss": 0.476,
+      "step": 750
+    },
+    {
+      "epoch": 0.6251387347391787,
+      "grad_norm": 0.3860422372817993,
+      "learning_rate": 9.647168145611385e-06,
+      "loss": 0.4808,
+      "step": 751
+    },
+    {
+      "epoch": 0.6259711431742508,
+      "grad_norm": 0.39864394068717957,
+      "learning_rate": 9.645378156813709e-06,
+      "loss": 0.4881,
+      "step": 752
+    },
+    {
+      "epoch": 0.626803551609323,
+      "grad_norm": 0.4221175014972687,
+      "learning_rate": 9.643583805924608e-06,
+      "loss": 0.4726,
+      "step": 753
+    },
+    {
+      "epoch": 0.6276359600443951,
+      "grad_norm": 0.40564703941345215,
+      "learning_rate": 9.641785094629008e-06,
+      "loss": 0.4507,
+      "step": 754
+    },
+    {
+      "epoch": 0.6284683684794673,
+      "grad_norm": 0.44357532262802124,
+      "learning_rate": 9.639982024615928e-06,
+      "loss": 0.4721,
+      "step": 755
+    },
+    {
+      "epoch": 0.6293007769145395,
+      "grad_norm": 0.4528225064277649,
+      "learning_rate": 9.638174597578486e-06,
+      "loss": 0.4774,
+      "step": 756
+    },
+    {
+      "epoch": 0.6301331853496115,
+      "grad_norm": 0.41333818435668945,
+      "learning_rate": 9.636362815213884e-06,
+      "loss": 0.4824,
+      "step": 757
+    },
+    {
+      "epoch": 0.6309655937846836,
+      "grad_norm": 0.41459783911705017,
+      "learning_rate": 9.63454667922342e-06,
+      "loss": 0.4808,
+      "step": 758
+    },
+    {
+      "epoch": 0.6317980022197558,
+      "grad_norm": 0.4365558624267578,
+      "learning_rate": 9.632726191312475e-06,
+      "loss": 0.4583,
+      "step": 759
+    },
+    {
+      "epoch": 0.632630410654828,
+      "grad_norm": 0.4275365471839905,
+      "learning_rate": 9.630901353190522e-06,
+      "loss": 0.4966,
+      "step": 760
+    },
+    {
+      "epoch": 0.6334628190899001,
+      "grad_norm": 0.4615970253944397,
+      "learning_rate": 9.629072166571114e-06,
+      "loss": 0.4945,
+      "step": 761
+    },
+    {
+      "epoch": 0.6342952275249722,
+      "grad_norm": 0.43543118238449097,
+      "learning_rate": 9.627238633171889e-06,
+      "loss": 0.4534,
+      "step": 762
+    },
+    {
+      "epoch": 0.6351276359600444,
+      "grad_norm": 0.37240052223205566,
+      "learning_rate": 9.625400754714568e-06,
+      "loss": 0.443,
+      "step": 763
+    },
+    {
+      "epoch": 0.6359600443951166,
+      "grad_norm": 0.47188568115234375,
+      "learning_rate": 9.623558532924952e-06,
+      "loss": 0.4559,
+      "step": 764
+    },
+    {
+      "epoch": 0.6367924528301887,
+      "grad_norm": 0.38750484585762024,
+      "learning_rate": 9.621711969532917e-06,
+      "loss": 0.4805,
+      "step": 765
+    },
+    {
+      "epoch": 0.6376248612652609,
+      "grad_norm": 0.40850406885147095,
+      "learning_rate": 9.61986106627242e-06,
+      "loss": 0.4768,
+      "step": 766
+    },
+    {
+      "epoch": 0.638457269700333,
+      "grad_norm": 0.38356107473373413,
+      "learning_rate": 9.618005824881491e-06,
+      "loss": 0.4671,
+      "step": 767
+    },
+    {
+      "epoch": 0.6392896781354052,
+      "grad_norm": 0.4014689028263092,
+      "learning_rate": 9.616146247102233e-06,
+      "loss": 0.4405,
+      "step": 768
+    },
+    {
+      "epoch": 0.6401220865704772,
+      "grad_norm": 0.3854086995124817,
+      "learning_rate": 9.614282334680827e-06,
+      "loss": 0.4562,
+      "step": 769
+    },
+    {
+      "epoch": 0.6409544950055494,
+      "grad_norm": 0.4454103708267212,
+      "learning_rate": 9.612414089367512e-06,
+      "loss": 0.4494,
+      "step": 770
+    },
+    {
+      "epoch": 0.6417869034406215,
+      "grad_norm": 0.4228101968765259,
+      "learning_rate": 9.61054151291661e-06,
+      "loss": 0.4675,
+      "step": 771
+    },
+    {
+      "epoch": 0.6426193118756937,
+      "grad_norm": 0.4053763151168823,
+      "learning_rate": 9.608664607086497e-06,
+      "loss": 0.4618,
+      "step": 772
+    },
+    {
+      "epoch": 0.6434517203107658,
+      "grad_norm": 0.4087989926338196,
+      "learning_rate": 9.606783373639626e-06,
+      "loss": 0.4618,
+      "step": 773
+    },
+    {
+      "epoch": 0.644284128745838,
+      "grad_norm": 0.4213400185108185,
+      "learning_rate": 9.604897814342504e-06,
+      "loss": 0.4909,
+      "step": 774
+    },
+    {
+      "epoch": 0.6451165371809101,
+      "grad_norm": 0.4600432217121124,
+      "learning_rate": 9.603007930965706e-06,
+      "loss": 0.5026,
+      "step": 775
+    },
+    {
+      "epoch": 0.6459489456159823,
+      "grad_norm": 0.3727787435054779,
+      "learning_rate": 9.601113725283864e-06,
+      "loss": 0.4797,
+      "step": 776
+    },
+    {
+      "epoch": 0.6467813540510544,
+      "grad_norm": 0.4441661834716797,
+      "learning_rate": 9.599215199075674e-06,
+      "loss": 0.4748,
+      "step": 777
+    },
+    {
+      "epoch": 0.6476137624861266,
+      "grad_norm": 0.386455237865448,
+      "learning_rate": 9.597312354123882e-06,
+      "loss": 0.4657,
+      "step": 778
+    },
+    {
+      "epoch": 0.6484461709211987,
+      "grad_norm": 0.4307407736778259,
+      "learning_rate": 9.595405192215293e-06,
+      "loss": 0.4416,
+      "step": 779
+    },
+    {
+      "epoch": 0.6492785793562708,
+      "grad_norm": 0.3854341208934784,
+      "learning_rate": 9.593493715140767e-06,
+      "loss": 0.4849,
+      "step": 780
+    },
+    {
+      "epoch": 0.6501109877913429,
+      "grad_norm": 0.38782092928886414,
+      "learning_rate": 9.591577924695213e-06,
+      "loss": 0.4506,
+      "step": 781
+    },
+    {
+      "epoch": 0.6509433962264151,
+      "grad_norm": 0.4059392809867859,
+      "learning_rate": 9.589657822677592e-06,
+      "loss": 0.4686,
+      "step": 782
+    },
+    {
+      "epoch": 0.6517758046614872,
+      "grad_norm": 0.39608749747276306,
+      "learning_rate": 9.587733410890916e-06,
+      "loss": 0.4789,
+      "step": 783
+    },
+    {
+      "epoch": 0.6526082130965594,
+      "grad_norm": 0.40410831570625305,
+      "learning_rate": 9.585804691142237e-06,
+      "loss": 0.4462,
+      "step": 784
+    },
+    {
+      "epoch": 0.6534406215316315,
+      "grad_norm": 0.40202003717422485,
+      "learning_rate": 9.583871665242659e-06,
+      "loss": 0.4329,
+      "step": 785
+    },
+    {
+      "epoch": 0.6542730299667037,
+      "grad_norm": 0.42759740352630615,
+      "learning_rate": 9.581934335007326e-06,
+      "loss": 0.4415,
+      "step": 786
+    },
+    {
+      "epoch": 0.6551054384017758,
+      "grad_norm": 0.44305333495140076,
+      "learning_rate": 9.579992702255428e-06,
+      "loss": 0.4938,
+      "step": 787
+    },
+    {
+      "epoch": 0.655937846836848,
+      "grad_norm": 0.427234947681427,
+      "learning_rate": 9.57804676881019e-06,
+      "loss": 0.4579,
+      "step": 788
+    },
+    {
+      "epoch": 0.6567702552719201,
+      "grad_norm": 0.45137763023376465,
+      "learning_rate": 9.576096536498875e-06,
+      "loss": 0.493,
+      "step": 789
+    },
+    {
+      "epoch": 0.6576026637069923,
+      "grad_norm": 0.47418224811553955,
+      "learning_rate": 9.574142007152789e-06,
+      "loss": 0.4789,
+      "step": 790
+    },
+    {
+      "epoch": 0.6584350721420644,
+      "grad_norm": 0.4159716069698334,
+      "learning_rate": 9.572183182607269e-06,
+      "loss": 0.4555,
+      "step": 791
+    },
+    {
+      "epoch": 0.6592674805771365,
+      "grad_norm": 0.40891221165657043,
+      "learning_rate": 9.570220064701686e-06,
+      "loss": 0.4561,
+      "step": 792
+    },
+    {
+      "epoch": 0.6600998890122086,
+      "grad_norm": 0.4291469156742096,
+      "learning_rate": 9.568252655279438e-06,
+      "loss": 0.4844,
+      "step": 793
+    },
+    {
+      "epoch": 0.6609322974472808,
+      "grad_norm": 0.41984859108924866,
+      "learning_rate": 9.566280956187961e-06,
+      "loss": 0.4797,
+      "step": 794
+    },
+    {
+      "epoch": 0.6617647058823529,
+      "grad_norm": 0.43832656741142273,
+      "learning_rate": 9.564304969278714e-06,
+      "loss": 0.4559,
+      "step": 795
+    },
+    {
+      "epoch": 0.6625971143174251,
+      "grad_norm": 0.42272278666496277,
+      "learning_rate": 9.562324696407181e-06,
+      "loss": 0.4788,
+      "step": 796
+    },
+    {
+      "epoch": 0.6634295227524972,
+      "grad_norm": 0.4183189868927002,
+      "learning_rate": 9.560340139432877e-06,
+      "loss": 0.4894,
+      "step": 797
+    },
+    {
+      "epoch": 0.6642619311875694,
+      "grad_norm": 0.3984436094760895,
+      "learning_rate": 9.558351300219335e-06,
+      "loss": 0.4917,
+      "step": 798
+    },
+    {
+      "epoch": 0.6650943396226415,
+      "grad_norm": 0.37821829319000244,
+      "learning_rate": 9.556358180634105e-06,
+      "loss": 0.4486,
+      "step": 799
+    },
+    {
+      "epoch": 0.6659267480577137,
+      "grad_norm": 0.40212714672088623,
+      "learning_rate": 9.554360782548766e-06,
+      "loss": 0.4569,
+      "step": 800
+    },
+    {
+      "epoch": 0.6667591564927858,
+      "grad_norm": 0.421878457069397,
+      "learning_rate": 9.55235910783891e-06,
+      "loss": 0.4667,
+      "step": 801
+    },
+    {
+      "epoch": 0.667591564927858,
+      "grad_norm": 0.4031965732574463,
+      "learning_rate": 9.550353158384142e-06,
+      "loss": 0.4908,
+      "step": 802
+    },
+    {
+      "epoch": 0.66842397336293,
+      "grad_norm": 0.3970634937286377,
+      "learning_rate": 9.548342936068085e-06,
+      "loss": 0.4644,
+      "step": 803
+    },
+    {
+      "epoch": 0.6692563817980022,
+      "grad_norm": 0.437465637922287,
+      "learning_rate": 9.54632844277837e-06,
+      "loss": 0.4628,
+      "step": 804
+    },
+    {
+      "epoch": 0.6700887902330743,
+      "grad_norm": 0.3883856236934662,
+      "learning_rate": 9.544309680406648e-06,
+      "loss": 0.4712,
+      "step": 805
+    },
+    {
+      "epoch": 0.6709211986681465,
+      "grad_norm": 0.40419673919677734,
+      "learning_rate": 9.542286650848567e-06,
+      "loss": 0.4819,
+      "step": 806
+    },
+    {
+      "epoch": 0.6717536071032186,
+      "grad_norm": 0.4148818254470825,
+      "learning_rate": 9.540259356003787e-06,
+      "loss": 0.4714,
+      "step": 807
+    },
+    {
+      "epoch": 0.6725860155382908,
+      "grad_norm": 0.4469767212867737,
+      "learning_rate": 9.538227797775976e-06,
+      "loss": 0.4601,
+      "step": 808
+    },
+    {
+      "epoch": 0.6734184239733629,
+      "grad_norm": 0.37725868821144104,
+      "learning_rate": 9.536191978072802e-06,
+      "loss": 0.4671,
+      "step": 809
+    },
+    {
+      "epoch": 0.6742508324084351,
+      "grad_norm": 0.4010414183139801,
+      "learning_rate": 9.534151898805934e-06,
+      "loss": 0.4641,
+      "step": 810
+    },
+    {
+      "epoch": 0.6750832408435072,
+      "grad_norm": 0.37603193521499634,
+      "learning_rate": 9.532107561891044e-06,
+      "loss": 0.4483,
+      "step": 811
+    },
+    {
+      "epoch": 0.6759156492785794,
+      "grad_norm": 0.4010283350944519,
+      "learning_rate": 9.5300589692478e-06,
+      "loss": 0.4779,
+      "step": 812
+    },
+    {
+      "epoch": 0.6767480577136515,
+      "grad_norm": 0.3795294463634491,
+      "learning_rate": 9.528006122799864e-06,
+      "loss": 0.4757,
+      "step": 813
+    },
+    {
+      "epoch": 0.6775804661487237,
+      "grad_norm": 0.42904961109161377,
+      "learning_rate": 9.525949024474897e-06,
+      "loss": 0.4592,
+      "step": 814
+    },
+    {
+      "epoch": 0.6784128745837957,
+      "grad_norm": 0.392860472202301,
+      "learning_rate": 9.52388767620455e-06,
+      "loss": 0.477,
+      "step": 815
+    },
+    {
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.4124400019645691,
+      "learning_rate": 9.521822079924465e-06,
+      "loss": 0.4737,
+      "step": 816
+    },
+    {
+      "epoch": 0.68007769145394,
+      "grad_norm": 0.4388231039047241,
+      "learning_rate": 9.519752237574273e-06,
+      "loss": 0.4866,
+      "step": 817
+    },
+    {
+      "epoch": 0.6809100998890122,
+      "grad_norm": 0.372005820274353,
+      "learning_rate": 9.517678151097591e-06,
+      "loss": 0.4705,
+      "step": 818
+    },
+    {
+      "epoch": 0.6817425083240843,
+      "grad_norm": 0.4569007158279419,
+      "learning_rate": 9.515599822442025e-06,
+      "loss": 0.4756,
+      "step": 819
+    },
+    {
+      "epoch": 0.6825749167591565,
+      "grad_norm": 0.4299950897693634,
+      "learning_rate": 9.51351725355916e-06,
+      "loss": 0.4807,
+      "step": 820
+    },
+    {
+      "epoch": 0.6834073251942286,
+      "grad_norm": 0.3495566248893738,
+      "learning_rate": 9.511430446404566e-06,
+      "loss": 0.4593,
+      "step": 821
+    },
+    {
+      "epoch": 0.6842397336293008,
+      "grad_norm": 0.43988698720932007,
+      "learning_rate": 9.50933940293779e-06,
+      "loss": 0.4946,
+      "step": 822
+    },
+    {
+      "epoch": 0.685072142064373,
+      "grad_norm": 0.4119141399860382,
+      "learning_rate": 9.507244125122358e-06,
+      "loss": 0.4565,
+      "step": 823
+    },
+    {
+      "epoch": 0.6859045504994451,
+      "grad_norm": 0.40885767340660095,
+      "learning_rate": 9.505144614925776e-06,
+      "loss": 0.4624,
+      "step": 824
+    },
+    {
+      "epoch": 0.6867369589345172,
+      "grad_norm": 0.4196932017803192,
+      "learning_rate": 9.503040874319519e-06,
+      "loss": 0.4623,
+      "step": 825
+    },
+    {
+      "epoch": 0.6875693673695893,
+      "grad_norm": 0.40707823634147644,
+      "learning_rate": 9.500932905279034e-06,
+      "loss": 0.4807,
+      "step": 826
+    },
+    {
+      "epoch": 0.6884017758046614,
+      "grad_norm": 0.43840739130973816,
+      "learning_rate": 9.498820709783743e-06,
+      "loss": 0.4698,
+      "step": 827
+    },
+    {
+      "epoch": 0.6892341842397336,
+      "grad_norm": 0.3804190158843994,
+      "learning_rate": 9.496704289817035e-06,
+      "loss": 0.4404,
+      "step": 828
+    },
+    {
+      "epoch": 0.6900665926748057,
+      "grad_norm": 0.44267645478248596,
+      "learning_rate": 9.494583647366264e-06,
+      "loss": 0.4905,
+      "step": 829
+    },
+    {
+      "epoch": 0.6908990011098779,
+      "grad_norm": 0.41382384300231934,
+      "learning_rate": 9.492458784422751e-06,
+      "loss": 0.4689,
+      "step": 830
+    },
+    {
+      "epoch": 0.69173140954495,
+      "grad_norm": 0.42739763855934143,
+      "learning_rate": 9.49032970298178e-06,
+      "loss": 0.4782,
+      "step": 831
+    },
+    {
+      "epoch": 0.6925638179800222,
+      "grad_norm": 0.4234493672847748,
+      "learning_rate": 9.488196405042596e-06,
+      "loss": 0.4639,
+      "step": 832
+    },
+    {
+      "epoch": 0.6933962264150944,
+      "grad_norm": 0.3597807288169861,
+      "learning_rate": 9.486058892608401e-06,
+      "loss": 0.4541,
+      "step": 833
+    },
+    {
+      "epoch": 0.6942286348501665,
+      "grad_norm": 0.38074445724487305,
+      "learning_rate": 9.483917167686358e-06,
+      "loss": 0.4801,
+      "step": 834
+    },
+    {
+      "epoch": 0.6950610432852387,
+      "grad_norm": 0.38729429244995117,
+      "learning_rate": 9.481771232287585e-06,
+      "loss": 0.482,
+      "step": 835
+    },
+    {
+      "epoch": 0.6958934517203108,
+      "grad_norm": 0.45910367369651794,
+      "learning_rate": 9.479621088427152e-06,
+      "loss": 0.4814,
+      "step": 836
+    },
+    {
+      "epoch": 0.696725860155383,
+      "grad_norm": 0.4074380397796631,
+      "learning_rate": 9.47746673812408e-06,
+      "loss": 0.4817,
+      "step": 837
+    },
+    {
+      "epoch": 0.697558268590455,
+      "grad_norm": 0.4241724908351898,
+      "learning_rate": 9.475308183401347e-06,
+      "loss": 0.4727,
+      "step": 838
+    },
+    {
+      "epoch": 0.6983906770255272,
+      "grad_norm": 0.38416868448257446,
+      "learning_rate": 9.473145426285869e-06,
+      "loss": 0.456,
+      "step": 839
+    },
+    {
+      "epoch": 0.6992230854605993,
+      "grad_norm": 0.3833445608615875,
+      "learning_rate": 9.470978468808514e-06,
+      "loss": 0.4428,
+      "step": 840
+    },
+    {
+      "epoch": 0.7000554938956715,
+      "grad_norm": 0.39522784948349,
+      "learning_rate": 9.46880731300409e-06,
+      "loss": 0.435,
+      "step": 841
+    },
+    {
+      "epoch": 0.7008879023307436,
+      "grad_norm": 0.37949830293655396,
+      "learning_rate": 9.466631960911358e-06,
+      "loss": 0.4574,
+      "step": 842
+    },
+    {
+      "epoch": 0.7017203107658158,
+      "grad_norm": 0.3668852746486664,
+      "learning_rate": 9.464452414573004e-06,
+      "loss": 0.4629,
+      "step": 843
+    },
+    {
+      "epoch": 0.7025527192008879,
+      "grad_norm": 0.4042564332485199,
+      "learning_rate": 9.462268676035664e-06,
+      "loss": 0.4774,
+      "step": 844
+    },
+    {
+      "epoch": 0.7033851276359601,
+      "grad_norm": 0.4077381491661072,
+      "learning_rate": 9.460080747349907e-06,
+      "loss": 0.4776,
+      "step": 845
+    },
+    {
+      "epoch": 0.7042175360710322,
+      "grad_norm": 0.38546618819236755,
+      "learning_rate": 9.457888630570234e-06,
+      "loss": 0.4547,
+      "step": 846
+    },
+    {
+      "epoch": 0.7050499445061044,
+      "grad_norm": 0.3926650583744049,
+      "learning_rate": 9.455692327755082e-06,
+      "loss": 0.4741,
+      "step": 847
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 0.4238012135028839,
+      "learning_rate": 9.45349184096682e-06,
+      "loss": 0.4618,
+      "step": 848
+    },
+    {
+      "epoch": 0.7067147613762487,
+      "grad_norm": 0.42878350615501404,
+      "learning_rate": 9.451287172271741e-06,
+      "loss": 0.4404,
+      "step": 849
+    },
+    {
+      "epoch": 0.7075471698113207,
+      "grad_norm": 0.4180905222892761,
+      "learning_rate": 9.449078323740066e-06,
+      "loss": 0.4795,
+      "step": 850
+    },
+    {
+      "epoch": 0.7083795782463929,
+      "grad_norm": 0.40476009249687195,
+      "learning_rate": 9.446865297445947e-06,
+      "loss": 0.459,
+      "step": 851
+    },
+    {
+      "epoch": 0.709211986681465,
+      "grad_norm": 0.3711640238761902,
+      "learning_rate": 9.444648095467453e-06,
+      "loss": 0.4731,
+      "step": 852
+    },
+    {
+      "epoch": 0.7100443951165372,
+      "grad_norm": 0.3749478757381439,
+      "learning_rate": 9.442426719886572e-06,
+      "loss": 0.4579,
+      "step": 853
+    },
+    {
+      "epoch": 0.7108768035516093,
+      "grad_norm": 0.381773442029953,
+      "learning_rate": 9.440201172789218e-06,
+      "loss": 0.4455,
+      "step": 854
+    },
+    {
+      "epoch": 0.7117092119866815,
+      "grad_norm": 0.4290565848350525,
+      "learning_rate": 9.437971456265218e-06,
+      "loss": 0.4741,
+      "step": 855
+    },
+    {
+      "epoch": 0.7125416204217536,
+      "grad_norm": 0.3964015245437622,
+      "learning_rate": 9.435737572408316e-06,
+      "loss": 0.4771,
+      "step": 856
+    },
+    {
+      "epoch": 0.7133740288568258,
+      "grad_norm": 0.39722374081611633,
+      "learning_rate": 9.433499523316165e-06,
+      "loss": 0.4639,
+      "step": 857
+    },
+    {
+      "epoch": 0.7142064372918979,
+      "grad_norm": 0.39079853892326355,
+      "learning_rate": 9.431257311090336e-06,
+      "loss": 0.438,
+      "step": 858
+    },
+    {
+      "epoch": 0.7150388457269701,
+      "grad_norm": 0.3988889157772064,
+      "learning_rate": 9.429010937836302e-06,
+      "loss": 0.4471,
+      "step": 859
+    },
+    {
+      "epoch": 0.7158712541620422,
+      "grad_norm": 0.39545944333076477,
+      "learning_rate": 9.426760405663448e-06,
+      "loss": 0.4542,
+      "step": 860
+    },
+    {
+      "epoch": 0.7167036625971143,
+      "grad_norm": 0.39783865213394165,
+      "learning_rate": 9.424505716685064e-06,
+      "loss": 0.4667,
+      "step": 861
+    },
+    {
+      "epoch": 0.7175360710321864,
+      "grad_norm": 0.38103538751602173,
+      "learning_rate": 9.422246873018343e-06,
+      "loss": 0.4689,
+      "step": 862
+    },
+    {
+      "epoch": 0.7183684794672586,
+      "grad_norm": 0.3842844069004059,
+      "learning_rate": 9.419983876784378e-06,
+      "loss": 0.4659,
+      "step": 863
+    },
+    {
+      "epoch": 0.7192008879023307,
+      "grad_norm": 0.36254891753196716,
+      "learning_rate": 9.41771673010816e-06,
+      "loss": 0.437,
+      "step": 864
+    },
+    {
+      "epoch": 0.7200332963374029,
+      "grad_norm": 0.411409467458725,
+      "learning_rate": 9.415445435118581e-06,
+      "loss": 0.4671,
+      "step": 865
+    },
+    {
+      "epoch": 0.720865704772475,
+      "grad_norm": 0.4189411401748657,
+      "learning_rate": 9.41316999394843e-06,
+      "loss": 0.507,
+      "step": 866
+    },
+    {
+      "epoch": 0.7216981132075472,
+      "grad_norm": 0.448301762342453,
+      "learning_rate": 9.410890408734381e-06,
+      "loss": 0.4789,
+      "step": 867
+    },
+    {
+      "epoch": 0.7225305216426193,
+      "grad_norm": 0.39381828904151917,
+      "learning_rate": 9.408606681617006e-06,
+      "loss": 0.4514,
+      "step": 868
+    },
+    {
+      "epoch": 0.7233629300776915,
+      "grad_norm": 0.42399051785469055,
+      "learning_rate": 9.406318814740767e-06,
+      "loss": 0.4513,
+      "step": 869
+    },
+    {
+      "epoch": 0.7241953385127636,
+      "grad_norm": 0.39758798480033875,
+      "learning_rate": 9.404026810254007e-06,
+      "loss": 0.4539,
+      "step": 870
+    },
+    {
+      "epoch": 0.7250277469478358,
+      "grad_norm": 0.4223097562789917,
+      "learning_rate": 9.401730670308963e-06,
+      "loss": 0.4824,
+      "step": 871
+    },
+    {
+      "epoch": 0.7258601553829079,
+      "grad_norm": 0.42844802141189575,
+      "learning_rate": 9.399430397061746e-06,
+      "loss": 0.4759,
+      "step": 872
+    },
+    {
+      "epoch": 0.72669256381798,
+      "grad_norm": 0.4464619755744934,
+      "learning_rate": 9.397125992672358e-06,
+      "loss": 0.4437,
+      "step": 873
+    },
+    {
+      "epoch": 0.7275249722530521,
+      "grad_norm": 0.3978061079978943,
+      "learning_rate": 9.394817459304671e-06,
+      "loss": 0.4828,
+      "step": 874
+    },
+    {
+      "epoch": 0.7283573806881243,
+      "grad_norm": 0.4394519329071045,
+      "learning_rate": 9.392504799126439e-06,
+      "loss": 0.4746,
+      "step": 875
+    },
+    {
+      "epoch": 0.7291897891231964,
+      "grad_norm": 0.40214264392852783,
+      "learning_rate": 9.39018801430929e-06,
+      "loss": 0.4673,
+      "step": 876
+    },
+    {
+      "epoch": 0.7300221975582686,
+      "grad_norm": 0.40048447251319885,
+      "learning_rate": 9.387867107028727e-06,
+      "loss": 0.4793,
+      "step": 877
+    },
+    {
+      "epoch": 0.7308546059933407,
+      "grad_norm": 0.4119110703468323,
+      "learning_rate": 9.385542079464123e-06,
+      "loss": 0.4615,
+      "step": 878
+    },
+    {
+      "epoch": 0.7316870144284129,
+      "grad_norm": 0.39967912435531616,
+      "learning_rate": 9.383212933798718e-06,
+      "loss": 0.4664,
+      "step": 879
+    },
+    {
+      "epoch": 0.732519422863485,
+      "grad_norm": 0.39798152446746826,
+      "learning_rate": 9.38087967221962e-06,
+      "loss": 0.4575,
+      "step": 880
+    },
+    {
+      "epoch": 0.7333518312985572,
+      "grad_norm": 0.38400161266326904,
+      "learning_rate": 9.378542296917804e-06,
+      "loss": 0.456,
+      "step": 881
+    },
+    {
+      "epoch": 0.7341842397336293,
+      "grad_norm": 0.4055246412754059,
+      "learning_rate": 9.376200810088108e-06,
+      "loss": 0.4613,
+      "step": 882
+    },
+    {
+      "epoch": 0.7350166481687015,
+      "grad_norm": 0.3930290639400482,
+      "learning_rate": 9.373855213929227e-06,
+      "loss": 0.4699,
+      "step": 883
+    },
+    {
+      "epoch": 0.7358490566037735,
+      "grad_norm": 0.39786094427108765,
+      "learning_rate": 9.371505510643714e-06,
+      "loss": 0.483,
+      "step": 884
+    },
+    {
+      "epoch": 0.7366814650388457,
+      "grad_norm": 0.3793211579322815,
+      "learning_rate": 9.369151702437987e-06,
+      "loss": 0.4762,
+      "step": 885
+    },
+    {
+      "epoch": 0.7375138734739178,
+      "grad_norm": 0.38085439801216125,
+      "learning_rate": 9.366793791522308e-06,
+      "loss": 0.4535,
+      "step": 886
+    },
+    {
+      "epoch": 0.73834628190899,
+      "grad_norm": 0.371055543422699,
+      "learning_rate": 9.364431780110801e-06,
+      "loss": 0.4722,
+      "step": 887
+    },
+    {
+      "epoch": 0.7391786903440621,
+      "grad_norm": 0.3615517318248749,
+      "learning_rate": 9.362065670421434e-06,
+      "loss": 0.4177,
+      "step": 888
+    },
+    {
+      "epoch": 0.7400110987791343,
+      "grad_norm": 0.3889777660369873,
+      "learning_rate": 9.359695464676025e-06,
+      "loss": 0.455,
+      "step": 889
+    },
+    {
+      "epoch": 0.7408435072142064,
+      "grad_norm": 0.4410359561443329,
+      "learning_rate": 9.35732116510024e-06,
+      "loss": 0.4966,
+      "step": 890
+    },
+    {
+      "epoch": 0.7416759156492786,
+      "grad_norm": 0.42662596702575684,
+      "learning_rate": 9.354942773923588e-06,
+      "loss": 0.46,
+      "step": 891
+    },
+    {
+      "epoch": 0.7425083240843507,
+      "grad_norm": 0.394815593957901,
+      "learning_rate": 9.352560293379417e-06,
+      "loss": 0.4762,
+      "step": 892
+    },
+    {
+      "epoch": 0.7433407325194229,
+      "grad_norm": 0.39383867383003235,
+      "learning_rate": 9.350173725704922e-06,
+      "loss": 0.4519,
+      "step": 893
+    },
+    {
+      "epoch": 0.744173140954495,
+      "grad_norm": 0.40083327889442444,
+      "learning_rate": 9.34778307314113e-06,
+      "loss": 0.4345,
+      "step": 894
+    },
+    {
+      "epoch": 0.7450055493895672,
+      "grad_norm": 0.386772096157074,
+      "learning_rate": 9.345388337932906e-06,
+      "loss": 0.4519,
+      "step": 895
+    },
+    {
+      "epoch": 0.7458379578246392,
+      "grad_norm": 0.4634164273738861,
+      "learning_rate": 9.342989522328947e-06,
+      "loss": 0.4256,
+      "step": 896
+    },
+    {
+      "epoch": 0.7466703662597114,
+      "grad_norm": 0.4075722396373749,
+      "learning_rate": 9.340586628581783e-06,
+      "loss": 0.4548,
+      "step": 897
+    },
+    {
+      "epoch": 0.7475027746947835,
+      "grad_norm": 0.4286724925041199,
+      "learning_rate": 9.338179658947774e-06,
+      "loss": 0.4737,
+      "step": 898
+    },
+    {
+      "epoch": 0.7483351831298557,
+      "grad_norm": 0.43397256731987,
+      "learning_rate": 9.335768615687108e-06,
+      "loss": 0.4543,
+      "step": 899
+    },
+    {
+      "epoch": 0.7491675915649278,
+      "grad_norm": 0.40175217390060425,
+      "learning_rate": 9.333353501063796e-06,
+      "loss": 0.4702,
+      "step": 900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3902556300163269,
+      "learning_rate": 9.330934317345673e-06,
+      "loss": 0.4734,
+      "step": 901
+    },
+    {
+      "epoch": 0.7508324084350722,
+      "grad_norm": 0.40193432569503784,
+      "learning_rate": 9.328511066804391e-06,
+      "loss": 0.4382,
+      "step": 902
+    },
+    {
+      "epoch": 0.7516648168701443,
+      "grad_norm": 0.4484616219997406,
+      "learning_rate": 9.32608375171543e-06,
+      "loss": 0.4704,
+      "step": 903
+    },
+    {
+      "epoch": 0.7524972253052165,
+      "grad_norm": 0.39739924669265747,
+      "learning_rate": 9.32365237435808e-06,
+      "loss": 0.4643,
+      "step": 904
+    },
+    {
+      "epoch": 0.7533296337402886,
+      "grad_norm": 0.42331647872924805,
+      "learning_rate": 9.321216937015446e-06,
+      "loss": 0.4584,
+      "step": 905
+    },
+    {
+      "epoch": 0.7541620421753608,
+      "grad_norm": 0.3581715524196625,
+      "learning_rate": 9.318777441974446e-06,
+      "loss": 0.467,
+      "step": 906
+    },
+    {
+      "epoch": 0.7549944506104328,
+      "grad_norm": 0.40118566155433655,
+      "learning_rate": 9.316333891525809e-06,
+      "loss": 0.443,
+      "step": 907
+    },
+    {
+      "epoch": 0.755826859045505,
+      "grad_norm": 0.4858897626399994,
+      "learning_rate": 9.313886287964072e-06,
+      "loss": 0.4666,
+      "step": 908
+    },
+    {
+      "epoch": 0.7566592674805771,
+      "grad_norm": 0.43074485659599304,
+      "learning_rate": 9.311434633587577e-06,
+      "loss": 0.4605,
+      "step": 909
+    },
+    {
+      "epoch": 0.7574916759156493,
+      "grad_norm": 0.4585159718990326,
+      "learning_rate": 9.308978930698472e-06,
+      "loss": 0.4605,
+      "step": 910
+    },
+    {
+      "epoch": 0.7583240843507214,
+      "grad_norm": 0.433880478143692,
+      "learning_rate": 9.306519181602704e-06,
+      "loss": 0.4644,
+      "step": 911
+    },
+    {
+      "epoch": 0.7591564927857936,
+      "grad_norm": 0.38200658559799194,
+      "learning_rate": 9.304055388610019e-06,
+      "loss": 0.4427,
+      "step": 912
+    },
+    {
+      "epoch": 0.7599889012208657,
+      "grad_norm": 0.49386703968048096,
+      "learning_rate": 9.301587554033965e-06,
+      "loss": 0.4637,
+      "step": 913
+    },
+    {
+      "epoch": 0.7608213096559379,
+      "grad_norm": 0.4780905544757843,
+      "learning_rate": 9.299115680191876e-06,
+      "loss": 0.4648,
+      "step": 914
+    },
+    {
+      "epoch": 0.76165371809101,
+      "grad_norm": 0.4210283160209656,
+      "learning_rate": 9.296639769404892e-06,
+      "loss": 0.4691,
+      "step": 915
+    },
+    {
+      "epoch": 0.7624861265260822,
+      "grad_norm": 0.4278091788291931,
+      "learning_rate": 9.294159823997933e-06,
+      "loss": 0.4551,
+      "step": 916
+    },
+    {
+      "epoch": 0.7633185349611543,
+      "grad_norm": 0.39724352955818176,
+      "learning_rate": 9.291675846299711e-06,
+      "loss": 0.4963,
+      "step": 917
+    },
+    {
+      "epoch": 0.7641509433962265,
+      "grad_norm": 0.4462328851222992,
+      "learning_rate": 9.289187838642724e-06,
+      "loss": 0.4781,
+      "step": 918
+    },
+    {
+      "epoch": 0.7649833518312985,
+      "grad_norm": 0.3846145570278168,
+      "learning_rate": 9.286695803363257e-06,
+      "loss": 0.442,
+      "step": 919
+    },
+    {
+      "epoch": 0.7658157602663707,
+      "grad_norm": 0.3947022557258606,
+      "learning_rate": 9.284199742801373e-06,
+      "loss": 0.4804,
+      "step": 920
+    },
+    {
+      "epoch": 0.7666481687014428,
+      "grad_norm": 0.4000520408153534,
+      "learning_rate": 9.281699659300917e-06,
+      "loss": 0.5051,
+      "step": 921
+    },
+    {
+      "epoch": 0.767480577136515,
+      "grad_norm": 0.4070630669593811,
+      "learning_rate": 9.279195555209513e-06,
+      "loss": 0.4547,
+      "step": 922
+    },
+    {
+      "epoch": 0.7683129855715871,
+      "grad_norm": 0.4421361982822418,
+      "learning_rate": 9.276687432878554e-06,
+      "loss": 0.4619,
+      "step": 923
+    },
+    {
+      "epoch": 0.7691453940066593,
+      "grad_norm": 0.404674232006073,
+      "learning_rate": 9.274175294663215e-06,
+      "loss": 0.462,
+      "step": 924
+    },
+    {
+      "epoch": 0.7699778024417314,
+      "grad_norm": 0.4667954444885254,
+      "learning_rate": 9.271659142922438e-06,
+      "loss": 0.4739,
+      "step": 925
+    },
+    {
+      "epoch": 0.7708102108768036,
+      "grad_norm": 0.3848412036895752,
+      "learning_rate": 9.26913898001893e-06,
+      "loss": 0.489,
+      "step": 926
+    },
+    {
+      "epoch": 0.7716426193118757,
+      "grad_norm": 0.4155692458152771,
+      "learning_rate": 9.26661480831917e-06,
+      "loss": 0.4522,
+      "step": 927
+    },
+    {
+      "epoch": 0.7724750277469479,
+      "grad_norm": 0.4222828447818756,
+      "learning_rate": 9.2640866301934e-06,
+      "loss": 0.4756,
+      "step": 928
+    },
+    {
+      "epoch": 0.77330743618202,
+      "grad_norm": 0.3686828911304474,
+      "learning_rate": 9.261554448015625e-06,
+      "loss": 0.4513,
+      "step": 929
+    },
+    {
+      "epoch": 0.7741398446170921,
+      "grad_norm": 0.4023358225822449,
+      "learning_rate": 9.259018264163604e-06,
+      "loss": 0.4447,
+      "step": 930
+    },
+    {
+      "epoch": 0.7749722530521642,
+      "grad_norm": 0.46268701553344727,
+      "learning_rate": 9.25647808101886e-06,
+      "loss": 0.4696,
+      "step": 931
+    },
+    {
+      "epoch": 0.7758046614872364,
+      "grad_norm": 0.39935389161109924,
+      "learning_rate": 9.253933900966672e-06,
+      "loss": 0.4549,
+      "step": 932
+    },
+    {
+      "epoch": 0.7766370699223085,
+      "grad_norm": 0.41673651337623596,
+      "learning_rate": 9.251385726396065e-06,
+      "loss": 0.4756,
+      "step": 933
+    },
+    {
+      "epoch": 0.7774694783573807,
+      "grad_norm": 0.47298237681388855,
+      "learning_rate": 9.248833559699824e-06,
+      "loss": 0.4617,
+      "step": 934
+    },
+    {
+      "epoch": 0.7783018867924528,
+      "grad_norm": 0.3737104833126068,
+      "learning_rate": 9.246277403274475e-06,
+      "loss": 0.437,
+      "step": 935
+    },
+    {
+      "epoch": 0.779134295227525,
+      "grad_norm": 0.47945085167884827,
+      "learning_rate": 9.243717259520296e-06,
+      "loss": 0.4657,
+      "step": 936
+    },
+    {
+      "epoch": 0.7799667036625971,
+      "grad_norm": 0.3635734021663666,
+      "learning_rate": 9.241153130841305e-06,
+      "loss": 0.4205,
+      "step": 937
+    },
+    {
+      "epoch": 0.7807991120976693,
+      "grad_norm": 0.4091576933860779,
+      "learning_rate": 9.238585019645265e-06,
+      "loss": 0.4579,
+      "step": 938
+    },
+    {
+      "epoch": 0.7816315205327414,
+      "grad_norm": 0.4312977194786072,
+      "learning_rate": 9.236012928343676e-06,
+      "loss": 0.4557,
+      "step": 939
+    },
+    {
+      "epoch": 0.7824639289678136,
+      "grad_norm": 0.44428691267967224,
+      "learning_rate": 9.233436859351778e-06,
+      "loss": 0.4538,
+      "step": 940
+    },
+    {
+      "epoch": 0.7832963374028857,
+      "grad_norm": 0.43025487661361694,
+      "learning_rate": 9.230856815088546e-06,
+      "loss": 0.4668,
+      "step": 941
+    },
+    {
+      "epoch": 0.7841287458379578,
+      "grad_norm": 0.42932865023612976,
+      "learning_rate": 9.228272797976685e-06,
+      "loss": 0.4588,
+      "step": 942
+    },
+    {
+      "epoch": 0.7849611542730299,
+      "grad_norm": 0.45379704236984253,
+      "learning_rate": 9.22568481044263e-06,
+      "loss": 0.4248,
+      "step": 943
+    },
+    {
+      "epoch": 0.7857935627081021,
+      "grad_norm": 0.5083206295967102,
+      "learning_rate": 9.223092854916552e-06,
+      "loss": 0.4797,
+      "step": 944
+    },
+    {
+      "epoch": 0.7866259711431742,
+      "grad_norm": 0.47013911604881287,
+      "learning_rate": 9.220496933832338e-06,
+      "loss": 0.4839,
+      "step": 945
+    },
+    {
+      "epoch": 0.7874583795782464,
+      "grad_norm": 0.38621267676353455,
+      "learning_rate": 9.217897049627605e-06,
+      "loss": 0.4352,
+      "step": 946
+    },
+    {
+      "epoch": 0.7882907880133185,
+      "grad_norm": 0.48977726697921753,
+      "learning_rate": 9.21529320474369e-06,
+      "loss": 0.4596,
+      "step": 947
+    },
+    {
+      "epoch": 0.7891231964483907,
+      "grad_norm": 0.4230138063430786,
+      "learning_rate": 9.212685401625649e-06,
+      "loss": 0.4623,
+      "step": 948
+    },
+    {
+      "epoch": 0.7899556048834628,
+      "grad_norm": 0.41531112790107727,
+      "learning_rate": 9.210073642722256e-06,
+      "loss": 0.4596,
+      "step": 949
+    },
+    {
+      "epoch": 0.790788013318535,
+      "grad_norm": 0.4705289602279663,
+      "learning_rate": 9.207457930485996e-06,
+      "loss": 0.4578,
+      "step": 950
+    },
+    {
+      "epoch": 0.7916204217536071,
+      "grad_norm": 0.4125988781452179,
+      "learning_rate": 9.20483826737307e-06,
+      "loss": 0.4498,
+      "step": 951
+    },
+    {
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.42193499207496643,
+      "learning_rate": 9.202214655843386e-06,
+      "loss": 0.447,
+      "step": 952
+    },
+    {
+      "epoch": 0.7932852386237513,
+      "grad_norm": 0.3790290057659149,
+      "learning_rate": 9.199587098360563e-06,
+      "loss": 0.4602,
+      "step": 953
+    },
+    {
+      "epoch": 0.7941176470588235,
+      "grad_norm": 0.37495550513267517,
+      "learning_rate": 9.196955597391923e-06,
+      "loss": 0.4458,
+      "step": 954
+    },
+    {
+      "epoch": 0.7949500554938956,
+      "grad_norm": 0.4026005268096924,
+      "learning_rate": 9.19432015540849e-06,
+      "loss": 0.4445,
+      "step": 955
+    },
+    {
+      "epoch": 0.7957824639289678,
+      "grad_norm": 0.40387800335884094,
+      "learning_rate": 9.191680774884992e-06,
+      "loss": 0.4688,
+      "step": 956
+    },
+    {
+      "epoch": 0.7966148723640399,
+      "grad_norm": 0.3699960708618164,
+      "learning_rate": 9.189037458299854e-06,
+      "loss": 0.4725,
+      "step": 957
+    },
+    {
+      "epoch": 0.7974472807991121,
+      "grad_norm": 0.41380584239959717,
+      "learning_rate": 9.186390208135194e-06,
+      "loss": 0.4589,
+      "step": 958
+    },
+    {
+      "epoch": 0.7982796892341842,
+      "grad_norm": 0.41292324662208557,
+      "learning_rate": 9.18373902687683e-06,
+      "loss": 0.4501,
+      "step": 959
+    },
+    {
+      "epoch": 0.7991120976692564,
+      "grad_norm": 0.3478350341320038,
+      "learning_rate": 9.181083917014262e-06,
+      "loss": 0.4391,
+      "step": 960
+    },
+    {
+      "epoch": 0.7999445061043285,
+      "grad_norm": 0.468945175409317,
+      "learning_rate": 9.17842488104069e-06,
+      "loss": 0.475,
+      "step": 961
+    },
+    {
+      "epoch": 0.8007769145394007,
+      "grad_norm": 0.40263715386390686,
+      "learning_rate": 9.175761921452992e-06,
+      "loss": 0.4416,
+      "step": 962
+    },
+    {
+      "epoch": 0.8016093229744728,
+      "grad_norm": 0.4122201204299927,
+      "learning_rate": 9.173095040751738e-06,
+      "loss": 0.4474,
+      "step": 963
+    },
+    {
+      "epoch": 0.802441731409545,
+      "grad_norm": 0.3991578221321106,
+      "learning_rate": 9.17042424144117e-06,
+      "loss": 0.4571,
+      "step": 964
+    },
+    {
+      "epoch": 0.803274139844617,
+      "grad_norm": 0.39560526609420776,
+      "learning_rate": 9.16774952602922e-06,
+      "loss": 0.4917,
+      "step": 965
+    },
+    {
+      "epoch": 0.8041065482796892,
+      "grad_norm": 0.40626710653305054,
+      "learning_rate": 9.165070897027487e-06,
+      "loss": 0.4676,
+      "step": 966
+    },
+    {
+      "epoch": 0.8049389567147613,
+      "grad_norm": 0.4202491343021393,
+      "learning_rate": 9.162388356951257e-06,
+      "loss": 0.454,
+      "step": 967
+    },
+    {
+      "epoch": 0.8057713651498335,
+      "grad_norm": 0.35746338963508606,
+      "learning_rate": 9.15970190831948e-06,
+      "loss": 0.4639,
+      "step": 968
+    },
+    {
+      "epoch": 0.8066037735849056,
+      "grad_norm": 0.3897421956062317,
+      "learning_rate": 9.157011553654776e-06,
+      "loss": 0.4548,
+      "step": 969
+    },
+    {
+      "epoch": 0.8074361820199778,
+      "grad_norm": 0.3858961760997772,
+      "learning_rate": 9.154317295483437e-06,
+      "loss": 0.4629,
+      "step": 970
+    },
+    {
+      "epoch": 0.80826859045505,
+      "grad_norm": 0.4071405827999115,
+      "learning_rate": 9.151619136335419e-06,
+      "loss": 0.4685,
+      "step": 971
+    },
+    {
+      "epoch": 0.8091009988901221,
+      "grad_norm": 0.41344863176345825,
+      "learning_rate": 9.14891707874434e-06,
+      "loss": 0.4563,
+      "step": 972
+    },
+    {
+      "epoch": 0.8099334073251943,
+      "grad_norm": 0.41553568840026855,
+      "learning_rate": 9.146211125247478e-06,
+      "loss": 0.4347,
+      "step": 973
+    },
+    {
+      "epoch": 0.8107658157602664,
+      "grad_norm": 0.44281312823295593,
+      "learning_rate": 9.143501278385773e-06,
+      "loss": 0.4563,
+      "step": 974
+    },
+    {
+      "epoch": 0.8115982241953386,
+      "grad_norm": 0.45483851432800293,
+      "learning_rate": 9.140787540703817e-06,
+      "loss": 0.4658,
+      "step": 975
+    },
+    {
+      "epoch": 0.8124306326304107,
+      "grad_norm": 0.382942795753479,
+      "learning_rate": 9.138069914749859e-06,
+      "loss": 0.4486,
+      "step": 976
+    },
+    {
+      "epoch": 0.8132630410654828,
+      "grad_norm": 0.42402827739715576,
+      "learning_rate": 9.135348403075795e-06,
+      "loss": 0.4853,
+      "step": 977
+    },
+    {
+      "epoch": 0.8140954495005549,
+      "grad_norm": 0.3964520990848541,
+      "learning_rate": 9.132623008237174e-06,
+      "loss": 0.4955,
+      "step": 978
+    },
+    {
+      "epoch": 0.814927857935627,
+      "grad_norm": 0.5112728476524353,
+      "learning_rate": 9.12989373279319e-06,
+      "loss": 0.4493,
+      "step": 979
+    },
+    {
+      "epoch": 0.8157602663706992,
+      "grad_norm": 0.4075373113155365,
+      "learning_rate": 9.127160579306678e-06,
+      "loss": 0.4551,
+      "step": 980
+    },
+    {
+      "epoch": 0.8165926748057714,
+      "grad_norm": 0.4562476873397827,
+      "learning_rate": 9.124423550344118e-06,
+      "loss": 0.4479,
+      "step": 981
+    },
+    {
+      "epoch": 0.8174250832408435,
+      "grad_norm": 0.4151056706905365,
+      "learning_rate": 9.12168264847563e-06,
+      "loss": 0.4721,
+      "step": 982
+    },
+    {
+      "epoch": 0.8182574916759157,
+      "grad_norm": 0.3973326086997986,
+      "learning_rate": 9.118937876274965e-06,
+      "loss": 0.4544,
+      "step": 983
+    },
+    {
+      "epoch": 0.8190899001109878,
+      "grad_norm": 0.4236065745353699,
+      "learning_rate": 9.116189236319515e-06,
+      "loss": 0.4596,
+      "step": 984
+    },
+    {
+      "epoch": 0.81992230854606,
+      "grad_norm": 0.3826847970485687,
+      "learning_rate": 9.113436731190302e-06,
+      "loss": 0.4686,
+      "step": 985
+    },
+    {
+      "epoch": 0.8207547169811321,
+      "grad_norm": 0.34965115785598755,
+      "learning_rate": 9.110680363471973e-06,
+      "loss": 0.4528,
+      "step": 986
+    },
+    {
+      "epoch": 0.8215871254162043,
+      "grad_norm": 0.40207812190055847,
+      "learning_rate": 9.10792013575281e-06,
+      "loss": 0.4559,
+      "step": 987
+    },
+    {
+      "epoch": 0.8224195338512763,
+      "grad_norm": 0.4016001522541046,
+      "learning_rate": 9.10515605062471e-06,
+      "loss": 0.4553,
+      "step": 988
+    },
+    {
+      "epoch": 0.8232519422863485,
+      "grad_norm": 0.44705718755722046,
+      "learning_rate": 9.102388110683201e-06,
+      "loss": 0.4915,
+      "step": 989
+    },
+    {
+      "epoch": 0.8240843507214206,
+      "grad_norm": 0.43472737073898315,
+      "learning_rate": 9.099616318527426e-06,
+      "loss": 0.4828,
+      "step": 990
+    },
+    {
+      "epoch": 0.8249167591564928,
+      "grad_norm": 0.42259252071380615,
+      "learning_rate": 9.096840676760146e-06,
+      "loss": 0.4427,
+      "step": 991
+    },
+    {
+      "epoch": 0.8257491675915649,
+      "grad_norm": 0.44350579380989075,
+      "learning_rate": 9.09406118798774e-06,
+      "loss": 0.4859,
+      "step": 992
+    },
+    {
+      "epoch": 0.8265815760266371,
+      "grad_norm": 0.39259374141693115,
+      "learning_rate": 9.091277854820191e-06,
+      "loss": 0.4398,
+      "step": 993
+    },
+    {
+      "epoch": 0.8274139844617092,
+      "grad_norm": 0.3800427317619324,
+      "learning_rate": 9.088490679871102e-06,
+      "loss": 0.4363,
+      "step": 994
+    },
+    {
+      "epoch": 0.8282463928967814,
+      "grad_norm": 0.4360864758491516,
+      "learning_rate": 9.085699665757679e-06,
+      "loss": 0.4651,
+      "step": 995
+    },
+    {
+      "epoch": 0.8290788013318535,
+      "grad_norm": 0.39429229497909546,
+      "learning_rate": 9.082904815100732e-06,
+      "loss": 0.4669,
+      "step": 996
+    },
+    {
+      "epoch": 0.8299112097669257,
+      "grad_norm": 0.4399055242538452,
+      "learning_rate": 9.080106130524675e-06,
+      "loss": 0.4396,
+      "step": 997
+    },
+    {
+      "epoch": 0.8307436182019978,
+      "grad_norm": 0.4866039454936981,
+      "learning_rate": 9.07730361465752e-06,
+      "loss": 0.4543,
+      "step": 998
+    },
+    {
+      "epoch": 0.83157602663707,
+      "grad_norm": 0.4812076687812805,
+      "learning_rate": 9.07449727013088e-06,
+      "loss": 0.4731,
+      "step": 999
+    },
+    {
+      "epoch": 0.832408435072142,
+      "grad_norm": 0.43150269985198975,
+      "learning_rate": 9.071687099579962e-06,
+      "loss": 0.4774,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8332408435072142,
+      "grad_norm": 0.3944185674190521,
+      "learning_rate": 9.068873105643565e-06,
+      "loss": 0.4399,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8340732519422863,
+      "grad_norm": 0.5351603031158447,
+      "learning_rate": 9.066055290964079e-06,
+      "loss": 0.4518,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8349056603773585,
+      "grad_norm": 0.38343068957328796,
+      "learning_rate": 9.063233658187482e-06,
+      "loss": 0.4843,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8357380688124306,
+      "grad_norm": 0.4362078607082367,
+      "learning_rate": 9.060408209963334e-06,
+      "loss": 0.4642,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8365704772475028,
+      "grad_norm": 0.4815027117729187,
+      "learning_rate": 9.057578948944783e-06,
+      "loss": 0.4497,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8374028856825749,
+      "grad_norm": 0.4579710066318512,
+      "learning_rate": 9.054745877788554e-06,
+      "loss": 0.4475,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8382352941176471,
+      "grad_norm": 0.49727219343185425,
+      "learning_rate": 9.051908999154948e-06,
+      "loss": 0.4707,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8390677025527192,
+      "grad_norm": 0.43605858087539673,
+      "learning_rate": 9.049068315707847e-06,
+      "loss": 0.4365,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8399001109877914,
+      "grad_norm": 0.5100065469741821,
+      "learning_rate": 9.0462238301147e-06,
+      "loss": 0.4332,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8407325194228635,
+      "grad_norm": 0.38812533020973206,
+      "learning_rate": 9.04337554504653e-06,
+      "loss": 0.4384,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8415649278579356,
+      "grad_norm": 0.384883314371109,
+      "learning_rate": 9.040523463177928e-06,
+      "loss": 0.4663,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8423973362930077,
+      "grad_norm": 0.44887587428092957,
+      "learning_rate": 9.037667587187045e-06,
+      "loss": 0.486,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8432297447280799,
+      "grad_norm": 0.46977436542510986,
+      "learning_rate": 9.034807919755602e-06,
+      "loss": 0.4575,
+      "step": 1013
+    },
+    {
+      "epoch": 0.844062153163152,
+      "grad_norm": 0.3983987867832184,
+      "learning_rate": 9.031944463568877e-06,
+      "loss": 0.4532,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8448945615982242,
+      "grad_norm": 0.4556075632572174,
+      "learning_rate": 9.029077221315703e-06,
+      "loss": 0.4685,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8457269700332963,
+      "grad_norm": 0.38745394349098206,
+      "learning_rate": 9.026206195688472e-06,
+      "loss": 0.4608,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8465593784683685,
+      "grad_norm": 0.4005940556526184,
+      "learning_rate": 9.023331389383126e-06,
+      "loss": 0.4628,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8473917869034406,
+      "grad_norm": 0.39503729343414307,
+      "learning_rate": 9.02045280509916e-06,
+      "loss": 0.4688,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8482241953385128,
+      "grad_norm": 0.4228137731552124,
+      "learning_rate": 9.017570445539616e-06,
+      "loss": 0.4594,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8490566037735849,
+      "grad_norm": 0.36277511715888977,
+      "learning_rate": 9.014684313411077e-06,
+      "loss": 0.4274,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8498890122086571,
+      "grad_norm": 0.3664073050022125,
+      "learning_rate": 9.011794411423675e-06,
+      "loss": 0.4592,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8507214206437292,
+      "grad_norm": 0.3959209620952606,
+      "learning_rate": 9.008900742291075e-06,
+      "loss": 0.4583,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8515538290788013,
+      "grad_norm": 0.3981209099292755,
+      "learning_rate": 9.006003308730487e-06,
+      "loss": 0.4543,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8523862375138734,
+      "grad_norm": 0.33361658453941345,
+      "learning_rate": 9.003102113462647e-06,
+      "loss": 0.4526,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8532186459489456,
+      "grad_norm": 0.37563925981521606,
+      "learning_rate": 9.000197159211834e-06,
+      "loss": 0.4657,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8540510543840177,
+      "grad_norm": 0.36745545268058777,
+      "learning_rate": 8.997288448705846e-06,
+      "loss": 0.4451,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8548834628190899,
+      "grad_norm": 0.3501058518886566,
+      "learning_rate": 8.994375984676014e-06,
+      "loss": 0.4413,
+      "step": 1027
+    },
+    {
+      "epoch": 0.855715871254162,
+      "grad_norm": 0.38974809646606445,
+      "learning_rate": 8.991459769857195e-06,
+      "loss": 0.4492,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8565482796892342,
+      "grad_norm": 0.3752514719963074,
+      "learning_rate": 8.988539806987764e-06,
+      "loss": 0.4533,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8573806881243063,
+      "grad_norm": 0.3805655241012573,
+      "learning_rate": 8.985616098809618e-06,
+      "loss": 0.4733,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8582130965593785,
+      "grad_norm": 0.46047642827033997,
+      "learning_rate": 8.982688648068169e-06,
+      "loss": 0.4593,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8590455049944506,
+      "grad_norm": 0.39984360337257385,
+      "learning_rate": 8.979757457512347e-06,
+      "loss": 0.4749,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8598779134295228,
+      "grad_norm": 0.3868159353733063,
+      "learning_rate": 8.976822529894588e-06,
+      "loss": 0.4577,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8607103218645948,
+      "grad_norm": 0.41874581575393677,
+      "learning_rate": 8.973883867970844e-06,
+      "loss": 0.4784,
+      "step": 1034
+    },
+    {
+      "epoch": 0.861542730299667,
+      "grad_norm": 0.3759986162185669,
+      "learning_rate": 8.970941474500565e-06,
+      "loss": 0.4554,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8623751387347391,
+      "grad_norm": 0.4303062856197357,
+      "learning_rate": 8.967995352246714e-06,
+      "loss": 0.4724,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8632075471698113,
+      "grad_norm": 0.420803040266037,
+      "learning_rate": 8.965045503975752e-06,
+      "loss": 0.486,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8640399556048834,
+      "grad_norm": 0.41059285402297974,
+      "learning_rate": 8.962091932457635e-06,
+      "loss": 0.4549,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8648723640399556,
+      "grad_norm": 0.3770391047000885,
+      "learning_rate": 8.959134640465821e-06,
+      "loss": 0.4783,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8657047724750278,
+      "grad_norm": 0.5017075538635254,
+      "learning_rate": 8.956173630777255e-06,
+      "loss": 0.4476,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8665371809100999,
+      "grad_norm": 0.33851176500320435,
+      "learning_rate": 8.953208906172384e-06,
+      "loss": 0.4677,
+      "step": 1041
+    },
+    {
+      "epoch": 0.867369589345172,
+      "grad_norm": 0.40573734045028687,
+      "learning_rate": 8.95024046943513e-06,
+      "loss": 0.4607,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8682019977802442,
+      "grad_norm": 0.40829798579216003,
+      "learning_rate": 8.947268323352909e-06,
+      "loss": 0.4613,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8690344062153164,
+      "grad_norm": 0.40306246280670166,
+      "learning_rate": 8.944292470716617e-06,
+      "loss": 0.487,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8698668146503885,
+      "grad_norm": 0.4743926227092743,
+      "learning_rate": 8.941312914320636e-06,
+      "loss": 0.4623,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8706992230854605,
+      "grad_norm": 0.388696551322937,
+      "learning_rate": 8.938329656962818e-06,
+      "loss": 0.4678,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8715316315205327,
+      "grad_norm": 0.3969860076904297,
+      "learning_rate": 8.935342701444495e-06,
+      "loss": 0.437,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8723640399556049,
+      "grad_norm": 0.3880145251750946,
+      "learning_rate": 8.932352050570467e-06,
+      "loss": 0.4702,
+      "step": 1048
+    },
+    {
+      "epoch": 0.873196448390677,
+      "grad_norm": 0.41687145829200745,
+      "learning_rate": 8.929357707149014e-06,
+      "loss": 0.4662,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8740288568257492,
+      "grad_norm": 0.3698790967464447,
+      "learning_rate": 8.926359673991874e-06,
+      "loss": 0.4579,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8748612652608213,
+      "grad_norm": 0.3788878917694092,
+      "learning_rate": 8.92335795391425e-06,
+      "loss": 0.4726,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8756936736958935,
+      "grad_norm": 0.43458738923072815,
+      "learning_rate": 8.920352549734812e-06,
+      "loss": 0.4771,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8765260821309656,
+      "grad_norm": 0.3463033139705658,
+      "learning_rate": 8.91734346427569e-06,
+      "loss": 0.4503,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8773584905660378,
+      "grad_norm": 0.3873491585254669,
+      "learning_rate": 8.914330700362461e-06,
+      "loss": 0.4513,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8781908990011099,
+      "grad_norm": 0.36356833577156067,
+      "learning_rate": 8.91131426082417e-06,
+      "loss": 0.4362,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8790233074361821,
+      "grad_norm": 0.3933038115501404,
+      "learning_rate": 8.908294148493303e-06,
+      "loss": 0.4602,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8798557158712541,
+      "grad_norm": 0.3321017026901245,
+      "learning_rate": 8.905270366205798e-06,
+      "loss": 0.4457,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8806881243063263,
+      "grad_norm": 0.42186474800109863,
+      "learning_rate": 8.902242916801043e-06,
+      "loss": 0.4743,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8815205327413984,
+      "grad_norm": 0.3869630694389343,
+      "learning_rate": 8.899211803121861e-06,
+      "loss": 0.432,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 0.3925645351409912,
+      "learning_rate": 8.896177028014524e-06,
+      "loss": 0.481,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8831853496115427,
+      "grad_norm": 0.35652878880500793,
+      "learning_rate": 8.893138594328738e-06,
+      "loss": 0.4576,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8840177580466149,
+      "grad_norm": 0.3799140453338623,
+      "learning_rate": 8.890096504917647e-06,
+      "loss": 0.4318,
+      "step": 1062
+    },
+    {
+      "epoch": 0.884850166481687,
+      "grad_norm": 0.43312981724739075,
+      "learning_rate": 8.887050762637825e-06,
+      "loss": 0.4647,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8856825749167592,
+      "grad_norm": 0.3973630666732788,
+      "learning_rate": 8.884001370349275e-06,
+      "loss": 0.4598,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8865149833518313,
+      "grad_norm": 0.3882988393306732,
+      "learning_rate": 8.880948330915435e-06,
+      "loss": 0.4827,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8873473917869035,
+      "grad_norm": 0.3990275263786316,
+      "learning_rate": 8.877891647203157e-06,
+      "loss": 0.4571,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8881798002219756,
+      "grad_norm": 0.38141873478889465,
+      "learning_rate": 8.874831322082725e-06,
+      "loss": 0.4471,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8890122086570478,
+      "grad_norm": 0.3606696128845215,
+      "learning_rate": 8.871767358427835e-06,
+      "loss": 0.4216,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8898446170921198,
+      "grad_norm": 0.41676607728004456,
+      "learning_rate": 8.868699759115604e-06,
+      "loss": 0.4574,
+      "step": 1069
+    },
+    {
+      "epoch": 0.890677025527192,
+      "grad_norm": 0.38336870074272156,
+      "learning_rate": 8.86562852702656e-06,
+      "loss": 0.4666,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8915094339622641,
+      "grad_norm": 0.41286566853523254,
+      "learning_rate": 8.862553665044644e-06,
+      "loss": 0.4788,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8923418423973363,
+      "grad_norm": 0.42297109961509705,
+      "learning_rate": 8.859475176057208e-06,
+      "loss": 0.4831,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8931742508324084,
+      "grad_norm": 0.41496801376342773,
+      "learning_rate": 8.856393062955003e-06,
+      "loss": 0.4696,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8940066592674806,
+      "grad_norm": 0.4308110177516937,
+      "learning_rate": 8.85330732863219e-06,
+      "loss": 0.452,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8948390677025527,
+      "grad_norm": 0.4222694933414459,
+      "learning_rate": 8.850217975986326e-06,
+      "loss": 0.4499,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8956714761376249,
+      "grad_norm": 0.43070679903030396,
+      "learning_rate": 8.84712500791837e-06,
+      "loss": 0.4717,
+      "step": 1076
+    },
+    {
+      "epoch": 0.896503884572697,
+      "grad_norm": 0.3741142153739929,
+      "learning_rate": 8.844028427332667e-06,
+      "loss": 0.4676,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8973362930077692,
+      "grad_norm": 0.4449978470802307,
+      "learning_rate": 8.840928237136967e-06,
+      "loss": 0.4547,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8981687014428413,
+      "grad_norm": 0.36339613795280457,
+      "learning_rate": 8.837824440242402e-06,
+      "loss": 0.4672,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8990011098779135,
+      "grad_norm": 0.385516494512558,
+      "learning_rate": 8.834717039563488e-06,
+      "loss": 0.4796,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8998335183129855,
+      "grad_norm": 0.3869156539440155,
+      "learning_rate": 8.83160603801813e-06,
+      "loss": 0.4498,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9006659267480577,
+      "grad_norm": 0.3644091784954071,
+      "learning_rate": 8.828491438527614e-06,
+      "loss": 0.4296,
+      "step": 1082
+    },
+    {
+      "epoch": 0.9014983351831298,
+      "grad_norm": 0.3746865689754486,
+      "learning_rate": 8.825373244016604e-06,
+      "loss": 0.4595,
+      "step": 1083
+    },
+    {
+      "epoch": 0.902330743618202,
+      "grad_norm": 0.36787107586860657,
+      "learning_rate": 8.822251457413138e-06,
+      "loss": 0.4635,
+      "step": 1084
+    },
+    {
+      "epoch": 0.9031631520532741,
+      "grad_norm": 0.3880446255207062,
+      "learning_rate": 8.819126081648627e-06,
+      "loss": 0.4619,
+      "step": 1085
+    },
+    {
+      "epoch": 0.9039955604883463,
+      "grad_norm": 0.3723534047603607,
+      "learning_rate": 8.815997119657856e-06,
+      "loss": 0.4545,
+      "step": 1086
+    },
+    {
+      "epoch": 0.9048279689234184,
+      "grad_norm": 0.3879696726799011,
+      "learning_rate": 8.812864574378974e-06,
+      "loss": 0.474,
+      "step": 1087
+    },
+    {
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.36703255772590637,
+      "learning_rate": 8.809728448753496e-06,
+      "loss": 0.4609,
+      "step": 1088
+    },
+    {
+      "epoch": 0.9064927857935627,
+      "grad_norm": 0.3635331094264984,
+      "learning_rate": 8.8065887457263e-06,
+      "loss": 0.4472,
+      "step": 1089
+    },
+    {
+      "epoch": 0.9073251942286349,
+      "grad_norm": 0.3628758192062378,
+      "learning_rate": 8.803445468245618e-06,
+      "loss": 0.4376,
+      "step": 1090
+    },
+    {
+      "epoch": 0.908157602663707,
+      "grad_norm": 0.41849446296691895,
+      "learning_rate": 8.800298619263047e-06,
+      "loss": 0.4313,
+      "step": 1091
+    },
+    {
+      "epoch": 0.9089900110987791,
+      "grad_norm": 0.4343433976173401,
+      "learning_rate": 8.797148201733533e-06,
+      "loss": 0.4294,
+      "step": 1092
+    },
+    {
+      "epoch": 0.9098224195338512,
+      "grad_norm": 0.37368395924568176,
+      "learning_rate": 8.793994218615371e-06,
+      "loss": 0.4416,
+      "step": 1093
+    },
+    {
+      "epoch": 0.9106548279689234,
+      "grad_norm": 0.4110583961009979,
+      "learning_rate": 8.79083667287021e-06,
+      "loss": 0.4457,
+      "step": 1094
+    },
+    {
+      "epoch": 0.9114872364039955,
+      "grad_norm": 0.43879762291908264,
+      "learning_rate": 8.787675567463034e-06,
+      "loss": 0.4837,
+      "step": 1095
+    },
+    {
+      "epoch": 0.9123196448390677,
+      "grad_norm": 0.44292664527893066,
+      "learning_rate": 8.784510905362185e-06,
+      "loss": 0.4603,
+      "step": 1096
+    },
+    {
+      "epoch": 0.9131520532741398,
+      "grad_norm": 0.4232296943664551,
+      "learning_rate": 8.781342689539329e-06,
+      "loss": 0.4736,
+      "step": 1097
+    },
+    {
+      "epoch": 0.913984461709212,
+      "grad_norm": 0.3918737769126892,
+      "learning_rate": 8.778170922969478e-06,
+      "loss": 0.4531,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9148168701442841,
+      "grad_norm": 0.4186196029186249,
+      "learning_rate": 8.774995608630979e-06,
+      "loss": 0.417,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9156492785793563,
+      "grad_norm": 0.41225314140319824,
+      "learning_rate": 8.771816749505504e-06,
+      "loss": 0.4499,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9164816870144284,
+      "grad_norm": 0.4248989522457123,
+      "learning_rate": 8.768634348578062e-06,
+      "loss": 0.4186,
+      "step": 1101
+    },
+    {
+      "epoch": 0.9173140954495006,
+      "grad_norm": 0.4327683746814728,
+      "learning_rate": 8.765448408836978e-06,
+      "loss": 0.4625,
+      "step": 1102
+    },
+    {
+      "epoch": 0.9181465038845728,
+      "grad_norm": 0.4192464053630829,
+      "learning_rate": 8.762258933273908e-06,
+      "loss": 0.4337,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9189789123196448,
+      "grad_norm": 0.4388251304626465,
+      "learning_rate": 8.759065924883827e-06,
+      "loss": 0.4489,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9198113207547169,
+      "grad_norm": 0.4132365882396698,
+      "learning_rate": 8.755869386665022e-06,
+      "loss": 0.4482,
+      "step": 1105
+    },
+    {
+      "epoch": 0.9206437291897891,
+      "grad_norm": 0.4240826964378357,
+      "learning_rate": 8.7526693216191e-06,
+      "loss": 0.4612,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9214761376248612,
+      "grad_norm": 0.4309800863265991,
+      "learning_rate": 8.749465732750982e-06,
+      "loss": 0.4827,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9223085460599334,
+      "grad_norm": 0.43188637495040894,
+      "learning_rate": 8.746258623068886e-06,
+      "loss": 0.4666,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9231409544950056,
+      "grad_norm": 0.4251968264579773,
+      "learning_rate": 8.74304799558435e-06,
+      "loss": 0.4654,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9239733629300777,
+      "grad_norm": 0.4286282956600189,
+      "learning_rate": 8.739833853312208e-06,
+      "loss": 0.4504,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9248057713651499,
+      "grad_norm": 0.39924654364585876,
+      "learning_rate": 8.736616199270595e-06,
+      "loss": 0.4432,
+      "step": 1111
+    },
+    {
+      "epoch": 0.925638179800222,
+      "grad_norm": 0.44831129908561707,
+      "learning_rate": 8.733395036480946e-06,
+      "loss": 0.4497,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9264705882352942,
+      "grad_norm": 0.38468286395072937,
+      "learning_rate": 8.73017036796799e-06,
+      "loss": 0.4564,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9273029966703663,
+      "grad_norm": 0.380112886428833,
+      "learning_rate": 8.726942196759744e-06,
+      "loss": 0.4557,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9281354051054383,
+      "grad_norm": 0.3891477584838867,
+      "learning_rate": 8.72371052588752e-06,
+      "loss": 0.4478,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9289678135405105,
+      "grad_norm": 0.41235268115997314,
+      "learning_rate": 8.720475358385912e-06,
+      "loss": 0.4429,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9298002219755827,
+      "grad_norm": 0.3957020938396454,
+      "learning_rate": 8.7172366972928e-06,
+      "loss": 0.4848,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9306326304106548,
+      "grad_norm": 0.42292919754981995,
+      "learning_rate": 8.713994545649343e-06,
+      "loss": 0.466,
+      "step": 1118
+    },
+    {
+      "epoch": 0.931465038845727,
+      "grad_norm": 0.40453973412513733,
+      "learning_rate": 8.710748906499977e-06,
+      "loss": 0.4424,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9322974472807991,
+      "grad_norm": 0.4147048592567444,
+      "learning_rate": 8.707499782892414e-06,
+      "loss": 0.471,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9331298557158713,
+      "grad_norm": 0.38894933462142944,
+      "learning_rate": 8.704247177877643e-06,
+      "loss": 0.4822,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9339622641509434,
+      "grad_norm": 0.39160868525505066,
+      "learning_rate": 8.700991094509909e-06,
+      "loss": 0.4814,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9347946725860156,
+      "grad_norm": 0.362416535615921,
+      "learning_rate": 8.697731535846739e-06,
+      "loss": 0.4605,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9356270810210877,
+      "grad_norm": 0.44123584032058716,
+      "learning_rate": 8.69446850494891e-06,
+      "loss": 0.4594,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9364594894561599,
+      "grad_norm": 0.3935483992099762,
+      "learning_rate": 8.691202004880468e-06,
+      "loss": 0.4643,
+      "step": 1125
+    },
+    {
+      "epoch": 0.937291897891232,
+      "grad_norm": 0.37961041927337646,
+      "learning_rate": 8.687932038708712e-06,
+      "loss": 0.4547,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9381243063263041,
+      "grad_norm": 0.4113844037055969,
+      "learning_rate": 8.684658609504199e-06,
+      "loss": 0.4683,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9389567147613762,
+      "grad_norm": 0.4119846820831299,
+      "learning_rate": 8.681381720340736e-06,
+      "loss": 0.4563,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9397891231964484,
+      "grad_norm": 0.3995191156864166,
+      "learning_rate": 8.67810137429538e-06,
+      "loss": 0.4544,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9406215316315205,
+      "grad_norm": 0.403024822473526,
+      "learning_rate": 8.674817574448431e-06,
+      "loss": 0.4732,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9414539400665927,
+      "grad_norm": 0.38387593626976013,
+      "learning_rate": 8.671530323883437e-06,
+      "loss": 0.4698,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9422863485016648,
+      "grad_norm": 0.3703984022140503,
+      "learning_rate": 8.668239625687183e-06,
+      "loss": 0.4346,
+      "step": 1132
+    },
+    {
+      "epoch": 0.943118756936737,
+      "grad_norm": 0.35403862595558167,
+      "learning_rate": 8.664945482949691e-06,
+      "loss": 0.4518,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9439511653718091,
+      "grad_norm": 0.3765757381916046,
+      "learning_rate": 8.661647898764221e-06,
+      "loss": 0.4547,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9447835738068813,
+      "grad_norm": 0.35943272709846497,
+      "learning_rate": 8.658346876227261e-06,
+      "loss": 0.4333,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9456159822419534,
+      "grad_norm": 0.36773252487182617,
+      "learning_rate": 8.655042418438529e-06,
+      "loss": 0.4498,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9464483906770256,
+      "grad_norm": 0.3745127022266388,
+      "learning_rate": 8.651734528500968e-06,
+      "loss": 0.461,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9472807991120976,
+      "grad_norm": 0.3643935024738312,
+      "learning_rate": 8.648423209520746e-06,
+      "loss": 0.4351,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9481132075471698,
+      "grad_norm": 0.4308556616306305,
+      "learning_rate": 8.64510846460725e-06,
+      "loss": 0.4828,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9489456159822419,
+      "grad_norm": 0.331564337015152,
+      "learning_rate": 8.641790296873081e-06,
+      "loss": 0.4513,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9497780244173141,
+      "grad_norm": 0.3650985658168793,
+      "learning_rate": 8.638468709434057e-06,
+      "loss": 0.4507,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9506104328523862,
+      "grad_norm": 0.4312916696071625,
+      "learning_rate": 8.63514370540921e-06,
+      "loss": 0.4469,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9514428412874584,
+      "grad_norm": 0.3546101748943329,
+      "learning_rate": 8.631815287920773e-06,
+      "loss": 0.4594,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9522752497225305,
+      "grad_norm": 0.3340010643005371,
+      "learning_rate": 8.62848346009419e-06,
+      "loss": 0.4506,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9531076581576027,
+      "grad_norm": 0.4434676468372345,
+      "learning_rate": 8.625148225058107e-06,
+      "loss": 0.4825,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9539400665926748,
+      "grad_norm": 0.39392799139022827,
+      "learning_rate": 8.621809585944366e-06,
+      "loss": 0.4532,
+      "step": 1146
+    },
+    {
+      "epoch": 0.954772475027747,
+      "grad_norm": 0.4085777699947357,
+      "learning_rate": 8.61846754588801e-06,
+      "loss": 0.4852,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9556048834628191,
+      "grad_norm": 0.42657431960105896,
+      "learning_rate": 8.61512210802727e-06,
+      "loss": 0.4647,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9564372918978913,
+      "grad_norm": 0.35785362124443054,
+      "learning_rate": 8.611773275503572e-06,
+      "loss": 0.4393,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9572697003329633,
+      "grad_norm": 0.37917736172676086,
+      "learning_rate": 8.608421051461529e-06,
+      "loss": 0.4496,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9581021087680355,
+      "grad_norm": 0.37420985102653503,
+      "learning_rate": 8.605065439048936e-06,
+      "loss": 0.4475,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9589345172031076,
+      "grad_norm": 0.38180387020111084,
+      "learning_rate": 8.601706441416776e-06,
+      "loss": 0.4694,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9597669256381798,
+      "grad_norm": 0.4098156988620758,
+      "learning_rate": 8.598344061719204e-06,
+      "loss": 0.4602,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9605993340732519,
+      "grad_norm": 0.38434961438179016,
+      "learning_rate": 8.594978303113552e-06,
+      "loss": 0.4214,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9614317425083241,
+      "grad_norm": 0.37275487184524536,
+      "learning_rate": 8.59160916876033e-06,
+      "loss": 0.4565,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9622641509433962,
+      "grad_norm": 0.4662790298461914,
+      "learning_rate": 8.588236661823209e-06,
+      "loss": 0.4807,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9630965593784684,
+      "grad_norm": 0.3571013808250427,
+      "learning_rate": 8.584860785469036e-06,
+      "loss": 0.4392,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9639289678135405,
+      "grad_norm": 0.3700477182865143,
+      "learning_rate": 8.581481542867818e-06,
+      "loss": 0.4548,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9647613762486127,
+      "grad_norm": 0.4246416687965393,
+      "learning_rate": 8.578098937192723e-06,
+      "loss": 0.4503,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9655937846836848,
+      "grad_norm": 0.4045102894306183,
+      "learning_rate": 8.574712971620075e-06,
+      "loss": 0.4749,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9664261931187569,
+      "grad_norm": 0.35602596402168274,
+      "learning_rate": 8.571323649329352e-06,
+      "loss": 0.4275,
+      "step": 1161
+    },
+    {
+      "epoch": 0.967258601553829,
+      "grad_norm": 0.49385765194892883,
+      "learning_rate": 8.567930973503196e-06,
+      "loss": 0.4623,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9680910099889012,
+      "grad_norm": 0.4097510576248169,
+      "learning_rate": 8.564534947327381e-06,
+      "loss": 0.46,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9689234184239733,
+      "grad_norm": 0.38444530963897705,
+      "learning_rate": 8.561135573990839e-06,
+      "loss": 0.441,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9697558268590455,
+      "grad_norm": 0.43890970945358276,
+      "learning_rate": 8.55773285668564e-06,
+      "loss": 0.4766,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9705882352941176,
+      "grad_norm": 0.37467607855796814,
+      "learning_rate": 8.554326798606994e-06,
+      "loss": 0.4743,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9714206437291898,
+      "grad_norm": 0.3442972004413605,
+      "learning_rate": 8.55091740295325e-06,
+      "loss": 0.4558,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9722530521642619,
+      "grad_norm": 0.3897690176963806,
+      "learning_rate": 8.547504672925892e-06,
+      "loss": 0.4647,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9730854605993341,
+      "grad_norm": 0.33819976449012756,
+      "learning_rate": 8.544088611729533e-06,
+      "loss": 0.4507,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9739178690344062,
+      "grad_norm": 0.3626435399055481,
+      "learning_rate": 8.540669222571911e-06,
+      "loss": 0.4365,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9747502774694784,
+      "grad_norm": 0.3824900686740875,
+      "learning_rate": 8.537246508663894e-06,
+      "loss": 0.4761,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9755826859045506,
+      "grad_norm": 0.3584446310997009,
+      "learning_rate": 8.533820473219472e-06,
+      "loss": 0.4595,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9764150943396226,
+      "grad_norm": 0.39771831035614014,
+      "learning_rate": 8.53039111945575e-06,
+      "loss": 0.4779,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9772475027746947,
+      "grad_norm": 0.37545955181121826,
+      "learning_rate": 8.526958450592952e-06,
+      "loss": 0.4357,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9780799112097669,
+      "grad_norm": 0.3663276731967926,
+      "learning_rate": 8.523522469854415e-06,
+      "loss": 0.4777,
+      "step": 1175
+    },
+    {
+      "epoch": 0.978912319644839,
+      "grad_norm": 0.3252856135368347,
+      "learning_rate": 8.520083180466585e-06,
+      "loss": 0.4317,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9797447280799112,
+      "grad_norm": 0.4126368463039398,
+      "learning_rate": 8.516640585659012e-06,
+      "loss": 0.4715,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9805771365149833,
+      "grad_norm": 0.35552918910980225,
+      "learning_rate": 8.513194688664356e-06,
+      "loss": 0.4821,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9814095449500555,
+      "grad_norm": 0.34943288564682007,
+      "learning_rate": 8.509745492718375e-06,
+      "loss": 0.4405,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9822419533851277,
+      "grad_norm": 0.3988153636455536,
+      "learning_rate": 8.506293001059922e-06,
+      "loss": 0.4575,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9830743618201998,
+      "grad_norm": 0.3627513647079468,
+      "learning_rate": 8.502837216930947e-06,
+      "loss": 0.4551,
+      "step": 1181
+    },
+    {
+      "epoch": 0.983906770255272,
+      "grad_norm": 0.3588287830352783,
+      "learning_rate": 8.499378143576496e-06,
+      "loss": 0.4544,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9847391786903441,
+      "grad_norm": 0.3817938566207886,
+      "learning_rate": 8.495915784244694e-06,
+      "loss": 0.458,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9855715871254163,
+      "grad_norm": 0.38958173990249634,
+      "learning_rate": 8.49245014218676e-06,
+      "loss": 0.4605,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9864039955604883,
+      "grad_norm": 0.3958968222141266,
+      "learning_rate": 8.488981220656993e-06,
+      "loss": 0.4407,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9872364039955605,
+      "grad_norm": 0.3889720141887665,
+      "learning_rate": 8.48550902291277e-06,
+      "loss": 0.4434,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9880688124306326,
+      "grad_norm": 0.3535591959953308,
+      "learning_rate": 8.482033552214546e-06,
+      "loss": 0.4687,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9889012208657048,
+      "grad_norm": 0.44984614849090576,
+      "learning_rate": 8.478554811825846e-06,
+      "loss": 0.4464,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9897336293007769,
+      "grad_norm": 0.3912472724914551,
+      "learning_rate": 8.475072805013274e-06,
+      "loss": 0.4488,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9905660377358491,
+      "grad_norm": 0.4023473858833313,
+      "learning_rate": 8.471587535046487e-06,
+      "loss": 0.4632,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9913984461709212,
+      "grad_norm": 0.3754764795303345,
+      "learning_rate": 8.468099005198224e-06,
+      "loss": 0.4345,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9922308546059934,
+      "grad_norm": 0.37986260652542114,
+      "learning_rate": 8.46460721874427e-06,
+      "loss": 0.4315,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9930632630410655,
+      "grad_norm": 0.37531065940856934,
+      "learning_rate": 8.461112178963475e-06,
+      "loss": 0.4342,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9938956714761377,
+      "grad_norm": 0.4208608865737915,
+      "learning_rate": 8.45761388913774e-06,
+      "loss": 0.4849,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9947280799112098,
+      "grad_norm": 0.3419855237007141,
+      "learning_rate": 8.454112352552025e-06,
+      "loss": 0.4543,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9955604883462819,
+      "grad_norm": 0.3961000144481659,
+      "learning_rate": 8.450607572494332e-06,
+      "loss": 0.4526,
+      "step": 1196
+    },
+    {
+      "epoch": 0.996392896781354,
+      "grad_norm": 0.42482149600982666,
+      "learning_rate": 8.447099552255708e-06,
+      "loss": 0.4498,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9972253052164262,
+      "grad_norm": 0.3759855329990387,
+      "learning_rate": 8.44358829513025e-06,
+      "loss": 0.4251,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9980577136514983,
+      "grad_norm": 0.3645038306713104,
+      "learning_rate": 8.44007380441509e-06,
+      "loss": 0.4439,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9988901220865705,
+      "grad_norm": 0.46094900369644165,
+      "learning_rate": 8.436556083410392e-06,
+      "loss": 0.4726,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9997225305216426,
+      "grad_norm": 0.42478302121162415,
+      "learning_rate": 8.433035135419358e-06,
+      "loss": 0.4445,
+      "step": 1201
+    },
+    {
+      "epoch": 1.0005549389567148,
+      "grad_norm": 0.7507914900779724,
+      "learning_rate": 8.429510963748224e-06,
+      "loss": 0.7647,
+      "step": 1202
+    },
+    {
+      "epoch": 1.001387347391787,
+      "grad_norm": 0.43571215867996216,
+      "learning_rate": 8.425983571706247e-06,
+      "loss": 0.4272,
+      "step": 1203
+    },
+    {
+      "epoch": 1.002219755826859,
+      "grad_norm": 0.4088655114173889,
+      "learning_rate": 8.422452962605709e-06,
+      "loss": 0.4526,
+      "step": 1204
+    },
+    {
+      "epoch": 1.0030521642619312,
+      "grad_norm": 0.34656473994255066,
+      "learning_rate": 8.418919139761914e-06,
+      "loss": 0.3969,
+      "step": 1205
+    },
+    {
+      "epoch": 1.0038845726970034,
+      "grad_norm": 0.4557936489582062,
+      "learning_rate": 8.415382106493183e-06,
+      "loss": 0.45,
+      "step": 1206
+    },
+    {
+      "epoch": 1.0047169811320755,
+      "grad_norm": 0.3934694528579712,
+      "learning_rate": 8.411841866120855e-06,
+      "loss": 0.4424,
+      "step": 1207
+    },
+    {
+      "epoch": 1.0055493895671477,
+      "grad_norm": 0.410007119178772,
+      "learning_rate": 8.408298421969275e-06,
+      "loss": 0.4463,
+      "step": 1208
+    },
+    {
+      "epoch": 1.0063817980022198,
+      "grad_norm": 0.3848797678947449,
+      "learning_rate": 8.4047517773658e-06,
+      "loss": 0.4238,
+      "step": 1209
+    },
+    {
+      "epoch": 1.007214206437292,
+      "grad_norm": 0.3674464821815491,
+      "learning_rate": 8.40120193564079e-06,
+      "loss": 0.3869,
+      "step": 1210
+    },
+    {
+      "epoch": 1.0080466148723641,
+      "grad_norm": 0.42686253786087036,
+      "learning_rate": 8.39764890012761e-06,
+      "loss": 0.4377,
+      "step": 1211
+    },
+    {
+      "epoch": 1.0088790233074363,
+      "grad_norm": 0.4299822151660919,
+      "learning_rate": 8.394092674162625e-06,
+      "loss": 0.4527,
+      "step": 1212
+    },
+    {
+      "epoch": 1.0097114317425082,
+      "grad_norm": 0.4144967496395111,
+      "learning_rate": 8.390533261085188e-06,
+      "loss": 0.4261,
+      "step": 1213
+    },
+    {
+      "epoch": 1.0105438401775804,
+      "grad_norm": 0.3909236192703247,
+      "learning_rate": 8.386970664237653e-06,
+      "loss": 0.4208,
+      "step": 1214
+    },
+    {
+      "epoch": 1.0113762486126525,
+      "grad_norm": 0.404340535402298,
+      "learning_rate": 8.383404886965361e-06,
+      "loss": 0.4307,
+      "step": 1215
+    },
+    {
+      "epoch": 1.0122086570477247,
+      "grad_norm": 0.39880749583244324,
+      "learning_rate": 8.37983593261664e-06,
+      "loss": 0.422,
+      "step": 1216
+    },
+    {
+      "epoch": 1.0130410654827968,
+      "grad_norm": 0.37474164366722107,
+      "learning_rate": 8.376263804542798e-06,
+      "loss": 0.4333,
+      "step": 1217
+    },
+    {
+      "epoch": 1.013873473917869,
+      "grad_norm": 0.39734119176864624,
+      "learning_rate": 8.372688506098128e-06,
+      "loss": 0.4147,
+      "step": 1218
+    },
+    {
+      "epoch": 1.0147058823529411,
+      "grad_norm": 0.3793570399284363,
+      "learning_rate": 8.369110040639899e-06,
+      "loss": 0.4257,
+      "step": 1219
+    },
+    {
+      "epoch": 1.0155382907880133,
+      "grad_norm": 0.38969290256500244,
+      "learning_rate": 8.365528411528348e-06,
+      "loss": 0.4657,
+      "step": 1220
+    },
+    {
+      "epoch": 1.0163706992230854,
+      "grad_norm": 0.33815810084342957,
+      "learning_rate": 8.361943622126694e-06,
+      "loss": 0.3868,
+      "step": 1221
+    },
+    {
+      "epoch": 1.0172031076581576,
+      "grad_norm": 0.4408543109893799,
+      "learning_rate": 8.358355675801112e-06,
+      "loss": 0.4481,
+      "step": 1222
+    },
+    {
+      "epoch": 1.0180355160932297,
+      "grad_norm": 0.3819434940814972,
+      "learning_rate": 8.354764575920747e-06,
+      "loss": 0.4484,
+      "step": 1223
+    },
+    {
+      "epoch": 1.0188679245283019,
+      "grad_norm": 0.3943055272102356,
+      "learning_rate": 8.351170325857705e-06,
+      "loss": 0.4066,
+      "step": 1224
+    },
+    {
+      "epoch": 1.019700332963374,
+      "grad_norm": 0.3979198634624481,
+      "learning_rate": 8.347572928987052e-06,
+      "loss": 0.4461,
+      "step": 1225
+    },
+    {
+      "epoch": 1.0205327413984462,
+      "grad_norm": 0.3459455370903015,
+      "learning_rate": 8.343972388686806e-06,
+      "loss": 0.3935,
+      "step": 1226
+    },
+    {
+      "epoch": 1.0213651498335183,
+      "grad_norm": 0.4697541892528534,
+      "learning_rate": 8.340368708337934e-06,
+      "loss": 0.478,
+      "step": 1227
+    },
+    {
+      "epoch": 1.0221975582685905,
+      "grad_norm": 0.392013818025589,
+      "learning_rate": 8.336761891324357e-06,
+      "loss": 0.4317,
+      "step": 1228
+    },
+    {
+      "epoch": 1.0230299667036626,
+      "grad_norm": 0.3422892987728119,
+      "learning_rate": 8.333151941032941e-06,
+      "loss": 0.3749,
+      "step": 1229
+    },
+    {
+      "epoch": 1.0238623751387348,
+      "grad_norm": 0.4656471312046051,
+      "learning_rate": 8.32953886085349e-06,
+      "loss": 0.4637,
+      "step": 1230
+    },
+    {
+      "epoch": 1.024694783573807,
+      "grad_norm": 0.4500276744365692,
+      "learning_rate": 8.325922654178752e-06,
+      "loss": 0.454,
+      "step": 1231
+    },
+    {
+      "epoch": 1.025527192008879,
+      "grad_norm": 0.3819878399372101,
+      "learning_rate": 8.322303324404408e-06,
+      "loss": 0.4158,
+      "step": 1232
+    },
+    {
+      "epoch": 1.0263596004439512,
+      "grad_norm": 0.4874471426010132,
+      "learning_rate": 8.318680874929068e-06,
+      "loss": 0.4479,
+      "step": 1233
+    },
+    {
+      "epoch": 1.0271920088790234,
+      "grad_norm": 0.39757636189460754,
+      "learning_rate": 8.315055309154283e-06,
+      "loss": 0.4527,
+      "step": 1234
+    },
+    {
+      "epoch": 1.0280244173140956,
+      "grad_norm": 0.3715905547142029,
+      "learning_rate": 8.311426630484513e-06,
+      "loss": 0.4181,
+      "step": 1235
+    },
+    {
+      "epoch": 1.0288568257491675,
+      "grad_norm": 0.44828590750694275,
+      "learning_rate": 8.30779484232716e-06,
+      "loss": 0.4454,
+      "step": 1236
+    },
+    {
+      "epoch": 1.0296892341842396,
+      "grad_norm": 0.4298873245716095,
+      "learning_rate": 8.304159948092532e-06,
+      "loss": 0.4382,
+      "step": 1237
+    },
+    {
+      "epoch": 1.0305216426193118,
+      "grad_norm": 0.36218270659446716,
+      "learning_rate": 8.30052195119386e-06,
+      "loss": 0.4058,
+      "step": 1238
+    },
+    {
+      "epoch": 1.031354051054384,
+      "grad_norm": 0.4292842149734497,
+      "learning_rate": 8.296880855047284e-06,
+      "loss": 0.4512,
+      "step": 1239
+    },
+    {
+      "epoch": 1.032186459489456,
+      "grad_norm": 0.3632907569408417,
+      "learning_rate": 8.293236663071859e-06,
+      "loss": 0.4331,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0330188679245282,
+      "grad_norm": 0.3891299366950989,
+      "learning_rate": 8.289589378689548e-06,
+      "loss": 0.4313,
+      "step": 1241
+    },
+    {
+      "epoch": 1.0338512763596004,
+      "grad_norm": 0.37545377016067505,
+      "learning_rate": 8.28593900532521e-06,
+      "loss": 0.375,
+      "step": 1242
+    },
+    {
+      "epoch": 1.0346836847946725,
+      "grad_norm": 0.3906904458999634,
+      "learning_rate": 8.28228554640661e-06,
+      "loss": 0.4785,
+      "step": 1243
+    },
+    {
+      "epoch": 1.0355160932297447,
+      "grad_norm": 0.3426535725593567,
+      "learning_rate": 8.278629005364412e-06,
+      "loss": 0.387,
+      "step": 1244
+    },
+    {
+      "epoch": 1.0363485016648168,
+      "grad_norm": 0.4440067410469055,
+      "learning_rate": 8.274969385632173e-06,
+      "loss": 0.4773,
+      "step": 1245
+    },
+    {
+      "epoch": 1.037180910099889,
+      "grad_norm": 0.36117398738861084,
+      "learning_rate": 8.271306690646336e-06,
+      "loss": 0.4375,
+      "step": 1246
+    },
+    {
+      "epoch": 1.0380133185349611,
+      "grad_norm": 0.4358084499835968,
+      "learning_rate": 8.267640923846242e-06,
+      "loss": 0.4346,
+      "step": 1247
+    },
+    {
+      "epoch": 1.0388457269700333,
+      "grad_norm": 0.3763819932937622,
+      "learning_rate": 8.263972088674103e-06,
+      "loss": 0.4179,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0396781354051055,
+      "grad_norm": 0.3596190810203552,
+      "learning_rate": 8.260300188575024e-06,
+      "loss": 0.4451,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0405105438401776,
+      "grad_norm": 0.42740023136138916,
+      "learning_rate": 8.256625226996981e-06,
+      "loss": 0.4607,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0413429522752498,
+      "grad_norm": 0.3612176775932312,
+      "learning_rate": 8.252947207390832e-06,
+      "loss": 0.4038,
+      "step": 1251
+    },
+    {
+      "epoch": 1.042175360710322,
+      "grad_norm": 0.3865748941898346,
+      "learning_rate": 8.249266133210296e-06,
+      "loss": 0.4395,
+      "step": 1252
+    },
+    {
+      "epoch": 1.043007769145394,
+      "grad_norm": 0.37287238240242004,
+      "learning_rate": 8.245582007911967e-06,
+      "loss": 0.4055,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0438401775804662,
+      "grad_norm": 0.4149041771888733,
+      "learning_rate": 8.241894834955306e-06,
+      "loss": 0.421,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0446725860155384,
+      "grad_norm": 0.42680850625038147,
+      "learning_rate": 8.238204617802633e-06,
+      "loss": 0.4522,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0455049944506105,
+      "grad_norm": 0.3849968910217285,
+      "learning_rate": 8.234511359919125e-06,
+      "loss": 0.3983,
+      "step": 1256
+    },
+    {
+      "epoch": 1.0463374028856827,
+      "grad_norm": 0.405513197183609,
+      "learning_rate": 8.230815064772815e-06,
+      "loss": 0.4165,
+      "step": 1257
+    },
+    {
+      "epoch": 1.0471698113207548,
+      "grad_norm": 0.4122266471385956,
+      "learning_rate": 8.22711573583459e-06,
+      "loss": 0.4478,
+      "step": 1258
+    },
+    {
+      "epoch": 1.0480022197558267,
+      "grad_norm": 0.3673398196697235,
+      "learning_rate": 8.223413376578182e-06,
+      "loss": 0.4134,
+      "step": 1259
+    },
+    {
+      "epoch": 1.048834628190899,
+      "grad_norm": 0.40100428462028503,
+      "learning_rate": 8.219707990480177e-06,
+      "loss": 0.4296,
+      "step": 1260
+    },
+    {
+      "epoch": 1.049667036625971,
+      "grad_norm": 0.34723445773124695,
+      "learning_rate": 8.215999581019993e-06,
+      "loss": 0.4036,
+      "step": 1261
+    },
+    {
+      "epoch": 1.0504994450610432,
+      "grad_norm": 0.4086048901081085,
+      "learning_rate": 8.212288151679892e-06,
+      "loss": 0.4462,
+      "step": 1262
+    },
+    {
+      "epoch": 1.0513318534961154,
+      "grad_norm": 0.37759000062942505,
+      "learning_rate": 8.208573705944972e-06,
+      "loss": 0.4191,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0521642619311875,
+      "grad_norm": 0.43619635701179504,
+      "learning_rate": 8.204856247303163e-06,
+      "loss": 0.4364,
+      "step": 1264
+    },
+    {
+      "epoch": 1.0529966703662597,
+      "grad_norm": 0.395222544670105,
+      "learning_rate": 8.201135779245222e-06,
+      "loss": 0.4511,
+      "step": 1265
+    },
+    {
+      "epoch": 1.0538290788013318,
+      "grad_norm": 0.4762343466281891,
+      "learning_rate": 8.197412305264735e-06,
+      "loss": 0.4423,
+      "step": 1266
+    },
+    {
+      "epoch": 1.054661487236404,
+      "grad_norm": 0.4282004237174988,
+      "learning_rate": 8.193685828858109e-06,
+      "loss": 0.4506,
+      "step": 1267
+    },
+    {
+      "epoch": 1.0554938956714761,
+      "grad_norm": 0.35303211212158203,
+      "learning_rate": 8.189956353524568e-06,
+      "loss": 0.4602,
+      "step": 1268
+    },
+    {
+      "epoch": 1.0563263041065483,
+      "grad_norm": 0.3769274652004242,
+      "learning_rate": 8.18622388276616e-06,
+      "loss": 0.4354,
+      "step": 1269
+    },
+    {
+      "epoch": 1.0571587125416204,
+      "grad_norm": 0.41085049510002136,
+      "learning_rate": 8.182488420087737e-06,
+      "loss": 0.4079,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0579911209766926,
+      "grad_norm": 0.40207546949386597,
+      "learning_rate": 8.178749968996965e-06,
+      "loss": 0.4262,
+      "step": 1271
+    },
+    {
+      "epoch": 1.0588235294117647,
+      "grad_norm": 0.4222968518733978,
+      "learning_rate": 8.175008533004312e-06,
+      "loss": 0.4536,
+      "step": 1272
+    },
+    {
+      "epoch": 1.0596559378468369,
+      "grad_norm": 0.4397706091403961,
+      "learning_rate": 8.171264115623056e-06,
+      "loss": 0.4143,
+      "step": 1273
+    },
+    {
+      "epoch": 1.060488346281909,
+      "grad_norm": 0.37260058522224426,
+      "learning_rate": 8.167516720369268e-06,
+      "loss": 0.3961,
+      "step": 1274
+    },
+    {
+      "epoch": 1.0613207547169812,
+      "grad_norm": 0.4061874449253082,
+      "learning_rate": 8.163766350761819e-06,
+      "loss": 0.4376,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0621531631520533,
+      "grad_norm": 0.4455544054508209,
+      "learning_rate": 8.160013010322372e-06,
+      "loss": 0.4101,
+      "step": 1276
+    },
+    {
+      "epoch": 1.0629855715871255,
+      "grad_norm": 0.36185914278030396,
+      "learning_rate": 8.156256702575378e-06,
+      "loss": 0.4202,
+      "step": 1277
+    },
+    {
+      "epoch": 1.0638179800221976,
+      "grad_norm": 0.4335435628890991,
+      "learning_rate": 8.152497431048076e-06,
+      "loss": 0.4351,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0646503884572698,
+      "grad_norm": 0.3470878303050995,
+      "learning_rate": 8.148735199270487e-06,
+      "loss": 0.3819,
+      "step": 1279
+    },
+    {
+      "epoch": 1.065482796892342,
+      "grad_norm": 0.3797052800655365,
+      "learning_rate": 8.144970010775417e-06,
+      "loss": 0.4304,
+      "step": 1280
+    },
+    {
+      "epoch": 1.066315205327414,
+      "grad_norm": 0.44937318563461304,
+      "learning_rate": 8.141201869098439e-06,
+      "loss": 0.4613,
+      "step": 1281
+    },
+    {
+      "epoch": 1.067147613762486,
+      "grad_norm": 0.3713797628879547,
+      "learning_rate": 8.137430777777904e-06,
+      "loss": 0.4366,
+      "step": 1282
+    },
+    {
+      "epoch": 1.0679800221975582,
+      "grad_norm": 0.3958394527435303,
+      "learning_rate": 8.133656740354936e-06,
+      "loss": 0.4576,
+      "step": 1283
+    },
+    {
+      "epoch": 1.0688124306326303,
+      "grad_norm": 0.4216252267360687,
+      "learning_rate": 8.129879760373419e-06,
+      "loss": 0.4375,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0696448390677025,
+      "grad_norm": 0.3756641745567322,
+      "learning_rate": 8.126099841380008e-06,
+      "loss": 0.442,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0704772475027746,
+      "grad_norm": 0.35558274388313293,
+      "learning_rate": 8.122316986924108e-06,
+      "loss": 0.4089,
+      "step": 1286
+    },
+    {
+      "epoch": 1.0713096559378468,
+      "grad_norm": 0.4253517985343933,
+      "learning_rate": 8.118531200557888e-06,
+      "loss": 0.479,
+      "step": 1287
+    },
+    {
+      "epoch": 1.072142064372919,
+      "grad_norm": 0.3373982012271881,
+      "learning_rate": 8.114742485836267e-06,
+      "loss": 0.404,
+      "step": 1288
+    },
+    {
+      "epoch": 1.072974472807991,
+      "grad_norm": 0.41005459427833557,
+      "learning_rate": 8.110950846316915e-06,
+      "loss": 0.4544,
+      "step": 1289
+    },
+    {
+      "epoch": 1.0738068812430632,
+      "grad_norm": 0.3480115532875061,
+      "learning_rate": 8.107156285560249e-06,
+      "loss": 0.4235,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0746392896781354,
+      "grad_norm": 0.373367041349411,
+      "learning_rate": 8.103358807129424e-06,
+      "loss": 0.4552,
+      "step": 1291
+    },
+    {
+      "epoch": 1.0754716981132075,
+      "grad_norm": 0.3816337585449219,
+      "learning_rate": 8.099558414590343e-06,
+      "loss": 0.4019,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0763041065482797,
+      "grad_norm": 0.3785970211029053,
+      "learning_rate": 8.09575511151164e-06,
+      "loss": 0.4361,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0771365149833518,
+      "grad_norm": 0.37990981340408325,
+      "learning_rate": 8.091948901464683e-06,
+      "loss": 0.4375,
+      "step": 1294
+    },
+    {
+      "epoch": 1.077968923418424,
+      "grad_norm": 0.36004897952079773,
+      "learning_rate": 8.088139788023568e-06,
+      "loss": 0.4331,
+      "step": 1295
+    },
+    {
+      "epoch": 1.0788013318534961,
+      "grad_norm": 0.38282451033592224,
+      "learning_rate": 8.084327774765121e-06,
+      "loss": 0.4659,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0796337402885683,
+      "grad_norm": 0.3787173330783844,
+      "learning_rate": 8.08051286526889e-06,
+      "loss": 0.4041,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0804661487236404,
+      "grad_norm": 0.366184264421463,
+      "learning_rate": 8.076695063117141e-06,
+      "loss": 0.4211,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0812985571587126,
+      "grad_norm": 0.3390844166278839,
+      "learning_rate": 8.072874371894856e-06,
+      "loss": 0.4288,
+      "step": 1299
+    },
+    {
+      "epoch": 1.0821309655937847,
+      "grad_norm": 0.35803496837615967,
+      "learning_rate": 8.069050795189732e-06,
+      "loss": 0.4241,
+      "step": 1300
+    },
+    {
+      "epoch": 1.082963374028857,
+      "grad_norm": 0.3761143982410431,
+      "learning_rate": 8.065224336592175e-06,
+      "loss": 0.4112,
+      "step": 1301
+    },
+    {
+      "epoch": 1.083795782463929,
+      "grad_norm": 0.3917039930820465,
+      "learning_rate": 8.061394999695295e-06,
+      "loss": 0.398,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0846281908990012,
+      "grad_norm": 0.35649728775024414,
+      "learning_rate": 8.057562788094909e-06,
+      "loss": 0.4079,
+      "step": 1303
+    },
+    {
+      "epoch": 1.0854605993340734,
+      "grad_norm": 0.4020186960697174,
+      "learning_rate": 8.053727705389527e-06,
+      "loss": 0.421,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0862930077691453,
+      "grad_norm": 0.3835377097129822,
+      "learning_rate": 8.049889755180363e-06,
+      "loss": 0.4015,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0871254162042177,
+      "grad_norm": 0.3527611196041107,
+      "learning_rate": 8.046048941071316e-06,
+      "loss": 0.4202,
+      "step": 1306
+    },
+    {
+      "epoch": 1.0879578246392896,
+      "grad_norm": 0.41720572113990784,
+      "learning_rate": 8.042205266668982e-06,
+      "loss": 0.494,
+      "step": 1307
+    },
+    {
+      "epoch": 1.0887902330743617,
+      "grad_norm": 0.4160442054271698,
+      "learning_rate": 8.038358735582632e-06,
+      "loss": 0.4654,
+      "step": 1308
+    },
+    {
+      "epoch": 1.0896226415094339,
+      "grad_norm": 0.3363052010536194,
+      "learning_rate": 8.034509351424231e-06,
+      "loss": 0.3651,
+      "step": 1309
+    },
+    {
+      "epoch": 1.090455049944506,
+      "grad_norm": 0.45190951228141785,
+      "learning_rate": 8.030657117808415e-06,
+      "loss": 0.4048,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0912874583795782,
+      "grad_norm": 0.3866722881793976,
+      "learning_rate": 8.026802038352503e-06,
+      "loss": 0.4191,
+      "step": 1311
+    },
+    {
+      "epoch": 1.0921198668146503,
+      "grad_norm": 0.4028158187866211,
+      "learning_rate": 8.02294411667648e-06,
+      "loss": 0.3982,
+      "step": 1312
+    },
+    {
+      "epoch": 1.0929522752497225,
+      "grad_norm": 0.4108687937259674,
+      "learning_rate": 8.019083356403002e-06,
+      "loss": 0.4326,
+      "step": 1313
+    },
+    {
+      "epoch": 1.0937846836847946,
+      "grad_norm": 0.417467325925827,
+      "learning_rate": 8.015219761157387e-06,
+      "loss": 0.4195,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0946170921198668,
+      "grad_norm": 0.3949446678161621,
+      "learning_rate": 8.011353334567625e-06,
+      "loss": 0.4035,
+      "step": 1315
+    },
+    {
+      "epoch": 1.095449500554939,
+      "grad_norm": 0.374812513589859,
+      "learning_rate": 8.007484080264355e-06,
+      "loss": 0.4318,
+      "step": 1316
+    },
+    {
+      "epoch": 1.096281908990011,
+      "grad_norm": 0.4128841459751129,
+      "learning_rate": 8.003612001880872e-06,
+      "loss": 0.436,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0971143174250833,
+      "grad_norm": 0.4258720278739929,
+      "learning_rate": 7.99973710305313e-06,
+      "loss": 0.4393,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0979467258601554,
+      "grad_norm": 0.38994690775871277,
+      "learning_rate": 7.995859387419726e-06,
+      "loss": 0.4135,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0987791342952276,
+      "grad_norm": 0.3956961929798126,
+      "learning_rate": 7.9919788586219e-06,
+      "loss": 0.4261,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0996115427302997,
+      "grad_norm": 0.4195059537887573,
+      "learning_rate": 7.988095520303539e-06,
+      "loss": 0.4351,
+      "step": 1321
+    },
+    {
+      "epoch": 1.1004439511653719,
+      "grad_norm": 0.3911471366882324,
+      "learning_rate": 7.984209376111165e-06,
+      "loss": 0.4434,
+      "step": 1322
+    },
+    {
+      "epoch": 1.101276359600444,
+      "grad_norm": 0.48853620886802673,
+      "learning_rate": 7.980320429693934e-06,
+      "loss": 0.4585,
+      "step": 1323
+    },
+    {
+      "epoch": 1.1021087680355162,
+      "grad_norm": 0.40332579612731934,
+      "learning_rate": 7.976428684703637e-06,
+      "loss": 0.3821,
+      "step": 1324
+    },
+    {
+      "epoch": 1.1029411764705883,
+      "grad_norm": 0.41588088870048523,
+      "learning_rate": 7.97253414479469e-06,
+      "loss": 0.469,
+      "step": 1325
+    },
+    {
+      "epoch": 1.1037735849056605,
+      "grad_norm": 0.405784010887146,
+      "learning_rate": 7.968636813624134e-06,
+      "loss": 0.3994,
+      "step": 1326
+    },
+    {
+      "epoch": 1.1046059933407326,
+      "grad_norm": 0.39917218685150146,
+      "learning_rate": 7.964736694851632e-06,
+      "loss": 0.4317,
+      "step": 1327
+    },
+    {
+      "epoch": 1.1054384017758045,
+      "grad_norm": 0.46947672963142395,
+      "learning_rate": 7.960833792139461e-06,
+      "loss": 0.4775,
+      "step": 1328
+    },
+    {
+      "epoch": 1.106270810210877,
+      "grad_norm": 0.42204639315605164,
+      "learning_rate": 7.95692810915252e-06,
+      "loss": 0.4228,
+      "step": 1329
+    },
+    {
+      "epoch": 1.1071032186459488,
+      "grad_norm": 0.41625940799713135,
+      "learning_rate": 7.953019649558309e-06,
+      "loss": 0.4016,
+      "step": 1330
+    },
+    {
+      "epoch": 1.107935627081021,
+      "grad_norm": 0.39135950803756714,
+      "learning_rate": 7.949108417026941e-06,
+      "loss": 0.4445,
+      "step": 1331
+    },
+    {
+      "epoch": 1.1087680355160932,
+      "grad_norm": 0.4039277136325836,
+      "learning_rate": 7.945194415231133e-06,
+      "loss": 0.4099,
+      "step": 1332
+    },
+    {
+      "epoch": 1.1096004439511653,
+      "grad_norm": 0.3570069670677185,
+      "learning_rate": 7.9412776478462e-06,
+      "loss": 0.4073,
+      "step": 1333
+    },
+    {
+      "epoch": 1.1104328523862375,
+      "grad_norm": 0.4042835235595703,
+      "learning_rate": 7.937358118550058e-06,
+      "loss": 0.4037,
+      "step": 1334
+    },
+    {
+      "epoch": 1.1112652608213096,
+      "grad_norm": 0.4284457266330719,
+      "learning_rate": 7.933435831023211e-06,
+      "loss": 0.4517,
+      "step": 1335
+    },
+    {
+      "epoch": 1.1120976692563818,
+      "grad_norm": 0.3967071771621704,
+      "learning_rate": 7.929510788948755e-06,
+      "loss": 0.4205,
+      "step": 1336
+    },
+    {
+      "epoch": 1.112930077691454,
+      "grad_norm": 0.3720296323299408,
+      "learning_rate": 7.925582996012375e-06,
+      "loss": 0.4472,
+      "step": 1337
+    },
+    {
+      "epoch": 1.113762486126526,
+      "grad_norm": 0.38390690088272095,
+      "learning_rate": 7.921652455902337e-06,
+      "loss": 0.3934,
+      "step": 1338
+    },
+    {
+      "epoch": 1.1145948945615982,
+      "grad_norm": 0.48500263690948486,
+      "learning_rate": 7.917719172309487e-06,
+      "loss": 0.4521,
+      "step": 1339
+    },
+    {
+      "epoch": 1.1154273029966704,
+      "grad_norm": 0.38441312313079834,
+      "learning_rate": 7.913783148927246e-06,
+      "loss": 0.4474,
+      "step": 1340
+    },
+    {
+      "epoch": 1.1162597114317425,
+      "grad_norm": 0.462807297706604,
+      "learning_rate": 7.909844389451611e-06,
+      "loss": 0.4397,
+      "step": 1341
+    },
+    {
+      "epoch": 1.1170921198668147,
+      "grad_norm": 0.38858792185783386,
+      "learning_rate": 7.905902897581145e-06,
+      "loss": 0.3972,
+      "step": 1342
+    },
+    {
+      "epoch": 1.1179245283018868,
+      "grad_norm": 0.3973788022994995,
+      "learning_rate": 7.901958677016977e-06,
+      "loss": 0.4265,
+      "step": 1343
+    },
+    {
+      "epoch": 1.118756936736959,
+      "grad_norm": 0.3690992593765259,
+      "learning_rate": 7.898011731462801e-06,
+      "loss": 0.4439,
+      "step": 1344
+    },
+    {
+      "epoch": 1.1195893451720311,
+      "grad_norm": 0.5391969680786133,
+      "learning_rate": 7.894062064624865e-06,
+      "loss": 0.4472,
+      "step": 1345
+    },
+    {
+      "epoch": 1.1204217536071033,
+      "grad_norm": 0.3407292664051056,
+      "learning_rate": 7.890109680211979e-06,
+      "loss": 0.3753,
+      "step": 1346
+    },
+    {
+      "epoch": 1.1212541620421754,
+      "grad_norm": 0.3949512541294098,
+      "learning_rate": 7.886154581935499e-06,
+      "loss": 0.459,
+      "step": 1347
+    },
+    {
+      "epoch": 1.1220865704772476,
+      "grad_norm": 0.3942648470401764,
+      "learning_rate": 7.88219677350933e-06,
+      "loss": 0.3809,
+      "step": 1348
+    },
+    {
+      "epoch": 1.1229189789123197,
+      "grad_norm": 0.4611280560493469,
+      "learning_rate": 7.878236258649927e-06,
+      "loss": 0.4762,
+      "step": 1349
+    },
+    {
+      "epoch": 1.1237513873473919,
+      "grad_norm": 0.37650808691978455,
+      "learning_rate": 7.874273041076283e-06,
+      "loss": 0.4166,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1245837957824638,
+      "grad_norm": 0.4160518944263458,
+      "learning_rate": 7.870307124509926e-06,
+      "loss": 0.3948,
+      "step": 1351
+    },
+    {
+      "epoch": 1.1254162042175362,
+      "grad_norm": 0.39588451385498047,
+      "learning_rate": 7.86633851267492e-06,
+      "loss": 0.4322,
+      "step": 1352
+    },
+    {
+      "epoch": 1.1262486126526081,
+      "grad_norm": 0.3730158507823944,
+      "learning_rate": 7.862367209297864e-06,
+      "loss": 0.4327,
+      "step": 1353
+    },
+    {
+      "epoch": 1.1270810210876803,
+      "grad_norm": 0.43521034717559814,
+      "learning_rate": 7.85839321810788e-06,
+      "loss": 0.392,
+      "step": 1354
+    },
+    {
+      "epoch": 1.1279134295227524,
+      "grad_norm": 0.4675239622592926,
+      "learning_rate": 7.854416542836617e-06,
+      "loss": 0.4395,
+      "step": 1355
+    },
+    {
+      "epoch": 1.1287458379578246,
+      "grad_norm": 0.38083502650260925,
+      "learning_rate": 7.85043718721824e-06,
+      "loss": 0.417,
+      "step": 1356
+    },
+    {
+      "epoch": 1.1295782463928967,
+      "grad_norm": 0.46890681982040405,
+      "learning_rate": 7.846455154989437e-06,
+      "loss": 0.4463,
+      "step": 1357
+    },
+    {
+      "epoch": 1.1304106548279689,
+      "grad_norm": 0.4093262851238251,
+      "learning_rate": 7.842470449889403e-06,
+      "loss": 0.394,
+      "step": 1358
+    },
+    {
+      "epoch": 1.131243063263041,
+      "grad_norm": 0.35766276717185974,
+      "learning_rate": 7.838483075659846e-06,
+      "loss": 0.4444,
+      "step": 1359
+    },
+    {
+      "epoch": 1.1320754716981132,
+      "grad_norm": 0.4069897532463074,
+      "learning_rate": 7.83449303604498e-06,
+      "loss": 0.4139,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1329078801331853,
+      "grad_norm": 0.4673629403114319,
+      "learning_rate": 7.830500334791525e-06,
+      "loss": 0.4268,
+      "step": 1361
+    },
+    {
+      "epoch": 1.1337402885682575,
+      "grad_norm": 0.4079434275627136,
+      "learning_rate": 7.826504975648696e-06,
+      "loss": 0.4431,
+      "step": 1362
+    },
+    {
+      "epoch": 1.1345726970033296,
+      "grad_norm": 0.4580193758010864,
+      "learning_rate": 7.822506962368204e-06,
+      "loss": 0.424,
+      "step": 1363
+    },
+    {
+      "epoch": 1.1354051054384018,
+      "grad_norm": 0.4982101023197174,
+      "learning_rate": 7.818506298704254e-06,
+      "loss": 0.4285,
+      "step": 1364
+    },
+    {
+      "epoch": 1.136237513873474,
+      "grad_norm": 0.3587871491909027,
+      "learning_rate": 7.814502988413539e-06,
+      "loss": 0.4058,
+      "step": 1365
+    },
+    {
+      "epoch": 1.137069922308546,
+      "grad_norm": 0.48973003029823303,
+      "learning_rate": 7.810497035255239e-06,
+      "loss": 0.4631,
+      "step": 1366
+    },
+    {
+      "epoch": 1.1379023307436182,
+      "grad_norm": 0.4257657825946808,
+      "learning_rate": 7.80648844299101e-06,
+      "loss": 0.4013,
+      "step": 1367
+    },
+    {
+      "epoch": 1.1387347391786904,
+      "grad_norm": 0.5077852010726929,
+      "learning_rate": 7.802477215384997e-06,
+      "loss": 0.4421,
+      "step": 1368
+    },
+    {
+      "epoch": 1.1395671476137625,
+      "grad_norm": 0.41671276092529297,
+      "learning_rate": 7.79846335620381e-06,
+      "loss": 0.4297,
+      "step": 1369
+    },
+    {
+      "epoch": 1.1403995560488347,
+      "grad_norm": 0.392742395401001,
+      "learning_rate": 7.794446869216527e-06,
+      "loss": 0.3575,
+      "step": 1370
+    },
+    {
+      "epoch": 1.1412319644839068,
+      "grad_norm": 0.504817545413971,
+      "learning_rate": 7.79042775819471e-06,
+      "loss": 0.4213,
+      "step": 1371
+    },
+    {
+      "epoch": 1.142064372918979,
+      "grad_norm": 0.4020238518714905,
+      "learning_rate": 7.786406026912368e-06,
+      "loss": 0.4363,
+      "step": 1372
+    },
+    {
+      "epoch": 1.1428967813540512,
+      "grad_norm": 0.49842900037765503,
+      "learning_rate": 7.782381679145979e-06,
+      "loss": 0.4386,
+      "step": 1373
+    },
+    {
+      "epoch": 1.143729189789123,
+      "grad_norm": 0.48727041482925415,
+      "learning_rate": 7.778354718674475e-06,
+      "loss": 0.463,
+      "step": 1374
+    },
+    {
+      "epoch": 1.1445615982241955,
+      "grad_norm": 0.39957672357559204,
+      "learning_rate": 7.774325149279243e-06,
+      "loss": 0.402,
+      "step": 1375
+    },
+    {
+      "epoch": 1.1453940066592674,
+      "grad_norm": 0.5904676914215088,
+      "learning_rate": 7.770292974744119e-06,
+      "loss": 0.4446,
+      "step": 1376
+    },
+    {
+      "epoch": 1.1462264150943395,
+      "grad_norm": 0.34512779116630554,
+      "learning_rate": 7.766258198855386e-06,
+      "loss": 0.4331,
+      "step": 1377
+    },
+    {
+      "epoch": 1.1470588235294117,
+      "grad_norm": 0.44855794310569763,
+      "learning_rate": 7.76222082540177e-06,
+      "loss": 0.4014,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1478912319644838,
+      "grad_norm": 0.5417425632476807,
+      "learning_rate": 7.758180858174434e-06,
+      "loss": 0.4453,
+      "step": 1379
+    },
+    {
+      "epoch": 1.148723640399556,
+      "grad_norm": 0.3855729401111603,
+      "learning_rate": 7.754138300966978e-06,
+      "loss": 0.4654,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1495560488346281,
+      "grad_norm": 0.4581325650215149,
+      "learning_rate": 7.750093157575433e-06,
+      "loss": 0.4398,
+      "step": 1381
+    },
+    {
+      "epoch": 1.1503884572697003,
+      "grad_norm": 0.4510229825973511,
+      "learning_rate": 7.746045431798264e-06,
+      "loss": 0.4312,
+      "step": 1382
+    },
+    {
+      "epoch": 1.1512208657047724,
+      "grad_norm": 0.3615362346172333,
+      "learning_rate": 7.74199512743635e-06,
+      "loss": 0.4297,
+      "step": 1383
+    },
+    {
+      "epoch": 1.1520532741398446,
+      "grad_norm": 0.4092762768268585,
+      "learning_rate": 7.737942248293001e-06,
+      "loss": 0.4472,
+      "step": 1384
+    },
+    {
+      "epoch": 1.1528856825749167,
+      "grad_norm": 0.46709978580474854,
+      "learning_rate": 7.733886798173945e-06,
+      "loss": 0.456,
+      "step": 1385
+    },
+    {
+      "epoch": 1.153718091009989,
+      "grad_norm": 0.3712662160396576,
+      "learning_rate": 7.729828780887313e-06,
+      "loss": 0.4129,
+      "step": 1386
+    },
+    {
+      "epoch": 1.154550499445061,
+      "grad_norm": 0.35037344694137573,
+      "learning_rate": 7.72576820024366e-06,
+      "loss": 0.4172,
+      "step": 1387
+    },
+    {
+      "epoch": 1.1553829078801332,
+      "grad_norm": 0.4096495807170868,
+      "learning_rate": 7.72170506005594e-06,
+      "loss": 0.4243,
+      "step": 1388
+    },
+    {
+      "epoch": 1.1562153163152054,
+      "grad_norm": 0.4149743914604187,
+      "learning_rate": 7.717639364139514e-06,
+      "loss": 0.4728,
+      "step": 1389
+    },
+    {
+      "epoch": 1.1570477247502775,
+      "grad_norm": 0.3861239552497864,
+      "learning_rate": 7.713571116312143e-06,
+      "loss": 0.4409,
+      "step": 1390
+    },
+    {
+      "epoch": 1.1578801331853497,
+      "grad_norm": 0.3818301558494568,
+      "learning_rate": 7.709500320393976e-06,
+      "loss": 0.415,
+      "step": 1391
+    },
+    {
+      "epoch": 1.1587125416204218,
+      "grad_norm": 0.3710886240005493,
+      "learning_rate": 7.70542698020757e-06,
+      "loss": 0.4251,
+      "step": 1392
+    },
+    {
+      "epoch": 1.159544950055494,
+      "grad_norm": 0.3728558421134949,
+      "learning_rate": 7.70135109957786e-06,
+      "loss": 0.4196,
+      "step": 1393
+    },
+    {
+      "epoch": 1.1603773584905661,
+      "grad_norm": 0.42186808586120605,
+      "learning_rate": 7.697272682332168e-06,
+      "loss": 0.4473,
+      "step": 1394
+    },
+    {
+      "epoch": 1.1612097669256383,
+      "grad_norm": 0.37886783480644226,
+      "learning_rate": 7.6931917323002e-06,
+      "loss": 0.4295,
+      "step": 1395
+    },
+    {
+      "epoch": 1.1620421753607104,
+      "grad_norm": 0.36047130823135376,
+      "learning_rate": 7.689108253314038e-06,
+      "loss": 0.4145,
+      "step": 1396
+    },
+    {
+      "epoch": 1.1628745837957823,
+      "grad_norm": 0.41782742738723755,
+      "learning_rate": 7.685022249208142e-06,
+      "loss": 0.4459,
+      "step": 1397
+    },
+    {
+      "epoch": 1.1637069922308547,
+      "grad_norm": 0.3524583578109741,
+      "learning_rate": 7.680933723819343e-06,
+      "loss": 0.4361,
+      "step": 1398
+    },
+    {
+      "epoch": 1.1645394006659266,
+      "grad_norm": 0.34524109959602356,
+      "learning_rate": 7.676842680986836e-06,
+      "loss": 0.405,
+      "step": 1399
+    },
+    {
+      "epoch": 1.1653718091009988,
+      "grad_norm": 0.40391451120376587,
+      "learning_rate": 7.67274912455218e-06,
+      "loss": 0.445,
+      "step": 1400
+    },
+    {
+      "epoch": 1.166204217536071,
+      "grad_norm": 0.3440330922603607,
+      "learning_rate": 7.6686530583593e-06,
+      "loss": 0.4126,
+      "step": 1401
+    },
+    {
+      "epoch": 1.167036625971143,
+      "grad_norm": 0.4131713807582855,
+      "learning_rate": 7.664554486254468e-06,
+      "loss": 0.4831,
+      "step": 1402
+    },
+    {
+      "epoch": 1.1678690344062153,
+      "grad_norm": 0.3954821527004242,
+      "learning_rate": 7.660453412086323e-06,
+      "loss": 0.4501,
+      "step": 1403
+    },
+    {
+      "epoch": 1.1687014428412874,
+      "grad_norm": 0.3479725420475006,
+      "learning_rate": 7.656349839705838e-06,
+      "loss": 0.4125,
+      "step": 1404
+    },
+    {
+      "epoch": 1.1695338512763596,
+      "grad_norm": 0.4041290581226349,
+      "learning_rate": 7.652243772966345e-06,
+      "loss": 0.3941,
+      "step": 1405
+    },
+    {
+      "epoch": 1.1703662597114317,
+      "grad_norm": 0.3935270607471466,
+      "learning_rate": 7.648135215723511e-06,
+      "loss": 0.4381,
+      "step": 1406
+    },
+    {
+      "epoch": 1.1711986681465039,
+      "grad_norm": 0.3363988399505615,
+      "learning_rate": 7.64402417183534e-06,
+      "loss": 0.3909,
+      "step": 1407
+    },
+    {
+      "epoch": 1.172031076581576,
+      "grad_norm": 0.358073353767395,
+      "learning_rate": 7.639910645162179e-06,
+      "loss": 0.4092,
+      "step": 1408
+    },
+    {
+      "epoch": 1.1728634850166482,
+      "grad_norm": 0.38340896368026733,
+      "learning_rate": 7.635794639566697e-06,
+      "loss": 0.4579,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1736958934517203,
+      "grad_norm": 0.3633081912994385,
+      "learning_rate": 7.631676158913899e-06,
+      "loss": 0.4123,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1745283018867925,
+      "grad_norm": 0.4083158075809479,
+      "learning_rate": 7.627555207071108e-06,
+      "loss": 0.451,
+      "step": 1411
+    },
+    {
+      "epoch": 1.1753607103218646,
+      "grad_norm": 0.3736763894557953,
+      "learning_rate": 7.623431787907971e-06,
+      "loss": 0.4061,
+      "step": 1412
+    },
+    {
+      "epoch": 1.1761931187569368,
+      "grad_norm": 0.3408348858356476,
+      "learning_rate": 7.61930590529645e-06,
+      "loss": 0.4175,
+      "step": 1413
+    },
+    {
+      "epoch": 1.177025527192009,
+      "grad_norm": 0.38261574506759644,
+      "learning_rate": 7.6151775631108245e-06,
+      "loss": 0.4458,
+      "step": 1414
+    },
+    {
+      "epoch": 1.177857935627081,
+      "grad_norm": 0.3570031225681305,
+      "learning_rate": 7.611046765227675e-06,
+      "loss": 0.3869,
+      "step": 1415
+    },
+    {
+      "epoch": 1.1786903440621532,
+      "grad_norm": 0.3958790600299835,
+      "learning_rate": 7.606913515525896e-06,
+      "loss": 0.4224,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1795227524972254,
+      "grad_norm": 0.43983784317970276,
+      "learning_rate": 7.602777817886678e-06,
+      "loss": 0.456,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1803551609322975,
+      "grad_norm": 0.3442942202091217,
+      "learning_rate": 7.59863967619352e-06,
+      "loss": 0.4149,
+      "step": 1418
+    },
+    {
+      "epoch": 1.1811875693673697,
+      "grad_norm": 0.3674313426017761,
+      "learning_rate": 7.594499094332204e-06,
+      "loss": 0.4029,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1820199778024416,
+      "grad_norm": 0.4117499589920044,
+      "learning_rate": 7.59035607619081e-06,
+      "loss": 0.434,
+      "step": 1420
+    },
+    {
+      "epoch": 1.182852386237514,
+      "grad_norm": 0.3603156507015228,
+      "learning_rate": 7.586210625659707e-06,
+      "loss": 0.4318,
+      "step": 1421
+    },
+    {
+      "epoch": 1.183684794672586,
+      "grad_norm": 0.38503581285476685,
+      "learning_rate": 7.582062746631542e-06,
+      "loss": 0.4139,
+      "step": 1422
+    },
+    {
+      "epoch": 1.184517203107658,
+      "grad_norm": 0.43392035365104675,
+      "learning_rate": 7.577912443001247e-06,
+      "loss": 0.4058,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1853496115427302,
+      "grad_norm": 0.3828240633010864,
+      "learning_rate": 7.573759718666031e-06,
+      "loss": 0.4342,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1861820199778024,
+      "grad_norm": 0.39743468165397644,
+      "learning_rate": 7.569604577525376e-06,
+      "loss": 0.4351,
+      "step": 1425
+    },
+    {
+      "epoch": 1.1870144284128745,
+      "grad_norm": 0.417901873588562,
+      "learning_rate": 7.56544702348103e-06,
+      "loss": 0.4358,
+      "step": 1426
+    },
+    {
+      "epoch": 1.1878468368479467,
+      "grad_norm": 0.4021197259426117,
+      "learning_rate": 7.5612870604370106e-06,
+      "loss": 0.4402,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1886792452830188,
+      "grad_norm": 0.32264193892478943,
+      "learning_rate": 7.557124692299593e-06,
+      "loss": 0.3862,
+      "step": 1428
+    },
+    {
+      "epoch": 1.189511653718091,
+      "grad_norm": 0.41591644287109375,
+      "learning_rate": 7.552959922977317e-06,
+      "loss": 0.4142,
+      "step": 1429
+    },
+    {
+      "epoch": 1.1903440621531631,
+      "grad_norm": 0.38096314668655396,
+      "learning_rate": 7.548792756380972e-06,
+      "loss": 0.4072,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1911764705882353,
+      "grad_norm": 0.3876970410346985,
+      "learning_rate": 7.5446231964236025e-06,
+      "loss": 0.4199,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1920088790233074,
+      "grad_norm": 0.36842256784439087,
+      "learning_rate": 7.540451247020495e-06,
+      "loss": 0.4071,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1928412874583796,
+      "grad_norm": 0.3903951644897461,
+      "learning_rate": 7.536276912089187e-06,
+      "loss": 0.4546,
+      "step": 1433
+    },
+    {
+      "epoch": 1.1936736958934517,
+      "grad_norm": 0.38267070055007935,
+      "learning_rate": 7.53210019554945e-06,
+      "loss": 0.4082,
+      "step": 1434
+    },
+    {
+      "epoch": 1.1945061043285239,
+      "grad_norm": 0.3489694893360138,
+      "learning_rate": 7.527921101323292e-06,
+      "loss": 0.4278,
+      "step": 1435
+    },
+    {
+      "epoch": 1.195338512763596,
+      "grad_norm": 0.36168771982192993,
+      "learning_rate": 7.523739633334959e-06,
+      "loss": 0.4372,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1961709211986682,
+      "grad_norm": 0.37979066371917725,
+      "learning_rate": 7.5195557955109225e-06,
+      "loss": 0.4514,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1970033296337403,
+      "grad_norm": 0.3345673382282257,
+      "learning_rate": 7.515369591779876e-06,
+      "loss": 0.4163,
+      "step": 1438
+    },
+    {
+      "epoch": 1.1978357380688125,
+      "grad_norm": 0.3836076855659485,
+      "learning_rate": 7.511181026072741e-06,
+      "loss": 0.4598,
+      "step": 1439
+    },
+    {
+      "epoch": 1.1986681465038846,
+      "grad_norm": 0.37280598282814026,
+      "learning_rate": 7.5069901023226545e-06,
+      "loss": 0.4072,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1995005549389568,
+      "grad_norm": 0.35891035199165344,
+      "learning_rate": 7.502796824464966e-06,
+      "loss": 0.4475,
+      "step": 1441
+    },
+    {
+      "epoch": 1.200332963374029,
+      "grad_norm": 0.3753512501716614,
+      "learning_rate": 7.498601196437238e-06,
+      "loss": 0.4583,
+      "step": 1442
+    },
+    {
+      "epoch": 1.2011653718091009,
+      "grad_norm": 0.33655253052711487,
+      "learning_rate": 7.494403222179235e-06,
+      "loss": 0.399,
+      "step": 1443
+    },
+    {
+      "epoch": 1.2019977802441733,
+      "grad_norm": 0.34640976786613464,
+      "learning_rate": 7.490202905632933e-06,
+      "loss": 0.401,
+      "step": 1444
+    },
+    {
+      "epoch": 1.2028301886792452,
+      "grad_norm": 0.34061723947525024,
+      "learning_rate": 7.4860002507425004e-06,
+      "loss": 0.408,
+      "step": 1445
+    },
+    {
+      "epoch": 1.2036625971143173,
+      "grad_norm": 0.36065739393234253,
+      "learning_rate": 7.481795261454304e-06,
+      "loss": 0.4472,
+      "step": 1446
+    },
+    {
+      "epoch": 1.2044950055493895,
+      "grad_norm": 0.33626216650009155,
+      "learning_rate": 7.477587941716904e-06,
+      "loss": 0.4088,
+      "step": 1447
+    },
+    {
+      "epoch": 1.2053274139844616,
+      "grad_norm": 0.3538858890533447,
+      "learning_rate": 7.4733782954810444e-06,
+      "loss": 0.4513,
+      "step": 1448
+    },
+    {
+      "epoch": 1.2061598224195338,
+      "grad_norm": 0.3653189539909363,
+      "learning_rate": 7.469166326699658e-06,
+      "loss": 0.4268,
+      "step": 1449
+    },
+    {
+      "epoch": 1.206992230854606,
+      "grad_norm": 0.3585223853588104,
+      "learning_rate": 7.4649520393278575e-06,
+      "loss": 0.3983,
+      "step": 1450
+    },
+    {
+      "epoch": 1.207824639289678,
+      "grad_norm": 0.38904839754104614,
+      "learning_rate": 7.460735437322933e-06,
+      "loss": 0.4459,
+      "step": 1451
+    },
+    {
+      "epoch": 1.2086570477247502,
+      "grad_norm": 0.34138140082359314,
+      "learning_rate": 7.456516524644347e-06,
+      "loss": 0.4143,
+      "step": 1452
+    },
+    {
+      "epoch": 1.2094894561598224,
+      "grad_norm": 0.3598451614379883,
+      "learning_rate": 7.452295305253731e-06,
+      "loss": 0.4108,
+      "step": 1453
+    },
+    {
+      "epoch": 1.2103218645948945,
+      "grad_norm": 0.35666725039482117,
+      "learning_rate": 7.448071783114887e-06,
+      "loss": 0.4157,
+      "step": 1454
+    },
+    {
+      "epoch": 1.2111542730299667,
+      "grad_norm": 0.42830565571784973,
+      "learning_rate": 7.443845962193775e-06,
+      "loss": 0.4488,
+      "step": 1455
+    },
+    {
+      "epoch": 1.2119866814650389,
+      "grad_norm": 0.3601066470146179,
+      "learning_rate": 7.439617846458513e-06,
+      "loss": 0.4221,
+      "step": 1456
+    },
+    {
+      "epoch": 1.212819089900111,
+      "grad_norm": 0.36113888025283813,
+      "learning_rate": 7.435387439879378e-06,
+      "loss": 0.368,
+      "step": 1457
+    },
+    {
+      "epoch": 1.2136514983351832,
+      "grad_norm": 0.4079122245311737,
+      "learning_rate": 7.431154746428794e-06,
+      "loss": 0.4567,
+      "step": 1458
+    },
+    {
+      "epoch": 1.2144839067702553,
+      "grad_norm": 0.3279145658016205,
+      "learning_rate": 7.4269197700813375e-06,
+      "loss": 0.401,
+      "step": 1459
+    },
+    {
+      "epoch": 1.2153163152053275,
+      "grad_norm": 0.3636506497859955,
+      "learning_rate": 7.4226825148137225e-06,
+      "loss": 0.4394,
+      "step": 1460
+    },
+    {
+      "epoch": 1.2161487236403996,
+      "grad_norm": 0.3573419153690338,
+      "learning_rate": 7.418442984604805e-06,
+      "loss": 0.3929,
+      "step": 1461
+    },
+    {
+      "epoch": 1.2169811320754718,
+      "grad_norm": 0.3526058793067932,
+      "learning_rate": 7.414201183435581e-06,
+      "loss": 0.4305,
+      "step": 1462
+    },
+    {
+      "epoch": 1.217813540510544,
+      "grad_norm": 0.34084025025367737,
+      "learning_rate": 7.409957115289175e-06,
+      "loss": 0.4018,
+      "step": 1463
+    },
+    {
+      "epoch": 1.218645948945616,
+      "grad_norm": 0.3763788342475891,
+      "learning_rate": 7.40571078415084e-06,
+      "loss": 0.4106,
+      "step": 1464
+    },
+    {
+      "epoch": 1.2194783573806882,
+      "grad_norm": 0.37996524572372437,
+      "learning_rate": 7.401462194007957e-06,
+      "loss": 0.4664,
+      "step": 1465
+    },
+    {
+      "epoch": 1.2203107658157601,
+      "grad_norm": 0.3957565724849701,
+      "learning_rate": 7.397211348850025e-06,
+      "loss": 0.4947,
+      "step": 1466
+    },
+    {
+      "epoch": 1.2211431742508325,
+      "grad_norm": 0.33967241644859314,
+      "learning_rate": 7.392958252668663e-06,
+      "loss": 0.3549,
+      "step": 1467
+    },
+    {
+      "epoch": 1.2219755826859044,
+      "grad_norm": 0.37266120314598083,
+      "learning_rate": 7.388702909457603e-06,
+      "loss": 0.4341,
+      "step": 1468
+    },
+    {
+      "epoch": 1.2228079911209766,
+      "grad_norm": 0.37723714113235474,
+      "learning_rate": 7.384445323212687e-06,
+      "loss": 0.4435,
+      "step": 1469
+    },
+    {
+      "epoch": 1.2236403995560488,
+      "grad_norm": 0.3732614815235138,
+      "learning_rate": 7.380185497931862e-06,
+      "loss": 0.4508,
+      "step": 1470
+    },
+    {
+      "epoch": 1.224472807991121,
+      "grad_norm": 0.35898712277412415,
+      "learning_rate": 7.375923437615179e-06,
+      "loss": 0.4299,
+      "step": 1471
+    },
+    {
+      "epoch": 1.225305216426193,
+      "grad_norm": 0.37605753540992737,
+      "learning_rate": 7.371659146264787e-06,
+      "loss": 0.4495,
+      "step": 1472
+    },
+    {
+      "epoch": 1.2261376248612652,
+      "grad_norm": 0.35970860719680786,
+      "learning_rate": 7.367392627884931e-06,
+      "loss": 0.3909,
+      "step": 1473
+    },
+    {
+      "epoch": 1.2269700332963374,
+      "grad_norm": 0.39131295680999756,
+      "learning_rate": 7.363123886481947e-06,
+      "loss": 0.4308,
+      "step": 1474
+    },
+    {
+      "epoch": 1.2278024417314095,
+      "grad_norm": 0.38385316729545593,
+      "learning_rate": 7.3588529260642564e-06,
+      "loss": 0.4483,
+      "step": 1475
+    },
+    {
+      "epoch": 1.2286348501664817,
+      "grad_norm": 0.35251525044441223,
+      "learning_rate": 7.3545797506423655e-06,
+      "loss": 0.4427,
+      "step": 1476
+    },
+    {
+      "epoch": 1.2294672586015538,
+      "grad_norm": 0.3673636317253113,
+      "learning_rate": 7.3503043642288614e-06,
+      "loss": 0.4065,
+      "step": 1477
+    },
+    {
+      "epoch": 1.230299667036626,
+      "grad_norm": 0.3953494727611542,
+      "learning_rate": 7.3460267708384084e-06,
+      "loss": 0.4341,
+      "step": 1478
+    },
+    {
+      "epoch": 1.2311320754716981,
+      "grad_norm": 0.34204229712486267,
+      "learning_rate": 7.3417469744877375e-06,
+      "loss": 0.4064,
+      "step": 1479
+    },
+    {
+      "epoch": 1.2319644839067703,
+      "grad_norm": 0.35544899106025696,
+      "learning_rate": 7.337464979195658e-06,
+      "loss": 0.4086,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2327968923418424,
+      "grad_norm": 0.37943002581596375,
+      "learning_rate": 7.333180788983034e-06,
+      "loss": 0.4459,
+      "step": 1481
+    },
+    {
+      "epoch": 1.2336293007769146,
+      "grad_norm": 0.33336302638053894,
+      "learning_rate": 7.328894407872797e-06,
+      "loss": 0.409,
+      "step": 1482
+    },
+    {
+      "epoch": 1.2344617092119867,
+      "grad_norm": 0.3566553294658661,
+      "learning_rate": 7.324605839889936e-06,
+      "loss": 0.4377,
+      "step": 1483
+    },
+    {
+      "epoch": 1.2352941176470589,
+      "grad_norm": 0.39889898896217346,
+      "learning_rate": 7.320315089061486e-06,
+      "loss": 0.4603,
+      "step": 1484
+    },
+    {
+      "epoch": 1.236126526082131,
+      "grad_norm": 0.4692656695842743,
+      "learning_rate": 7.3160221594165415e-06,
+      "loss": 0.4108,
+      "step": 1485
+    },
+    {
+      "epoch": 1.2369589345172032,
+      "grad_norm": 0.39253804087638855,
+      "learning_rate": 7.3117270549862385e-06,
+      "loss": 0.4393,
+      "step": 1486
+    },
+    {
+      "epoch": 1.2377913429522753,
+      "grad_norm": 0.4060220718383789,
+      "learning_rate": 7.3074297798037515e-06,
+      "loss": 0.423,
+      "step": 1487
+    },
+    {
+      "epoch": 1.2386237513873475,
+      "grad_norm": 0.32819664478302,
+      "learning_rate": 7.303130337904303e-06,
+      "loss": 0.3855,
+      "step": 1488
+    },
+    {
+      "epoch": 1.2394561598224194,
+      "grad_norm": 0.36876818537712097,
+      "learning_rate": 7.298828733325138e-06,
+      "loss": 0.4056,
+      "step": 1489
+    },
+    {
+      "epoch": 1.2402885682574918,
+      "grad_norm": 0.37898144125938416,
+      "learning_rate": 7.294524970105543e-06,
+      "loss": 0.3985,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2411209766925637,
+      "grad_norm": 0.3846915662288666,
+      "learning_rate": 7.290219052286826e-06,
+      "loss": 0.4752,
+      "step": 1491
+    },
+    {
+      "epoch": 1.2419533851276359,
+      "grad_norm": 0.4225359857082367,
+      "learning_rate": 7.285910983912317e-06,
+      "loss": 0.4152,
+      "step": 1492
+    },
+    {
+      "epoch": 1.242785793562708,
+      "grad_norm": 0.3616348206996918,
+      "learning_rate": 7.281600769027371e-06,
+      "loss": 0.4244,
+      "step": 1493
+    },
+    {
+      "epoch": 1.2436182019977802,
+      "grad_norm": 0.35701385140419006,
+      "learning_rate": 7.277288411679352e-06,
+      "loss": 0.4284,
+      "step": 1494
+    },
+    {
+      "epoch": 1.2444506104328523,
+      "grad_norm": 0.37644630670547485,
+      "learning_rate": 7.272973915917642e-06,
+      "loss": 0.4185,
+      "step": 1495
+    },
+    {
+      "epoch": 1.2452830188679245,
+      "grad_norm": 0.3375144898891449,
+      "learning_rate": 7.268657285793625e-06,
+      "loss": 0.3967,
+      "step": 1496
+    },
+    {
+      "epoch": 1.2461154273029966,
+      "grad_norm": 0.36483046412467957,
+      "learning_rate": 7.264338525360695e-06,
+      "loss": 0.4346,
+      "step": 1497
+    },
+    {
+      "epoch": 1.2469478357380688,
+      "grad_norm": 0.364069402217865,
+      "learning_rate": 7.260017638674244e-06,
+      "loss": 0.4308,
+      "step": 1498
+    },
+    {
+      "epoch": 1.247780244173141,
+      "grad_norm": 0.3628464639186859,
+      "learning_rate": 7.255694629791659e-06,
+      "loss": 0.4345,
+      "step": 1499
+    },
+    {
+      "epoch": 1.248612652608213,
+      "grad_norm": 0.33808740973472595,
+      "learning_rate": 7.251369502772318e-06,
+      "loss": 0.405,
+      "step": 1500
+    },
+    {
+      "epoch": 1.2494450610432852,
+      "grad_norm": 0.4051111042499542,
+      "learning_rate": 7.247042261677597e-06,
+      "loss": 0.4147,
+      "step": 1501
+    },
+    {
+      "epoch": 1.2502774694783574,
+      "grad_norm": 0.34634676575660706,
+      "learning_rate": 7.242712910570846e-06,
+      "loss": 0.4635,
+      "step": 1502
+    },
+    {
+      "epoch": 1.2511098779134295,
+      "grad_norm": 0.42578983306884766,
+      "learning_rate": 7.238381453517405e-06,
+      "loss": 0.4309,
+      "step": 1503
+    },
+    {
+      "epoch": 1.2519422863485017,
+      "grad_norm": 0.38502374291419983,
+      "learning_rate": 7.234047894584586e-06,
+      "loss": 0.4128,
+      "step": 1504
+    },
+    {
+      "epoch": 1.2527746947835738,
+      "grad_norm": 0.37332308292388916,
+      "learning_rate": 7.229712237841679e-06,
+      "loss": 0.4229,
+      "step": 1505
+    },
+    {
+      "epoch": 1.253607103218646,
+      "grad_norm": 0.3640212118625641,
+      "learning_rate": 7.225374487359937e-06,
+      "loss": 0.4239,
+      "step": 1506
+    },
+    {
+      "epoch": 1.2544395116537181,
+      "grad_norm": 0.3998686671257019,
+      "learning_rate": 7.221034647212588e-06,
+      "loss": 0.3917,
+      "step": 1507
+    },
+    {
+      "epoch": 1.2552719200887903,
+      "grad_norm": 0.3780185282230377,
+      "learning_rate": 7.216692721474816e-06,
+      "loss": 0.4168,
+      "step": 1508
+    },
+    {
+      "epoch": 1.2561043285238624,
+      "grad_norm": 0.3652763068675995,
+      "learning_rate": 7.212348714223767e-06,
+      "loss": 0.4179,
+      "step": 1509
+    },
+    {
+      "epoch": 1.2569367369589346,
+      "grad_norm": 0.3468812108039856,
+      "learning_rate": 7.208002629538537e-06,
+      "loss": 0.3956,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2577691453940067,
+      "grad_norm": 0.46192142367362976,
+      "learning_rate": 7.203654471500179e-06,
+      "loss": 0.4355,
+      "step": 1511
+    },
+    {
+      "epoch": 1.2586015538290787,
+      "grad_norm": 0.336422324180603,
+      "learning_rate": 7.199304244191687e-06,
+      "loss": 0.4085,
+      "step": 1512
+    },
+    {
+      "epoch": 1.259433962264151,
+      "grad_norm": 0.3533170521259308,
+      "learning_rate": 7.1949519516980005e-06,
+      "loss": 0.4076,
+      "step": 1513
+    },
+    {
+      "epoch": 1.260266370699223,
+      "grad_norm": 0.34538713097572327,
+      "learning_rate": 7.190597598106001e-06,
+      "loss": 0.3822,
+      "step": 1514
+    },
+    {
+      "epoch": 1.2610987791342954,
+      "grad_norm": 0.35250788927078247,
+      "learning_rate": 7.186241187504499e-06,
+      "loss": 0.4416,
+      "step": 1515
+    },
+    {
+      "epoch": 1.2619311875693673,
+      "grad_norm": 0.38682812452316284,
+      "learning_rate": 7.1818827239842446e-06,
+      "loss": 0.4295,
+      "step": 1516
+    },
+    {
+      "epoch": 1.2627635960044394,
+      "grad_norm": 0.3738081455230713,
+      "learning_rate": 7.177522211637906e-06,
+      "loss": 0.4203,
+      "step": 1517
+    },
+    {
+      "epoch": 1.2635960044395116,
+      "grad_norm": 0.39243027567863464,
+      "learning_rate": 7.173159654560087e-06,
+      "loss": 0.4994,
+      "step": 1518
+    },
+    {
+      "epoch": 1.2644284128745837,
+      "grad_norm": 0.337700754404068,
+      "learning_rate": 7.168795056847301e-06,
+      "loss": 0.407,
+      "step": 1519
+    },
+    {
+      "epoch": 1.265260821309656,
+      "grad_norm": 0.4204353094100952,
+      "learning_rate": 7.164428422597982e-06,
+      "loss": 0.4189,
+      "step": 1520
+    },
+    {
+      "epoch": 1.266093229744728,
+      "grad_norm": 0.38597655296325684,
+      "learning_rate": 7.1600597559124765e-06,
+      "loss": 0.4476,
+      "step": 1521
+    },
+    {
+      "epoch": 1.2669256381798002,
+      "grad_norm": 0.32296887040138245,
+      "learning_rate": 7.155689060893038e-06,
+      "loss": 0.3669,
+      "step": 1522
+    },
+    {
+      "epoch": 1.2677580466148723,
+      "grad_norm": 0.35379377007484436,
+      "learning_rate": 7.151316341643828e-06,
+      "loss": 0.4346,
+      "step": 1523
+    },
+    {
+      "epoch": 1.2685904550499445,
+      "grad_norm": 0.5642543435096741,
+      "learning_rate": 7.146941602270905e-06,
+      "loss": 0.4237,
+      "step": 1524
+    },
+    {
+      "epoch": 1.2694228634850167,
+      "grad_norm": 0.4288339614868164,
+      "learning_rate": 7.142564846882227e-06,
+      "loss": 0.3797,
+      "step": 1525
+    },
+    {
+      "epoch": 1.2702552719200888,
+      "grad_norm": 0.36099711060523987,
+      "learning_rate": 7.1381860795876415e-06,
+      "loss": 0.4519,
+      "step": 1526
+    },
+    {
+      "epoch": 1.271087680355161,
+      "grad_norm": 0.4087908864021301,
+      "learning_rate": 7.13380530449889e-06,
+      "loss": 0.4499,
+      "step": 1527
+    },
+    {
+      "epoch": 1.271920088790233,
+      "grad_norm": 0.37486109137535095,
+      "learning_rate": 7.129422525729594e-06,
+      "loss": 0.4141,
+      "step": 1528
+    },
+    {
+      "epoch": 1.2727524972253053,
+      "grad_norm": 0.3277265429496765,
+      "learning_rate": 7.125037747395264e-06,
+      "loss": 0.3747,
+      "step": 1529
+    },
+    {
+      "epoch": 1.2735849056603774,
+      "grad_norm": 0.36414340138435364,
+      "learning_rate": 7.120650973613279e-06,
+      "loss": 0.4181,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2744173140954496,
+      "grad_norm": 0.3803533911705017,
+      "learning_rate": 7.116262208502901e-06,
+      "loss": 0.429,
+      "step": 1531
+    },
+    {
+      "epoch": 1.2752497225305217,
+      "grad_norm": 0.3974792957305908,
+      "learning_rate": 7.111871456185253e-06,
+      "loss": 0.4555,
+      "step": 1532
+    },
+    {
+      "epoch": 1.2760821309655939,
+      "grad_norm": 0.35720840096473694,
+      "learning_rate": 7.107478720783332e-06,
+      "loss": 0.4415,
+      "step": 1533
+    },
+    {
+      "epoch": 1.276914539400666,
+      "grad_norm": 0.4027242362499237,
+      "learning_rate": 7.1030840064219906e-06,
+      "loss": 0.4108,
+      "step": 1534
+    },
+    {
+      "epoch": 1.277746947835738,
+      "grad_norm": 0.38121697306632996,
+      "learning_rate": 7.098687317227943e-06,
+      "loss": 0.4387,
+      "step": 1535
+    },
+    {
+      "epoch": 1.2785793562708103,
+      "grad_norm": 0.37340235710144043,
+      "learning_rate": 7.09428865732976e-06,
+      "loss": 0.4174,
+      "step": 1536
+    },
+    {
+      "epoch": 1.2794117647058822,
+      "grad_norm": 0.4037769138813019,
+      "learning_rate": 7.089888030857857e-06,
+      "loss": 0.4613,
+      "step": 1537
+    },
+    {
+      "epoch": 1.2802441731409546,
+      "grad_norm": 0.33645448088645935,
+      "learning_rate": 7.0854854419445e-06,
+      "loss": 0.3608,
+      "step": 1538
+    },
+    {
+      "epoch": 1.2810765815760266,
+      "grad_norm": 0.4041195511817932,
+      "learning_rate": 7.0810808947237975e-06,
+      "loss": 0.4305,
+      "step": 1539
+    },
+    {
+      "epoch": 1.2819089900110987,
+      "grad_norm": 0.34097573161125183,
+      "learning_rate": 7.076674393331697e-06,
+      "loss": 0.4156,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2827413984461709,
+      "grad_norm": 0.40564876794815063,
+      "learning_rate": 7.0722659419059806e-06,
+      "loss": 0.4365,
+      "step": 1541
+    },
+    {
+      "epoch": 1.283573806881243,
+      "grad_norm": 0.39337435364723206,
+      "learning_rate": 7.0678555445862605e-06,
+      "loss": 0.4482,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2844062153163152,
+      "grad_norm": 0.3427252173423767,
+      "learning_rate": 7.063443205513975e-06,
+      "loss": 0.3806,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2852386237513873,
+      "grad_norm": 0.4486899673938751,
+      "learning_rate": 7.059028928832394e-06,
+      "loss": 0.4238,
+      "step": 1544
+    },
+    {
+      "epoch": 1.2860710321864595,
+      "grad_norm": 0.37656182050704956,
+      "learning_rate": 7.054612718686593e-06,
+      "loss": 0.4165,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2869034406215316,
+      "grad_norm": 0.37508389353752136,
+      "learning_rate": 7.0501945792234776e-06,
+      "loss": 0.4368,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2877358490566038,
+      "grad_norm": 0.385955274105072,
+      "learning_rate": 7.045774514591753e-06,
+      "loss": 0.4263,
+      "step": 1547
+    },
+    {
+      "epoch": 1.288568257491676,
+      "grad_norm": 0.3990999162197113,
+      "learning_rate": 7.041352528941939e-06,
+      "loss": 0.4219,
+      "step": 1548
+    },
+    {
+      "epoch": 1.289400665926748,
+      "grad_norm": 0.43571391701698303,
+      "learning_rate": 7.036928626426358e-06,
+      "loss": 0.4525,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2902330743618202,
+      "grad_norm": 0.40208232402801514,
+      "learning_rate": 7.0325028111991325e-06,
+      "loss": 0.4166,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2910654827968924,
+      "grad_norm": 0.5420854687690735,
+      "learning_rate": 7.02807508741618e-06,
+      "loss": 0.4908,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2918978912319645,
+      "grad_norm": 0.3404446840286255,
+      "learning_rate": 7.0236454592352065e-06,
+      "loss": 0.3513,
+      "step": 1552
+    },
+    {
+      "epoch": 1.2927302996670367,
+      "grad_norm": 0.37210413813591003,
+      "learning_rate": 7.019213930815718e-06,
+      "loss": 0.4292,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2935627081021088,
+      "grad_norm": 0.4521826207637787,
+      "learning_rate": 7.01478050631899e-06,
+      "loss": 0.4314,
+      "step": 1554
+    },
+    {
+      "epoch": 1.294395116537181,
+      "grad_norm": 0.3963296413421631,
+      "learning_rate": 7.010345189908092e-06,
+      "loss": 0.4345,
+      "step": 1555
+    },
+    {
+      "epoch": 1.2952275249722531,
+      "grad_norm": 0.3517422378063202,
+      "learning_rate": 7.0059079857478596e-06,
+      "loss": 0.4088,
+      "step": 1556
+    },
+    {
+      "epoch": 1.2960599334073253,
+      "grad_norm": 0.4621788263320923,
+      "learning_rate": 7.001468898004907e-06,
+      "loss": 0.4385,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2968923418423972,
+      "grad_norm": 0.41421958804130554,
+      "learning_rate": 6.997027930847614e-06,
+      "loss": 0.4428,
+      "step": 1558
+    },
+    {
+      "epoch": 1.2977247502774696,
+      "grad_norm": 0.32892701029777527,
+      "learning_rate": 6.992585088446129e-06,
+      "loss": 0.4213,
+      "step": 1559
+    },
+    {
+      "epoch": 1.2985571587125415,
+      "grad_norm": 0.37421393394470215,
+      "learning_rate": 6.988140374972357e-06,
+      "loss": 0.3801,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2993895671476139,
+      "grad_norm": 0.4647720456123352,
+      "learning_rate": 6.983693794599959e-06,
+      "loss": 0.4472,
+      "step": 1561
+    },
+    {
+      "epoch": 1.3002219755826858,
+      "grad_norm": 0.3581669330596924,
+      "learning_rate": 6.979245351504358e-06,
+      "loss": 0.4224,
+      "step": 1562
+    },
+    {
+      "epoch": 1.301054384017758,
+      "grad_norm": 0.37526553869247437,
+      "learning_rate": 6.974795049862715e-06,
+      "loss": 0.4071,
+      "step": 1563
+    },
+    {
+      "epoch": 1.3018867924528301,
+      "grad_norm": 0.3938126266002655,
+      "learning_rate": 6.970342893853943e-06,
+      "loss": 0.4474,
+      "step": 1564
+    },
+    {
+      "epoch": 1.3027192008879023,
+      "grad_norm": 0.33351486921310425,
+      "learning_rate": 6.965888887658695e-06,
+      "loss": 0.3736,
+      "step": 1565
+    },
+    {
+      "epoch": 1.3035516093229744,
+      "grad_norm": 0.4295842945575714,
+      "learning_rate": 6.961433035459361e-06,
+      "loss": 0.4689,
+      "step": 1566
+    },
+    {
+      "epoch": 1.3043840177580466,
+      "grad_norm": 0.3872586786746979,
+      "learning_rate": 6.956975341440061e-06,
+      "loss": 0.4328,
+      "step": 1567
+    },
+    {
+      "epoch": 1.3052164261931187,
+      "grad_norm": 0.3561748266220093,
+      "learning_rate": 6.952515809786652e-06,
+      "loss": 0.4492,
+      "step": 1568
+    },
+    {
+      "epoch": 1.3060488346281909,
+      "grad_norm": 0.41043657064437866,
+      "learning_rate": 6.948054444686709e-06,
+      "loss": 0.4037,
+      "step": 1569
+    },
+    {
+      "epoch": 1.306881243063263,
+      "grad_norm": 0.4166451394557953,
+      "learning_rate": 6.943591250329534e-06,
+      "loss": 0.4192,
+      "step": 1570
+    },
+    {
+      "epoch": 1.3077136514983352,
+      "grad_norm": 0.37019088864326477,
+      "learning_rate": 6.939126230906144e-06,
+      "loss": 0.4187,
+      "step": 1571
+    },
+    {
+      "epoch": 1.3085460599334073,
+      "grad_norm": 0.3775978684425354,
+      "learning_rate": 6.934659390609271e-06,
+      "loss": 0.4589,
+      "step": 1572
+    },
+    {
+      "epoch": 1.3093784683684795,
+      "grad_norm": 0.4076300859451294,
+      "learning_rate": 6.930190733633355e-06,
+      "loss": 0.437,
+      "step": 1573
+    },
+    {
+      "epoch": 1.3102108768035516,
+      "grad_norm": 0.36930978298187256,
+      "learning_rate": 6.925720264174543e-06,
+      "loss": 0.3679,
+      "step": 1574
+    },
+    {
+      "epoch": 1.3110432852386238,
+      "grad_norm": 0.40303748846054077,
+      "learning_rate": 6.921247986430686e-06,
+      "loss": 0.4646,
+      "step": 1575
+    },
+    {
+      "epoch": 1.311875693673696,
+      "grad_norm": 0.3812013566493988,
+      "learning_rate": 6.9167739046013305e-06,
+      "loss": 0.4285,
+      "step": 1576
+    },
+    {
+      "epoch": 1.312708102108768,
+      "grad_norm": 0.4709984064102173,
+      "learning_rate": 6.912298022887716e-06,
+      "loss": 0.4492,
+      "step": 1577
+    },
+    {
+      "epoch": 1.3135405105438402,
+      "grad_norm": 0.33343759179115295,
+      "learning_rate": 6.907820345492775e-06,
+      "loss": 0.3764,
+      "step": 1578
+    },
+    {
+      "epoch": 1.3143729189789124,
+      "grad_norm": 0.4059264361858368,
+      "learning_rate": 6.903340876621125e-06,
+      "loss": 0.4234,
+      "step": 1579
+    },
+    {
+      "epoch": 1.3152053274139845,
+      "grad_norm": 0.39759019017219543,
+      "learning_rate": 6.8988596204790655e-06,
+      "loss": 0.4593,
+      "step": 1580
+    },
+    {
+      "epoch": 1.3160377358490565,
+      "grad_norm": 0.3702462315559387,
+      "learning_rate": 6.894376581274578e-06,
+      "loss": 0.4284,
+      "step": 1581
+    },
+    {
+      "epoch": 1.3168701442841289,
+      "grad_norm": 0.3988160490989685,
+      "learning_rate": 6.889891763217307e-06,
+      "loss": 0.4043,
+      "step": 1582
+    },
+    {
+      "epoch": 1.3177025527192008,
+      "grad_norm": 0.3909897208213806,
+      "learning_rate": 6.8854051705185825e-06,
+      "loss": 0.4282,
+      "step": 1583
+    },
+    {
+      "epoch": 1.3185349611542732,
+      "grad_norm": 0.37735849618911743,
+      "learning_rate": 6.880916807391388e-06,
+      "loss": 0.4185,
+      "step": 1584
+    },
+    {
+      "epoch": 1.319367369589345,
+      "grad_norm": 0.4028033912181854,
+      "learning_rate": 6.876426678050379e-06,
+      "loss": 0.427,
+      "step": 1585
+    },
+    {
+      "epoch": 1.3201997780244172,
+      "grad_norm": 0.3585635721683502,
+      "learning_rate": 6.871934786711866e-06,
+      "loss": 0.401,
+      "step": 1586
+    },
+    {
+      "epoch": 1.3210321864594894,
+      "grad_norm": 0.36207467317581177,
+      "learning_rate": 6.86744113759381e-06,
+      "loss": 0.4119,
+      "step": 1587
+    },
+    {
+      "epoch": 1.3218645948945615,
+      "grad_norm": 0.3779536485671997,
+      "learning_rate": 6.862945734915829e-06,
+      "loss": 0.4531,
+      "step": 1588
+    },
+    {
+      "epoch": 1.3226970033296337,
+      "grad_norm": 0.38031184673309326,
+      "learning_rate": 6.858448582899183e-06,
+      "loss": 0.4332,
+      "step": 1589
+    },
+    {
+      "epoch": 1.3235294117647058,
+      "grad_norm": 0.3429410755634308,
+      "learning_rate": 6.8539496857667785e-06,
+      "loss": 0.3944,
+      "step": 1590
+    },
+    {
+      "epoch": 1.324361820199778,
+      "grad_norm": 0.3546883463859558,
+      "learning_rate": 6.849449047743158e-06,
+      "loss": 0.4342,
+      "step": 1591
+    },
+    {
+      "epoch": 1.3251942286348501,
+      "grad_norm": 0.39290136098861694,
+      "learning_rate": 6.844946673054498e-06,
+      "loss": 0.431,
+      "step": 1592
+    },
+    {
+      "epoch": 1.3260266370699223,
+      "grad_norm": 0.3893829584121704,
+      "learning_rate": 6.840442565928609e-06,
+      "loss": 0.4844,
+      "step": 1593
+    },
+    {
+      "epoch": 1.3268590455049944,
+      "grad_norm": 0.344115287065506,
+      "learning_rate": 6.8359367305949256e-06,
+      "loss": 0.4036,
+      "step": 1594
+    },
+    {
+      "epoch": 1.3276914539400666,
+      "grad_norm": 0.3980282247066498,
+      "learning_rate": 6.831429171284506e-06,
+      "loss": 0.4888,
+      "step": 1595
+    },
+    {
+      "epoch": 1.3285238623751388,
+      "grad_norm": 0.36405783891677856,
+      "learning_rate": 6.8269198922300274e-06,
+      "loss": 0.4067,
+      "step": 1596
+    },
+    {
+      "epoch": 1.329356270810211,
+      "grad_norm": 0.3426407277584076,
+      "learning_rate": 6.822408897665782e-06,
+      "loss": 0.3926,
+      "step": 1597
+    },
+    {
+      "epoch": 1.330188679245283,
+      "grad_norm": 0.35977932810783386,
+      "learning_rate": 6.817896191827673e-06,
+      "loss": 0.4423,
+      "step": 1598
+    },
+    {
+      "epoch": 1.3310210876803552,
+      "grad_norm": 0.34656643867492676,
+      "learning_rate": 6.81338177895321e-06,
+      "loss": 0.4234,
+      "step": 1599
+    },
+    {
+      "epoch": 1.3318534961154274,
+      "grad_norm": 0.4227568507194519,
+      "learning_rate": 6.808865663281504e-06,
+      "loss": 0.4866,
+      "step": 1600
+    },
+    {
+      "epoch": 1.3326859045504995,
+      "grad_norm": 0.3399163782596588,
+      "learning_rate": 6.8043478490532695e-06,
+      "loss": 0.4099,
+      "step": 1601
+    },
+    {
+      "epoch": 1.3335183129855717,
+      "grad_norm": 0.37561044096946716,
+      "learning_rate": 6.799828340510811e-06,
+      "loss": 0.4149,
+      "step": 1602
+    },
+    {
+      "epoch": 1.3343507214206438,
+      "grad_norm": 0.40371450781822205,
+      "learning_rate": 6.795307141898027e-06,
+      "loss": 0.3866,
+      "step": 1603
+    },
+    {
+      "epoch": 1.3351831298557157,
+      "grad_norm": 0.4308793842792511,
+      "learning_rate": 6.790784257460403e-06,
+      "loss": 0.4635,
+      "step": 1604
+    },
+    {
+      "epoch": 1.3360155382907881,
+      "grad_norm": 0.33165696263313293,
+      "learning_rate": 6.786259691445005e-06,
+      "loss": 0.3694,
+      "step": 1605
+    },
+    {
+      "epoch": 1.33684794672586,
+      "grad_norm": 0.37500032782554626,
+      "learning_rate": 6.781733448100482e-06,
+      "loss": 0.4279,
+      "step": 1606
+    },
+    {
+      "epoch": 1.3376803551609324,
+      "grad_norm": 0.3972361981868744,
+      "learning_rate": 6.777205531677052e-06,
+      "loss": 0.4096,
+      "step": 1607
+    },
+    {
+      "epoch": 1.3385127635960044,
+      "grad_norm": 0.41537413001060486,
+      "learning_rate": 6.772675946426511e-06,
+      "loss": 0.4562,
+      "step": 1608
+    },
+    {
+      "epoch": 1.3393451720310765,
+      "grad_norm": 0.3932049870491028,
+      "learning_rate": 6.768144696602219e-06,
+      "loss": 0.4244,
+      "step": 1609
+    },
+    {
+      "epoch": 1.3401775804661487,
+      "grad_norm": 0.3874600827693939,
+      "learning_rate": 6.763611786459097e-06,
+      "loss": 0.4485,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3410099889012208,
+      "grad_norm": 0.3644678294658661,
+      "learning_rate": 6.759077220253628e-06,
+      "loss": 0.3804,
+      "step": 1611
+    },
+    {
+      "epoch": 1.341842397336293,
+      "grad_norm": 0.36026453971862793,
+      "learning_rate": 6.7545410022438495e-06,
+      "loss": 0.4498,
+      "step": 1612
+    },
+    {
+      "epoch": 1.342674805771365,
+      "grad_norm": 0.3394271433353424,
+      "learning_rate": 6.750003136689349e-06,
+      "loss": 0.4059,
+      "step": 1613
+    },
+    {
+      "epoch": 1.3435072142064373,
+      "grad_norm": 0.3929471969604492,
+      "learning_rate": 6.745463627851261e-06,
+      "loss": 0.4259,
+      "step": 1614
+    },
+    {
+      "epoch": 1.3443396226415094,
+      "grad_norm": 0.34686771035194397,
+      "learning_rate": 6.740922479992264e-06,
+      "loss": 0.4158,
+      "step": 1615
+    },
+    {
+      "epoch": 1.3451720310765816,
+      "grad_norm": 0.37497010827064514,
+      "learning_rate": 6.736379697376578e-06,
+      "loss": 0.3876,
+      "step": 1616
+    },
+    {
+      "epoch": 1.3460044395116537,
+      "grad_norm": 0.4034978449344635,
+      "learning_rate": 6.731835284269952e-06,
+      "loss": 0.4623,
+      "step": 1617
+    },
+    {
+      "epoch": 1.3468368479467259,
+      "grad_norm": 0.37284499406814575,
+      "learning_rate": 6.727289244939671e-06,
+      "loss": 0.3982,
+      "step": 1618
+    },
+    {
+      "epoch": 1.347669256381798,
+      "grad_norm": 0.37045595049858093,
+      "learning_rate": 6.722741583654545e-06,
+      "loss": 0.4012,
+      "step": 1619
+    },
+    {
+      "epoch": 1.3485016648168702,
+      "grad_norm": 0.39600858092308044,
+      "learning_rate": 6.718192304684909e-06,
+      "loss": 0.4241,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3493340732519423,
+      "grad_norm": 0.35542434453964233,
+      "learning_rate": 6.713641412302614e-06,
+      "loss": 0.4276,
+      "step": 1621
+    },
+    {
+      "epoch": 1.3501664816870145,
+      "grad_norm": 0.40306901931762695,
+      "learning_rate": 6.7090889107810275e-06,
+      "loss": 0.4232,
+      "step": 1622
+    },
+    {
+      "epoch": 1.3509988901220866,
+      "grad_norm": 0.4040415287017822,
+      "learning_rate": 6.704534804395029e-06,
+      "loss": 0.428,
+      "step": 1623
+    },
+    {
+      "epoch": 1.3518312985571588,
+      "grad_norm": 0.41643235087394714,
+      "learning_rate": 6.699979097421004e-06,
+      "loss": 0.4166,
+      "step": 1624
+    },
+    {
+      "epoch": 1.352663706992231,
+      "grad_norm": 0.38821861147880554,
+      "learning_rate": 6.695421794136843e-06,
+      "loss": 0.4237,
+      "step": 1625
+    },
+    {
+      "epoch": 1.353496115427303,
+      "grad_norm": 0.42019572854042053,
+      "learning_rate": 6.690862898821928e-06,
+      "loss": 0.4018,
+      "step": 1626
+    },
+    {
+      "epoch": 1.354328523862375,
+      "grad_norm": 0.41116052865982056,
+      "learning_rate": 6.686302415757149e-06,
+      "loss": 0.4343,
+      "step": 1627
+    },
+    {
+      "epoch": 1.3551609322974474,
+      "grad_norm": 0.346463143825531,
+      "learning_rate": 6.681740349224873e-06,
+      "loss": 0.4361,
+      "step": 1628
+    },
+    {
+      "epoch": 1.3559933407325193,
+      "grad_norm": 0.4325784742832184,
+      "learning_rate": 6.677176703508963e-06,
+      "loss": 0.4472,
+      "step": 1629
+    },
+    {
+      "epoch": 1.3568257491675917,
+      "grad_norm": 0.3932659924030304,
+      "learning_rate": 6.672611482894763e-06,
+      "loss": 0.4421,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3576581576026636,
+      "grad_norm": 0.3726898729801178,
+      "learning_rate": 6.668044691669094e-06,
+      "loss": 0.4374,
+      "step": 1631
+    },
+    {
+      "epoch": 1.3584905660377358,
+      "grad_norm": 0.3914749324321747,
+      "learning_rate": 6.663476334120254e-06,
+      "loss": 0.4275,
+      "step": 1632
+    },
+    {
+      "epoch": 1.359322974472808,
+      "grad_norm": 0.3771442174911499,
+      "learning_rate": 6.658906414538009e-06,
+      "loss": 0.4356,
+      "step": 1633
+    },
+    {
+      "epoch": 1.36015538290788,
+      "grad_norm": 0.41853395104408264,
+      "learning_rate": 6.6543349372135946e-06,
+      "loss": 0.4489,
+      "step": 1634
+    },
+    {
+      "epoch": 1.3609877913429522,
+      "grad_norm": 0.3918192982673645,
+      "learning_rate": 6.649761906439708e-06,
+      "loss": 0.4469,
+      "step": 1635
+    },
+    {
+      "epoch": 1.3618201997780244,
+      "grad_norm": 0.3729172348976135,
+      "learning_rate": 6.6451873265105045e-06,
+      "loss": 0.4394,
+      "step": 1636
+    },
+    {
+      "epoch": 1.3626526082130965,
+      "grad_norm": 0.45446979999542236,
+      "learning_rate": 6.6406112017215966e-06,
+      "loss": 0.4286,
+      "step": 1637
+    },
+    {
+      "epoch": 1.3634850166481687,
+      "grad_norm": 0.4050742983818054,
+      "learning_rate": 6.6360335363700435e-06,
+      "loss": 0.4214,
+      "step": 1638
+    },
+    {
+      "epoch": 1.3643174250832408,
+      "grad_norm": 0.3428737223148346,
+      "learning_rate": 6.631454334754353e-06,
+      "loss": 0.4248,
+      "step": 1639
+    },
+    {
+      "epoch": 1.365149833518313,
+      "grad_norm": 0.44034427404403687,
+      "learning_rate": 6.626873601174478e-06,
+      "loss": 0.4406,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3659822419533851,
+      "grad_norm": 0.38823384046554565,
+      "learning_rate": 6.622291339931806e-06,
+      "loss": 0.4313,
+      "step": 1641
+    },
+    {
+      "epoch": 1.3668146503884573,
+      "grad_norm": 0.36457860469818115,
+      "learning_rate": 6.61770755532916e-06,
+      "loss": 0.4313,
+      "step": 1642
+    },
+    {
+      "epoch": 1.3676470588235294,
+      "grad_norm": 0.3769315481185913,
+      "learning_rate": 6.613122251670795e-06,
+      "loss": 0.4295,
+      "step": 1643
+    },
+    {
+      "epoch": 1.3684794672586016,
+      "grad_norm": 0.35854166746139526,
+      "learning_rate": 6.608535433262391e-06,
+      "loss": 0.4386,
+      "step": 1644
+    },
+    {
+      "epoch": 1.3693118756936737,
+      "grad_norm": 0.33545204997062683,
+      "learning_rate": 6.60394710441105e-06,
+      "loss": 0.3737,
+      "step": 1645
+    },
+    {
+      "epoch": 1.370144284128746,
+      "grad_norm": 0.37274548411369324,
+      "learning_rate": 6.599357269425294e-06,
+      "loss": 0.4362,
+      "step": 1646
+    },
+    {
+      "epoch": 1.370976692563818,
+      "grad_norm": 0.3416080176830292,
+      "learning_rate": 6.594765932615059e-06,
+      "loss": 0.4346,
+      "step": 1647
+    },
+    {
+      "epoch": 1.3718091009988902,
+      "grad_norm": 0.3786942660808563,
+      "learning_rate": 6.59017309829169e-06,
+      "loss": 0.4237,
+      "step": 1648
+    },
+    {
+      "epoch": 1.3726415094339623,
+      "grad_norm": 0.36071640253067017,
+      "learning_rate": 6.585578770767939e-06,
+      "loss": 0.4231,
+      "step": 1649
+    },
+    {
+      "epoch": 1.3734739178690343,
+      "grad_norm": 0.3784758746623993,
+      "learning_rate": 6.5809829543579595e-06,
+      "loss": 0.458,
+      "step": 1650
+    },
+    {
+      "epoch": 1.3743063263041067,
+      "grad_norm": 0.31035900115966797,
+      "learning_rate": 6.576385653377303e-06,
+      "loss": 0.3688,
+      "step": 1651
+    },
+    {
+      "epoch": 1.3751387347391786,
+      "grad_norm": 0.3701719641685486,
+      "learning_rate": 6.5717868721429175e-06,
+      "loss": 0.4114,
+      "step": 1652
+    },
+    {
+      "epoch": 1.375971143174251,
+      "grad_norm": 0.42909175157546997,
+      "learning_rate": 6.56718661497314e-06,
+      "loss": 0.4538,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3768035516093229,
+      "grad_norm": 0.3627909719944,
+      "learning_rate": 6.562584886187687e-06,
+      "loss": 0.4097,
+      "step": 1654
+    },
+    {
+      "epoch": 1.377635960044395,
+      "grad_norm": 0.36822018027305603,
+      "learning_rate": 6.557981690107669e-06,
+      "loss": 0.4246,
+      "step": 1655
+    },
+    {
+      "epoch": 1.3784683684794672,
+      "grad_norm": 0.3948631286621094,
+      "learning_rate": 6.553377031055564e-06,
+      "loss": 0.4232,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3793007769145393,
+      "grad_norm": 0.4063318371772766,
+      "learning_rate": 6.5487709133552275e-06,
+      "loss": 0.4494,
+      "step": 1657
+    },
+    {
+      "epoch": 1.3801331853496115,
+      "grad_norm": 0.36222216486930847,
+      "learning_rate": 6.544163341331886e-06,
+      "loss": 0.3822,
+      "step": 1658
+    },
+    {
+      "epoch": 1.3809655937846836,
+      "grad_norm": 0.37411925196647644,
+      "learning_rate": 6.539554319312129e-06,
+      "loss": 0.4213,
+      "step": 1659
+    },
+    {
+      "epoch": 1.3817980022197558,
+      "grad_norm": 0.39759203791618347,
+      "learning_rate": 6.534943851623911e-06,
+      "loss": 0.4124,
+      "step": 1660
+    },
+    {
+      "epoch": 1.382630410654828,
+      "grad_norm": 0.3889487087726593,
+      "learning_rate": 6.530331942596539e-06,
+      "loss": 0.4449,
+      "step": 1661
+    },
+    {
+      "epoch": 1.3834628190899,
+      "grad_norm": 0.3533934950828552,
+      "learning_rate": 6.525718596560679e-06,
+      "loss": 0.4068,
+      "step": 1662
+    },
+    {
+      "epoch": 1.3842952275249722,
+      "grad_norm": 0.4298970699310303,
+      "learning_rate": 6.521103817848342e-06,
+      "loss": 0.4447,
+      "step": 1663
+    },
+    {
+      "epoch": 1.3851276359600444,
+      "grad_norm": 0.3471847176551819,
+      "learning_rate": 6.516487610792888e-06,
+      "loss": 0.4082,
+      "step": 1664
+    },
+    {
+      "epoch": 1.3859600443951166,
+      "grad_norm": 0.3483141362667084,
+      "learning_rate": 6.511869979729013e-06,
+      "loss": 0.4342,
+      "step": 1665
+    },
+    {
+      "epoch": 1.3867924528301887,
+      "grad_norm": 0.34657421708106995,
+      "learning_rate": 6.507250928992757e-06,
+      "loss": 0.3625,
+      "step": 1666
+    },
+    {
+      "epoch": 1.3876248612652609,
+      "grad_norm": 0.3735605478286743,
+      "learning_rate": 6.5026304629214846e-06,
+      "loss": 0.4006,
+      "step": 1667
+    },
+    {
+      "epoch": 1.388457269700333,
+      "grad_norm": 0.35017845034599304,
+      "learning_rate": 6.498008585853901e-06,
+      "loss": 0.4373,
+      "step": 1668
+    },
+    {
+      "epoch": 1.3892896781354052,
+      "grad_norm": 0.3881807029247284,
+      "learning_rate": 6.493385302130023e-06,
+      "loss": 0.3704,
+      "step": 1669
+    },
+    {
+      "epoch": 1.3901220865704773,
+      "grad_norm": 0.3898671567440033,
+      "learning_rate": 6.488760616091201e-06,
+      "loss": 0.4166,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3909544950055495,
+      "grad_norm": 0.3844543695449829,
+      "learning_rate": 6.484134532080091e-06,
+      "loss": 0.444,
+      "step": 1671
+    },
+    {
+      "epoch": 1.3917869034406216,
+      "grad_norm": 0.3733096718788147,
+      "learning_rate": 6.479507054440671e-06,
+      "loss": 0.4312,
+      "step": 1672
+    },
+    {
+      "epoch": 1.3926193118756935,
+      "grad_norm": 0.3414270281791687,
+      "learning_rate": 6.474878187518221e-06,
+      "loss": 0.4285,
+      "step": 1673
+    },
+    {
+      "epoch": 1.393451720310766,
+      "grad_norm": 0.3758377134799957,
+      "learning_rate": 6.470247935659328e-06,
+      "loss": 0.4341,
+      "step": 1674
+    },
+    {
+      "epoch": 1.3942841287458378,
+      "grad_norm": 0.35513973236083984,
+      "learning_rate": 6.465616303211881e-06,
+      "loss": 0.4031,
+      "step": 1675
+    },
+    {
+      "epoch": 1.3951165371809102,
+      "grad_norm": 0.36078712344169617,
+      "learning_rate": 6.460983294525064e-06,
+      "loss": 0.4193,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3959489456159822,
+      "grad_norm": 0.3634653091430664,
+      "learning_rate": 6.456348913949352e-06,
+      "loss": 0.4201,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3967813540510543,
+      "grad_norm": 0.36076661944389343,
+      "learning_rate": 6.451713165836511e-06,
+      "loss": 0.3709,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3976137624861265,
+      "grad_norm": 0.3891527056694031,
+      "learning_rate": 6.447076054539588e-06,
+      "loss": 0.4665,
+      "step": 1679
+    },
+    {
+      "epoch": 1.3984461709211986,
+      "grad_norm": 0.32691043615341187,
+      "learning_rate": 6.442437584412912e-06,
+      "loss": 0.3864,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3992785793562708,
+      "grad_norm": 0.37780606746673584,
+      "learning_rate": 6.43779775981209e-06,
+      "loss": 0.4385,
+      "step": 1681
+    },
+    {
+      "epoch": 1.400110987791343,
+      "grad_norm": 0.3391205966472626,
+      "learning_rate": 6.433156585093994e-06,
+      "loss": 0.4281,
+      "step": 1682
+    },
+    {
+      "epoch": 1.400943396226415,
+      "grad_norm": 0.32850903272628784,
+      "learning_rate": 6.4285140646167735e-06,
+      "loss": 0.4434,
+      "step": 1683
+    },
+    {
+      "epoch": 1.4017758046614872,
+      "grad_norm": 0.3246311843395233,
+      "learning_rate": 6.423870202739831e-06,
+      "loss": 0.3899,
+      "step": 1684
+    },
+    {
+      "epoch": 1.4026082130965594,
+      "grad_norm": 0.36959534883499146,
+      "learning_rate": 6.41922500382384e-06,
+      "loss": 0.4113,
+      "step": 1685
+    },
+    {
+      "epoch": 1.4034406215316315,
+      "grad_norm": 0.3844015598297119,
+      "learning_rate": 6.414578472230719e-06,
+      "loss": 0.4143,
+      "step": 1686
+    },
+    {
+      "epoch": 1.4042730299667037,
+      "grad_norm": 0.3275579512119293,
+      "learning_rate": 6.409930612323646e-06,
+      "loss": 0.4199,
+      "step": 1687
+    },
+    {
+      "epoch": 1.4051054384017758,
+      "grad_norm": 0.388668030500412,
+      "learning_rate": 6.405281428467041e-06,
+      "loss": 0.4404,
+      "step": 1688
+    },
+    {
+      "epoch": 1.405937846836848,
+      "grad_norm": 0.375386506319046,
+      "learning_rate": 6.400630925026568e-06,
+      "loss": 0.4471,
+      "step": 1689
+    },
+    {
+      "epoch": 1.4067702552719201,
+      "grad_norm": 0.3668792247772217,
+      "learning_rate": 6.395979106369132e-06,
+      "loss": 0.4422,
+      "step": 1690
+    },
+    {
+      "epoch": 1.4076026637069923,
+      "grad_norm": 0.34209975600242615,
+      "learning_rate": 6.391325976862872e-06,
+      "loss": 0.3815,
+      "step": 1691
+    },
+    {
+      "epoch": 1.4084350721420644,
+      "grad_norm": 0.33450645208358765,
+      "learning_rate": 6.386671540877162e-06,
+      "loss": 0.3906,
+      "step": 1692
+    },
+    {
+      "epoch": 1.4092674805771366,
+      "grad_norm": 0.34266212582588196,
+      "learning_rate": 6.382015802782592e-06,
+      "loss": 0.3992,
+      "step": 1693
+    },
+    {
+      "epoch": 1.4100998890122087,
+      "grad_norm": 0.38996613025665283,
+      "learning_rate": 6.377358766950987e-06,
+      "loss": 0.4352,
+      "step": 1694
+    },
+    {
+      "epoch": 1.4109322974472809,
+      "grad_norm": 0.3618418872356415,
+      "learning_rate": 6.372700437755381e-06,
+      "loss": 0.4297,
+      "step": 1695
+    },
+    {
+      "epoch": 1.4117647058823528,
+      "grad_norm": 0.36019423604011536,
+      "learning_rate": 6.368040819570032e-06,
+      "loss": 0.4046,
+      "step": 1696
+    },
+    {
+      "epoch": 1.4125971143174252,
+      "grad_norm": 0.3539135158061981,
+      "learning_rate": 6.3633799167703954e-06,
+      "loss": 0.4039,
+      "step": 1697
+    },
+    {
+      "epoch": 1.4134295227524971,
+      "grad_norm": 0.37196677923202515,
+      "learning_rate": 6.35871773373315e-06,
+      "loss": 0.4241,
+      "step": 1698
+    },
+    {
+      "epoch": 1.4142619311875695,
+      "grad_norm": 0.33772680163383484,
+      "learning_rate": 6.3540542748361585e-06,
+      "loss": 0.4144,
+      "step": 1699
+    },
+    {
+      "epoch": 1.4150943396226414,
+      "grad_norm": 0.36471793055534363,
+      "learning_rate": 6.349389544458497e-06,
+      "loss": 0.4414,
+      "step": 1700
+    },
+    {
+      "epoch": 1.4159267480577136,
+      "grad_norm": 0.39760175347328186,
+      "learning_rate": 6.3447235469804255e-06,
+      "loss": 0.4318,
+      "step": 1701
+    },
+    {
+      "epoch": 1.4167591564927857,
+      "grad_norm": 0.34777477383613586,
+      "learning_rate": 6.3400562867833984e-06,
+      "loss": 0.406,
+      "step": 1702
+    },
+    {
+      "epoch": 1.4175915649278579,
+      "grad_norm": 0.37017935514450073,
+      "learning_rate": 6.335387768250054e-06,
+      "loss": 0.4616,
+      "step": 1703
+    },
+    {
+      "epoch": 1.41842397336293,
+      "grad_norm": 0.36705178022384644,
+      "learning_rate": 6.330717995764215e-06,
+      "loss": 0.4275,
+      "step": 1704
+    },
+    {
+      "epoch": 1.4192563817980022,
+      "grad_norm": 0.34226226806640625,
+      "learning_rate": 6.326046973710878e-06,
+      "loss": 0.428,
+      "step": 1705
+    },
+    {
+      "epoch": 1.4200887902330743,
+      "grad_norm": 0.3641026020050049,
+      "learning_rate": 6.321374706476212e-06,
+      "loss": 0.3847,
+      "step": 1706
+    },
+    {
+      "epoch": 1.4209211986681465,
+      "grad_norm": 0.39205509424209595,
+      "learning_rate": 6.316701198447562e-06,
+      "loss": 0.4306,
+      "step": 1707
+    },
+    {
+      "epoch": 1.4217536071032186,
+      "grad_norm": 0.32389676570892334,
+      "learning_rate": 6.312026454013431e-06,
+      "loss": 0.4076,
+      "step": 1708
+    },
+    {
+      "epoch": 1.4225860155382908,
+      "grad_norm": 0.35079190135002136,
+      "learning_rate": 6.3073504775634884e-06,
+      "loss": 0.4088,
+      "step": 1709
+    },
+    {
+      "epoch": 1.423418423973363,
+      "grad_norm": 0.3681551218032837,
+      "learning_rate": 6.302673273488556e-06,
+      "loss": 0.4128,
+      "step": 1710
+    },
+    {
+      "epoch": 1.424250832408435,
+      "grad_norm": 0.3662969470024109,
+      "learning_rate": 6.297994846180611e-06,
+      "loss": 0.4279,
+      "step": 1711
+    },
+    {
+      "epoch": 1.4250832408435072,
+      "grad_norm": 0.5505783557891846,
+      "learning_rate": 6.293315200032777e-06,
+      "loss": 0.4447,
+      "step": 1712
+    },
+    {
+      "epoch": 1.4259156492785794,
+      "grad_norm": 0.35179439187049866,
+      "learning_rate": 6.288634339439328e-06,
+      "loss": 0.4202,
+      "step": 1713
+    },
+    {
+      "epoch": 1.4267480577136515,
+      "grad_norm": 0.3551366329193115,
+      "learning_rate": 6.283952268795669e-06,
+      "loss": 0.435,
+      "step": 1714
+    },
+    {
+      "epoch": 1.4275804661487237,
+      "grad_norm": 0.4332524240016937,
+      "learning_rate": 6.279268992498349e-06,
+      "loss": 0.4654,
+      "step": 1715
+    },
+    {
+      "epoch": 1.4284128745837958,
+      "grad_norm": 0.35650166869163513,
+      "learning_rate": 6.274584514945046e-06,
+      "loss": 0.4416,
+      "step": 1716
+    },
+    {
+      "epoch": 1.429245283018868,
+      "grad_norm": 0.3940298855304718,
+      "learning_rate": 6.269898840534566e-06,
+      "loss": 0.4315,
+      "step": 1717
+    },
+    {
+      "epoch": 1.4300776914539401,
+      "grad_norm": 0.3638506829738617,
+      "learning_rate": 6.26521197366684e-06,
+      "loss": 0.4081,
+      "step": 1718
+    },
+    {
+      "epoch": 1.430910099889012,
+      "grad_norm": 0.4078425168991089,
+      "learning_rate": 6.2605239187429175e-06,
+      "loss": 0.472,
+      "step": 1719
+    },
+    {
+      "epoch": 1.4317425083240845,
+      "grad_norm": 0.42004722356796265,
+      "learning_rate": 6.255834680164966e-06,
+      "loss": 0.3941,
+      "step": 1720
+    },
+    {
+      "epoch": 1.4325749167591564,
+      "grad_norm": 0.38885217905044556,
+      "learning_rate": 6.2511442623362585e-06,
+      "loss": 0.4549,
+      "step": 1721
+    },
+    {
+      "epoch": 1.4334073251942288,
+      "grad_norm": 0.3824394643306732,
+      "learning_rate": 6.246452669661184e-06,
+      "loss": 0.4207,
+      "step": 1722
+    },
+    {
+      "epoch": 1.4342397336293007,
+      "grad_norm": 0.39993953704833984,
+      "learning_rate": 6.241759906545226e-06,
+      "loss": 0.4386,
+      "step": 1723
+    },
+    {
+      "epoch": 1.435072142064373,
+      "grad_norm": 0.34289178252220154,
+      "learning_rate": 6.237065977394976e-06,
+      "loss": 0.4224,
+      "step": 1724
+    },
+    {
+      "epoch": 1.435904550499445,
+      "grad_norm": 0.3350009024143219,
+      "learning_rate": 6.23237088661811e-06,
+      "loss": 0.4408,
+      "step": 1725
+    },
+    {
+      "epoch": 1.4367369589345171,
+      "grad_norm": 0.337027907371521,
+      "learning_rate": 6.227674638623406e-06,
+      "loss": 0.3893,
+      "step": 1726
+    },
+    {
+      "epoch": 1.4375693673695893,
+      "grad_norm": 0.37013328075408936,
+      "learning_rate": 6.22297723782072e-06,
+      "loss": 0.4376,
+      "step": 1727
+    },
+    {
+      "epoch": 1.4384017758046614,
+      "grad_norm": 0.3607780337333679,
+      "learning_rate": 6.218278688620994e-06,
+      "loss": 0.4236,
+      "step": 1728
+    },
+    {
+      "epoch": 1.4392341842397336,
+      "grad_norm": 0.3762837052345276,
+      "learning_rate": 6.213578995436248e-06,
+      "loss": 0.4209,
+      "step": 1729
+    },
+    {
+      "epoch": 1.4400665926748057,
+      "grad_norm": 0.365888774394989,
+      "learning_rate": 6.208878162679577e-06,
+      "loss": 0.4778,
+      "step": 1730
+    },
+    {
+      "epoch": 1.440899001109878,
+      "grad_norm": 0.37897226214408875,
+      "learning_rate": 6.204176194765143e-06,
+      "loss": 0.4331,
+      "step": 1731
+    },
+    {
+      "epoch": 1.44173140954495,
+      "grad_norm": 0.3936547040939331,
+      "learning_rate": 6.199473096108179e-06,
+      "loss": 0.4007,
+      "step": 1732
+    },
+    {
+      "epoch": 1.4425638179800222,
+      "grad_norm": 0.33423808217048645,
+      "learning_rate": 6.194768871124976e-06,
+      "loss": 0.3973,
+      "step": 1733
+    },
+    {
+      "epoch": 1.4433962264150944,
+      "grad_norm": 0.395235538482666,
+      "learning_rate": 6.190063524232883e-06,
+      "loss": 0.4326,
+      "step": 1734
+    },
+    {
+      "epoch": 1.4442286348501665,
+      "grad_norm": 0.35545146465301514,
+      "learning_rate": 6.1853570598503045e-06,
+      "loss": 0.3881,
+      "step": 1735
+    },
+    {
+      "epoch": 1.4450610432852387,
+      "grad_norm": 0.36360153555870056,
+      "learning_rate": 6.18064948239669e-06,
+      "loss": 0.3702,
+      "step": 1736
+    },
+    {
+      "epoch": 1.4458934517203108,
+      "grad_norm": 0.4225115478038788,
+      "learning_rate": 6.175940796292541e-06,
+      "loss": 0.4497,
+      "step": 1737
+    },
+    {
+      "epoch": 1.446725860155383,
+      "grad_norm": 0.3660155236721039,
+      "learning_rate": 6.171231005959393e-06,
+      "loss": 0.4226,
+      "step": 1738
+    },
+    {
+      "epoch": 1.447558268590455,
+      "grad_norm": 0.38432952761650085,
+      "learning_rate": 6.166520115819825e-06,
+      "loss": 0.4009,
+      "step": 1739
+    },
+    {
+      "epoch": 1.4483906770255273,
+      "grad_norm": 0.3980556130409241,
+      "learning_rate": 6.161808130297442e-06,
+      "loss": 0.4006,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4492230854605994,
+      "grad_norm": 0.34962958097457886,
+      "learning_rate": 6.157095053816882e-06,
+      "loss": 0.4289,
+      "step": 1741
+    },
+    {
+      "epoch": 1.4500554938956713,
+      "grad_norm": 0.3962990939617157,
+      "learning_rate": 6.152380890803806e-06,
+      "loss": 0.4005,
+      "step": 1742
+    },
+    {
+      "epoch": 1.4508879023307437,
+      "grad_norm": 0.3911704123020172,
+      "learning_rate": 6.147665645684897e-06,
+      "loss": 0.4301,
+      "step": 1743
+    },
+    {
+      "epoch": 1.4517203107658156,
+      "grad_norm": 0.35797592997550964,
+      "learning_rate": 6.142949322887852e-06,
+      "loss": 0.4265,
+      "step": 1744
+    },
+    {
+      "epoch": 1.452552719200888,
+      "grad_norm": 0.39295870065689087,
+      "learning_rate": 6.138231926841381e-06,
+      "loss": 0.3893,
+      "step": 1745
+    },
+    {
+      "epoch": 1.45338512763596,
+      "grad_norm": 0.387833833694458,
+      "learning_rate": 6.1335134619751994e-06,
+      "loss": 0.4233,
+      "step": 1746
+    },
+    {
+      "epoch": 1.4542175360710323,
+      "grad_norm": 0.3874453902244568,
+      "learning_rate": 6.128793932720031e-06,
+      "loss": 0.4878,
+      "step": 1747
+    },
+    {
+      "epoch": 1.4550499445061043,
+      "grad_norm": 0.4330272078514099,
+      "learning_rate": 6.1240733435075946e-06,
+      "loss": 0.4074,
+      "step": 1748
+    },
+    {
+      "epoch": 1.4558823529411764,
+      "grad_norm": 0.3374917507171631,
+      "learning_rate": 6.119351698770607e-06,
+      "loss": 0.4121,
+      "step": 1749
+    },
+    {
+      "epoch": 1.4567147613762486,
+      "grad_norm": 0.37870532274246216,
+      "learning_rate": 6.1146290029427755e-06,
+      "loss": 0.4317,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4575471698113207,
+      "grad_norm": 0.42150020599365234,
+      "learning_rate": 6.1099052604587935e-06,
+      "loss": 0.4376,
+      "step": 1751
+    },
+    {
+      "epoch": 1.4583795782463929,
+      "grad_norm": 0.3680340051651001,
+      "learning_rate": 6.105180475754341e-06,
+      "loss": 0.4167,
+      "step": 1752
+    },
+    {
+      "epoch": 1.459211986681465,
+      "grad_norm": 0.32261571288108826,
+      "learning_rate": 6.100454653266068e-06,
+      "loss": 0.3828,
+      "step": 1753
+    },
+    {
+      "epoch": 1.4600443951165372,
+      "grad_norm": 0.3715158700942993,
+      "learning_rate": 6.095727797431607e-06,
+      "loss": 0.4435,
+      "step": 1754
+    },
+    {
+      "epoch": 1.4608768035516093,
+      "grad_norm": 0.3581849932670593,
+      "learning_rate": 6.0909999126895605e-06,
+      "loss": 0.431,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4617092119866815,
+      "grad_norm": 0.3825138509273529,
+      "learning_rate": 6.086271003479492e-06,
+      "loss": 0.4651,
+      "step": 1756
+    },
+    {
+      "epoch": 1.4625416204217536,
+      "grad_norm": 0.38319095969200134,
+      "learning_rate": 6.081541074241932e-06,
+      "loss": 0.4446,
+      "step": 1757
+    },
+    {
+      "epoch": 1.4633740288568258,
+      "grad_norm": 0.34215807914733887,
+      "learning_rate": 6.076810129418367e-06,
+      "loss": 0.3892,
+      "step": 1758
+    },
+    {
+      "epoch": 1.464206437291898,
+      "grad_norm": 0.3696172833442688,
+      "learning_rate": 6.072078173451235e-06,
+      "loss": 0.4422,
+      "step": 1759
+    },
+    {
+      "epoch": 1.46503884572697,
+      "grad_norm": 0.42305469512939453,
+      "learning_rate": 6.067345210783927e-06,
+      "loss": 0.4343,
+      "step": 1760
+    },
+    {
+      "epoch": 1.4658712541620422,
+      "grad_norm": 0.4272174537181854,
+      "learning_rate": 6.062611245860778e-06,
+      "loss": 0.416,
+      "step": 1761
+    },
+    {
+      "epoch": 1.4667036625971144,
+      "grad_norm": 0.33553504943847656,
+      "learning_rate": 6.057876283127062e-06,
+      "loss": 0.3888,
+      "step": 1762
+    },
+    {
+      "epoch": 1.4675360710321865,
+      "grad_norm": 0.4211355149745941,
+      "learning_rate": 6.053140327028996e-06,
+      "loss": 0.4565,
+      "step": 1763
+    },
+    {
+      "epoch": 1.4683684794672587,
+      "grad_norm": 0.4059634208679199,
+      "learning_rate": 6.048403382013721e-06,
+      "loss": 0.4007,
+      "step": 1764
+    },
+    {
+      "epoch": 1.4692008879023306,
+      "grad_norm": 0.3843512535095215,
+      "learning_rate": 6.043665452529315e-06,
+      "loss": 0.4307,
+      "step": 1765
+    },
+    {
+      "epoch": 1.470033296337403,
+      "grad_norm": 0.44470739364624023,
+      "learning_rate": 6.038926543024774e-06,
+      "loss": 0.4408,
+      "step": 1766
+    },
+    {
+      "epoch": 1.470865704772475,
+      "grad_norm": 0.3952719271183014,
+      "learning_rate": 6.034186657950019e-06,
+      "loss": 0.3943,
+      "step": 1767
+    },
+    {
+      "epoch": 1.4716981132075473,
+      "grad_norm": 0.4131404161453247,
+      "learning_rate": 6.029445801755884e-06,
+      "loss": 0.4505,
+      "step": 1768
+    },
+    {
+      "epoch": 1.4725305216426192,
+      "grad_norm": 0.4458027184009552,
+      "learning_rate": 6.024703978894118e-06,
+      "loss": 0.4549,
+      "step": 1769
+    },
+    {
+      "epoch": 1.4733629300776916,
+      "grad_norm": 0.3414630889892578,
+      "learning_rate": 6.019961193817371e-06,
+      "loss": 0.3777,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4741953385127635,
+      "grad_norm": 0.3965328633785248,
+      "learning_rate": 6.015217450979206e-06,
+      "loss": 0.4192,
+      "step": 1771
+    },
+    {
+      "epoch": 1.4750277469478357,
+      "grad_norm": 0.4391860365867615,
+      "learning_rate": 6.010472754834078e-06,
+      "loss": 0.471,
+      "step": 1772
+    },
+    {
+      "epoch": 1.4758601553829078,
+      "grad_norm": 0.34298470616340637,
+      "learning_rate": 6.00572710983734e-06,
+      "loss": 0.4105,
+      "step": 1773
+    },
+    {
+      "epoch": 1.47669256381798,
+      "grad_norm": 0.37611621618270874,
+      "learning_rate": 6.000980520445237e-06,
+      "loss": 0.3869,
+      "step": 1774
+    },
+    {
+      "epoch": 1.4775249722530521,
+      "grad_norm": 0.3933986723423004,
+      "learning_rate": 5.9962329911148985e-06,
+      "loss": 0.4365,
+      "step": 1775
+    },
+    {
+      "epoch": 1.4783573806881243,
+      "grad_norm": 0.36345475912094116,
+      "learning_rate": 5.991484526304338e-06,
+      "loss": 0.401,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4791897891231964,
+      "grad_norm": 0.4031011462211609,
+      "learning_rate": 5.986735130472449e-06,
+      "loss": 0.4051,
+      "step": 1777
+    },
+    {
+      "epoch": 1.4800221975582686,
+      "grad_norm": 0.3900188207626343,
+      "learning_rate": 5.981984808078993e-06,
+      "loss": 0.4398,
+      "step": 1778
+    },
+    {
+      "epoch": 1.4808546059933407,
+      "grad_norm": 0.3524153232574463,
+      "learning_rate": 5.97723356358461e-06,
+      "loss": 0.4145,
+      "step": 1779
+    },
+    {
+      "epoch": 1.4816870144284129,
+      "grad_norm": 0.3924466669559479,
+      "learning_rate": 5.972481401450798e-06,
+      "loss": 0.4317,
+      "step": 1780
+    },
+    {
+      "epoch": 1.482519422863485,
+      "grad_norm": 0.33220553398132324,
+      "learning_rate": 5.967728326139926e-06,
+      "loss": 0.4181,
+      "step": 1781
+    },
+    {
+      "epoch": 1.4833518312985572,
+      "grad_norm": 0.39167076349258423,
+      "learning_rate": 5.962974342115209e-06,
+      "loss": 0.4228,
+      "step": 1782
+    },
+    {
+      "epoch": 1.4841842397336293,
+      "grad_norm": 0.3629782497882843,
+      "learning_rate": 5.9582194538407235e-06,
+      "loss": 0.4085,
+      "step": 1783
+    },
+    {
+      "epoch": 1.4850166481687015,
+      "grad_norm": 0.3289967179298401,
+      "learning_rate": 5.9534636657813935e-06,
+      "loss": 0.4189,
+      "step": 1784
+    },
+    {
+      "epoch": 1.4858490566037736,
+      "grad_norm": 0.36576929688453674,
+      "learning_rate": 5.948706982402987e-06,
+      "loss": 0.452,
+      "step": 1785
+    },
+    {
+      "epoch": 1.4866814650388458,
+      "grad_norm": 0.3725149631500244,
+      "learning_rate": 5.9439494081721125e-06,
+      "loss": 0.4378,
+      "step": 1786
+    },
+    {
+      "epoch": 1.487513873473918,
+      "grad_norm": 0.3603670299053192,
+      "learning_rate": 5.939190947556216e-06,
+      "loss": 0.4281,
+      "step": 1787
+    },
+    {
+      "epoch": 1.4883462819089899,
+      "grad_norm": 0.35488083958625793,
+      "learning_rate": 5.934431605023575e-06,
+      "loss": 0.4333,
+      "step": 1788
+    },
+    {
+      "epoch": 1.4891786903440623,
+      "grad_norm": 0.33910173177719116,
+      "learning_rate": 5.929671385043296e-06,
+      "loss": 0.3941,
+      "step": 1789
+    },
+    {
+      "epoch": 1.4900110987791342,
+      "grad_norm": 0.4183632433414459,
+      "learning_rate": 5.924910292085308e-06,
+      "loss": 0.4207,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4908435072142066,
+      "grad_norm": 0.3865908086299896,
+      "learning_rate": 5.920148330620362e-06,
+      "loss": 0.4192,
+      "step": 1791
+    },
+    {
+      "epoch": 1.4916759156492785,
+      "grad_norm": 0.34254950284957886,
+      "learning_rate": 5.915385505120024e-06,
+      "loss": 0.4404,
+      "step": 1792
+    },
+    {
+      "epoch": 1.4925083240843509,
+      "grad_norm": 0.399553120136261,
+      "learning_rate": 5.9106218200566646e-06,
+      "loss": 0.4519,
+      "step": 1793
+    },
+    {
+      "epoch": 1.4933407325194228,
+      "grad_norm": 0.33408740162849426,
+      "learning_rate": 5.905857279903475e-06,
+      "loss": 0.3971,
+      "step": 1794
+    },
+    {
+      "epoch": 1.494173140954495,
+      "grad_norm": 0.3779832422733307,
+      "learning_rate": 5.9010918891344375e-06,
+      "loss": 0.4478,
+      "step": 1795
+    },
+    {
+      "epoch": 1.495005549389567,
+      "grad_norm": 0.3475707471370697,
+      "learning_rate": 5.896325652224339e-06,
+      "loss": 0.4329,
+      "step": 1796
+    },
+    {
+      "epoch": 1.4958379578246392,
+      "grad_norm": 0.3472834527492523,
+      "learning_rate": 5.891558573648759e-06,
+      "loss": 0.4036,
+      "step": 1797
+    },
+    {
+      "epoch": 1.4966703662597114,
+      "grad_norm": 0.36704128980636597,
+      "learning_rate": 5.886790657884067e-06,
+      "loss": 0.4479,
+      "step": 1798
+    },
+    {
+      "epoch": 1.4975027746947835,
+      "grad_norm": 0.39827099442481995,
+      "learning_rate": 5.8820219094074215e-06,
+      "loss": 0.4401,
+      "step": 1799
+    },
+    {
+      "epoch": 1.4983351831298557,
+      "grad_norm": 0.37675169110298157,
+      "learning_rate": 5.877252332696759e-06,
+      "loss": 0.439,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4991675915649278,
+      "grad_norm": 0.3367178738117218,
+      "learning_rate": 5.8724819322307955e-06,
+      "loss": 0.4434,
+      "step": 1801
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3833068013191223,
+      "learning_rate": 5.8677107124890206e-06,
+      "loss": 0.4348,
+      "step": 1802
+    },
+    {
+      "epoch": 1.5008324084350722,
+      "grad_norm": 0.333822101354599,
+      "learning_rate": 5.862938677951695e-06,
+      "loss": 0.4245,
+      "step": 1803
+    },
+    {
+      "epoch": 1.5016648168701443,
+      "grad_norm": 0.33005255460739136,
+      "learning_rate": 5.85816583309984e-06,
+      "loss": 0.4149,
+      "step": 1804
+    },
+    {
+      "epoch": 1.5024972253052165,
+      "grad_norm": 0.3507244288921356,
+      "learning_rate": 5.853392182415244e-06,
+      "loss": 0.4029,
+      "step": 1805
+    },
+    {
+      "epoch": 1.5033296337402886,
+      "grad_norm": 0.34437042474746704,
+      "learning_rate": 5.848617730380444e-06,
+      "loss": 0.3865,
+      "step": 1806
+    },
+    {
+      "epoch": 1.5041620421753608,
+      "grad_norm": 0.33490440249443054,
+      "learning_rate": 5.843842481478739e-06,
+      "loss": 0.4468,
+      "step": 1807
+    },
+    {
+      "epoch": 1.504994450610433,
+      "grad_norm": 0.34904393553733826,
+      "learning_rate": 5.839066440194165e-06,
+      "loss": 0.4286,
+      "step": 1808
+    },
+    {
+      "epoch": 1.5058268590455048,
+      "grad_norm": 0.3498096168041229,
+      "learning_rate": 5.834289611011515e-06,
+      "loss": 0.4244,
+      "step": 1809
+    },
+    {
+      "epoch": 1.5066592674805772,
+      "grad_norm": 0.3397802710533142,
+      "learning_rate": 5.82951199841631e-06,
+      "loss": 0.4203,
+      "step": 1810
+    },
+    {
+      "epoch": 1.5074916759156491,
+      "grad_norm": 0.34732288122177124,
+      "learning_rate": 5.824733606894818e-06,
+      "loss": 0.4276,
+      "step": 1811
+    },
+    {
+      "epoch": 1.5083240843507215,
+      "grad_norm": 0.3793588876724243,
+      "learning_rate": 5.819954440934026e-06,
+      "loss": 0.4587,
+      "step": 1812
+    },
+    {
+      "epoch": 1.5091564927857934,
+      "grad_norm": 0.339600145816803,
+      "learning_rate": 5.815174505021659e-06,
+      "loss": 0.3988,
+      "step": 1813
+    },
+    {
+      "epoch": 1.5099889012208658,
+      "grad_norm": 0.37940147519111633,
+      "learning_rate": 5.810393803646157e-06,
+      "loss": 0.4525,
+      "step": 1814
+    },
+    {
+      "epoch": 1.5108213096559377,
+      "grad_norm": 0.36495909094810486,
+      "learning_rate": 5.805612341296685e-06,
+      "loss": 0.3835,
+      "step": 1815
+    },
+    {
+      "epoch": 1.5116537180910101,
+      "grad_norm": 0.38348937034606934,
+      "learning_rate": 5.800830122463117e-06,
+      "loss": 0.4352,
+      "step": 1816
+    },
+    {
+      "epoch": 1.512486126526082,
+      "grad_norm": 0.35472238063812256,
+      "learning_rate": 5.7960471516360435e-06,
+      "loss": 0.4223,
+      "step": 1817
+    },
+    {
+      "epoch": 1.5133185349611544,
+      "grad_norm": 0.4020749032497406,
+      "learning_rate": 5.791263433306758e-06,
+      "loss": 0.4726,
+      "step": 1818
+    },
+    {
+      "epoch": 1.5141509433962264,
+      "grad_norm": 0.3420332372188568,
+      "learning_rate": 5.786478971967249e-06,
+      "loss": 0.3763,
+      "step": 1819
+    },
+    {
+      "epoch": 1.5149833518312985,
+      "grad_norm": 0.35677722096443176,
+      "learning_rate": 5.781693772110219e-06,
+      "loss": 0.4016,
+      "step": 1820
+    },
+    {
+      "epoch": 1.5158157602663707,
+      "grad_norm": 0.3244923949241638,
+      "learning_rate": 5.776907838229049e-06,
+      "loss": 0.4264,
+      "step": 1821
+    },
+    {
+      "epoch": 1.5166481687014428,
+      "grad_norm": 0.3919678032398224,
+      "learning_rate": 5.772121174817816e-06,
+      "loss": 0.451,
+      "step": 1822
+    },
+    {
+      "epoch": 1.517480577136515,
+      "grad_norm": 0.3128311336040497,
+      "learning_rate": 5.767333786371279e-06,
+      "loss": 0.3684,
+      "step": 1823
+    },
+    {
+      "epoch": 1.5183129855715871,
+      "grad_norm": 0.3211652338504791,
+      "learning_rate": 5.762545677384884e-06,
+      "loss": 0.4176,
+      "step": 1824
+    },
+    {
+      "epoch": 1.5191453940066593,
+      "grad_norm": 0.3597281277179718,
+      "learning_rate": 5.757756852354743e-06,
+      "loss": 0.4235,
+      "step": 1825
+    },
+    {
+      "epoch": 1.5199778024417314,
+      "grad_norm": 0.3358439803123474,
+      "learning_rate": 5.752967315777653e-06,
+      "loss": 0.3894,
+      "step": 1826
+    },
+    {
+      "epoch": 1.5208102108768036,
+      "grad_norm": 0.3780466616153717,
+      "learning_rate": 5.748177072151068e-06,
+      "loss": 0.4414,
+      "step": 1827
+    },
+    {
+      "epoch": 1.5216426193118757,
+      "grad_norm": 0.341908723115921,
+      "learning_rate": 5.743386125973112e-06,
+      "loss": 0.4439,
+      "step": 1828
+    },
+    {
+      "epoch": 1.5224750277469479,
+      "grad_norm": 0.3324446976184845,
+      "learning_rate": 5.738594481742568e-06,
+      "loss": 0.4143,
+      "step": 1829
+    },
+    {
+      "epoch": 1.52330743618202,
+      "grad_norm": 0.37258070707321167,
+      "learning_rate": 5.733802143958872e-06,
+      "loss": 0.4369,
+      "step": 1830
+    },
+    {
+      "epoch": 1.5241398446170922,
+      "grad_norm": 0.34372997283935547,
+      "learning_rate": 5.729009117122117e-06,
+      "loss": 0.4175,
+      "step": 1831
+    },
+    {
+      "epoch": 1.524972253052164,
+      "grad_norm": 0.3553544580936432,
+      "learning_rate": 5.724215405733033e-06,
+      "loss": 0.4233,
+      "step": 1832
+    },
+    {
+      "epoch": 1.5258046614872365,
+      "grad_norm": 0.34760257601737976,
+      "learning_rate": 5.7194210142930065e-06,
+      "loss": 0.4174,
+      "step": 1833
+    },
+    {
+      "epoch": 1.5266370699223084,
+      "grad_norm": 0.37886732816696167,
+      "learning_rate": 5.714625947304048e-06,
+      "loss": 0.4328,
+      "step": 1834
+    },
+    {
+      "epoch": 1.5274694783573808,
+      "grad_norm": 0.35653242468833923,
+      "learning_rate": 5.709830209268814e-06,
+      "loss": 0.4064,
+      "step": 1835
+    },
+    {
+      "epoch": 1.5283018867924527,
+      "grad_norm": 0.4173387885093689,
+      "learning_rate": 5.705033804690583e-06,
+      "loss": 0.4177,
+      "step": 1836
+    },
+    {
+      "epoch": 1.529134295227525,
+      "grad_norm": 0.33839157223701477,
+      "learning_rate": 5.7002367380732685e-06,
+      "loss": 0.3887,
+      "step": 1837
+    },
+    {
+      "epoch": 1.529966703662597,
+      "grad_norm": 0.35368192195892334,
+      "learning_rate": 5.695439013921391e-06,
+      "loss": 0.3985,
+      "step": 1838
+    },
+    {
+      "epoch": 1.5307991120976694,
+      "grad_norm": 0.3887537121772766,
+      "learning_rate": 5.6906406367401075e-06,
+      "loss": 0.4437,
+      "step": 1839
+    },
+    {
+      "epoch": 1.5316315205327413,
+      "grad_norm": 0.3811011016368866,
+      "learning_rate": 5.6858416110351715e-06,
+      "loss": 0.4423,
+      "step": 1840
+    },
+    {
+      "epoch": 1.5324639289678137,
+      "grad_norm": 0.33723312616348267,
+      "learning_rate": 5.681041941312954e-06,
+      "loss": 0.411,
+      "step": 1841
+    },
+    {
+      "epoch": 1.5332963374028856,
+      "grad_norm": 0.4133073389530182,
+      "learning_rate": 5.676241632080429e-06,
+      "loss": 0.4312,
+      "step": 1842
+    },
+    {
+      "epoch": 1.5341287458379578,
+      "grad_norm": 0.3302142918109894,
+      "learning_rate": 5.6714406878451715e-06,
+      "loss": 0.3994,
+      "step": 1843
+    },
+    {
+      "epoch": 1.53496115427303,
+      "grad_norm": 0.36849483847618103,
+      "learning_rate": 5.666639113115351e-06,
+      "loss": 0.4616,
+      "step": 1844
+    },
+    {
+      "epoch": 1.535793562708102,
+      "grad_norm": 0.38487333059310913,
+      "learning_rate": 5.661836912399731e-06,
+      "loss": 0.4145,
+      "step": 1845
+    },
+    {
+      "epoch": 1.5366259711431742,
+      "grad_norm": 0.37898313999176025,
+      "learning_rate": 5.657034090207663e-06,
+      "loss": 0.4573,
+      "step": 1846
+    },
+    {
+      "epoch": 1.5374583795782464,
+      "grad_norm": 0.35560891032218933,
+      "learning_rate": 5.652230651049077e-06,
+      "loss": 0.4162,
+      "step": 1847
+    },
+    {
+      "epoch": 1.5382907880133185,
+      "grad_norm": 0.3300834894180298,
+      "learning_rate": 5.647426599434493e-06,
+      "loss": 0.4577,
+      "step": 1848
+    },
+    {
+      "epoch": 1.5391231964483907,
+      "grad_norm": 0.32240939140319824,
+      "learning_rate": 5.642621939874995e-06,
+      "loss": 0.3774,
+      "step": 1849
+    },
+    {
+      "epoch": 1.5399556048834628,
+      "grad_norm": 0.41031232476234436,
+      "learning_rate": 5.637816676882244e-06,
+      "loss": 0.4475,
+      "step": 1850
+    },
+    {
+      "epoch": 1.540788013318535,
+      "grad_norm": 0.32196757197380066,
+      "learning_rate": 5.633010814968465e-06,
+      "loss": 0.4096,
+      "step": 1851
+    },
+    {
+      "epoch": 1.5416204217536071,
+      "grad_norm": 0.38092145323753357,
+      "learning_rate": 5.628204358646448e-06,
+      "loss": 0.4481,
+      "step": 1852
+    },
+    {
+      "epoch": 1.5424528301886793,
+      "grad_norm": 0.34755241870880127,
+      "learning_rate": 5.623397312429537e-06,
+      "loss": 0.431,
+      "step": 1853
+    },
+    {
+      "epoch": 1.5432852386237514,
+      "grad_norm": 0.33400028944015503,
+      "learning_rate": 5.618589680831636e-06,
+      "loss": 0.3853,
+      "step": 1854
+    },
+    {
+      "epoch": 1.5441176470588234,
+      "grad_norm": 0.35710418224334717,
+      "learning_rate": 5.6137814683671935e-06,
+      "loss": 0.4172,
+      "step": 1855
+    },
+    {
+      "epoch": 1.5449500554938957,
+      "grad_norm": 0.37034469842910767,
+      "learning_rate": 5.608972679551205e-06,
+      "loss": 0.3994,
+      "step": 1856
+    },
+    {
+      "epoch": 1.5457824639289677,
+      "grad_norm": 0.33173617720603943,
+      "learning_rate": 5.604163318899207e-06,
+      "loss": 0.4299,
+      "step": 1857
+    },
+    {
+      "epoch": 1.54661487236404,
+      "grad_norm": 0.3156852424144745,
+      "learning_rate": 5.599353390927275e-06,
+      "loss": 0.4213,
+      "step": 1858
+    },
+    {
+      "epoch": 1.547447280799112,
+      "grad_norm": 0.3290148377418518,
+      "learning_rate": 5.594542900152015e-06,
+      "loss": 0.4192,
+      "step": 1859
+    },
+    {
+      "epoch": 1.5482796892341844,
+      "grad_norm": 0.3706801235675812,
+      "learning_rate": 5.589731851090559e-06,
+      "loss": 0.4677,
+      "step": 1860
+    },
+    {
+      "epoch": 1.5491120976692563,
+      "grad_norm": 0.33532100915908813,
+      "learning_rate": 5.584920248260572e-06,
+      "loss": 0.4131,
+      "step": 1861
+    },
+    {
+      "epoch": 1.5499445061043287,
+      "grad_norm": 0.3689619302749634,
+      "learning_rate": 5.580108096180229e-06,
+      "loss": 0.4205,
+      "step": 1862
+    },
+    {
+      "epoch": 1.5507769145394006,
+      "grad_norm": 0.32412606477737427,
+      "learning_rate": 5.575295399368228e-06,
+      "loss": 0.4127,
+      "step": 1863
+    },
+    {
+      "epoch": 1.551609322974473,
+      "grad_norm": 0.363091379404068,
+      "learning_rate": 5.570482162343772e-06,
+      "loss": 0.4362,
+      "step": 1864
+    },
+    {
+      "epoch": 1.552441731409545,
+      "grad_norm": 0.37017905712127686,
+      "learning_rate": 5.5656683896265786e-06,
+      "loss": 0.4014,
+      "step": 1865
+    },
+    {
+      "epoch": 1.553274139844617,
+      "grad_norm": 0.3684568405151367,
+      "learning_rate": 5.560854085736861e-06,
+      "loss": 0.4703,
+      "step": 1866
+    },
+    {
+      "epoch": 1.5541065482796892,
+      "grad_norm": 0.32582852244377136,
+      "learning_rate": 5.556039255195338e-06,
+      "loss": 0.4202,
+      "step": 1867
+    },
+    {
+      "epoch": 1.5549389567147613,
+      "grad_norm": 0.37376418709754944,
+      "learning_rate": 5.551223902523218e-06,
+      "loss": 0.4237,
+      "step": 1868
+    },
+    {
+      "epoch": 1.5557713651498335,
+      "grad_norm": 0.3530517518520355,
+      "learning_rate": 5.546408032242202e-06,
+      "loss": 0.4311,
+      "step": 1869
+    },
+    {
+      "epoch": 1.5566037735849056,
+      "grad_norm": 0.34218358993530273,
+      "learning_rate": 5.541591648874476e-06,
+      "loss": 0.4404,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5574361820199778,
+      "grad_norm": 0.31669047474861145,
+      "learning_rate": 5.53677475694271e-06,
+      "loss": 0.4103,
+      "step": 1871
+    },
+    {
+      "epoch": 1.55826859045505,
+      "grad_norm": 0.3490392863750458,
+      "learning_rate": 5.531957360970048e-06,
+      "loss": 0.3915,
+      "step": 1872
+    },
+    {
+      "epoch": 1.559100998890122,
+      "grad_norm": 0.3820657730102539,
+      "learning_rate": 5.527139465480109e-06,
+      "loss": 0.4277,
+      "step": 1873
+    },
+    {
+      "epoch": 1.5599334073251943,
+      "grad_norm": 0.35189470648765564,
+      "learning_rate": 5.5223210749969845e-06,
+      "loss": 0.4014,
+      "step": 1874
+    },
+    {
+      "epoch": 1.5607658157602664,
+      "grad_norm": 0.36156129837036133,
+      "learning_rate": 5.5175021940452225e-06,
+      "loss": 0.413,
+      "step": 1875
+    },
+    {
+      "epoch": 1.5615982241953386,
+      "grad_norm": 0.32655417919158936,
+      "learning_rate": 5.512682827149841e-06,
+      "loss": 0.4177,
+      "step": 1876
+    },
+    {
+      "epoch": 1.5624306326304107,
+      "grad_norm": 0.3407149314880371,
+      "learning_rate": 5.507862978836306e-06,
+      "loss": 0.4112,
+      "step": 1877
+    },
+    {
+      "epoch": 1.5632630410654826,
+      "grad_norm": 0.41151124238967896,
+      "learning_rate": 5.503042653630543e-06,
+      "loss": 0.4631,
+      "step": 1878
+    },
+    {
+      "epoch": 1.564095449500555,
+      "grad_norm": 0.3386502265930176,
+      "learning_rate": 5.49822185605892e-06,
+      "loss": 0.3822,
+      "step": 1879
+    },
+    {
+      "epoch": 1.564927857935627,
+      "grad_norm": 0.3886801302433014,
+      "learning_rate": 5.4934005906482525e-06,
+      "loss": 0.4602,
+      "step": 1880
+    },
+    {
+      "epoch": 1.5657602663706993,
+      "grad_norm": 0.3777741491794586,
+      "learning_rate": 5.488578861925788e-06,
+      "loss": 0.4215,
+      "step": 1881
+    },
+    {
+      "epoch": 1.5665926748057712,
+      "grad_norm": 0.3248192369937897,
+      "learning_rate": 5.4837566744192196e-06,
+      "loss": 0.3973,
+      "step": 1882
+    },
+    {
+      "epoch": 1.5674250832408436,
+      "grad_norm": 0.3611052930355072,
+      "learning_rate": 5.478934032656663e-06,
+      "loss": 0.4408,
+      "step": 1883
+    },
+    {
+      "epoch": 1.5682574916759155,
+      "grad_norm": 0.35995787382125854,
+      "learning_rate": 5.4741109411666635e-06,
+      "loss": 0.401,
+      "step": 1884
+    },
+    {
+      "epoch": 1.569089900110988,
+      "grad_norm": 0.39710691571235657,
+      "learning_rate": 5.46928740447819e-06,
+      "loss": 0.4711,
+      "step": 1885
+    },
+    {
+      "epoch": 1.5699223085460599,
+      "grad_norm": 0.34556522965431213,
+      "learning_rate": 5.464463427120626e-06,
+      "loss": 0.4347,
+      "step": 1886
+    },
+    {
+      "epoch": 1.5707547169811322,
+      "grad_norm": 0.3444367051124573,
+      "learning_rate": 5.459639013623772e-06,
+      "loss": 0.3928,
+      "step": 1887
+    },
+    {
+      "epoch": 1.5715871254162042,
+      "grad_norm": 0.3429131805896759,
+      "learning_rate": 5.454814168517836e-06,
+      "loss": 0.4362,
+      "step": 1888
+    },
+    {
+      "epoch": 1.5724195338512763,
+      "grad_norm": 0.3870256841182709,
+      "learning_rate": 5.449988896333431e-06,
+      "loss": 0.4537,
+      "step": 1889
+    },
+    {
+      "epoch": 1.5732519422863485,
+      "grad_norm": 0.3829275071620941,
+      "learning_rate": 5.445163201601575e-06,
+      "loss": 0.4134,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5740843507214206,
+      "grad_norm": 0.37084445357322693,
+      "learning_rate": 5.440337088853679e-06,
+      "loss": 0.3903,
+      "step": 1891
+    },
+    {
+      "epoch": 1.5749167591564928,
+      "grad_norm": 0.36003977060317993,
+      "learning_rate": 5.435510562621544e-06,
+      "loss": 0.4641,
+      "step": 1892
+    },
+    {
+      "epoch": 1.575749167591565,
+      "grad_norm": 0.4320586025714874,
+      "learning_rate": 5.4306836274373675e-06,
+      "loss": 0.4261,
+      "step": 1893
+    },
+    {
+      "epoch": 1.576581576026637,
+      "grad_norm": 0.35853341221809387,
+      "learning_rate": 5.425856287833723e-06,
+      "loss": 0.4363,
+      "step": 1894
+    },
+    {
+      "epoch": 1.5774139844617092,
+      "grad_norm": 0.3714522123336792,
+      "learning_rate": 5.421028548343568e-06,
+      "loss": 0.4157,
+      "step": 1895
+    },
+    {
+      "epoch": 1.5782463928967814,
+      "grad_norm": 0.37098610401153564,
+      "learning_rate": 5.4162004135002336e-06,
+      "loss": 0.4233,
+      "step": 1896
+    },
+    {
+      "epoch": 1.5790788013318535,
+      "grad_norm": 0.35609301924705505,
+      "learning_rate": 5.411371887837423e-06,
+      "loss": 0.4317,
+      "step": 1897
+    },
+    {
+      "epoch": 1.5799112097669257,
+      "grad_norm": 0.31282979249954224,
+      "learning_rate": 5.406542975889209e-06,
+      "loss": 0.3921,
+      "step": 1898
+    },
+    {
+      "epoch": 1.5807436182019978,
+      "grad_norm": 0.34338200092315674,
+      "learning_rate": 5.40171368219002e-06,
+      "loss": 0.4239,
+      "step": 1899
+    },
+    {
+      "epoch": 1.58157602663707,
+      "grad_norm": 0.357689768075943,
+      "learning_rate": 5.396884011274651e-06,
+      "loss": 0.4402,
+      "step": 1900
+    },
+    {
+      "epoch": 1.582408435072142,
+      "grad_norm": 0.3389754891395569,
+      "learning_rate": 5.3920539676782455e-06,
+      "loss": 0.3947,
+      "step": 1901
+    },
+    {
+      "epoch": 1.5832408435072143,
+      "grad_norm": 0.348406046628952,
+      "learning_rate": 5.387223555936301e-06,
+      "loss": 0.4213,
+      "step": 1902
+    },
+    {
+      "epoch": 1.5840732519422862,
+      "grad_norm": 0.3109308183193207,
+      "learning_rate": 5.382392780584655e-06,
+      "loss": 0.3912,
+      "step": 1903
+    },
+    {
+      "epoch": 1.5849056603773586,
+      "grad_norm": 0.3686921298503876,
+      "learning_rate": 5.377561646159495e-06,
+      "loss": 0.4632,
+      "step": 1904
+    },
+    {
+      "epoch": 1.5857380688124305,
+      "grad_norm": 0.33451026678085327,
+      "learning_rate": 5.372730157197338e-06,
+      "loss": 0.3807,
+      "step": 1905
+    },
+    {
+      "epoch": 1.5865704772475029,
+      "grad_norm": 0.3415146768093109,
+      "learning_rate": 5.367898318235037e-06,
+      "loss": 0.4282,
+      "step": 1906
+    },
+    {
+      "epoch": 1.5874028856825748,
+      "grad_norm": 0.3252672553062439,
+      "learning_rate": 5.363066133809773e-06,
+      "loss": 0.4125,
+      "step": 1907
+    },
+    {
+      "epoch": 1.5882352941176472,
+      "grad_norm": 0.35067319869995117,
+      "learning_rate": 5.3582336084590535e-06,
+      "loss": 0.399,
+      "step": 1908
+    },
+    {
+      "epoch": 1.5890677025527191,
+      "grad_norm": 0.33104994893074036,
+      "learning_rate": 5.3534007467207024e-06,
+      "loss": 0.4203,
+      "step": 1909
+    },
+    {
+      "epoch": 1.5899001109877915,
+      "grad_norm": 0.33715182542800903,
+      "learning_rate": 5.348567553132862e-06,
+      "loss": 0.4237,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5907325194228634,
+      "grad_norm": 0.3445633053779602,
+      "learning_rate": 5.343734032233986e-06,
+      "loss": 0.4466,
+      "step": 1911
+    },
+    {
+      "epoch": 1.5915649278579356,
+      "grad_norm": 0.3408707082271576,
+      "learning_rate": 5.338900188562836e-06,
+      "loss": 0.3845,
+      "step": 1912
+    },
+    {
+      "epoch": 1.5923973362930077,
+      "grad_norm": 0.3361123204231262,
+      "learning_rate": 5.334066026658475e-06,
+      "loss": 0.4134,
+      "step": 1913
+    },
+    {
+      "epoch": 1.5932297447280799,
+      "grad_norm": 0.3556373119354248,
+      "learning_rate": 5.329231551060264e-06,
+      "loss": 0.416,
+      "step": 1914
+    },
+    {
+      "epoch": 1.594062153163152,
+      "grad_norm": 0.34920355677604675,
+      "learning_rate": 5.324396766307863e-06,
+      "loss": 0.4377,
+      "step": 1915
+    },
+    {
+      "epoch": 1.5948945615982242,
+      "grad_norm": 0.3253536820411682,
+      "learning_rate": 5.31956167694122e-06,
+      "loss": 0.4011,
+      "step": 1916
+    },
+    {
+      "epoch": 1.5957269700332963,
+      "grad_norm": 0.35277408361434937,
+      "learning_rate": 5.314726287500565e-06,
+      "loss": 0.4428,
+      "step": 1917
+    },
+    {
+      "epoch": 1.5965593784683685,
+      "grad_norm": 0.36568132042884827,
+      "learning_rate": 5.309890602526416e-06,
+      "loss": 0.438,
+      "step": 1918
+    },
+    {
+      "epoch": 1.5973917869034406,
+      "grad_norm": 0.31992262601852417,
+      "learning_rate": 5.305054626559565e-06,
+      "loss": 0.3983,
+      "step": 1919
+    },
+    {
+      "epoch": 1.5982241953385128,
+      "grad_norm": 0.3432292640209198,
+      "learning_rate": 5.30021836414108e-06,
+      "loss": 0.446,
+      "step": 1920
+    },
+    {
+      "epoch": 1.599056603773585,
+      "grad_norm": 0.3381594121456146,
+      "learning_rate": 5.295381819812293e-06,
+      "loss": 0.43,
+      "step": 1921
+    },
+    {
+      "epoch": 1.599889012208657,
+      "grad_norm": 0.32569068670272827,
+      "learning_rate": 5.290544998114805e-06,
+      "loss": 0.4224,
+      "step": 1922
+    },
+    {
+      "epoch": 1.6007214206437292,
+      "grad_norm": 0.32701462507247925,
+      "learning_rate": 5.2857079035904764e-06,
+      "loss": 0.4351,
+      "step": 1923
+    },
+    {
+      "epoch": 1.6015538290788012,
+      "grad_norm": 0.32411783933639526,
+      "learning_rate": 5.280870540781425e-06,
+      "loss": 0.4015,
+      "step": 1924
+    },
+    {
+      "epoch": 1.6023862375138735,
+      "grad_norm": 0.35999277234077454,
+      "learning_rate": 5.2760329142300174e-06,
+      "loss": 0.4268,
+      "step": 1925
+    },
+    {
+      "epoch": 1.6032186459489455,
+      "grad_norm": 0.3214718997478485,
+      "learning_rate": 5.271195028478871e-06,
+      "loss": 0.3986,
+      "step": 1926
+    },
+    {
+      "epoch": 1.6040510543840178,
+      "grad_norm": 0.3498198986053467,
+      "learning_rate": 5.266356888070843e-06,
+      "loss": 0.4462,
+      "step": 1927
+    },
+    {
+      "epoch": 1.6048834628190898,
+      "grad_norm": 0.3669768273830414,
+      "learning_rate": 5.261518497549033e-06,
+      "loss": 0.4052,
+      "step": 1928
+    },
+    {
+      "epoch": 1.6057158712541622,
+      "grad_norm": 0.3158254623413086,
+      "learning_rate": 5.256679861456776e-06,
+      "loss": 0.3936,
+      "step": 1929
+    },
+    {
+      "epoch": 1.606548279689234,
+      "grad_norm": 0.33665063977241516,
+      "learning_rate": 5.251840984337634e-06,
+      "loss": 0.4338,
+      "step": 1930
+    },
+    {
+      "epoch": 1.6073806881243065,
+      "grad_norm": 0.37514346837997437,
+      "learning_rate": 5.247001870735398e-06,
+      "loss": 0.464,
+      "step": 1931
+    },
+    {
+      "epoch": 1.6082130965593784,
+      "grad_norm": 0.33054375648498535,
+      "learning_rate": 5.242162525194082e-06,
+      "loss": 0.3726,
+      "step": 1932
+    },
+    {
+      "epoch": 1.6090455049944508,
+      "grad_norm": 0.3735671937465668,
+      "learning_rate": 5.237322952257915e-06,
+      "loss": 0.4846,
+      "step": 1933
+    },
+    {
+      "epoch": 1.6098779134295227,
+      "grad_norm": 0.30594101548194885,
+      "learning_rate": 5.232483156471339e-06,
+      "loss": 0.3673,
+      "step": 1934
+    },
+    {
+      "epoch": 1.6107103218645948,
+      "grad_norm": 0.3897792100906372,
+      "learning_rate": 5.227643142379009e-06,
+      "loss": 0.4665,
+      "step": 1935
+    },
+    {
+      "epoch": 1.611542730299667,
+      "grad_norm": 0.3239308297634125,
+      "learning_rate": 5.222802914525782e-06,
+      "loss": 0.4004,
+      "step": 1936
+    },
+    {
+      "epoch": 1.6123751387347391,
+      "grad_norm": 0.33833977580070496,
+      "learning_rate": 5.217962477456718e-06,
+      "loss": 0.4278,
+      "step": 1937
+    },
+    {
+      "epoch": 1.6132075471698113,
+      "grad_norm": 0.3144145905971527,
+      "learning_rate": 5.21312183571707e-06,
+      "loss": 0.3737,
+      "step": 1938
+    },
+    {
+      "epoch": 1.6140399556048834,
+      "grad_norm": 0.34893402457237244,
+      "learning_rate": 5.208280993852287e-06,
+      "loss": 0.4249,
+      "step": 1939
+    },
+    {
+      "epoch": 1.6148723640399556,
+      "grad_norm": 0.32780539989471436,
+      "learning_rate": 5.203439956408005e-06,
+      "loss": 0.4224,
+      "step": 1940
+    },
+    {
+      "epoch": 1.6157047724750278,
+      "grad_norm": 0.3463350534439087,
+      "learning_rate": 5.198598727930041e-06,
+      "loss": 0.395,
+      "step": 1941
+    },
+    {
+      "epoch": 1.6165371809101,
+      "grad_norm": 0.3848700225353241,
+      "learning_rate": 5.193757312964394e-06,
+      "loss": 0.3988,
+      "step": 1942
+    },
+    {
+      "epoch": 1.617369589345172,
+      "grad_norm": 0.3813340961933136,
+      "learning_rate": 5.188915716057238e-06,
+      "loss": 0.4296,
+      "step": 1943
+    },
+    {
+      "epoch": 1.6182019977802442,
+      "grad_norm": 0.396963894367218,
+      "learning_rate": 5.184073941754916e-06,
+      "loss": 0.473,
+      "step": 1944
+    },
+    {
+      "epoch": 1.6190344062153164,
+      "grad_norm": 0.34140685200691223,
+      "learning_rate": 5.1792319946039405e-06,
+      "loss": 0.3853,
+      "step": 1945
+    },
+    {
+      "epoch": 1.6198668146503885,
+      "grad_norm": 0.3750185966491699,
+      "learning_rate": 5.174389879150985e-06,
+      "loss": 0.4811,
+      "step": 1946
+    },
+    {
+      "epoch": 1.6206992230854604,
+      "grad_norm": 0.28287845849990845,
+      "learning_rate": 5.169547599942877e-06,
+      "loss": 0.3697,
+      "step": 1947
+    },
+    {
+      "epoch": 1.6215316315205328,
+      "grad_norm": 0.3522551357746124,
+      "learning_rate": 5.164705161526605e-06,
+      "loss": 0.482,
+      "step": 1948
+    },
+    {
+      "epoch": 1.6223640399556047,
+      "grad_norm": 0.35294002294540405,
+      "learning_rate": 5.159862568449302e-06,
+      "loss": 0.4473,
+      "step": 1949
+    },
+    {
+      "epoch": 1.6231964483906771,
+      "grad_norm": 0.34549230337142944,
+      "learning_rate": 5.155019825258251e-06,
+      "loss": 0.3954,
+      "step": 1950
+    },
+    {
+      "epoch": 1.624028856825749,
+      "grad_norm": 0.3237864673137665,
+      "learning_rate": 5.1501769365008654e-06,
+      "loss": 0.4288,
+      "step": 1951
+    },
+    {
+      "epoch": 1.6248612652608214,
+      "grad_norm": 0.3312610685825348,
+      "learning_rate": 5.14533390672471e-06,
+      "loss": 0.4135,
+      "step": 1952
+    },
+    {
+      "epoch": 1.6256936736958933,
+      "grad_norm": 0.39398959279060364,
+      "learning_rate": 5.140490740477471e-06,
+      "loss": 0.4262,
+      "step": 1953
+    },
+    {
+      "epoch": 1.6265260821309657,
+      "grad_norm": 0.30289697647094727,
+      "learning_rate": 5.135647442306966e-06,
+      "loss": 0.3799,
+      "step": 1954
+    },
+    {
+      "epoch": 1.6273584905660377,
+      "grad_norm": 0.3318493366241455,
+      "learning_rate": 5.130804016761138e-06,
+      "loss": 0.436,
+      "step": 1955
+    },
+    {
+      "epoch": 1.62819089900111,
+      "grad_norm": 0.3526514172554016,
+      "learning_rate": 5.1259604683880485e-06,
+      "loss": 0.4636,
+      "step": 1956
+    },
+    {
+      "epoch": 1.629023307436182,
+      "grad_norm": 0.3857317864894867,
+      "learning_rate": 5.121116801735873e-06,
+      "loss": 0.443,
+      "step": 1957
+    },
+    {
+      "epoch": 1.629855715871254,
+      "grad_norm": 0.3597477078437805,
+      "learning_rate": 5.1162730213529e-06,
+      "loss": 0.4177,
+      "step": 1958
+    },
+    {
+      "epoch": 1.6306881243063263,
+      "grad_norm": 0.34153300523757935,
+      "learning_rate": 5.1114291317875244e-06,
+      "loss": 0.4612,
+      "step": 1959
+    },
+    {
+      "epoch": 1.6315205327413984,
+      "grad_norm": 0.3484551012516022,
+      "learning_rate": 5.1065851375882425e-06,
+      "loss": 0.4088,
+      "step": 1960
+    },
+    {
+      "epoch": 1.6323529411764706,
+      "grad_norm": 0.36641940474510193,
+      "learning_rate": 5.101741043303651e-06,
+      "loss": 0.4182,
+      "step": 1961
+    },
+    {
+      "epoch": 1.6331853496115427,
+      "grad_norm": 0.3681751787662506,
+      "learning_rate": 5.096896853482437e-06,
+      "loss": 0.4161,
+      "step": 1962
+    },
+    {
+      "epoch": 1.6340177580466149,
+      "grad_norm": 0.3790402114391327,
+      "learning_rate": 5.092052572673383e-06,
+      "loss": 0.4297,
+      "step": 1963
+    },
+    {
+      "epoch": 1.634850166481687,
+      "grad_norm": 0.3269292116165161,
+      "learning_rate": 5.087208205425349e-06,
+      "loss": 0.3941,
+      "step": 1964
+    },
+    {
+      "epoch": 1.6356825749167592,
+      "grad_norm": 0.3894573748111725,
+      "learning_rate": 5.082363756287285e-06,
+      "loss": 0.4084,
+      "step": 1965
+    },
+    {
+      "epoch": 1.6365149833518313,
+      "grad_norm": 0.39526158571243286,
+      "learning_rate": 5.077519229808211e-06,
+      "loss": 0.4229,
+      "step": 1966
+    },
+    {
+      "epoch": 1.6373473917869035,
+      "grad_norm": 0.3372214436531067,
+      "learning_rate": 5.072674630537223e-06,
+      "loss": 0.4109,
+      "step": 1967
+    },
+    {
+      "epoch": 1.6381798002219756,
+      "grad_norm": 0.4343195855617523,
+      "learning_rate": 5.067829963023485e-06,
+      "loss": 0.4377,
+      "step": 1968
+    },
+    {
+      "epoch": 1.6390122086570478,
+      "grad_norm": 0.3801560699939728,
+      "learning_rate": 5.062985231816225e-06,
+      "loss": 0.4452,
+      "step": 1969
+    },
+    {
+      "epoch": 1.6398446170921197,
+      "grad_norm": 0.3717063367366791,
+      "learning_rate": 5.0581404414647276e-06,
+      "loss": 0.4361,
+      "step": 1970
+    },
+    {
+      "epoch": 1.640677025527192,
+      "grad_norm": 0.3857610523700714,
+      "learning_rate": 5.053295596518337e-06,
+      "loss": 0.4063,
+      "step": 1971
+    },
+    {
+      "epoch": 1.641509433962264,
+      "grad_norm": 0.3806234300136566,
+      "learning_rate": 5.04845070152645e-06,
+      "loss": 0.4423,
+      "step": 1972
+    },
+    {
+      "epoch": 1.6423418423973364,
+      "grad_norm": 0.39662259817123413,
+      "learning_rate": 5.043605761038505e-06,
+      "loss": 0.4482,
+      "step": 1973
+    },
+    {
+      "epoch": 1.6431742508324083,
+      "grad_norm": 0.3193778395652771,
+      "learning_rate": 5.038760779603989e-06,
+      "loss": 0.3858,
+      "step": 1974
+    },
+    {
+      "epoch": 1.6440066592674807,
+      "grad_norm": 0.35678938031196594,
+      "learning_rate": 5.033915761772419e-06,
+      "loss": 0.4691,
+      "step": 1975
+    },
+    {
+      "epoch": 1.6448390677025526,
+      "grad_norm": 0.3585291802883148,
+      "learning_rate": 5.029070712093357e-06,
+      "loss": 0.3971,
+      "step": 1976
+    },
+    {
+      "epoch": 1.645671476137625,
+      "grad_norm": 0.3424661457538605,
+      "learning_rate": 5.024225635116386e-06,
+      "loss": 0.4139,
+      "step": 1977
+    },
+    {
+      "epoch": 1.646503884572697,
+      "grad_norm": 0.35479363799095154,
+      "learning_rate": 5.01938053539112e-06,
+      "loss": 0.4225,
+      "step": 1978
+    },
+    {
+      "epoch": 1.6473362930077693,
+      "grad_norm": 0.3378525972366333,
+      "learning_rate": 5.014535417467191e-06,
+      "loss": 0.4122,
+      "step": 1979
+    },
+    {
+      "epoch": 1.6481687014428412,
+      "grad_norm": 0.36042535305023193,
+      "learning_rate": 5.009690285894252e-06,
+      "loss": 0.4601,
+      "step": 1980
+    },
+    {
+      "epoch": 1.6490011098779136,
+      "grad_norm": 0.3910295069217682,
+      "learning_rate": 5.004845145221965e-06,
+      "loss": 0.4279,
+      "step": 1981
+    },
+    {
+      "epoch": 1.6498335183129855,
+      "grad_norm": 0.3552689552307129,
+      "learning_rate": 5e-06,
+      "loss": 0.4187,
+      "step": 1982
+    },
+    {
+      "epoch": 1.6506659267480577,
+      "grad_norm": 0.33062514662742615,
+      "learning_rate": 4.995154854778036e-06,
+      "loss": 0.4125,
+      "step": 1983
+    },
+    {
+      "epoch": 1.6514983351831298,
+      "grad_norm": 0.3373256325721741,
+      "learning_rate": 4.99030971410575e-06,
+      "loss": 0.444,
+      "step": 1984
+    },
+    {
+      "epoch": 1.652330743618202,
+      "grad_norm": 0.3146938383579254,
+      "learning_rate": 4.9854645825328096e-06,
+      "loss": 0.3551,
+      "step": 1985
+    },
+    {
+      "epoch": 1.6531631520532741,
+      "grad_norm": 0.34619879722595215,
+      "learning_rate": 4.980619464608881e-06,
+      "loss": 0.4253,
+      "step": 1986
+    },
+    {
+      "epoch": 1.6539955604883463,
+      "grad_norm": 0.3549228012561798,
+      "learning_rate": 4.975774364883617e-06,
+      "loss": 0.434,
+      "step": 1987
+    },
+    {
+      "epoch": 1.6548279689234184,
+      "grad_norm": 0.3266083896160126,
+      "learning_rate": 4.9709292879066464e-06,
+      "loss": 0.4061,
+      "step": 1988
+    },
+    {
+      "epoch": 1.6556603773584906,
+      "grad_norm": 0.30864620208740234,
+      "learning_rate": 4.966084238227582e-06,
+      "loss": 0.4607,
+      "step": 1989
+    },
+    {
+      "epoch": 1.6564927857935627,
+      "grad_norm": 0.357719361782074,
+      "learning_rate": 4.961239220396014e-06,
+      "loss": 0.4443,
+      "step": 1990
+    },
+    {
+      "epoch": 1.657325194228635,
+      "grad_norm": 0.36556276679039,
+      "learning_rate": 4.956394238961497e-06,
+      "loss": 0.3715,
+      "step": 1991
+    },
+    {
+      "epoch": 1.658157602663707,
+      "grad_norm": 0.3319959342479706,
+      "learning_rate": 4.951549298473552e-06,
+      "loss": 0.4252,
+      "step": 1992
+    },
+    {
+      "epoch": 1.658990011098779,
+      "grad_norm": 0.35500237345695496,
+      "learning_rate": 4.946704403481663e-06,
+      "loss": 0.4447,
+      "step": 1993
+    },
+    {
+      "epoch": 1.6598224195338513,
+      "grad_norm": 0.38562414050102234,
+      "learning_rate": 4.941859558535275e-06,
+      "loss": 0.4528,
+      "step": 1994
+    },
+    {
+      "epoch": 1.6606548279689233,
+      "grad_norm": 0.3319378197193146,
+      "learning_rate": 4.937014768183778e-06,
+      "loss": 0.3608,
+      "step": 1995
+    },
+    {
+      "epoch": 1.6614872364039956,
+      "grad_norm": 0.34083935618400574,
+      "learning_rate": 4.9321700369765165e-06,
+      "loss": 0.4559,
+      "step": 1996
+    },
+    {
+      "epoch": 1.6623196448390676,
+      "grad_norm": 0.3207685947418213,
+      "learning_rate": 4.927325369462777e-06,
+      "loss": 0.375,
+      "step": 1997
+    },
+    {
+      "epoch": 1.66315205327414,
+      "grad_norm": 0.3647385835647583,
+      "learning_rate": 4.92248077019179e-06,
+      "loss": 0.4085,
+      "step": 1998
+    },
+    {
+      "epoch": 1.6639844617092119,
+      "grad_norm": 0.3710618317127228,
+      "learning_rate": 4.917636243712716e-06,
+      "loss": 0.4605,
+      "step": 1999
+    },
+    {
+      "epoch": 1.6648168701442843,
+      "grad_norm": 0.35661581158638,
+      "learning_rate": 4.912791794574653e-06,
+      "loss": 0.407,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6656492785793562,
+      "grad_norm": 0.39443153142929077,
+      "learning_rate": 4.9079474273266195e-06,
+      "loss": 0.4512,
+      "step": 2001
+    },
+    {
+      "epoch": 1.6664816870144286,
+      "grad_norm": 0.36028945446014404,
+      "learning_rate": 4.903103146517564e-06,
+      "loss": 0.4333,
+      "step": 2002
+    },
+    {
+      "epoch": 1.6673140954495005,
+      "grad_norm": 0.3345014154911041,
+      "learning_rate": 4.898258956696351e-06,
+      "loss": 0.3976,
+      "step": 2003
+    },
+    {
+      "epoch": 1.6681465038845729,
+      "grad_norm": 0.3746415376663208,
+      "learning_rate": 4.893414862411759e-06,
+      "loss": 0.4256,
+      "step": 2004
+    },
+    {
+      "epoch": 1.6689789123196448,
+      "grad_norm": 0.35627481341362,
+      "learning_rate": 4.888570868212478e-06,
+      "loss": 0.4167,
+      "step": 2005
+    },
+    {
+      "epoch": 1.669811320754717,
+      "grad_norm": 0.30849483609199524,
+      "learning_rate": 4.883726978647101e-06,
+      "loss": 0.3915,
+      "step": 2006
+    },
+    {
+      "epoch": 1.670643729189789,
+      "grad_norm": 0.3932332396507263,
+      "learning_rate": 4.878883198264129e-06,
+      "loss": 0.461,
+      "step": 2007
+    },
+    {
+      "epoch": 1.6714761376248612,
+      "grad_norm": 0.32576408982276917,
+      "learning_rate": 4.874039531611954e-06,
+      "loss": 0.4585,
+      "step": 2008
+    },
+    {
+      "epoch": 1.6723085460599334,
+      "grad_norm": 0.3283080756664276,
+      "learning_rate": 4.8691959832388635e-06,
+      "loss": 0.3801,
+      "step": 2009
+    },
+    {
+      "epoch": 1.6731409544950056,
+      "grad_norm": 0.40403053164482117,
+      "learning_rate": 4.864352557693035e-06,
+      "loss": 0.4624,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6739733629300777,
+      "grad_norm": 0.3406950831413269,
+      "learning_rate": 4.859509259522531e-06,
+      "loss": 0.4116,
+      "step": 2011
+    },
+    {
+      "epoch": 1.6748057713651499,
+      "grad_norm": 0.3636447787284851,
+      "learning_rate": 4.854666093275291e-06,
+      "loss": 0.4214,
+      "step": 2012
+    },
+    {
+      "epoch": 1.675638179800222,
+      "grad_norm": 0.34086188673973083,
+      "learning_rate": 4.849823063499136e-06,
+      "loss": 0.4077,
+      "step": 2013
+    },
+    {
+      "epoch": 1.6764705882352942,
+      "grad_norm": 0.345948189496994,
+      "learning_rate": 4.844980174741752e-06,
+      "loss": 0.4133,
+      "step": 2014
+    },
+    {
+      "epoch": 1.6773029966703663,
+      "grad_norm": 0.339345782995224,
+      "learning_rate": 4.840137431550698e-06,
+      "loss": 0.431,
+      "step": 2015
+    },
+    {
+      "epoch": 1.6781354051054382,
+      "grad_norm": 0.3828071057796478,
+      "learning_rate": 4.835294838473396e-06,
+      "loss": 0.4574,
+      "step": 2016
+    },
+    {
+      "epoch": 1.6789678135405106,
+      "grad_norm": 0.3178552985191345,
+      "learning_rate": 4.8304524000571255e-06,
+      "loss": 0.4198,
+      "step": 2017
+    },
+    {
+      "epoch": 1.6798002219755825,
+      "grad_norm": 0.35455217957496643,
+      "learning_rate": 4.825610120849018e-06,
+      "loss": 0.4129,
+      "step": 2018
+    },
+    {
+      "epoch": 1.680632630410655,
+      "grad_norm": 0.35618969798088074,
+      "learning_rate": 4.8207680053960594e-06,
+      "loss": 0.4373,
+      "step": 2019
+    },
+    {
+      "epoch": 1.6814650388457268,
+      "grad_norm": 0.33656421303749084,
+      "learning_rate": 4.815926058245085e-06,
+      "loss": 0.4209,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6822974472807992,
+      "grad_norm": 0.34791067242622375,
+      "learning_rate": 4.811084283942764e-06,
+      "loss": 0.4398,
+      "step": 2021
+    },
+    {
+      "epoch": 1.6831298557158711,
+      "grad_norm": 0.32604023814201355,
+      "learning_rate": 4.806242687035608e-06,
+      "loss": 0.4346,
+      "step": 2022
+    },
+    {
+      "epoch": 1.6839622641509435,
+      "grad_norm": 0.3848741352558136,
+      "learning_rate": 4.80140127206996e-06,
+      "loss": 0.4152,
+      "step": 2023
+    },
+    {
+      "epoch": 1.6847946725860155,
+      "grad_norm": 0.34885889291763306,
+      "learning_rate": 4.796560043591996e-06,
+      "loss": 0.3928,
+      "step": 2024
+    },
+    {
+      "epoch": 1.6856270810210878,
+      "grad_norm": 0.3154467046260834,
+      "learning_rate": 4.791719006147714e-06,
+      "loss": 0.4268,
+      "step": 2025
+    },
+    {
+      "epoch": 1.6864594894561598,
+      "grad_norm": 0.34713122248649597,
+      "learning_rate": 4.7868781642829326e-06,
+      "loss": 0.4561,
+      "step": 2026
+    },
+    {
+      "epoch": 1.6872918978912321,
+      "grad_norm": 0.2982744872570038,
+      "learning_rate": 4.782037522543283e-06,
+      "loss": 0.3945,
+      "step": 2027
+    },
+    {
+      "epoch": 1.688124306326304,
+      "grad_norm": 0.33279117941856384,
+      "learning_rate": 4.777197085474219e-06,
+      "loss": 0.4445,
+      "step": 2028
+    },
+    {
+      "epoch": 1.6889567147613762,
+      "grad_norm": 0.3618778586387634,
+      "learning_rate": 4.772356857620992e-06,
+      "loss": 0.4938,
+      "step": 2029
+    },
+    {
+      "epoch": 1.6897891231964484,
+      "grad_norm": 0.311985582113266,
+      "learning_rate": 4.767516843528664e-06,
+      "loss": 0.4027,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6906215316315205,
+      "grad_norm": 0.30986541509628296,
+      "learning_rate": 4.762677047742088e-06,
+      "loss": 0.3541,
+      "step": 2031
+    },
+    {
+      "epoch": 1.6914539400665927,
+      "grad_norm": 0.3394172191619873,
+      "learning_rate": 4.757837474805918e-06,
+      "loss": 0.4316,
+      "step": 2032
+    },
+    {
+      "epoch": 1.6922863485016648,
+      "grad_norm": 0.331486314535141,
+      "learning_rate": 4.7529981292646025e-06,
+      "loss": 0.4357,
+      "step": 2033
+    },
+    {
+      "epoch": 1.693118756936737,
+      "grad_norm": 0.3562561869621277,
+      "learning_rate": 4.748159015662367e-06,
+      "loss": 0.4528,
+      "step": 2034
+    },
+    {
+      "epoch": 1.6939511653718091,
+      "grad_norm": 0.3469091057777405,
+      "learning_rate": 4.743320138543225e-06,
+      "loss": 0.4298,
+      "step": 2035
+    },
+    {
+      "epoch": 1.6947835738068813,
+      "grad_norm": 0.3128523528575897,
+      "learning_rate": 4.738481502450967e-06,
+      "loss": 0.43,
+      "step": 2036
+    },
+    {
+      "epoch": 1.6956159822419534,
+      "grad_norm": 0.36427655816078186,
+      "learning_rate": 4.733643111929159e-06,
+      "loss": 0.4474,
+      "step": 2037
+    },
+    {
+      "epoch": 1.6964483906770256,
+      "grad_norm": 0.32562559843063354,
+      "learning_rate": 4.728804971521132e-06,
+      "loss": 0.4002,
+      "step": 2038
+    },
+    {
+      "epoch": 1.6972807991120975,
+      "grad_norm": 0.3482353091239929,
+      "learning_rate": 4.723967085769985e-06,
+      "loss": 0.3841,
+      "step": 2039
+    },
+    {
+      "epoch": 1.6981132075471699,
+      "grad_norm": 0.3613423705101013,
+      "learning_rate": 4.719129459218575e-06,
+      "loss": 0.4263,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6989456159822418,
+      "grad_norm": 0.3004484474658966,
+      "learning_rate": 4.714292096409524e-06,
+      "loss": 0.389,
+      "step": 2041
+    },
+    {
+      "epoch": 1.6997780244173142,
+      "grad_norm": 0.3427788317203522,
+      "learning_rate": 4.709455001885196e-06,
+      "loss": 0.4658,
+      "step": 2042
+    },
+    {
+      "epoch": 1.7006104328523861,
+      "grad_norm": 0.3731909394264221,
+      "learning_rate": 4.704618180187709e-06,
+      "loss": 0.4452,
+      "step": 2043
+    },
+    {
+      "epoch": 1.7014428412874585,
+      "grad_norm": 0.34222519397735596,
+      "learning_rate": 4.699781635858923e-06,
+      "loss": 0.4187,
+      "step": 2044
+    },
+    {
+      "epoch": 1.7022752497225304,
+      "grad_norm": 0.32763800024986267,
+      "learning_rate": 4.694945373440435e-06,
+      "loss": 0.4081,
+      "step": 2045
+    },
+    {
+      "epoch": 1.7031076581576028,
+      "grad_norm": 0.3098640739917755,
+      "learning_rate": 4.690109397473586e-06,
+      "loss": 0.3837,
+      "step": 2046
+    },
+    {
+      "epoch": 1.7039400665926747,
+      "grad_norm": 0.3192623555660248,
+      "learning_rate": 4.685273712499436e-06,
+      "loss": 0.429,
+      "step": 2047
+    },
+    {
+      "epoch": 1.704772475027747,
+      "grad_norm": 0.34456512331962585,
+      "learning_rate": 4.680438323058783e-06,
+      "loss": 0.4263,
+      "step": 2048
+    },
+    {
+      "epoch": 1.705604883462819,
+      "grad_norm": 0.3312320411205292,
+      "learning_rate": 4.675603233692137e-06,
+      "loss": 0.4054,
+      "step": 2049
+    },
+    {
+      "epoch": 1.7064372918978914,
+      "grad_norm": 0.33208805322647095,
+      "learning_rate": 4.670768448939737e-06,
+      "loss": 0.3949,
+      "step": 2050
+    },
+    {
+      "epoch": 1.7072697003329633,
+      "grad_norm": 0.3596723973751068,
+      "learning_rate": 4.665933973341527e-06,
+      "loss": 0.4395,
+      "step": 2051
+    },
+    {
+      "epoch": 1.7081021087680355,
+      "grad_norm": 0.3322398066520691,
+      "learning_rate": 4.661099811437166e-06,
+      "loss": 0.399,
+      "step": 2052
+    },
+    {
+      "epoch": 1.7089345172031076,
+      "grad_norm": 0.3125096261501312,
+      "learning_rate": 4.656265967766014e-06,
+      "loss": 0.4142,
+      "step": 2053
+    },
+    {
+      "epoch": 1.7097669256381798,
+      "grad_norm": 0.3498634696006775,
+      "learning_rate": 4.651432446867139e-06,
+      "loss": 0.4149,
+      "step": 2054
+    },
+    {
+      "epoch": 1.710599334073252,
+      "grad_norm": 0.34815841913223267,
+      "learning_rate": 4.646599253279299e-06,
+      "loss": 0.4006,
+      "step": 2055
+    },
+    {
+      "epoch": 1.711431742508324,
+      "grad_norm": 0.3256421983242035,
+      "learning_rate": 4.641766391540949e-06,
+      "loss": 0.4225,
+      "step": 2056
+    },
+    {
+      "epoch": 1.7122641509433962,
+      "grad_norm": 0.3041287362575531,
+      "learning_rate": 4.636933866190228e-06,
+      "loss": 0.4198,
+      "step": 2057
+    },
+    {
+      "epoch": 1.7130965593784684,
+      "grad_norm": 0.3252321481704712,
+      "learning_rate": 4.632101681764964e-06,
+      "loss": 0.4309,
+      "step": 2058
+    },
+    {
+      "epoch": 1.7139289678135405,
+      "grad_norm": 0.3307684361934662,
+      "learning_rate": 4.627269842802664e-06,
+      "loss": 0.4331,
+      "step": 2059
+    },
+    {
+      "epoch": 1.7147613762486127,
+      "grad_norm": 0.34094369411468506,
+      "learning_rate": 4.622438353840506e-06,
+      "loss": 0.4036,
+      "step": 2060
+    },
+    {
+      "epoch": 1.7155937846836848,
+      "grad_norm": 0.31734779477119446,
+      "learning_rate": 4.617607219415346e-06,
+      "loss": 0.4096,
+      "step": 2061
+    },
+    {
+      "epoch": 1.7164261931187568,
+      "grad_norm": 0.3557928502559662,
+      "learning_rate": 4.6127764440637e-06,
+      "loss": 0.4343,
+      "step": 2062
+    },
+    {
+      "epoch": 1.7172586015538291,
+      "grad_norm": 0.35101959109306335,
+      "learning_rate": 4.607946032321755e-06,
+      "loss": 0.418,
+      "step": 2063
+    },
+    {
+      "epoch": 1.718091009988901,
+      "grad_norm": 0.35029613971710205,
+      "learning_rate": 4.603115988725351e-06,
+      "loss": 0.4141,
+      "step": 2064
+    },
+    {
+      "epoch": 1.7189234184239734,
+      "grad_norm": 0.333061546087265,
+      "learning_rate": 4.598286317809983e-06,
+      "loss": 0.4064,
+      "step": 2065
+    },
+    {
+      "epoch": 1.7197558268590454,
+      "grad_norm": 0.32927843928337097,
+      "learning_rate": 4.593457024110792e-06,
+      "loss": 0.4149,
+      "step": 2066
+    },
+    {
+      "epoch": 1.7205882352941178,
+      "grad_norm": 0.33346185088157654,
+      "learning_rate": 4.588628112162578e-06,
+      "loss": 0.4092,
+      "step": 2067
+    },
+    {
+      "epoch": 1.7214206437291897,
+      "grad_norm": 0.32560572028160095,
+      "learning_rate": 4.583799586499768e-06,
+      "loss": 0.4182,
+      "step": 2068
+    },
+    {
+      "epoch": 1.722253052164262,
+      "grad_norm": 0.3583771586418152,
+      "learning_rate": 4.578971451656435e-06,
+      "loss": 0.4472,
+      "step": 2069
+    },
+    {
+      "epoch": 1.723085460599334,
+      "grad_norm": 0.3493313491344452,
+      "learning_rate": 4.574143712166279e-06,
+      "loss": 0.4249,
+      "step": 2070
+    },
+    {
+      "epoch": 1.7239178690344064,
+      "grad_norm": 0.3568575978279114,
+      "learning_rate": 4.569316372562634e-06,
+      "loss": 0.4193,
+      "step": 2071
+    },
+    {
+      "epoch": 1.7247502774694783,
+      "grad_norm": 0.33443883061408997,
+      "learning_rate": 4.564489437378457e-06,
+      "loss": 0.4556,
+      "step": 2072
+    },
+    {
+      "epoch": 1.7255826859045507,
+      "grad_norm": 0.31167763471603394,
+      "learning_rate": 4.559662911146324e-06,
+      "loss": 0.3944,
+      "step": 2073
+    },
+    {
+      "epoch": 1.7264150943396226,
+      "grad_norm": 0.32042884826660156,
+      "learning_rate": 4.554836798398425e-06,
+      "loss": 0.4126,
+      "step": 2074
+    },
+    {
+      "epoch": 1.7272475027746947,
+      "grad_norm": 0.33394762873649597,
+      "learning_rate": 4.550011103666568e-06,
+      "loss": 0.4307,
+      "step": 2075
+    },
+    {
+      "epoch": 1.728079911209767,
+      "grad_norm": 0.3465491831302643,
+      "learning_rate": 4.545185831482166e-06,
+      "loss": 0.4582,
+      "step": 2076
+    },
+    {
+      "epoch": 1.728912319644839,
+      "grad_norm": 0.30535024404525757,
+      "learning_rate": 4.5403609863762295e-06,
+      "loss": 0.3633,
+      "step": 2077
+    },
+    {
+      "epoch": 1.7297447280799112,
+      "grad_norm": 0.3142050802707672,
+      "learning_rate": 4.535536572879376e-06,
+      "loss": 0.4242,
+      "step": 2078
+    },
+    {
+      "epoch": 1.7305771365149833,
+      "grad_norm": 0.3482121229171753,
+      "learning_rate": 4.53071259552181e-06,
+      "loss": 0.447,
+      "step": 2079
+    },
+    {
+      "epoch": 1.7314095449500555,
+      "grad_norm": 0.3179130256175995,
+      "learning_rate": 4.525889058833337e-06,
+      "loss": 0.4052,
+      "step": 2080
+    },
+    {
+      "epoch": 1.7322419533851277,
+      "grad_norm": 0.2993527352809906,
+      "learning_rate": 4.5210659673433386e-06,
+      "loss": 0.3627,
+      "step": 2081
+    },
+    {
+      "epoch": 1.7330743618201998,
+      "grad_norm": 0.3593730032444,
+      "learning_rate": 4.516243325580782e-06,
+      "loss": 0.4697,
+      "step": 2082
+    },
+    {
+      "epoch": 1.733906770255272,
+      "grad_norm": 0.3211425542831421,
+      "learning_rate": 4.511421138074213e-06,
+      "loss": 0.401,
+      "step": 2083
+    },
+    {
+      "epoch": 1.734739178690344,
+      "grad_norm": 0.33504122495651245,
+      "learning_rate": 4.50659940935175e-06,
+      "loss": 0.4107,
+      "step": 2084
+    },
+    {
+      "epoch": 1.7355715871254163,
+      "grad_norm": 0.36016252636909485,
+      "learning_rate": 4.5017781439410806e-06,
+      "loss": 0.4275,
+      "step": 2085
+    },
+    {
+      "epoch": 1.7364039955604884,
+      "grad_norm": 0.3467753827571869,
+      "learning_rate": 4.496957346369458e-06,
+      "loss": 0.4255,
+      "step": 2086
+    },
+    {
+      "epoch": 1.7372364039955603,
+      "grad_norm": 0.37390729784965515,
+      "learning_rate": 4.492137021163694e-06,
+      "loss": 0.4342,
+      "step": 2087
+    },
+    {
+      "epoch": 1.7380688124306327,
+      "grad_norm": 0.38485151529312134,
+      "learning_rate": 4.4873171728501604e-06,
+      "loss": 0.3993,
+      "step": 2088
+    },
+    {
+      "epoch": 1.7389012208657046,
+      "grad_norm": 0.37274548411369324,
+      "learning_rate": 4.482497805954779e-06,
+      "loss": 0.4139,
+      "step": 2089
+    },
+    {
+      "epoch": 1.739733629300777,
+      "grad_norm": 0.3811771273612976,
+      "learning_rate": 4.477678925003018e-06,
+      "loss": 0.4109,
+      "step": 2090
+    },
+    {
+      "epoch": 1.740566037735849,
+      "grad_norm": 0.37200403213500977,
+      "learning_rate": 4.472860534519893e-06,
+      "loss": 0.4366,
+      "step": 2091
+    },
+    {
+      "epoch": 1.7413984461709213,
+      "grad_norm": 0.3533194065093994,
+      "learning_rate": 4.468042639029952e-06,
+      "loss": 0.3646,
+      "step": 2092
+    },
+    {
+      "epoch": 1.7422308546059933,
+      "grad_norm": 0.3777799904346466,
+      "learning_rate": 4.463225243057292e-06,
+      "loss": 0.4576,
+      "step": 2093
+    },
+    {
+      "epoch": 1.7430632630410656,
+      "grad_norm": 0.34515514969825745,
+      "learning_rate": 4.458408351125525e-06,
+      "loss": 0.421,
+      "step": 2094
+    },
+    {
+      "epoch": 1.7438956714761376,
+      "grad_norm": 0.340119868516922,
+      "learning_rate": 4.453591967757801e-06,
+      "loss": 0.41,
+      "step": 2095
+    },
+    {
+      "epoch": 1.74472807991121,
+      "grad_norm": 0.3491295576095581,
+      "learning_rate": 4.4487760974767835e-06,
+      "loss": 0.4169,
+      "step": 2096
+    },
+    {
+      "epoch": 1.7455604883462819,
+      "grad_norm": 0.35444924235343933,
+      "learning_rate": 4.4439607448046636e-06,
+      "loss": 0.4326,
+      "step": 2097
+    },
+    {
+      "epoch": 1.746392896781354,
+      "grad_norm": 0.36455461382865906,
+      "learning_rate": 4.43914591426314e-06,
+      "loss": 0.3829,
+      "step": 2098
+    },
+    {
+      "epoch": 1.7472253052164262,
+      "grad_norm": 0.3415604531764984,
+      "learning_rate": 4.434331610373424e-06,
+      "loss": 0.4204,
+      "step": 2099
+    },
+    {
+      "epoch": 1.7480577136514983,
+      "grad_norm": 0.3621603846549988,
+      "learning_rate": 4.4295178376562285e-06,
+      "loss": 0.4484,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7488901220865705,
+      "grad_norm": 0.3243803381919861,
+      "learning_rate": 4.424704600631774e-06,
+      "loss": 0.3944,
+      "step": 2101
+    },
+    {
+      "epoch": 1.7497225305216426,
+      "grad_norm": 0.32423099875450134,
+      "learning_rate": 4.419891903819773e-06,
+      "loss": 0.3849,
+      "step": 2102
+    },
+    {
+      "epoch": 1.7505549389567148,
+      "grad_norm": 0.35436517000198364,
+      "learning_rate": 4.4150797517394295e-06,
+      "loss": 0.4817,
+      "step": 2103
+    },
+    {
+      "epoch": 1.751387347391787,
+      "grad_norm": 0.3532561659812927,
+      "learning_rate": 4.410268148909441e-06,
+      "loss": 0.448,
+      "step": 2104
+    },
+    {
+      "epoch": 1.752219755826859,
+      "grad_norm": 0.360914409160614,
+      "learning_rate": 4.405457099847986e-06,
+      "loss": 0.4412,
+      "step": 2105
+    },
+    {
+      "epoch": 1.7530521642619312,
+      "grad_norm": 0.34743252396583557,
+      "learning_rate": 4.400646609072727e-06,
+      "loss": 0.4094,
+      "step": 2106
+    },
+    {
+      "epoch": 1.7538845726970034,
+      "grad_norm": 0.3823837339878082,
+      "learning_rate": 4.395836681100794e-06,
+      "loss": 0.4252,
+      "step": 2107
+    },
+    {
+      "epoch": 1.7547169811320755,
+      "grad_norm": 0.36400553584098816,
+      "learning_rate": 4.391027320448798e-06,
+      "loss": 0.4383,
+      "step": 2108
+    },
+    {
+      "epoch": 1.7555493895671477,
+      "grad_norm": 0.3543497920036316,
+      "learning_rate": 4.386218531632808e-06,
+      "loss": 0.3985,
+      "step": 2109
+    },
+    {
+      "epoch": 1.7563817980022196,
+      "grad_norm": 0.33355042338371277,
+      "learning_rate": 4.3814103191683655e-06,
+      "loss": 0.447,
+      "step": 2110
+    },
+    {
+      "epoch": 1.757214206437292,
+      "grad_norm": 0.31347766518592834,
+      "learning_rate": 4.376602687570464e-06,
+      "loss": 0.4034,
+      "step": 2111
+    },
+    {
+      "epoch": 1.758046614872364,
+      "grad_norm": 0.3570155203342438,
+      "learning_rate": 4.371795641353555e-06,
+      "loss": 0.3879,
+      "step": 2112
+    },
+    {
+      "epoch": 1.7588790233074363,
+      "grad_norm": 0.38342198729515076,
+      "learning_rate": 4.366989185031536e-06,
+      "loss": 0.4377,
+      "step": 2113
+    },
+    {
+      "epoch": 1.7597114317425082,
+      "grad_norm": 0.3486819565296173,
+      "learning_rate": 4.362183323117757e-06,
+      "loss": 0.4352,
+      "step": 2114
+    },
+    {
+      "epoch": 1.7605438401775806,
+      "grad_norm": 0.3399284780025482,
+      "learning_rate": 4.357378060125007e-06,
+      "loss": 0.4193,
+      "step": 2115
+    },
+    {
+      "epoch": 1.7613762486126525,
+      "grad_norm": 0.36731529235839844,
+      "learning_rate": 4.3525734005655085e-06,
+      "loss": 0.4648,
+      "step": 2116
+    },
+    {
+      "epoch": 1.762208657047725,
+      "grad_norm": 0.3027511537075043,
+      "learning_rate": 4.347769348950922e-06,
+      "loss": 0.4082,
+      "step": 2117
+    },
+    {
+      "epoch": 1.7630410654827968,
+      "grad_norm": 0.31376221776008606,
+      "learning_rate": 4.342965909792338e-06,
+      "loss": 0.4031,
+      "step": 2118
+    },
+    {
+      "epoch": 1.7638734739178692,
+      "grad_norm": 0.31292369961738586,
+      "learning_rate": 4.338163087600271e-06,
+      "loss": 0.4102,
+      "step": 2119
+    },
+    {
+      "epoch": 1.7647058823529411,
+      "grad_norm": 0.29499831795692444,
+      "learning_rate": 4.33336088688465e-06,
+      "loss": 0.3865,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7655382907880133,
+      "grad_norm": 0.34691059589385986,
+      "learning_rate": 4.328559312154831e-06,
+      "loss": 0.4631,
+      "step": 2121
+    },
+    {
+      "epoch": 1.7663706992230854,
+      "grad_norm": 0.32868635654449463,
+      "learning_rate": 4.323758367919572e-06,
+      "loss": 0.4256,
+      "step": 2122
+    },
+    {
+      "epoch": 1.7672031076581576,
+      "grad_norm": 0.33149829506874084,
+      "learning_rate": 4.318958058687047e-06,
+      "loss": 0.4242,
+      "step": 2123
+    },
+    {
+      "epoch": 1.7680355160932297,
+      "grad_norm": 0.3130408823490143,
+      "learning_rate": 4.31415838896483e-06,
+      "loss": 0.4296,
+      "step": 2124
+    },
+    {
+      "epoch": 1.7688679245283019,
+      "grad_norm": 0.3201256990432739,
+      "learning_rate": 4.309359363259895e-06,
+      "loss": 0.4092,
+      "step": 2125
+    },
+    {
+      "epoch": 1.769700332963374,
+      "grad_norm": 0.3389187455177307,
+      "learning_rate": 4.304560986078609e-06,
+      "loss": 0.4557,
+      "step": 2126
+    },
+    {
+      "epoch": 1.7705327413984462,
+      "grad_norm": 0.32760846614837646,
+      "learning_rate": 4.299763261926734e-06,
+      "loss": 0.4082,
+      "step": 2127
+    },
+    {
+      "epoch": 1.7713651498335183,
+      "grad_norm": 0.3498232960700989,
+      "learning_rate": 4.294966195309418e-06,
+      "loss": 0.4329,
+      "step": 2128
+    },
+    {
+      "epoch": 1.7721975582685905,
+      "grad_norm": 0.3230178952217102,
+      "learning_rate": 4.2901697907311876e-06,
+      "loss": 0.4315,
+      "step": 2129
+    },
+    {
+      "epoch": 1.7730299667036626,
+      "grad_norm": 0.3200167715549469,
+      "learning_rate": 4.285374052695953e-06,
+      "loss": 0.3921,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7738623751387348,
+      "grad_norm": 0.3524058163166046,
+      "learning_rate": 4.280578985706995e-06,
+      "loss": 0.4731,
+      "step": 2131
+    },
+    {
+      "epoch": 1.774694783573807,
+      "grad_norm": 0.2892106771469116,
+      "learning_rate": 4.2757845942669674e-06,
+      "loss": 0.3912,
+      "step": 2132
+    },
+    {
+      "epoch": 1.7755271920088789,
+      "grad_norm": 0.3491550087928772,
+      "learning_rate": 4.270990882877885e-06,
+      "loss": 0.4164,
+      "step": 2133
+    },
+    {
+      "epoch": 1.7763596004439512,
+      "grad_norm": 0.3432190418243408,
+      "learning_rate": 4.2661978560411274e-06,
+      "loss": 0.4305,
+      "step": 2134
+    },
+    {
+      "epoch": 1.7771920088790232,
+      "grad_norm": 0.2854191064834595,
+      "learning_rate": 4.261405518257434e-06,
+      "loss": 0.3578,
+      "step": 2135
+    },
+    {
+      "epoch": 1.7780244173140956,
+      "grad_norm": 0.36918261647224426,
+      "learning_rate": 4.25661387402689e-06,
+      "loss": 0.4849,
+      "step": 2136
+    },
+    {
+      "epoch": 1.7788568257491675,
+      "grad_norm": 0.3176991939544678,
+      "learning_rate": 4.251822927848934e-06,
+      "loss": 0.3844,
+      "step": 2137
+    },
+    {
+      "epoch": 1.7796892341842399,
+      "grad_norm": 0.3359120488166809,
+      "learning_rate": 4.24703268422235e-06,
+      "loss": 0.4387,
+      "step": 2138
+    },
+    {
+      "epoch": 1.7805216426193118,
+      "grad_norm": 0.3274317681789398,
+      "learning_rate": 4.242243147645257e-06,
+      "loss": 0.4009,
+      "step": 2139
+    },
+    {
+      "epoch": 1.7813540510543842,
+      "grad_norm": 0.3508370518684387,
+      "learning_rate": 4.237454322615118e-06,
+      "loss": 0.4086,
+      "step": 2140
+    },
+    {
+      "epoch": 1.782186459489456,
+      "grad_norm": 0.3434849977493286,
+      "learning_rate": 4.232666213628722e-06,
+      "loss": 0.4478,
+      "step": 2141
+    },
+    {
+      "epoch": 1.7830188679245285,
+      "grad_norm": 0.32151880860328674,
+      "learning_rate": 4.227878825182186e-06,
+      "loss": 0.3955,
+      "step": 2142
+    },
+    {
+      "epoch": 1.7838512763596004,
+      "grad_norm": 0.40656089782714844,
+      "learning_rate": 4.223092161770952e-06,
+      "loss": 0.4138,
+      "step": 2143
+    },
+    {
+      "epoch": 1.7846836847946725,
+      "grad_norm": 0.3283717930316925,
+      "learning_rate": 4.218306227889782e-06,
+      "loss": 0.4156,
+      "step": 2144
+    },
+    {
+      "epoch": 1.7855160932297447,
+      "grad_norm": 0.32961905002593994,
+      "learning_rate": 4.213521028032751e-06,
+      "loss": 0.3817,
+      "step": 2145
+    },
+    {
+      "epoch": 1.7863485016648168,
+      "grad_norm": 0.35188883543014526,
+      "learning_rate": 4.2087365666932456e-06,
+      "loss": 0.4086,
+      "step": 2146
+    },
+    {
+      "epoch": 1.787180910099889,
+      "grad_norm": 0.36771219968795776,
+      "learning_rate": 4.203952848363957e-06,
+      "loss": 0.4582,
+      "step": 2147
+    },
+    {
+      "epoch": 1.7880133185349611,
+      "grad_norm": 0.305984765291214,
+      "learning_rate": 4.199169877536884e-06,
+      "loss": 0.371,
+      "step": 2148
+    },
+    {
+      "epoch": 1.7888457269700333,
+      "grad_norm": 0.3949022591114044,
+      "learning_rate": 4.194387658703317e-06,
+      "loss": 0.4562,
+      "step": 2149
+    },
+    {
+      "epoch": 1.7896781354051055,
+      "grad_norm": 0.3683740496635437,
+      "learning_rate": 4.189606196353844e-06,
+      "loss": 0.4009,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7905105438401776,
+      "grad_norm": 0.32549092173576355,
+      "learning_rate": 4.184825494978342e-06,
+      "loss": 0.4445,
+      "step": 2151
+    },
+    {
+      "epoch": 1.7913429522752498,
+      "grad_norm": 0.35574018955230713,
+      "learning_rate": 4.180045559065974e-06,
+      "loss": 0.4337,
+      "step": 2152
+    },
+    {
+      "epoch": 1.792175360710322,
+      "grad_norm": 0.3322107195854187,
+      "learning_rate": 4.175266393105183e-06,
+      "loss": 0.3973,
+      "step": 2153
+    },
+    {
+      "epoch": 1.793007769145394,
+      "grad_norm": 0.3085898458957672,
+      "learning_rate": 4.1704880015836905e-06,
+      "loss": 0.4016,
+      "step": 2154
+    },
+    {
+      "epoch": 1.7938401775804662,
+      "grad_norm": 0.3286598026752472,
+      "learning_rate": 4.165710388988487e-06,
+      "loss": 0.4377,
+      "step": 2155
+    },
+    {
+      "epoch": 1.7946725860155381,
+      "grad_norm": 0.32184141874313354,
+      "learning_rate": 4.1609335598058355e-06,
+      "loss": 0.3827,
+      "step": 2156
+    },
+    {
+      "epoch": 1.7955049944506105,
+      "grad_norm": 0.3248361051082611,
+      "learning_rate": 4.156157518521264e-06,
+      "loss": 0.4447,
+      "step": 2157
+    },
+    {
+      "epoch": 1.7963374028856824,
+      "grad_norm": 0.35561513900756836,
+      "learning_rate": 4.151382269619558e-06,
+      "loss": 0.4322,
+      "step": 2158
+    },
+    {
+      "epoch": 1.7971698113207548,
+      "grad_norm": 0.33568158745765686,
+      "learning_rate": 4.146607817584759e-06,
+      "loss": 0.417,
+      "step": 2159
+    },
+    {
+      "epoch": 1.7980022197558267,
+      "grad_norm": 0.325347363948822,
+      "learning_rate": 4.14183416690016e-06,
+      "loss": 0.4263,
+      "step": 2160
+    },
+    {
+      "epoch": 1.7988346281908991,
+      "grad_norm": 0.31573471426963806,
+      "learning_rate": 4.137061322048307e-06,
+      "loss": 0.4142,
+      "step": 2161
+    },
+    {
+      "epoch": 1.799667036625971,
+      "grad_norm": 0.33735018968582153,
+      "learning_rate": 4.13228928751098e-06,
+      "loss": 0.4247,
+      "step": 2162
+    },
+    {
+      "epoch": 1.8004994450610434,
+      "grad_norm": 0.3238171339035034,
+      "learning_rate": 4.127518067769206e-06,
+      "loss": 0.3978,
+      "step": 2163
+    },
+    {
+      "epoch": 1.8013318534961154,
+      "grad_norm": 0.33231449127197266,
+      "learning_rate": 4.122747667303242e-06,
+      "loss": 0.4361,
+      "step": 2164
+    },
+    {
+      "epoch": 1.8021642619311877,
+      "grad_norm": 0.32602575421333313,
+      "learning_rate": 4.11797809059258e-06,
+      "loss": 0.3914,
+      "step": 2165
+    },
+    {
+      "epoch": 1.8029966703662597,
+      "grad_norm": 0.29990747570991516,
+      "learning_rate": 4.1132093421159335e-06,
+      "loss": 0.3839,
+      "step": 2166
+    },
+    {
+      "epoch": 1.8038290788013318,
+      "grad_norm": 0.3527059257030487,
+      "learning_rate": 4.108441426351243e-06,
+      "loss": 0.4338,
+      "step": 2167
+    },
+    {
+      "epoch": 1.804661487236404,
+      "grad_norm": 0.3696962893009186,
+      "learning_rate": 4.103674347775663e-06,
+      "loss": 0.4063,
+      "step": 2168
+    },
+    {
+      "epoch": 1.8054938956714761,
+      "grad_norm": 0.3006901144981384,
+      "learning_rate": 4.098908110865563e-06,
+      "loss": 0.3799,
+      "step": 2169
+    },
+    {
+      "epoch": 1.8063263041065483,
+      "grad_norm": 0.350769579410553,
+      "learning_rate": 4.094142720096526e-06,
+      "loss": 0.4381,
+      "step": 2170
+    },
+    {
+      "epoch": 1.8071587125416204,
+      "grad_norm": 0.3283845782279968,
+      "learning_rate": 4.089378179943336e-06,
+      "loss": 0.4216,
+      "step": 2171
+    },
+    {
+      "epoch": 1.8079911209766926,
+      "grad_norm": 0.2863730490207672,
+      "learning_rate": 4.084614494879979e-06,
+      "loss": 0.4155,
+      "step": 2172
+    },
+    {
+      "epoch": 1.8088235294117647,
+      "grad_norm": 0.3665149211883545,
+      "learning_rate": 4.079851669379638e-06,
+      "loss": 0.4351,
+      "step": 2173
+    },
+    {
+      "epoch": 1.8096559378468369,
+      "grad_norm": 0.39417773485183716,
+      "learning_rate": 4.0750897079146924e-06,
+      "loss": 0.4559,
+      "step": 2174
+    },
+    {
+      "epoch": 1.810488346281909,
+      "grad_norm": 0.3096025288105011,
+      "learning_rate": 4.070328614956705e-06,
+      "loss": 0.4012,
+      "step": 2175
+    },
+    {
+      "epoch": 1.8113207547169812,
+      "grad_norm": 0.3567507863044739,
+      "learning_rate": 4.065568394976426e-06,
+      "loss": 0.416,
+      "step": 2176
+    },
+    {
+      "epoch": 1.8121531631520533,
+      "grad_norm": 0.29146334528923035,
+      "learning_rate": 4.060809052443784e-06,
+      "loss": 0.3917,
+      "step": 2177
+    },
+    {
+      "epoch": 1.8129855715871255,
+      "grad_norm": 0.3747054636478424,
+      "learning_rate": 4.056050591827888e-06,
+      "loss": 0.4189,
+      "step": 2178
+    },
+    {
+      "epoch": 1.8138179800221974,
+      "grad_norm": 0.3492047190666199,
+      "learning_rate": 4.051293017597014e-06,
+      "loss": 0.4179,
+      "step": 2179
+    },
+    {
+      "epoch": 1.8146503884572698,
+      "grad_norm": 0.3406948745250702,
+      "learning_rate": 4.046536334218609e-06,
+      "loss": 0.421,
+      "step": 2180
+    },
+    {
+      "epoch": 1.8154827968923417,
+      "grad_norm": 0.34447532892227173,
+      "learning_rate": 4.0417805461592764e-06,
+      "loss": 0.4349,
+      "step": 2181
+    },
+    {
+      "epoch": 1.816315205327414,
+      "grad_norm": 0.3566656708717346,
+      "learning_rate": 4.037025657884793e-06,
+      "loss": 0.4326,
+      "step": 2182
+    },
+    {
+      "epoch": 1.817147613762486,
+      "grad_norm": 0.3476259112358093,
+      "learning_rate": 4.032271673860077e-06,
+      "loss": 0.4038,
+      "step": 2183
+    },
+    {
+      "epoch": 1.8179800221975584,
+      "grad_norm": 0.3235064446926117,
+      "learning_rate": 4.0275185985492025e-06,
+      "loss": 0.412,
+      "step": 2184
+    },
+    {
+      "epoch": 1.8188124306326303,
+      "grad_norm": 0.37167060375213623,
+      "learning_rate": 4.022766436415392e-06,
+      "loss": 0.4514,
+      "step": 2185
+    },
+    {
+      "epoch": 1.8196448390677027,
+      "grad_norm": 0.31849905848503113,
+      "learning_rate": 4.018015191921008e-06,
+      "loss": 0.4254,
+      "step": 2186
+    },
+    {
+      "epoch": 1.8204772475027746,
+      "grad_norm": 0.2981938123703003,
+      "learning_rate": 4.013264869527553e-06,
+      "loss": 0.379,
+      "step": 2187
+    },
+    {
+      "epoch": 1.821309655937847,
+      "grad_norm": 0.36852484941482544,
+      "learning_rate": 4.008515473695663e-06,
+      "loss": 0.4213,
+      "step": 2188
+    },
+    {
+      "epoch": 1.822142064372919,
+      "grad_norm": 0.37913084030151367,
+      "learning_rate": 4.003767008885102e-06,
+      "loss": 0.4507,
+      "step": 2189
+    },
+    {
+      "epoch": 1.822974472807991,
+      "grad_norm": 0.30263346433639526,
+      "learning_rate": 3.999019479554764e-06,
+      "loss": 0.3763,
+      "step": 2190
+    },
+    {
+      "epoch": 1.8238068812430632,
+      "grad_norm": 0.33267539739608765,
+      "learning_rate": 3.9942728901626605e-06,
+      "loss": 0.4024,
+      "step": 2191
+    },
+    {
+      "epoch": 1.8246392896781354,
+      "grad_norm": 0.32982137799263,
+      "learning_rate": 3.989527245165924e-06,
+      "loss": 0.4079,
+      "step": 2192
+    },
+    {
+      "epoch": 1.8254716981132075,
+      "grad_norm": 0.35067951679229736,
+      "learning_rate": 3.984782549020797e-06,
+      "loss": 0.4182,
+      "step": 2193
+    },
+    {
+      "epoch": 1.8263041065482797,
+      "grad_norm": 0.328118234872818,
+      "learning_rate": 3.980038806182629e-06,
+      "loss": 0.4588,
+      "step": 2194
+    },
+    {
+      "epoch": 1.8271365149833518,
+      "grad_norm": 0.37364038825035095,
+      "learning_rate": 3.975296021105885e-06,
+      "loss": 0.4546,
+      "step": 2195
+    },
+    {
+      "epoch": 1.827968923418424,
+      "grad_norm": 0.3203776180744171,
+      "learning_rate": 3.970554198244116e-06,
+      "loss": 0.4091,
+      "step": 2196
+    },
+    {
+      "epoch": 1.8288013318534961,
+      "grad_norm": 0.36616021394729614,
+      "learning_rate": 3.965813342049983e-06,
+      "loss": 0.4341,
+      "step": 2197
+    },
+    {
+      "epoch": 1.8296337402885683,
+      "grad_norm": 0.3346719741821289,
+      "learning_rate": 3.961073456975227e-06,
+      "loss": 0.4067,
+      "step": 2198
+    },
+    {
+      "epoch": 1.8304661487236404,
+      "grad_norm": 0.3459213674068451,
+      "learning_rate": 3.956334547470686e-06,
+      "loss": 0.3825,
+      "step": 2199
+    },
+    {
+      "epoch": 1.8312985571587126,
+      "grad_norm": 0.336300253868103,
+      "learning_rate": 3.95159661798628e-06,
+      "loss": 0.4056,
+      "step": 2200
+    },
+    {
+      "epoch": 1.8321309655937847,
+      "grad_norm": 0.3746432662010193,
+      "learning_rate": 3.946859672971006e-06,
+      "loss": 0.4649,
+      "step": 2201
+    },
+    {
+      "epoch": 1.8329633740288567,
+      "grad_norm": 0.3622547388076782,
+      "learning_rate": 3.9421237168729386e-06,
+      "loss": 0.3967,
+      "step": 2202
+    },
+    {
+      "epoch": 1.833795782463929,
+      "grad_norm": 0.3507779836654663,
+      "learning_rate": 3.937388754139223e-06,
+      "loss": 0.4041,
+      "step": 2203
+    },
+    {
+      "epoch": 1.834628190899001,
+      "grad_norm": 0.37171730399131775,
+      "learning_rate": 3.9326547892160746e-06,
+      "loss": 0.4587,
+      "step": 2204
+    },
+    {
+      "epoch": 1.8354605993340734,
+      "grad_norm": 0.3501371443271637,
+      "learning_rate": 3.927921826548767e-06,
+      "loss": 0.4174,
+      "step": 2205
+    },
+    {
+      "epoch": 1.8362930077691453,
+      "grad_norm": 0.3328426778316498,
+      "learning_rate": 3.923189870581636e-06,
+      "loss": 0.4201,
+      "step": 2206
+    },
+    {
+      "epoch": 1.8371254162042177,
+      "grad_norm": 0.3560123145580292,
+      "learning_rate": 3.918458925758068e-06,
+      "loss": 0.4023,
+      "step": 2207
+    },
+    {
+      "epoch": 1.8379578246392896,
+      "grad_norm": 0.35025811195373535,
+      "learning_rate": 3.9137289965205086e-06,
+      "loss": 0.4353,
+      "step": 2208
+    },
+    {
+      "epoch": 1.838790233074362,
+      "grad_norm": 0.29967784881591797,
+      "learning_rate": 3.909000087310441e-06,
+      "loss": 0.4074,
+      "step": 2209
+    },
+    {
+      "epoch": 1.8396226415094339,
+      "grad_norm": 0.35869383811950684,
+      "learning_rate": 3.9042722025683945e-06,
+      "loss": 0.3998,
+      "step": 2210
+    },
+    {
+      "epoch": 1.8404550499445063,
+      "grad_norm": 0.3381917178630829,
+      "learning_rate": 3.899545346733933e-06,
+      "loss": 0.4623,
+      "step": 2211
+    },
+    {
+      "epoch": 1.8412874583795782,
+      "grad_norm": 0.30779141187667847,
+      "learning_rate": 3.894819524245661e-06,
+      "loss": 0.395,
+      "step": 2212
+    },
+    {
+      "epoch": 1.8421198668146503,
+      "grad_norm": 0.35666483640670776,
+      "learning_rate": 3.890094739541207e-06,
+      "loss": 0.3997,
+      "step": 2213
+    },
+    {
+      "epoch": 1.8429522752497225,
+      "grad_norm": 0.3682236969470978,
+      "learning_rate": 3.885370997057225e-06,
+      "loss": 0.4708,
+      "step": 2214
+    },
+    {
+      "epoch": 1.8437846836847946,
+      "grad_norm": 0.3033082187175751,
+      "learning_rate": 3.880648301229394e-06,
+      "loss": 0.368,
+      "step": 2215
+    },
+    {
+      "epoch": 1.8446170921198668,
+      "grad_norm": 0.335415244102478,
+      "learning_rate": 3.875926656492406e-06,
+      "loss": 0.4432,
+      "step": 2216
+    },
+    {
+      "epoch": 1.845449500554939,
+      "grad_norm": 0.35151904821395874,
+      "learning_rate": 3.871206067279971e-06,
+      "loss": 0.4582,
+      "step": 2217
+    },
+    {
+      "epoch": 1.846281908990011,
+      "grad_norm": 0.32559242844581604,
+      "learning_rate": 3.866486538024802e-06,
+      "loss": 0.4067,
+      "step": 2218
+    },
+    {
+      "epoch": 1.8471143174250833,
+      "grad_norm": 0.3076154589653015,
+      "learning_rate": 3.861768073158623e-06,
+      "loss": 0.4061,
+      "step": 2219
+    },
+    {
+      "epoch": 1.8479467258601554,
+      "grad_norm": 0.29312387108802795,
+      "learning_rate": 3.8570506771121484e-06,
+      "loss": 0.3832,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8487791342952276,
+      "grad_norm": 0.3544323742389679,
+      "learning_rate": 3.852334354315104e-06,
+      "loss": 0.4602,
+      "step": 2221
+    },
+    {
+      "epoch": 1.8496115427302997,
+      "grad_norm": 0.3405257761478424,
+      "learning_rate": 3.847619109196195e-06,
+      "loss": 0.4202,
+      "step": 2222
+    },
+    {
+      "epoch": 1.8504439511653719,
+      "grad_norm": 0.35631346702575684,
+      "learning_rate": 3.842904946183121e-06,
+      "loss": 0.4349,
+      "step": 2223
+    },
+    {
+      "epoch": 1.851276359600444,
+      "grad_norm": 0.35186129808425903,
+      "learning_rate": 3.83819186970256e-06,
+      "loss": 0.4067,
+      "step": 2224
+    },
+    {
+      "epoch": 1.852108768035516,
+      "grad_norm": 0.3434910476207733,
+      "learning_rate": 3.833479884180177e-06,
+      "loss": 0.4053,
+      "step": 2225
+    },
+    {
+      "epoch": 1.8529411764705883,
+      "grad_norm": 0.3569307029247284,
+      "learning_rate": 3.828768994040608e-06,
+      "loss": 0.4356,
+      "step": 2226
+    },
+    {
+      "epoch": 1.8537735849056602,
+      "grad_norm": 0.33176884055137634,
+      "learning_rate": 3.824059203707461e-06,
+      "loss": 0.4422,
+      "step": 2227
+    },
+    {
+      "epoch": 1.8546059933407326,
+      "grad_norm": 0.2953312397003174,
+      "learning_rate": 3.81935051760331e-06,
+      "loss": 0.3875,
+      "step": 2228
+    },
+    {
+      "epoch": 1.8554384017758045,
+      "grad_norm": 0.336987167596817,
+      "learning_rate": 3.8146429401496963e-06,
+      "loss": 0.4396,
+      "step": 2229
+    },
+    {
+      "epoch": 1.856270810210877,
+      "grad_norm": 0.3004775643348694,
+      "learning_rate": 3.8099364757671188e-06,
+      "loss": 0.4045,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8571032186459488,
+      "grad_norm": 0.3074621856212616,
+      "learning_rate": 3.8052311288750255e-06,
+      "loss": 0.3944,
+      "step": 2231
+    },
+    {
+      "epoch": 1.8579356270810212,
+      "grad_norm": 0.31900620460510254,
+      "learning_rate": 3.800526903891823e-06,
+      "loss": 0.4074,
+      "step": 2232
+    },
+    {
+      "epoch": 1.8587680355160932,
+      "grad_norm": 0.3192499876022339,
+      "learning_rate": 3.795823805234857e-06,
+      "loss": 0.4026,
+      "step": 2233
+    },
+    {
+      "epoch": 1.8596004439511655,
+      "grad_norm": 0.3406783640384674,
+      "learning_rate": 3.791121837320425e-06,
+      "loss": 0.4335,
+      "step": 2234
+    },
+    {
+      "epoch": 1.8604328523862375,
+      "grad_norm": 0.33335891366004944,
+      "learning_rate": 3.786421004563753e-06,
+      "loss": 0.4305,
+      "step": 2235
+    },
+    {
+      "epoch": 1.8612652608213096,
+      "grad_norm": 0.30208635330200195,
+      "learning_rate": 3.7817213113790088e-06,
+      "loss": 0.4136,
+      "step": 2236
+    },
+    {
+      "epoch": 1.8620976692563818,
+      "grad_norm": 0.3383508026599884,
+      "learning_rate": 3.7770227621792815e-06,
+      "loss": 0.4167,
+      "step": 2237
+    },
+    {
+      "epoch": 1.862930077691454,
+      "grad_norm": 0.32437726855278015,
+      "learning_rate": 3.7723253613765954e-06,
+      "loss": 0.4547,
+      "step": 2238
+    },
+    {
+      "epoch": 1.863762486126526,
+      "grad_norm": 0.3173195719718933,
+      "learning_rate": 3.767629113381891e-06,
+      "loss": 0.4056,
+      "step": 2239
+    },
+    {
+      "epoch": 1.8645948945615982,
+      "grad_norm": 0.3287826180458069,
+      "learning_rate": 3.762934022605027e-06,
+      "loss": 0.4125,
+      "step": 2240
+    },
+    {
+      "epoch": 1.8654273029966704,
+      "grad_norm": 0.33371874690055847,
+      "learning_rate": 3.758240093454775e-06,
+      "loss": 0.4262,
+      "step": 2241
+    },
+    {
+      "epoch": 1.8662597114317425,
+      "grad_norm": 0.3588714599609375,
+      "learning_rate": 3.7535473303388175e-06,
+      "loss": 0.3972,
+      "step": 2242
+    },
+    {
+      "epoch": 1.8670921198668147,
+      "grad_norm": 0.32598569989204407,
+      "learning_rate": 3.7488557376637436e-06,
+      "loss": 0.3772,
+      "step": 2243
+    },
+    {
+      "epoch": 1.8679245283018868,
+      "grad_norm": 0.3502293825149536,
+      "learning_rate": 3.744165319835037e-06,
+      "loss": 0.4538,
+      "step": 2244
+    },
+    {
+      "epoch": 1.868756936736959,
+      "grad_norm": 0.31929466128349304,
+      "learning_rate": 3.739476081257085e-06,
+      "loss": 0.4287,
+      "step": 2245
+    },
+    {
+      "epoch": 1.8695893451720311,
+      "grad_norm": 0.366178959608078,
+      "learning_rate": 3.7347880263331603e-06,
+      "loss": 0.442,
+      "step": 2246
+    },
+    {
+      "epoch": 1.8704217536071033,
+      "grad_norm": 0.40710756182670593,
+      "learning_rate": 3.730101159465435e-06,
+      "loss": 0.441,
+      "step": 2247
+    },
+    {
+      "epoch": 1.8712541620421752,
+      "grad_norm": 0.3121664226055145,
+      "learning_rate": 3.725415485054955e-06,
+      "loss": 0.3975,
+      "step": 2248
+    },
+    {
+      "epoch": 1.8720865704772476,
+      "grad_norm": 0.3380068242549896,
+      "learning_rate": 3.7207310075016533e-06,
+      "loss": 0.4533,
+      "step": 2249
+    },
+    {
+      "epoch": 1.8729189789123195,
+      "grad_norm": 0.35721099376678467,
+      "learning_rate": 3.716047731204332e-06,
+      "loss": 0.4422,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8737513873473919,
+      "grad_norm": 0.35127249360084534,
+      "learning_rate": 3.711365660560674e-06,
+      "loss": 0.4263,
+      "step": 2251
+    },
+    {
+      "epoch": 1.8745837957824638,
+      "grad_norm": 0.36538636684417725,
+      "learning_rate": 3.706684799967224e-06,
+      "loss": 0.4003,
+      "step": 2252
+    },
+    {
+      "epoch": 1.8754162042175362,
+      "grad_norm": 0.3314541280269623,
+      "learning_rate": 3.702005153819391e-06,
+      "loss": 0.369,
+      "step": 2253
+    },
+    {
+      "epoch": 1.8762486126526081,
+      "grad_norm": 0.35317155718803406,
+      "learning_rate": 3.6973267265114456e-06,
+      "loss": 0.4318,
+      "step": 2254
+    },
+    {
+      "epoch": 1.8770810210876805,
+      "grad_norm": 0.397045373916626,
+      "learning_rate": 3.6926495224365124e-06,
+      "loss": 0.4304,
+      "step": 2255
+    },
+    {
+      "epoch": 1.8779134295227524,
+      "grad_norm": 0.34217485785484314,
+      "learning_rate": 3.6879735459865708e-06,
+      "loss": 0.3961,
+      "step": 2256
+    },
+    {
+      "epoch": 1.8787458379578248,
+      "grad_norm": 0.3433960974216461,
+      "learning_rate": 3.68329880155244e-06,
+      "loss": 0.4528,
+      "step": 2257
+    },
+    {
+      "epoch": 1.8795782463928967,
+      "grad_norm": 0.36350932717323303,
+      "learning_rate": 3.6786252935237886e-06,
+      "loss": 0.4229,
+      "step": 2258
+    },
+    {
+      "epoch": 1.8804106548279689,
+      "grad_norm": 0.3405548334121704,
+      "learning_rate": 3.6739530262891245e-06,
+      "loss": 0.4055,
+      "step": 2259
+    },
+    {
+      "epoch": 1.881243063263041,
+      "grad_norm": 0.3336378037929535,
+      "learning_rate": 3.669282004235787e-06,
+      "loss": 0.3886,
+      "step": 2260
+    },
+    {
+      "epoch": 1.8820754716981132,
+      "grad_norm": 0.3233022093772888,
+      "learning_rate": 3.6646122317499465e-06,
+      "loss": 0.4265,
+      "step": 2261
+    },
+    {
+      "epoch": 1.8829078801331853,
+      "grad_norm": 0.3442718982696533,
+      "learning_rate": 3.6599437132166036e-06,
+      "loss": 0.4322,
+      "step": 2262
+    },
+    {
+      "epoch": 1.8837402885682575,
+      "grad_norm": 0.3712242841720581,
+      "learning_rate": 3.655276453019575e-06,
+      "loss": 0.4382,
+      "step": 2263
+    },
+    {
+      "epoch": 1.8845726970033296,
+      "grad_norm": 0.3307363986968994,
+      "learning_rate": 3.650610455541504e-06,
+      "loss": 0.387,
+      "step": 2264
+    },
+    {
+      "epoch": 1.8854051054384018,
+      "grad_norm": 0.3208552896976471,
+      "learning_rate": 3.6459457251638423e-06,
+      "loss": 0.4071,
+      "step": 2265
+    },
+    {
+      "epoch": 1.886237513873474,
+      "grad_norm": 0.345205157995224,
+      "learning_rate": 3.641282266266853e-06,
+      "loss": 0.4206,
+      "step": 2266
+    },
+    {
+      "epoch": 1.887069922308546,
+      "grad_norm": 0.4148430824279785,
+      "learning_rate": 3.636620083229604e-06,
+      "loss": 0.445,
+      "step": 2267
+    },
+    {
+      "epoch": 1.8879023307436182,
+      "grad_norm": 0.33523353934288025,
+      "learning_rate": 3.6319591804299703e-06,
+      "loss": 0.389,
+      "step": 2268
+    },
+    {
+      "epoch": 1.8887347391786904,
+      "grad_norm": 0.3195513188838959,
+      "learning_rate": 3.6272995622446204e-06,
+      "loss": 0.4133,
+      "step": 2269
+    },
+    {
+      "epoch": 1.8895671476137625,
+      "grad_norm": 0.36299553513526917,
+      "learning_rate": 3.622641233049016e-06,
+      "loss": 0.4297,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8903995560488345,
+      "grad_norm": 0.3698355555534363,
+      "learning_rate": 3.617984197217409e-06,
+      "loss": 0.4338,
+      "step": 2271
+    },
+    {
+      "epoch": 1.8912319644839068,
+      "grad_norm": 0.332303524017334,
+      "learning_rate": 3.6133284591228403e-06,
+      "loss": 0.3874,
+      "step": 2272
+    },
+    {
+      "epoch": 1.8920643729189788,
+      "grad_norm": 0.34952232241630554,
+      "learning_rate": 3.608674023137129e-06,
+      "loss": 0.425,
+      "step": 2273
+    },
+    {
+      "epoch": 1.8928967813540512,
+      "grad_norm": 0.3578547239303589,
+      "learning_rate": 3.6040208936308697e-06,
+      "loss": 0.4576,
+      "step": 2274
+    },
+    {
+      "epoch": 1.893729189789123,
+      "grad_norm": 0.3056720793247223,
+      "learning_rate": 3.599369074973433e-06,
+      "loss": 0.4044,
+      "step": 2275
+    },
+    {
+      "epoch": 1.8945615982241955,
+      "grad_norm": 0.3444676101207733,
+      "learning_rate": 3.5947185715329614e-06,
+      "loss": 0.3809,
+      "step": 2276
+    },
+    {
+      "epoch": 1.8953940066592674,
+      "grad_norm": 0.3490372896194458,
+      "learning_rate": 3.5900693876763556e-06,
+      "loss": 0.4221,
+      "step": 2277
+    },
+    {
+      "epoch": 1.8962264150943398,
+      "grad_norm": 0.3269912600517273,
+      "learning_rate": 3.585421527769283e-06,
+      "loss": 0.4139,
+      "step": 2278
+    },
+    {
+      "epoch": 1.8970588235294117,
+      "grad_norm": 0.3209238350391388,
+      "learning_rate": 3.580774996176162e-06,
+      "loss": 0.3649,
+      "step": 2279
+    },
+    {
+      "epoch": 1.897891231964484,
+      "grad_norm": 0.3572242259979248,
+      "learning_rate": 3.5761297972601695e-06,
+      "loss": 0.4546,
+      "step": 2280
+    },
+    {
+      "epoch": 1.898723640399556,
+      "grad_norm": 0.31195253133773804,
+      "learning_rate": 3.5714859353832286e-06,
+      "loss": 0.4141,
+      "step": 2281
+    },
+    {
+      "epoch": 1.8995560488346281,
+      "grad_norm": 0.3187641501426697,
+      "learning_rate": 3.5668434149060076e-06,
+      "loss": 0.4276,
+      "step": 2282
+    },
+    {
+      "epoch": 1.9003884572697003,
+      "grad_norm": 0.34391510486602783,
+      "learning_rate": 3.562202240187913e-06,
+      "loss": 0.4086,
+      "step": 2283
+    },
+    {
+      "epoch": 1.9012208657047724,
+      "grad_norm": 0.35638102889060974,
+      "learning_rate": 3.5575624155870885e-06,
+      "loss": 0.4511,
+      "step": 2284
+    },
+    {
+      "epoch": 1.9020532741398446,
+      "grad_norm": 0.30889418721199036,
+      "learning_rate": 3.552923945460413e-06,
+      "loss": 0.417,
+      "step": 2285
+    },
+    {
+      "epoch": 1.9028856825749167,
+      "grad_norm": 0.3231523036956787,
+      "learning_rate": 3.548286834163491e-06,
+      "loss": 0.3931,
+      "step": 2286
+    },
+    {
+      "epoch": 1.903718091009989,
+      "grad_norm": 0.3281930983066559,
+      "learning_rate": 3.543651086050649e-06,
+      "loss": 0.3971,
+      "step": 2287
+    },
+    {
+      "epoch": 1.904550499445061,
+      "grad_norm": 0.3393603563308716,
+      "learning_rate": 3.5390167054749363e-06,
+      "loss": 0.3878,
+      "step": 2288
+    },
+    {
+      "epoch": 1.9053829078801332,
+      "grad_norm": 0.33854469656944275,
+      "learning_rate": 3.5343836967881194e-06,
+      "loss": 0.4232,
+      "step": 2289
+    },
+    {
+      "epoch": 1.9062153163152054,
+      "grad_norm": 0.3317667245864868,
+      "learning_rate": 3.529752064340673e-06,
+      "loss": 0.3732,
+      "step": 2290
+    },
+    {
+      "epoch": 1.9070477247502775,
+      "grad_norm": 0.31549349427223206,
+      "learning_rate": 3.5251218124817803e-06,
+      "loss": 0.4203,
+      "step": 2291
+    },
+    {
+      "epoch": 1.9078801331853497,
+      "grad_norm": 0.3274786174297333,
+      "learning_rate": 3.5204929455593316e-06,
+      "loss": 0.4142,
+      "step": 2292
+    },
+    {
+      "epoch": 1.9087125416204218,
+      "grad_norm": 0.3583471477031708,
+      "learning_rate": 3.51586546791991e-06,
+      "loss": 0.4494,
+      "step": 2293
+    },
+    {
+      "epoch": 1.9095449500554937,
+      "grad_norm": 0.34534788131713867,
+      "learning_rate": 3.511239383908801e-06,
+      "loss": 0.4386,
+      "step": 2294
+    },
+    {
+      "epoch": 1.9103773584905661,
+      "grad_norm": 0.34133318066596985,
+      "learning_rate": 3.5066146978699785e-06,
+      "loss": 0.4227,
+      "step": 2295
+    },
+    {
+      "epoch": 1.911209766925638,
+      "grad_norm": 0.3159574568271637,
+      "learning_rate": 3.501991414146102e-06,
+      "loss": 0.42,
+      "step": 2296
+    },
+    {
+      "epoch": 1.9120421753607104,
+      "grad_norm": 0.321329265832901,
+      "learning_rate": 3.4973695370785154e-06,
+      "loss": 0.391,
+      "step": 2297
+    },
+    {
+      "epoch": 1.9128745837957823,
+      "grad_norm": 0.3425721526145935,
+      "learning_rate": 3.4927490710072454e-06,
+      "loss": 0.427,
+      "step": 2298
+    },
+    {
+      "epoch": 1.9137069922308547,
+      "grad_norm": 0.34851303696632385,
+      "learning_rate": 3.488130020270989e-06,
+      "loss": 0.4335,
+      "step": 2299
+    },
+    {
+      "epoch": 1.9145394006659266,
+      "grad_norm": 0.26636067032814026,
+      "learning_rate": 3.4835123892071145e-06,
+      "loss": 0.3935,
+      "step": 2300
+    },
+    {
+      "epoch": 1.915371809100999,
+      "grad_norm": 0.30461832880973816,
+      "learning_rate": 3.4788961821516576e-06,
+      "loss": 0.3859,
+      "step": 2301
+    },
+    {
+      "epoch": 1.916204217536071,
+      "grad_norm": 0.3343258500099182,
+      "learning_rate": 3.4742814034393224e-06,
+      "loss": 0.4512,
+      "step": 2302
+    },
+    {
+      "epoch": 1.9170366259711433,
+      "grad_norm": 0.3263416588306427,
+      "learning_rate": 3.4696680574034613e-06,
+      "loss": 0.4181,
+      "step": 2303
+    },
+    {
+      "epoch": 1.9178690344062153,
+      "grad_norm": 0.30046147108078003,
+      "learning_rate": 3.46505614837609e-06,
+      "loss": 0.3963,
+      "step": 2304
+    },
+    {
+      "epoch": 1.9187014428412874,
+      "grad_norm": 0.3026406764984131,
+      "learning_rate": 3.4604456806878704e-06,
+      "loss": 0.4206,
+      "step": 2305
+    },
+    {
+      "epoch": 1.9195338512763596,
+      "grad_norm": 0.30180978775024414,
+      "learning_rate": 3.4558366586681152e-06,
+      "loss": 0.4089,
+      "step": 2306
+    },
+    {
+      "epoch": 1.9203662597114317,
+      "grad_norm": 0.2909837067127228,
+      "learning_rate": 3.451229086644774e-06,
+      "loss": 0.3858,
+      "step": 2307
+    },
+    {
+      "epoch": 1.9211986681465039,
+      "grad_norm": 0.30994731187820435,
+      "learning_rate": 3.4466229689444384e-06,
+      "loss": 0.3947,
+      "step": 2308
+    },
+    {
+      "epoch": 1.922031076581576,
+      "grad_norm": 0.3551662862300873,
+      "learning_rate": 3.442018309892333e-06,
+      "loss": 0.4318,
+      "step": 2309
+    },
+    {
+      "epoch": 1.9228634850166482,
+      "grad_norm": 0.3257605731487274,
+      "learning_rate": 3.4374151138123135e-06,
+      "loss": 0.4569,
+      "step": 2310
+    },
+    {
+      "epoch": 1.9236958934517203,
+      "grad_norm": 0.28167983889579773,
+      "learning_rate": 3.432813385026862e-06,
+      "loss": 0.3689,
+      "step": 2311
+    },
+    {
+      "epoch": 1.9245283018867925,
+      "grad_norm": 0.36538660526275635,
+      "learning_rate": 3.4282131278570833e-06,
+      "loss": 0.4454,
+      "step": 2312
+    },
+    {
+      "epoch": 1.9253607103218646,
+      "grad_norm": 0.33179739117622375,
+      "learning_rate": 3.423614346622698e-06,
+      "loss": 0.3962,
+      "step": 2313
+    },
+    {
+      "epoch": 1.9261931187569368,
+      "grad_norm": 0.31999471783638,
+      "learning_rate": 3.4190170456420413e-06,
+      "loss": 0.3943,
+      "step": 2314
+    },
+    {
+      "epoch": 1.927025527192009,
+      "grad_norm": 0.3487468659877777,
+      "learning_rate": 3.4144212292320634e-06,
+      "loss": 0.4468,
+      "step": 2315
+    },
+    {
+      "epoch": 1.927857935627081,
+      "grad_norm": 0.3570094406604767,
+      "learning_rate": 3.409826901708312e-06,
+      "loss": 0.4241,
+      "step": 2316
+    },
+    {
+      "epoch": 1.928690344062153,
+      "grad_norm": 0.354623407125473,
+      "learning_rate": 3.4052340673849426e-06,
+      "loss": 0.4656,
+      "step": 2317
+    },
+    {
+      "epoch": 1.9295227524972254,
+      "grad_norm": 0.29796338081359863,
+      "learning_rate": 3.400642730574706e-06,
+      "loss": 0.422,
+      "step": 2318
+    },
+    {
+      "epoch": 1.9303551609322973,
+      "grad_norm": 0.32531046867370605,
+      "learning_rate": 3.3960528955889516e-06,
+      "loss": 0.417,
+      "step": 2319
+    },
+    {
+      "epoch": 1.9311875693673697,
+      "grad_norm": 0.3640810251235962,
+      "learning_rate": 3.391464566737611e-06,
+      "loss": 0.4625,
+      "step": 2320
+    },
+    {
+      "epoch": 1.9320199778024416,
+      "grad_norm": 0.3448679745197296,
+      "learning_rate": 3.386877748329208e-06,
+      "loss": 0.4093,
+      "step": 2321
+    },
+    {
+      "epoch": 1.932852386237514,
+      "grad_norm": 0.3299039900302887,
+      "learning_rate": 3.382292444670843e-06,
+      "loss": 0.4068,
+      "step": 2322
+    },
+    {
+      "epoch": 1.933684794672586,
+      "grad_norm": 0.3216043710708618,
+      "learning_rate": 3.3777086600681954e-06,
+      "loss": 0.4023,
+      "step": 2323
+    },
+    {
+      "epoch": 1.9345172031076583,
+      "grad_norm": 0.3004588484764099,
+      "learning_rate": 3.3731263988255223e-06,
+      "loss": 0.3881,
+      "step": 2324
+    },
+    {
+      "epoch": 1.9353496115427302,
+      "grad_norm": 0.3696260452270508,
+      "learning_rate": 3.3685456652456484e-06,
+      "loss": 0.4373,
+      "step": 2325
+    },
+    {
+      "epoch": 1.9361820199778026,
+      "grad_norm": 0.35806453227996826,
+      "learning_rate": 3.3639664636299586e-06,
+      "loss": 0.4515,
+      "step": 2326
+    },
+    {
+      "epoch": 1.9370144284128745,
+      "grad_norm": 0.28217509388923645,
+      "learning_rate": 3.3593887982784047e-06,
+      "loss": 0.3911,
+      "step": 2327
+    },
+    {
+      "epoch": 1.9378468368479467,
+      "grad_norm": 0.3391430974006653,
+      "learning_rate": 3.354812673489497e-06,
+      "loss": 0.4178,
+      "step": 2328
+    },
+    {
+      "epoch": 1.9386792452830188,
+      "grad_norm": 0.3683750629425049,
+      "learning_rate": 3.3502380935602942e-06,
+      "loss": 0.4342,
+      "step": 2329
+    },
+    {
+      "epoch": 1.939511653718091,
+      "grad_norm": 0.34491217136383057,
+      "learning_rate": 3.3456650627864075e-06,
+      "loss": 0.389,
+      "step": 2330
+    },
+    {
+      "epoch": 1.9403440621531631,
+      "grad_norm": 0.3198521137237549,
+      "learning_rate": 3.341093585461992e-06,
+      "loss": 0.3983,
+      "step": 2331
+    },
+    {
+      "epoch": 1.9411764705882353,
+      "grad_norm": 0.34554436802864075,
+      "learning_rate": 3.336523665879748e-06,
+      "loss": 0.4338,
+      "step": 2332
+    },
+    {
+      "epoch": 1.9420088790233074,
+      "grad_norm": 0.32669246196746826,
+      "learning_rate": 3.331955308330907e-06,
+      "loss": 0.3832,
+      "step": 2333
+    },
+    {
+      "epoch": 1.9428412874583796,
+      "grad_norm": 0.3620488941669464,
+      "learning_rate": 3.327388517105239e-06,
+      "loss": 0.493,
+      "step": 2334
+    },
+    {
+      "epoch": 1.9436736958934517,
+      "grad_norm": 0.2980594038963318,
+      "learning_rate": 3.3228232964910377e-06,
+      "loss": 0.4098,
+      "step": 2335
+    },
+    {
+      "epoch": 1.9445061043285239,
+      "grad_norm": 0.32015350461006165,
+      "learning_rate": 3.3182596507751288e-06,
+      "loss": 0.3994,
+      "step": 2336
+    },
+    {
+      "epoch": 1.945338512763596,
+      "grad_norm": 0.32040566205978394,
+      "learning_rate": 3.313697584242853e-06,
+      "loss": 0.4373,
+      "step": 2337
+    },
+    {
+      "epoch": 1.9461709211986682,
+      "grad_norm": 0.3319530785083771,
+      "learning_rate": 3.309137101178073e-06,
+      "loss": 0.4209,
+      "step": 2338
+    },
+    {
+      "epoch": 1.9470033296337403,
+      "grad_norm": 0.3502316474914551,
+      "learning_rate": 3.3045782058631597e-06,
+      "loss": 0.4313,
+      "step": 2339
+    },
+    {
+      "epoch": 1.9478357380688123,
+      "grad_norm": 0.3295128047466278,
+      "learning_rate": 3.3000209025789965e-06,
+      "loss": 0.429,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9486681465038846,
+      "grad_norm": 0.30817556381225586,
+      "learning_rate": 3.295465195604972e-06,
+      "loss": 0.3908,
+      "step": 2341
+    },
+    {
+      "epoch": 1.9495005549389566,
+      "grad_norm": 0.34200671315193176,
+      "learning_rate": 3.2909110892189745e-06,
+      "loss": 0.4588,
+      "step": 2342
+    },
+    {
+      "epoch": 1.950332963374029,
+      "grad_norm": 0.3436935842037201,
+      "learning_rate": 3.286358587697388e-06,
+      "loss": 0.4262,
+      "step": 2343
+    },
+    {
+      "epoch": 1.9511653718091009,
+      "grad_norm": 0.29214903712272644,
+      "learning_rate": 3.2818076953150917e-06,
+      "loss": 0.4161,
+      "step": 2344
+    },
+    {
+      "epoch": 1.9519977802441733,
+      "grad_norm": 0.2840372920036316,
+      "learning_rate": 3.277258416345456e-06,
+      "loss": 0.3877,
+      "step": 2345
+    },
+    {
+      "epoch": 1.9528301886792452,
+      "grad_norm": 0.34386762976646423,
+      "learning_rate": 3.2727107550603305e-06,
+      "loss": 0.4524,
+      "step": 2346
+    },
+    {
+      "epoch": 1.9536625971143176,
+      "grad_norm": 0.3042285740375519,
+      "learning_rate": 3.26816471573005e-06,
+      "loss": 0.4206,
+      "step": 2347
+    },
+    {
+      "epoch": 1.9544950055493895,
+      "grad_norm": 0.28492388129234314,
+      "learning_rate": 3.2636203026234236e-06,
+      "loss": 0.3565,
+      "step": 2348
+    },
+    {
+      "epoch": 1.9553274139844619,
+      "grad_norm": 0.33080917596817017,
+      "learning_rate": 3.2590775200077364e-06,
+      "loss": 0.4431,
+      "step": 2349
+    },
+    {
+      "epoch": 1.9561598224195338,
+      "grad_norm": 0.3305113613605499,
+      "learning_rate": 3.25453637214874e-06,
+      "loss": 0.4194,
+      "step": 2350
+    },
+    {
+      "epoch": 1.956992230854606,
+      "grad_norm": 0.3372589647769928,
+      "learning_rate": 3.249996863310654e-06,
+      "loss": 0.4155,
+      "step": 2351
+    },
+    {
+      "epoch": 1.957824639289678,
+      "grad_norm": 0.33676356077194214,
+      "learning_rate": 3.2454589977561513e-06,
+      "loss": 0.3825,
+      "step": 2352
+    },
+    {
+      "epoch": 1.9586570477247502,
+      "grad_norm": 0.37880995869636536,
+      "learning_rate": 3.2409227797463727e-06,
+      "loss": 0.4373,
+      "step": 2353
+    },
+    {
+      "epoch": 1.9594894561598224,
+      "grad_norm": 0.3162456750869751,
+      "learning_rate": 3.236388213540904e-06,
+      "loss": 0.3984,
+      "step": 2354
+    },
+    {
+      "epoch": 1.9603218645948945,
+      "grad_norm": 0.33563530445098877,
+      "learning_rate": 3.231855303397783e-06,
+      "loss": 0.4313,
+      "step": 2355
+    },
+    {
+      "epoch": 1.9611542730299667,
+      "grad_norm": 0.31390097737312317,
+      "learning_rate": 3.2273240535734895e-06,
+      "loss": 0.4177,
+      "step": 2356
+    },
+    {
+      "epoch": 1.9619866814650389,
+      "grad_norm": 0.3237724006175995,
+      "learning_rate": 3.2227944683229484e-06,
+      "loss": 0.4041,
+      "step": 2357
+    },
+    {
+      "epoch": 1.962819089900111,
+      "grad_norm": 0.32047805190086365,
+      "learning_rate": 3.2182665518995203e-06,
+      "loss": 0.4324,
+      "step": 2358
+    },
+    {
+      "epoch": 1.9636514983351832,
+      "grad_norm": 0.2876075506210327,
+      "learning_rate": 3.2137403085549962e-06,
+      "loss": 0.3837,
+      "step": 2359
+    },
+    {
+      "epoch": 1.9644839067702553,
+      "grad_norm": 0.2909439504146576,
+      "learning_rate": 3.2092157425395996e-06,
+      "loss": 0.4206,
+      "step": 2360
+    },
+    {
+      "epoch": 1.9653163152053275,
+      "grad_norm": 0.29555991291999817,
+      "learning_rate": 3.2046928581019744e-06,
+      "loss": 0.3688,
+      "step": 2361
+    },
+    {
+      "epoch": 1.9661487236403996,
+      "grad_norm": 0.35118407011032104,
+      "learning_rate": 3.20017165948919e-06,
+      "loss": 0.4504,
+      "step": 2362
+    },
+    {
+      "epoch": 1.9669811320754715,
+      "grad_norm": 0.309773325920105,
+      "learning_rate": 3.195652150946732e-06,
+      "loss": 0.3804,
+      "step": 2363
+    },
+    {
+      "epoch": 1.967813540510544,
+      "grad_norm": 0.32229289412498474,
+      "learning_rate": 3.1911343367184977e-06,
+      "loss": 0.4497,
+      "step": 2364
+    },
+    {
+      "epoch": 1.9686459489456158,
+      "grad_norm": 0.30695047974586487,
+      "learning_rate": 3.1866182210467923e-06,
+      "loss": 0.4304,
+      "step": 2365
+    },
+    {
+      "epoch": 1.9694783573806882,
+      "grad_norm": 0.33170729875564575,
+      "learning_rate": 3.1821038081723283e-06,
+      "loss": 0.4071,
+      "step": 2366
+    },
+    {
+      "epoch": 1.9703107658157601,
+      "grad_norm": 0.30925998091697693,
+      "learning_rate": 3.1775911023342197e-06,
+      "loss": 0.3772,
+      "step": 2367
+    },
+    {
+      "epoch": 1.9711431742508325,
+      "grad_norm": 0.3299291729927063,
+      "learning_rate": 3.1730801077699747e-06,
+      "loss": 0.4137,
+      "step": 2368
+    },
+    {
+      "epoch": 1.9719755826859044,
+      "grad_norm": 0.3230315148830414,
+      "learning_rate": 3.168570828715496e-06,
+      "loss": 0.4289,
+      "step": 2369
+    },
+    {
+      "epoch": 1.9728079911209768,
+      "grad_norm": 0.33906981348991394,
+      "learning_rate": 3.1640632694050753e-06,
+      "loss": 0.4349,
+      "step": 2370
+    },
+    {
+      "epoch": 1.9736403995560488,
+      "grad_norm": 0.31247228384017944,
+      "learning_rate": 3.159557434071393e-06,
+      "loss": 0.4172,
+      "step": 2371
+    },
+    {
+      "epoch": 1.9744728079911211,
+      "grad_norm": 0.31131434440612793,
+      "learning_rate": 3.155053326945503e-06,
+      "loss": 0.4167,
+      "step": 2372
+    },
+    {
+      "epoch": 1.975305216426193,
+      "grad_norm": 0.3348797857761383,
+      "learning_rate": 3.1505509522568444e-06,
+      "loss": 0.4398,
+      "step": 2373
+    },
+    {
+      "epoch": 1.9761376248612652,
+      "grad_norm": 0.33107826113700867,
+      "learning_rate": 3.1460503142332227e-06,
+      "loss": 0.394,
+      "step": 2374
+    },
+    {
+      "epoch": 1.9769700332963374,
+      "grad_norm": 0.30621546506881714,
+      "learning_rate": 3.1415514171008176e-06,
+      "loss": 0.4005,
+      "step": 2375
+    },
+    {
+      "epoch": 1.9778024417314095,
+      "grad_norm": 0.3266296088695526,
+      "learning_rate": 3.137054265084173e-06,
+      "loss": 0.4249,
+      "step": 2376
+    },
+    {
+      "epoch": 1.9786348501664817,
+      "grad_norm": 0.3151954412460327,
+      "learning_rate": 3.1325588624061925e-06,
+      "loss": 0.3982,
+      "step": 2377
+    },
+    {
+      "epoch": 1.9794672586015538,
+      "grad_norm": 0.32240432500839233,
+      "learning_rate": 3.128065213288136e-06,
+      "loss": 0.3918,
+      "step": 2378
+    },
+    {
+      "epoch": 1.980299667036626,
+      "grad_norm": 0.3490481376647949,
+      "learning_rate": 3.123573321949621e-06,
+      "loss": 0.4313,
+      "step": 2379
+    },
+    {
+      "epoch": 1.9811320754716981,
+      "grad_norm": 0.3069087862968445,
+      "learning_rate": 3.119083192608614e-06,
+      "loss": 0.3813,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9819644839067703,
+      "grad_norm": 0.3291850984096527,
+      "learning_rate": 3.114594829481421e-06,
+      "loss": 0.4728,
+      "step": 2381
+    },
+    {
+      "epoch": 1.9827968923418424,
+      "grad_norm": 0.33388224244117737,
+      "learning_rate": 3.110108236782694e-06,
+      "loss": 0.4257,
+      "step": 2382
+    },
+    {
+      "epoch": 1.9836293007769146,
+      "grad_norm": 0.31437182426452637,
+      "learning_rate": 3.105623418725424e-06,
+      "loss": 0.3869,
+      "step": 2383
+    },
+    {
+      "epoch": 1.9844617092119867,
+      "grad_norm": 0.31633225083351135,
+      "learning_rate": 3.101140379520935e-06,
+      "loss": 0.3808,
+      "step": 2384
+    },
+    {
+      "epoch": 1.9852941176470589,
+      "grad_norm": 0.3501662015914917,
+      "learning_rate": 3.0966591233788757e-06,
+      "loss": 0.4576,
+      "step": 2385
+    },
+    {
+      "epoch": 1.9861265260821308,
+      "grad_norm": 0.3176516592502594,
+      "learning_rate": 3.092179654507227e-06,
+      "loss": 0.4181,
+      "step": 2386
+    },
+    {
+      "epoch": 1.9869589345172032,
+      "grad_norm": 0.354187935590744,
+      "learning_rate": 3.0877019771122848e-06,
+      "loss": 0.4359,
+      "step": 2387
+    },
+    {
+      "epoch": 1.987791342952275,
+      "grad_norm": 0.32240673899650574,
+      "learning_rate": 3.0832260953986716e-06,
+      "loss": 0.3794,
+      "step": 2388
+    },
+    {
+      "epoch": 1.9886237513873475,
+      "grad_norm": 0.3612145781517029,
+      "learning_rate": 3.078752013569315e-06,
+      "loss": 0.4573,
+      "step": 2389
+    },
+    {
+      "epoch": 1.9894561598224194,
+      "grad_norm": 0.29033204913139343,
+      "learning_rate": 3.0742797358254584e-06,
+      "loss": 0.4075,
+      "step": 2390
+    },
+    {
+      "epoch": 1.9902885682574918,
+      "grad_norm": 0.3590858280658722,
+      "learning_rate": 3.069809266366647e-06,
+      "loss": 0.4718,
+      "step": 2391
+    },
+    {
+      "epoch": 1.9911209766925637,
+      "grad_norm": 0.3385516405105591,
+      "learning_rate": 3.06534060939073e-06,
+      "loss": 0.4209,
+      "step": 2392
+    },
+    {
+      "epoch": 1.991953385127636,
+      "grad_norm": 0.32453739643096924,
+      "learning_rate": 3.060873769093858e-06,
+      "loss": 0.3889,
+      "step": 2393
+    },
+    {
+      "epoch": 1.992785793562708,
+      "grad_norm": 0.3326762616634369,
+      "learning_rate": 3.0564087496704676e-06,
+      "loss": 0.4476,
+      "step": 2394
+    },
+    {
+      "epoch": 1.9936182019977804,
+      "grad_norm": 0.30887943506240845,
+      "learning_rate": 3.0519455553132914e-06,
+      "loss": 0.422,
+      "step": 2395
+    },
+    {
+      "epoch": 1.9944506104328523,
+      "grad_norm": 0.32724782824516296,
+      "learning_rate": 3.047484190213349e-06,
+      "loss": 0.4304,
+      "step": 2396
+    },
+    {
+      "epoch": 1.9952830188679245,
+      "grad_norm": 0.3449188768863678,
+      "learning_rate": 3.0430246585599402e-06,
+      "loss": 0.4361,
+      "step": 2397
+    },
+    {
+      "epoch": 1.9961154273029966,
+      "grad_norm": 0.3099921941757202,
+      "learning_rate": 3.0385669645406413e-06,
+      "loss": 0.378,
+      "step": 2398
+    },
+    {
+      "epoch": 1.9969478357380688,
+      "grad_norm": 0.3242183327674866,
+      "learning_rate": 3.034111112341307e-06,
+      "loss": 0.4265,
+      "step": 2399
+    },
+    {
+      "epoch": 1.997780244173141,
+      "grad_norm": 0.3651365637779236,
+      "learning_rate": 3.029657106146057e-06,
+      "loss": 0.4341,
+      "step": 2400
+    },
+    {
+      "epoch": 1.998612652608213,
+      "grad_norm": 0.32321080565452576,
+      "learning_rate": 3.025204950137286e-06,
+      "loss": 0.4153,
+      "step": 2401
+    },
+    {
+      "epoch": 1.9994450610432852,
+      "grad_norm": 0.3187124729156494,
+      "learning_rate": 3.020754648495644e-06,
+      "loss": 0.3873,
+      "step": 2402
+    },
+    {
+      "epoch": 2.000277469478357,
+      "grad_norm": 0.69718337059021,
+      "learning_rate": 3.0163062054000424e-06,
+      "loss": 0.683,
+      "step": 2403
+    },
+    {
+      "epoch": 2.0011098779134295,
+      "grad_norm": 0.33480337262153625,
+      "learning_rate": 3.0118596250276453e-06,
+      "loss": 0.4026,
+      "step": 2404
+    },
+    {
+      "epoch": 2.0019422863485015,
+      "grad_norm": 0.3858184814453125,
+      "learning_rate": 3.0074149115538725e-06,
+      "loss": 0.4126,
+      "step": 2405
+    },
+    {
+      "epoch": 2.002774694783574,
+      "grad_norm": 0.3335935175418854,
+      "learning_rate": 3.0029720691523873e-06,
+      "loss": 0.372,
+      "step": 2406
+    },
+    {
+      "epoch": 2.0036071032186458,
+      "grad_norm": 0.3692198693752289,
+      "learning_rate": 2.9985311019950945e-06,
+      "loss": 0.4148,
+      "step": 2407
+    },
+    {
+      "epoch": 2.004439511653718,
+      "grad_norm": 0.333741158246994,
+      "learning_rate": 2.9940920142521413e-06,
+      "loss": 0.3529,
+      "step": 2408
+    },
+    {
+      "epoch": 2.00527192008879,
+      "grad_norm": 0.3616425395011902,
+      "learning_rate": 2.9896548100919087e-06,
+      "loss": 0.3824,
+      "step": 2409
+    },
+    {
+      "epoch": 2.0061043285238624,
+      "grad_norm": 0.34554699063301086,
+      "learning_rate": 2.985219493681011e-06,
+      "loss": 0.3984,
+      "step": 2410
+    },
+    {
+      "epoch": 2.0069367369589344,
+      "grad_norm": 0.35826876759529114,
+      "learning_rate": 2.980786069184285e-06,
+      "loss": 0.4494,
+      "step": 2411
+    },
+    {
+      "epoch": 2.0077691453940067,
+      "grad_norm": 0.31827691197395325,
+      "learning_rate": 2.976354540764793e-06,
+      "loss": 0.3831,
+      "step": 2412
+    },
+    {
+      "epoch": 2.0086015538290787,
+      "grad_norm": 0.3482878506183624,
+      "learning_rate": 2.971924912583822e-06,
+      "loss": 0.3803,
+      "step": 2413
+    },
+    {
+      "epoch": 2.009433962264151,
+      "grad_norm": 0.3719305098056793,
+      "learning_rate": 2.9674971888008696e-06,
+      "loss": 0.4542,
+      "step": 2414
+    },
+    {
+      "epoch": 2.010266370699223,
+      "grad_norm": 0.30256187915802,
+      "learning_rate": 2.9630713735736428e-06,
+      "loss": 0.3519,
+      "step": 2415
+    },
+    {
+      "epoch": 2.0110987791342954,
+      "grad_norm": 0.3034214973449707,
+      "learning_rate": 2.9586474710580627e-06,
+      "loss": 0.3865,
+      "step": 2416
+    },
+    {
+      "epoch": 2.0119311875693673,
+      "grad_norm": 0.3587876260280609,
+      "learning_rate": 2.954225485408248e-06,
+      "loss": 0.4371,
+      "step": 2417
+    },
+    {
+      "epoch": 2.0127635960044397,
+      "grad_norm": 0.35651448369026184,
+      "learning_rate": 2.9498054207765237e-06,
+      "loss": 0.3977,
+      "step": 2418
+    },
+    {
+      "epoch": 2.0135960044395116,
+      "grad_norm": 0.2968187630176544,
+      "learning_rate": 2.945387281313408e-06,
+      "loss": 0.3755,
+      "step": 2419
+    },
+    {
+      "epoch": 2.014428412874584,
+      "grad_norm": 0.34130021929740906,
+      "learning_rate": 2.940971071167608e-06,
+      "loss": 0.4274,
+      "step": 2420
+    },
+    {
+      "epoch": 2.015260821309656,
+      "grad_norm": 0.3059341609477997,
+      "learning_rate": 2.936556794486024e-06,
+      "loss": 0.3419,
+      "step": 2421
+    },
+    {
+      "epoch": 2.0160932297447283,
+      "grad_norm": 0.3298993408679962,
+      "learning_rate": 2.932144455413741e-06,
+      "loss": 0.4018,
+      "step": 2422
+    },
+    {
+      "epoch": 2.0169256381798,
+      "grad_norm": 0.3346230387687683,
+      "learning_rate": 2.9277340580940215e-06,
+      "loss": 0.4095,
+      "step": 2423
+    },
+    {
+      "epoch": 2.0177580466148726,
+      "grad_norm": 0.35374805331230164,
+      "learning_rate": 2.9233256066683047e-06,
+      "loss": 0.3878,
+      "step": 2424
+    },
+    {
+      "epoch": 2.0185904550499445,
+      "grad_norm": 0.2975861430168152,
+      "learning_rate": 2.9189191052762038e-06,
+      "loss": 0.3523,
+      "step": 2425
+    },
+    {
+      "epoch": 2.0194228634850164,
+      "grad_norm": 0.3865160048007965,
+      "learning_rate": 2.914514558055502e-06,
+      "loss": 0.4404,
+      "step": 2426
+    },
+    {
+      "epoch": 2.020255271920089,
+      "grad_norm": 0.2803787887096405,
+      "learning_rate": 2.9101119691421453e-06,
+      "loss": 0.3509,
+      "step": 2427
+    },
+    {
+      "epoch": 2.0210876803551607,
+      "grad_norm": 0.35009345412254333,
+      "learning_rate": 2.905711342670242e-06,
+      "loss": 0.3977,
+      "step": 2428
+    },
+    {
+      "epoch": 2.021920088790233,
+      "grad_norm": 0.3857073485851288,
+      "learning_rate": 2.901312682772058e-06,
+      "loss": 0.4167,
+      "step": 2429
+    },
+    {
+      "epoch": 2.022752497225305,
+      "grad_norm": 0.3246917426586151,
+      "learning_rate": 2.896915993578011e-06,
+      "loss": 0.3457,
+      "step": 2430
+    },
+    {
+      "epoch": 2.0235849056603774,
+      "grad_norm": 0.32466015219688416,
+      "learning_rate": 2.8925212792166694e-06,
+      "loss": 0.3807,
+      "step": 2431
+    },
+    {
+      "epoch": 2.0244173140954493,
+      "grad_norm": 0.3205777704715729,
+      "learning_rate": 2.8881285438147477e-06,
+      "loss": 0.3801,
+      "step": 2432
+    },
+    {
+      "epoch": 2.0252497225305217,
+      "grad_norm": 0.31383705139160156,
+      "learning_rate": 2.8837377914971003e-06,
+      "loss": 0.3692,
+      "step": 2433
+    },
+    {
+      "epoch": 2.0260821309655936,
+      "grad_norm": 0.36327627301216125,
+      "learning_rate": 2.8793490263867212e-06,
+      "loss": 0.395,
+      "step": 2434
+    },
+    {
+      "epoch": 2.026914539400666,
+      "grad_norm": 0.33775612711906433,
+      "learning_rate": 2.8749622526047373e-06,
+      "loss": 0.3764,
+      "step": 2435
+    },
+    {
+      "epoch": 2.027746947835738,
+      "grad_norm": 0.31958720088005066,
+      "learning_rate": 2.8705774742704063e-06,
+      "loss": 0.4473,
+      "step": 2436
+    },
+    {
+      "epoch": 2.0285793562708103,
+      "grad_norm": 0.3061386048793793,
+      "learning_rate": 2.8661946955011145e-06,
+      "loss": 0.4101,
+      "step": 2437
+    },
+    {
+      "epoch": 2.0294117647058822,
+      "grad_norm": 0.27900204062461853,
+      "learning_rate": 2.8618139204123597e-06,
+      "loss": 0.3726,
+      "step": 2438
+    },
+    {
+      "epoch": 2.0302441731409546,
+      "grad_norm": 0.32336297631263733,
+      "learning_rate": 2.8574351531177747e-06,
+      "loss": 0.403,
+      "step": 2439
+    },
+    {
+      "epoch": 2.0310765815760266,
+      "grad_norm": 0.3574887216091156,
+      "learning_rate": 2.853058397729095e-06,
+      "loss": 0.354,
+      "step": 2440
+    },
+    {
+      "epoch": 2.031908990011099,
+      "grad_norm": 0.3319253623485565,
+      "learning_rate": 2.8486836583561737e-06,
+      "loss": 0.4185,
+      "step": 2441
+    },
+    {
+      "epoch": 2.032741398446171,
+      "grad_norm": 0.31709638237953186,
+      "learning_rate": 2.8443109391069616e-06,
+      "loss": 0.3966,
+      "step": 2442
+    },
+    {
+      "epoch": 2.0335738068812432,
+      "grad_norm": 0.28898531198501587,
+      "learning_rate": 2.8399402440875248e-06,
+      "loss": 0.3537,
+      "step": 2443
+    },
+    {
+      "epoch": 2.034406215316315,
+      "grad_norm": 0.32134294509887695,
+      "learning_rate": 2.835571577402021e-06,
+      "loss": 0.3991,
+      "step": 2444
+    },
+    {
+      "epoch": 2.0352386237513875,
+      "grad_norm": 0.3410552144050598,
+      "learning_rate": 2.831204943152701e-06,
+      "loss": 0.4234,
+      "step": 2445
+    },
+    {
+      "epoch": 2.0360710321864595,
+      "grad_norm": 0.31080541014671326,
+      "learning_rate": 2.8268403454399154e-06,
+      "loss": 0.3578,
+      "step": 2446
+    },
+    {
+      "epoch": 2.036903440621532,
+      "grad_norm": 0.3171520531177521,
+      "learning_rate": 2.8224777883620926e-06,
+      "loss": 0.3696,
+      "step": 2447
+    },
+    {
+      "epoch": 2.0377358490566038,
+      "grad_norm": 0.3251802921295166,
+      "learning_rate": 2.8181172760157575e-06,
+      "loss": 0.4128,
+      "step": 2448
+    },
+    {
+      "epoch": 2.0385682574916757,
+      "grad_norm": 0.3506574034690857,
+      "learning_rate": 2.8137588124955017e-06,
+      "loss": 0.3945,
+      "step": 2449
+    },
+    {
+      "epoch": 2.039400665926748,
+      "grad_norm": 0.32733920216560364,
+      "learning_rate": 2.8094024018940012e-06,
+      "loss": 0.4038,
+      "step": 2450
+    },
+    {
+      "epoch": 2.04023307436182,
+      "grad_norm": 0.3144792914390564,
+      "learning_rate": 2.8050480483020003e-06,
+      "loss": 0.4005,
+      "step": 2451
+    },
+    {
+      "epoch": 2.0410654827968924,
+      "grad_norm": 0.27931925654411316,
+      "learning_rate": 2.8006957558083147e-06,
+      "loss": 0.3664,
+      "step": 2452
+    },
+    {
+      "epoch": 2.0418978912319643,
+      "grad_norm": 0.3300316333770752,
+      "learning_rate": 2.7963455284998225e-06,
+      "loss": 0.438,
+      "step": 2453
+    },
+    {
+      "epoch": 2.0427302996670367,
+      "grad_norm": 0.31316685676574707,
+      "learning_rate": 2.7919973704614632e-06,
+      "loss": 0.3973,
+      "step": 2454
+    },
+    {
+      "epoch": 2.0435627081021086,
+      "grad_norm": 0.3359510600566864,
+      "learning_rate": 2.7876512857762343e-06,
+      "loss": 0.3725,
+      "step": 2455
+    },
+    {
+      "epoch": 2.044395116537181,
+      "grad_norm": 0.3232194483280182,
+      "learning_rate": 2.7833072785251846e-06,
+      "loss": 0.4057,
+      "step": 2456
+    },
+    {
+      "epoch": 2.045227524972253,
+      "grad_norm": 0.31120437383651733,
+      "learning_rate": 2.778965352787413e-06,
+      "loss": 0.3684,
+      "step": 2457
+    },
+    {
+      "epoch": 2.0460599334073253,
+      "grad_norm": 0.32518357038497925,
+      "learning_rate": 2.774625512640064e-06,
+      "loss": 0.4007,
+      "step": 2458
+    },
+    {
+      "epoch": 2.046892341842397,
+      "grad_norm": 0.3096177875995636,
+      "learning_rate": 2.7702877621583234e-06,
+      "loss": 0.3879,
+      "step": 2459
+    },
+    {
+      "epoch": 2.0477247502774696,
+      "grad_norm": 0.3457062840461731,
+      "learning_rate": 2.7659521054154147e-06,
+      "loss": 0.4057,
+      "step": 2460
+    },
+    {
+      "epoch": 2.0485571587125415,
+      "grad_norm": 0.33439356088638306,
+      "learning_rate": 2.7616185464825963e-06,
+      "loss": 0.4092,
+      "step": 2461
+    },
+    {
+      "epoch": 2.049389567147614,
+      "grad_norm": 0.33163490891456604,
+      "learning_rate": 2.7572870894291542e-06,
+      "loss": 0.4207,
+      "step": 2462
+    },
+    {
+      "epoch": 2.050221975582686,
+      "grad_norm": 0.3037160336971283,
+      "learning_rate": 2.752957738322406e-06,
+      "loss": 0.3703,
+      "step": 2463
+    },
+    {
+      "epoch": 2.051054384017758,
+      "grad_norm": 0.3286396563053131,
+      "learning_rate": 2.748630497227682e-06,
+      "loss": 0.4138,
+      "step": 2464
+    },
+    {
+      "epoch": 2.05188679245283,
+      "grad_norm": 0.3129012882709503,
+      "learning_rate": 2.744305370208342e-06,
+      "loss": 0.3747,
+      "step": 2465
+    },
+    {
+      "epoch": 2.0527192008879025,
+      "grad_norm": 0.3447505831718445,
+      "learning_rate": 2.7399823613257565e-06,
+      "loss": 0.3802,
+      "step": 2466
+    },
+    {
+      "epoch": 2.0535516093229744,
+      "grad_norm": 0.32165050506591797,
+      "learning_rate": 2.7356614746393063e-06,
+      "loss": 0.4049,
+      "step": 2467
+    },
+    {
+      "epoch": 2.054384017758047,
+      "grad_norm": 0.31354621052742004,
+      "learning_rate": 2.7313427142063742e-06,
+      "loss": 0.3912,
+      "step": 2468
+    },
+    {
+      "epoch": 2.0552164261931187,
+      "grad_norm": 0.3174445331096649,
+      "learning_rate": 2.7270260840823588e-06,
+      "loss": 0.3831,
+      "step": 2469
+    },
+    {
+      "epoch": 2.056048834628191,
+      "grad_norm": 0.3428478538990021,
+      "learning_rate": 2.72271158832065e-06,
+      "loss": 0.3866,
+      "step": 2470
+    },
+    {
+      "epoch": 2.056881243063263,
+      "grad_norm": 0.35326361656188965,
+      "learning_rate": 2.718399230972632e-06,
+      "loss": 0.4147,
+      "step": 2471
+    },
+    {
+      "epoch": 2.057713651498335,
+      "grad_norm": 0.31062963604927063,
+      "learning_rate": 2.714089016087683e-06,
+      "loss": 0.3806,
+      "step": 2472
+    },
+    {
+      "epoch": 2.0585460599334073,
+      "grad_norm": 0.3279687166213989,
+      "learning_rate": 2.7097809477131754e-06,
+      "loss": 0.4389,
+      "step": 2473
+    },
+    {
+      "epoch": 2.0593784683684793,
+      "grad_norm": 0.33881494402885437,
+      "learning_rate": 2.705475029894459e-06,
+      "loss": 0.4214,
+      "step": 2474
+    },
+    {
+      "epoch": 2.0602108768035516,
+      "grad_norm": 0.3012351989746094,
+      "learning_rate": 2.7011712666748636e-06,
+      "loss": 0.3647,
+      "step": 2475
+    },
+    {
+      "epoch": 2.0610432852386236,
+      "grad_norm": 0.36581099033355713,
+      "learning_rate": 2.696869662095698e-06,
+      "loss": 0.4215,
+      "step": 2476
+    },
+    {
+      "epoch": 2.061875693673696,
+      "grad_norm": 0.3162044286727905,
+      "learning_rate": 2.6925702201962493e-06,
+      "loss": 0.4061,
+      "step": 2477
+    },
+    {
+      "epoch": 2.062708102108768,
+      "grad_norm": 0.30521532893180847,
+      "learning_rate": 2.6882729450137636e-06,
+      "loss": 0.368,
+      "step": 2478
+    },
+    {
+      "epoch": 2.0635405105438402,
+      "grad_norm": 0.36388954520225525,
+      "learning_rate": 2.6839778405834593e-06,
+      "loss": 0.3992,
+      "step": 2479
+    },
+    {
+      "epoch": 2.064372918978912,
+      "grad_norm": 0.3033269941806793,
+      "learning_rate": 2.6796849109385147e-06,
+      "loss": 0.3603,
+      "step": 2480
+    },
+    {
+      "epoch": 2.0652053274139845,
+      "grad_norm": 0.32970061898231506,
+      "learning_rate": 2.6753941601100662e-06,
+      "loss": 0.4103,
+      "step": 2481
+    },
+    {
+      "epoch": 2.0660377358490565,
+      "grad_norm": 0.3365088105201721,
+      "learning_rate": 2.6711055921272033e-06,
+      "loss": 0.43,
+      "step": 2482
+    },
+    {
+      "epoch": 2.066870144284129,
+      "grad_norm": 0.3210076689720154,
+      "learning_rate": 2.6668192110169664e-06,
+      "loss": 0.3518,
+      "step": 2483
+    },
+    {
+      "epoch": 2.067702552719201,
+      "grad_norm": 0.37223634123802185,
+      "learning_rate": 2.6625350208043432e-06,
+      "loss": 0.3965,
+      "step": 2484
+    },
+    {
+      "epoch": 2.068534961154273,
+      "grad_norm": 0.3405874967575073,
+      "learning_rate": 2.658253025512263e-06,
+      "loss": 0.401,
+      "step": 2485
+    },
+    {
+      "epoch": 2.069367369589345,
+      "grad_norm": 0.32502299547195435,
+      "learning_rate": 2.6539732291615937e-06,
+      "loss": 0.365,
+      "step": 2486
+    },
+    {
+      "epoch": 2.0701997780244175,
+      "grad_norm": 0.36859750747680664,
+      "learning_rate": 2.6496956357711402e-06,
+      "loss": 0.4403,
+      "step": 2487
+    },
+    {
+      "epoch": 2.0710321864594894,
+      "grad_norm": 0.3416479527950287,
+      "learning_rate": 2.6454202493576366e-06,
+      "loss": 0.3544,
+      "step": 2488
+    },
+    {
+      "epoch": 2.0718645948945618,
+      "grad_norm": 0.3147336542606354,
+      "learning_rate": 2.641147073935746e-06,
+      "loss": 0.3804,
+      "step": 2489
+    },
+    {
+      "epoch": 2.0726970033296337,
+      "grad_norm": 0.3625737428665161,
+      "learning_rate": 2.6368761135180544e-06,
+      "loss": 0.4079,
+      "step": 2490
+    },
+    {
+      "epoch": 2.073529411764706,
+      "grad_norm": 0.3262125849723816,
+      "learning_rate": 2.632607372115069e-06,
+      "loss": 0.4015,
+      "step": 2491
+    },
+    {
+      "epoch": 2.074361820199778,
+      "grad_norm": 0.3019062280654907,
+      "learning_rate": 2.628340853735213e-06,
+      "loss": 0.3562,
+      "step": 2492
+    },
+    {
+      "epoch": 2.0751942286348504,
+      "grad_norm": 0.35128647089004517,
+      "learning_rate": 2.624076562384823e-06,
+      "loss": 0.416,
+      "step": 2493
+    },
+    {
+      "epoch": 2.0760266370699223,
+      "grad_norm": 0.3422330915927887,
+      "learning_rate": 2.619814502068139e-06,
+      "loss": 0.3876,
+      "step": 2494
+    },
+    {
+      "epoch": 2.0768590455049942,
+      "grad_norm": 0.3407348692417145,
+      "learning_rate": 2.6155546767873136e-06,
+      "loss": 0.4153,
+      "step": 2495
+    },
+    {
+      "epoch": 2.0776914539400666,
+      "grad_norm": 0.3248727023601532,
+      "learning_rate": 2.611297090542399e-06,
+      "loss": 0.3867,
+      "step": 2496
+    },
+    {
+      "epoch": 2.0785238623751385,
+      "grad_norm": 0.2963961958885193,
+      "learning_rate": 2.607041747331339e-06,
+      "loss": 0.372,
+      "step": 2497
+    },
+    {
+      "epoch": 2.079356270810211,
+      "grad_norm": 0.3318076431751251,
+      "learning_rate": 2.6027886511499756e-06,
+      "loss": 0.3884,
+      "step": 2498
+    },
+    {
+      "epoch": 2.080188679245283,
+      "grad_norm": 0.32221075892448425,
+      "learning_rate": 2.598537805992044e-06,
+      "loss": 0.3999,
+      "step": 2499
+    },
+    {
+      "epoch": 2.081021087680355,
+      "grad_norm": 0.3247624635696411,
+      "learning_rate": 2.5942892158491626e-06,
+      "loss": 0.4089,
+      "step": 2500
+    },
+    {
+      "epoch": 2.081853496115427,
+      "grad_norm": 0.33710792660713196,
+      "learning_rate": 2.590042884710828e-06,
+      "loss": 0.4056,
+      "step": 2501
+    },
+    {
+      "epoch": 2.0826859045504995,
+      "grad_norm": 0.3118056058883667,
+      "learning_rate": 2.585798816564419e-06,
+      "loss": 0.4116,
+      "step": 2502
+    },
+    {
+      "epoch": 2.0835183129855714,
+      "grad_norm": 0.2979108691215515,
+      "learning_rate": 2.5815570153951942e-06,
+      "loss": 0.3746,
+      "step": 2503
+    },
+    {
+      "epoch": 2.084350721420644,
+      "grad_norm": 0.33021363615989685,
+      "learning_rate": 2.5773174851862796e-06,
+      "loss": 0.394,
+      "step": 2504
+    },
+    {
+      "epoch": 2.0851831298557157,
+      "grad_norm": 0.32653099298477173,
+      "learning_rate": 2.573080229918664e-06,
+      "loss": 0.4189,
+      "step": 2505
+    },
+    {
+      "epoch": 2.086015538290788,
+      "grad_norm": 0.3247196674346924,
+      "learning_rate": 2.568845253571204e-06,
+      "loss": 0.3851,
+      "step": 2506
+    },
+    {
+      "epoch": 2.08684794672586,
+      "grad_norm": 0.3410961329936981,
+      "learning_rate": 2.564612560120623e-06,
+      "loss": 0.4209,
+      "step": 2507
+    },
+    {
+      "epoch": 2.0876803551609324,
+      "grad_norm": 0.3214920163154602,
+      "learning_rate": 2.5603821535414874e-06,
+      "loss": 0.3618,
+      "step": 2508
+    },
+    {
+      "epoch": 2.0885127635960044,
+      "grad_norm": 0.34245097637176514,
+      "learning_rate": 2.556154037806226e-06,
+      "loss": 0.4046,
+      "step": 2509
+    },
+    {
+      "epoch": 2.0893451720310767,
+      "grad_norm": 0.3048022389411926,
+      "learning_rate": 2.5519282168851134e-06,
+      "loss": 0.3755,
+      "step": 2510
+    },
+    {
+      "epoch": 2.0901775804661487,
+      "grad_norm": 0.3066607415676117,
+      "learning_rate": 2.547704694746269e-06,
+      "loss": 0.3606,
+      "step": 2511
+    },
+    {
+      "epoch": 2.091009988901221,
+      "grad_norm": 0.30674123764038086,
+      "learning_rate": 2.543483475355654e-06,
+      "loss": 0.3706,
+      "step": 2512
+    },
+    {
+      "epoch": 2.091842397336293,
+      "grad_norm": 0.32365816831588745,
+      "learning_rate": 2.5392645626770686e-06,
+      "loss": 0.3973,
+      "step": 2513
+    },
+    {
+      "epoch": 2.0926748057713653,
+      "grad_norm": 0.3154730498790741,
+      "learning_rate": 2.5350479606721433e-06,
+      "loss": 0.4019,
+      "step": 2514
+    },
+    {
+      "epoch": 2.0935072142064373,
+      "grad_norm": 0.28480201959609985,
+      "learning_rate": 2.5308336733003435e-06,
+      "loss": 0.362,
+      "step": 2515
+    },
+    {
+      "epoch": 2.0943396226415096,
+      "grad_norm": 0.3131362497806549,
+      "learning_rate": 2.5266217045189572e-06,
+      "loss": 0.4336,
+      "step": 2516
+    },
+    {
+      "epoch": 2.0951720310765816,
+      "grad_norm": 0.3055063784122467,
+      "learning_rate": 2.522412058283098e-06,
+      "loss": 0.3658,
+      "step": 2517
+    },
+    {
+      "epoch": 2.0960044395116535,
+      "grad_norm": 0.3128129839897156,
+      "learning_rate": 2.5182047385456967e-06,
+      "loss": 0.3963,
+      "step": 2518
+    },
+    {
+      "epoch": 2.096836847946726,
+      "grad_norm": 0.3318641483783722,
+      "learning_rate": 2.513999749257501e-06,
+      "loss": 0.4023,
+      "step": 2519
+    },
+    {
+      "epoch": 2.097669256381798,
+      "grad_norm": 0.30515703558921814,
+      "learning_rate": 2.509797094367068e-06,
+      "loss": 0.3823,
+      "step": 2520
+    },
+    {
+      "epoch": 2.09850166481687,
+      "grad_norm": 0.3311365842819214,
+      "learning_rate": 2.505596777820766e-06,
+      "loss": 0.3941,
+      "step": 2521
+    },
+    {
+      "epoch": 2.099334073251942,
+      "grad_norm": 0.3067466914653778,
+      "learning_rate": 2.5013988035627656e-06,
+      "loss": 0.4081,
+      "step": 2522
+    },
+    {
+      "epoch": 2.1001664816870145,
+      "grad_norm": 0.344066858291626,
+      "learning_rate": 2.4972031755350366e-06,
+      "loss": 0.3707,
+      "step": 2523
+    },
+    {
+      "epoch": 2.1009988901220864,
+      "grad_norm": 0.3539952039718628,
+      "learning_rate": 2.493009897677346e-06,
+      "loss": 0.4235,
+      "step": 2524
+    },
+    {
+      "epoch": 2.101831298557159,
+      "grad_norm": 0.34413817524909973,
+      "learning_rate": 2.4888189739272587e-06,
+      "loss": 0.382,
+      "step": 2525
+    },
+    {
+      "epoch": 2.1026637069922307,
+      "grad_norm": 0.30042779445648193,
+      "learning_rate": 2.484630408220126e-06,
+      "loss": 0.3718,
+      "step": 2526
+    },
+    {
+      "epoch": 2.103496115427303,
+      "grad_norm": 0.3354708254337311,
+      "learning_rate": 2.480444204489081e-06,
+      "loss": 0.4273,
+      "step": 2527
+    },
+    {
+      "epoch": 2.104328523862375,
+      "grad_norm": 0.31707751750946045,
+      "learning_rate": 2.476260366665041e-06,
+      "loss": 0.4271,
+      "step": 2528
+    },
+    {
+      "epoch": 2.1051609322974474,
+      "grad_norm": 0.28172051906585693,
+      "learning_rate": 2.472078898676708e-06,
+      "loss": 0.3713,
+      "step": 2529
+    },
+    {
+      "epoch": 2.1059933407325193,
+      "grad_norm": 0.31706756353378296,
+      "learning_rate": 2.467899804450553e-06,
+      "loss": 0.4136,
+      "step": 2530
+    },
+    {
+      "epoch": 2.1068257491675917,
+      "grad_norm": 0.31650689244270325,
+      "learning_rate": 2.463723087910815e-06,
+      "loss": 0.3989,
+      "step": 2531
+    },
+    {
+      "epoch": 2.1076581576026636,
+      "grad_norm": 0.3059511184692383,
+      "learning_rate": 2.4595487529795044e-06,
+      "loss": 0.3801,
+      "step": 2532
+    },
+    {
+      "epoch": 2.108490566037736,
+      "grad_norm": 0.30133718252182007,
+      "learning_rate": 2.4553768035763996e-06,
+      "loss": 0.3809,
+      "step": 2533
+    },
+    {
+      "epoch": 2.109322974472808,
+      "grad_norm": 0.3258849084377289,
+      "learning_rate": 2.451207243619029e-06,
+      "loss": 0.4455,
+      "step": 2534
+    },
+    {
+      "epoch": 2.1101553829078803,
+      "grad_norm": 0.29617834091186523,
+      "learning_rate": 2.447040077022685e-06,
+      "loss": 0.3728,
+      "step": 2535
+    },
+    {
+      "epoch": 2.1109877913429522,
+      "grad_norm": 0.3328491747379303,
+      "learning_rate": 2.4428753077004067e-06,
+      "loss": 0.3945,
+      "step": 2536
+    },
+    {
+      "epoch": 2.1118201997780246,
+      "grad_norm": 0.3035510778427124,
+      "learning_rate": 2.438712939562992e-06,
+      "loss": 0.3755,
+      "step": 2537
+    },
+    {
+      "epoch": 2.1126526082130965,
+      "grad_norm": 0.31253474950790405,
+      "learning_rate": 2.434552976518971e-06,
+      "loss": 0.3796,
+      "step": 2538
+    },
+    {
+      "epoch": 2.113485016648169,
+      "grad_norm": 0.35178011655807495,
+      "learning_rate": 2.430395422474625e-06,
+      "loss": 0.4061,
+      "step": 2539
+    },
+    {
+      "epoch": 2.114317425083241,
+      "grad_norm": 0.3211081326007843,
+      "learning_rate": 2.426240281333969e-06,
+      "loss": 0.3668,
+      "step": 2540
+    },
+    {
+      "epoch": 2.1151498335183128,
+      "grad_norm": 0.35287511348724365,
+      "learning_rate": 2.422087556998754e-06,
+      "loss": 0.4332,
+      "step": 2541
+    },
+    {
+      "epoch": 2.115982241953385,
+      "grad_norm": 0.31688159704208374,
+      "learning_rate": 2.41793725336846e-06,
+      "loss": 0.392,
+      "step": 2542
+    },
+    {
+      "epoch": 2.116814650388457,
+      "grad_norm": 0.3495129644870758,
+      "learning_rate": 2.4137893743402954e-06,
+      "loss": 0.3871,
+      "step": 2543
+    },
+    {
+      "epoch": 2.1176470588235294,
+      "grad_norm": 0.29781919717788696,
+      "learning_rate": 2.409643923809191e-06,
+      "loss": 0.366,
+      "step": 2544
+    },
+    {
+      "epoch": 2.1184794672586014,
+      "grad_norm": 0.3296952545642853,
+      "learning_rate": 2.4055009056677977e-06,
+      "loss": 0.4213,
+      "step": 2545
+    },
+    {
+      "epoch": 2.1193118756936737,
+      "grad_norm": 0.3212566077709198,
+      "learning_rate": 2.4013603238064814e-06,
+      "loss": 0.3989,
+      "step": 2546
+    },
+    {
+      "epoch": 2.1201442841287457,
+      "grad_norm": 0.31362348794937134,
+      "learning_rate": 2.397222182113322e-06,
+      "loss": 0.364,
+      "step": 2547
+    },
+    {
+      "epoch": 2.120976692563818,
+      "grad_norm": 0.3248312473297119,
+      "learning_rate": 2.393086484474108e-06,
+      "loss": 0.3903,
+      "step": 2548
+    },
+    {
+      "epoch": 2.12180910099889,
+      "grad_norm": 0.3590147793292999,
+      "learning_rate": 2.3889532347723266e-06,
+      "loss": 0.418,
+      "step": 2549
+    },
+    {
+      "epoch": 2.1226415094339623,
+      "grad_norm": 0.3201029300689697,
+      "learning_rate": 2.384822436889177e-06,
+      "loss": 0.4164,
+      "step": 2550
+    },
+    {
+      "epoch": 2.1234739178690343,
+      "grad_norm": 0.2909408211708069,
+      "learning_rate": 2.3806940947035497e-06,
+      "loss": 0.35,
+      "step": 2551
+    },
+    {
+      "epoch": 2.1243063263041067,
+      "grad_norm": 0.29795750975608826,
+      "learning_rate": 2.3765682120920315e-06,
+      "loss": 0.3996,
+      "step": 2552
+    },
+    {
+      "epoch": 2.1251387347391786,
+      "grad_norm": 0.31154143810272217,
+      "learning_rate": 2.3724447929288925e-06,
+      "loss": 0.3876,
+      "step": 2553
+    },
+    {
+      "epoch": 2.125971143174251,
+      "grad_norm": 0.33874204754829407,
+      "learning_rate": 2.368323841086102e-06,
+      "loss": 0.3969,
+      "step": 2554
+    },
+    {
+      "epoch": 2.126803551609323,
+      "grad_norm": 0.3370445966720581,
+      "learning_rate": 2.3642053604333032e-06,
+      "loss": 0.4318,
+      "step": 2555
+    },
+    {
+      "epoch": 2.1276359600443953,
+      "grad_norm": 0.30151233077049255,
+      "learning_rate": 2.3600893548378238e-06,
+      "loss": 0.3805,
+      "step": 2556
+    },
+    {
+      "epoch": 2.128468368479467,
+      "grad_norm": 0.3622559607028961,
+      "learning_rate": 2.3559758281646615e-06,
+      "loss": 0.418,
+      "step": 2557
+    },
+    {
+      "epoch": 2.1293007769145396,
+      "grad_norm": 0.307041734457016,
+      "learning_rate": 2.35186478427649e-06,
+      "loss": 0.3494,
+      "step": 2558
+    },
+    {
+      "epoch": 2.1301331853496115,
+      "grad_norm": 0.2808810770511627,
+      "learning_rate": 2.3477562270336564e-06,
+      "loss": 0.4012,
+      "step": 2559
+    },
+    {
+      "epoch": 2.130965593784684,
+      "grad_norm": 0.2891574800014496,
+      "learning_rate": 2.343650160294163e-06,
+      "loss": 0.358,
+      "step": 2560
+    },
+    {
+      "epoch": 2.131798002219756,
+      "grad_norm": 0.32515814900398254,
+      "learning_rate": 2.3395465879136795e-06,
+      "loss": 0.4128,
+      "step": 2561
+    },
+    {
+      "epoch": 2.132630410654828,
+      "grad_norm": 0.30551859736442566,
+      "learning_rate": 2.3354455137455312e-06,
+      "loss": 0.3673,
+      "step": 2562
+    },
+    {
+      "epoch": 2.1334628190899,
+      "grad_norm": 0.3107419013977051,
+      "learning_rate": 2.3313469416407037e-06,
+      "loss": 0.3966,
+      "step": 2563
+    },
+    {
+      "epoch": 2.134295227524972,
+      "grad_norm": 0.3009784519672394,
+      "learning_rate": 2.3272508754478224e-06,
+      "loss": 0.3692,
+      "step": 2564
+    },
+    {
+      "epoch": 2.1351276359600444,
+      "grad_norm": 0.3275810778141022,
+      "learning_rate": 2.3231573190131666e-06,
+      "loss": 0.4089,
+      "step": 2565
+    },
+    {
+      "epoch": 2.1359600443951163,
+      "grad_norm": 0.32291144132614136,
+      "learning_rate": 2.3190662761806586e-06,
+      "loss": 0.363,
+      "step": 2566
+    },
+    {
+      "epoch": 2.1367924528301887,
+      "grad_norm": 0.32055914402008057,
+      "learning_rate": 2.3149777507918587e-06,
+      "loss": 0.4111,
+      "step": 2567
+    },
+    {
+      "epoch": 2.1376248612652606,
+      "grad_norm": 0.31367000937461853,
+      "learning_rate": 2.310891746685963e-06,
+      "loss": 0.3907,
+      "step": 2568
+    },
+    {
+      "epoch": 2.138457269700333,
+      "grad_norm": 0.31795892119407654,
+      "learning_rate": 2.3068082676998022e-06,
+      "loss": 0.4011,
+      "step": 2569
+    },
+    {
+      "epoch": 2.139289678135405,
+      "grad_norm": 0.30362680554389954,
+      "learning_rate": 2.3027273176678337e-06,
+      "loss": 0.3924,
+      "step": 2570
+    },
+    {
+      "epoch": 2.1401220865704773,
+      "grad_norm": 0.3192872405052185,
+      "learning_rate": 2.298648900422141e-06,
+      "loss": 0.382,
+      "step": 2571
+    },
+    {
+      "epoch": 2.1409544950055492,
+      "grad_norm": 0.33818933367729187,
+      "learning_rate": 2.2945730197924303e-06,
+      "loss": 0.4072,
+      "step": 2572
+    },
+    {
+      "epoch": 2.1417869034406216,
+      "grad_norm": 0.3056696951389313,
+      "learning_rate": 2.2904996796060243e-06,
+      "loss": 0.3397,
+      "step": 2573
+    },
+    {
+      "epoch": 2.1426193118756935,
+      "grad_norm": 0.3521409034729004,
+      "learning_rate": 2.2864288836878616e-06,
+      "loss": 0.4124,
+      "step": 2574
+    },
+    {
+      "epoch": 2.143451720310766,
+      "grad_norm": 0.3235545754432678,
+      "learning_rate": 2.2823606358604868e-06,
+      "loss": 0.3854,
+      "step": 2575
+    },
+    {
+      "epoch": 2.144284128745838,
+      "grad_norm": 0.33744242787361145,
+      "learning_rate": 2.278294939944061e-06,
+      "loss": 0.3948,
+      "step": 2576
+    },
+    {
+      "epoch": 2.14511653718091,
+      "grad_norm": 0.30026137828826904,
+      "learning_rate": 2.2742317997563407e-06,
+      "loss": 0.3687,
+      "step": 2577
+    },
+    {
+      "epoch": 2.145948945615982,
+      "grad_norm": 0.3329930007457733,
+      "learning_rate": 2.2701712191126895e-06,
+      "loss": 0.4255,
+      "step": 2578
+    },
+    {
+      "epoch": 2.1467813540510545,
+      "grad_norm": 0.3148747682571411,
+      "learning_rate": 2.266113201826057e-06,
+      "loss": 0.3728,
+      "step": 2579
+    },
+    {
+      "epoch": 2.1476137624861265,
+      "grad_norm": 0.300738126039505,
+      "learning_rate": 2.2620577517069986e-06,
+      "loss": 0.3788,
+      "step": 2580
+    },
+    {
+      "epoch": 2.148446170921199,
+      "grad_norm": 0.31569600105285645,
+      "learning_rate": 2.2580048725636506e-06,
+      "loss": 0.4349,
+      "step": 2581
+    },
+    {
+      "epoch": 2.1492785793562708,
+      "grad_norm": 0.3264215290546417,
+      "learning_rate": 2.2539545682017394e-06,
+      "loss": 0.3995,
+      "step": 2582
+    },
+    {
+      "epoch": 2.150110987791343,
+      "grad_norm": 0.3219297230243683,
+      "learning_rate": 2.2499068424245667e-06,
+      "loss": 0.3659,
+      "step": 2583
+    },
+    {
+      "epoch": 2.150943396226415,
+      "grad_norm": 0.32074615359306335,
+      "learning_rate": 2.245861699033023e-06,
+      "loss": 0.3935,
+      "step": 2584
+    },
+    {
+      "epoch": 2.1517758046614874,
+      "grad_norm": 0.31240931153297424,
+      "learning_rate": 2.2418191418255684e-06,
+      "loss": 0.4093,
+      "step": 2585
+    },
+    {
+      "epoch": 2.1526082130965594,
+      "grad_norm": 0.30113837122917175,
+      "learning_rate": 2.2377791745982323e-06,
+      "loss": 0.367,
+      "step": 2586
+    },
+    {
+      "epoch": 2.1534406215316313,
+      "grad_norm": 0.3172069489955902,
+      "learning_rate": 2.2337418011446154e-06,
+      "loss": 0.3933,
+      "step": 2587
+    },
+    {
+      "epoch": 2.1542730299667037,
+      "grad_norm": 0.3283001780509949,
+      "learning_rate": 2.229707025255881e-06,
+      "loss": 0.4171,
+      "step": 2588
+    },
+    {
+      "epoch": 2.1551054384017756,
+      "grad_norm": 0.30055010318756104,
+      "learning_rate": 2.225674850720759e-06,
+      "loss": 0.3517,
+      "step": 2589
+    },
+    {
+      "epoch": 2.155937846836848,
+      "grad_norm": 0.33835846185684204,
+      "learning_rate": 2.2216452813255273e-06,
+      "loss": 0.3955,
+      "step": 2590
+    },
+    {
+      "epoch": 2.15677025527192,
+      "grad_norm": 0.3244904577732086,
+      "learning_rate": 2.2176183208540236e-06,
+      "loss": 0.4064,
+      "step": 2591
+    },
+    {
+      "epoch": 2.1576026637069923,
+      "grad_norm": 0.32495352625846863,
+      "learning_rate": 2.2135939730876344e-06,
+      "loss": 0.4063,
+      "step": 2592
+    },
+    {
+      "epoch": 2.158435072142064,
+      "grad_norm": 0.32691752910614014,
+      "learning_rate": 2.2095722418052916e-06,
+      "loss": 0.3824,
+      "step": 2593
+    },
+    {
+      "epoch": 2.1592674805771366,
+      "grad_norm": 0.33834582567214966,
+      "learning_rate": 2.2055531307834734e-06,
+      "loss": 0.3959,
+      "step": 2594
+    },
+    {
+      "epoch": 2.1600998890122085,
+      "grad_norm": 0.3010924756526947,
+      "learning_rate": 2.2015366437961932e-06,
+      "loss": 0.3838,
+      "step": 2595
+    },
+    {
+      "epoch": 2.160932297447281,
+      "grad_norm": 0.3164057731628418,
+      "learning_rate": 2.197522784615004e-06,
+      "loss": 0.3656,
+      "step": 2596
+    },
+    {
+      "epoch": 2.161764705882353,
+      "grad_norm": 0.3379717171192169,
+      "learning_rate": 2.1935115570089897e-06,
+      "loss": 0.4121,
+      "step": 2597
+    },
+    {
+      "epoch": 2.162597114317425,
+      "grad_norm": 0.2782793939113617,
+      "learning_rate": 2.189502964744763e-06,
+      "loss": 0.3191,
+      "step": 2598
+    },
+    {
+      "epoch": 2.163429522752497,
+      "grad_norm": 0.2974027991294861,
+      "learning_rate": 2.1854970115864623e-06,
+      "loss": 0.4204,
+      "step": 2599
+    },
+    {
+      "epoch": 2.1642619311875695,
+      "grad_norm": 0.2943418622016907,
+      "learning_rate": 2.1814937012957476e-06,
+      "loss": 0.4057,
+      "step": 2600
+    },
+    {
+      "epoch": 2.1650943396226414,
+      "grad_norm": 0.30796945095062256,
+      "learning_rate": 2.1774930376317976e-06,
+      "loss": 0.3434,
+      "step": 2601
+    },
+    {
+      "epoch": 2.165926748057714,
+      "grad_norm": 0.35572606325149536,
+      "learning_rate": 2.1734950243513054e-06,
+      "loss": 0.4329,
+      "step": 2602
+    },
+    {
+      "epoch": 2.1667591564927857,
+      "grad_norm": 0.30681103467941284,
+      "learning_rate": 2.1694996652084752e-06,
+      "loss": 0.3559,
+      "step": 2603
+    },
+    {
+      "epoch": 2.167591564927858,
+      "grad_norm": 0.30400606989860535,
+      "learning_rate": 2.165506963955022e-06,
+      "loss": 0.3995,
+      "step": 2604
+    },
+    {
+      "epoch": 2.16842397336293,
+      "grad_norm": 0.30469441413879395,
+      "learning_rate": 2.1615169243401557e-06,
+      "loss": 0.3993,
+      "step": 2605
+    },
+    {
+      "epoch": 2.1692563817980024,
+      "grad_norm": 0.3207082748413086,
+      "learning_rate": 2.1575295501105987e-06,
+      "loss": 0.4184,
+      "step": 2606
+    },
+    {
+      "epoch": 2.1700887902330743,
+      "grad_norm": 0.30796340107917786,
+      "learning_rate": 2.1535448450105644e-06,
+      "loss": 0.3651,
+      "step": 2607
+    },
+    {
+      "epoch": 2.1709211986681467,
+      "grad_norm": 0.3026013672351837,
+      "learning_rate": 2.1495628127817618e-06,
+      "loss": 0.3816,
+      "step": 2608
+    },
+    {
+      "epoch": 2.1717536071032186,
+      "grad_norm": 0.33026769757270813,
+      "learning_rate": 2.1455834571633836e-06,
+      "loss": 0.4104,
+      "step": 2609
+    },
+    {
+      "epoch": 2.1725860155382906,
+      "grad_norm": 0.32119134068489075,
+      "learning_rate": 2.14160678189212e-06,
+      "loss": 0.4247,
+      "step": 2610
+    },
+    {
+      "epoch": 2.173418423973363,
+      "grad_norm": 0.30723845958709717,
+      "learning_rate": 2.1376327907021385e-06,
+      "loss": 0.3703,
+      "step": 2611
+    },
+    {
+      "epoch": 2.1742508324084353,
+      "grad_norm": 0.29472365975379944,
+      "learning_rate": 2.133661487325082e-06,
+      "loss": 0.3878,
+      "step": 2612
+    },
+    {
+      "epoch": 2.1750832408435072,
+      "grad_norm": 0.3001120090484619,
+      "learning_rate": 2.1296928754900753e-06,
+      "loss": 0.3795,
+      "step": 2613
+    },
+    {
+      "epoch": 2.175915649278579,
+      "grad_norm": 0.31931760907173157,
+      "learning_rate": 2.125726958923718e-06,
+      "loss": 0.3901,
+      "step": 2614
+    },
+    {
+      "epoch": 2.1767480577136515,
+      "grad_norm": 0.3011782765388489,
+      "learning_rate": 2.1217637413500735e-06,
+      "loss": 0.3744,
+      "step": 2615
+    },
+    {
+      "epoch": 2.1775804661487235,
+      "grad_norm": 0.3152015507221222,
+      "learning_rate": 2.1178032264906704e-06,
+      "loss": 0.382,
+      "step": 2616
+    },
+    {
+      "epoch": 2.178412874583796,
+      "grad_norm": 0.3104303479194641,
+      "learning_rate": 2.1138454180645035e-06,
+      "loss": 0.3745,
+      "step": 2617
+    },
+    {
+      "epoch": 2.1792452830188678,
+      "grad_norm": 0.32811030745506287,
+      "learning_rate": 2.109890319788023e-06,
+      "loss": 0.3892,
+      "step": 2618
+    },
+    {
+      "epoch": 2.18007769145394,
+      "grad_norm": 0.32053300738334656,
+      "learning_rate": 2.105937935375136e-06,
+      "loss": 0.372,
+      "step": 2619
+    },
+    {
+      "epoch": 2.180910099889012,
+      "grad_norm": 0.3680615723133087,
+      "learning_rate": 2.1019882685372016e-06,
+      "loss": 0.4414,
+      "step": 2620
+    },
+    {
+      "epoch": 2.1817425083240845,
+      "grad_norm": 0.29054370522499084,
+      "learning_rate": 2.0980413229830248e-06,
+      "loss": 0.3299,
+      "step": 2621
+    },
+    {
+      "epoch": 2.1825749167591564,
+      "grad_norm": 0.31951627135276794,
+      "learning_rate": 2.094097102418857e-06,
+      "loss": 0.3982,
+      "step": 2622
+    },
+    {
+      "epoch": 2.1834073251942288,
+      "grad_norm": 0.3075707256793976,
+      "learning_rate": 2.09015561054839e-06,
+      "loss": 0.3976,
+      "step": 2623
+    },
+    {
+      "epoch": 2.1842397336293007,
+      "grad_norm": 0.28511252999305725,
+      "learning_rate": 2.0862168510727545e-06,
+      "loss": 0.3907,
+      "step": 2624
+    },
+    {
+      "epoch": 2.185072142064373,
+      "grad_norm": 0.2891198694705963,
+      "learning_rate": 2.0822808276905144e-06,
+      "loss": 0.3904,
+      "step": 2625
+    },
+    {
+      "epoch": 2.185904550499445,
+      "grad_norm": 0.3299737870693207,
+      "learning_rate": 2.0783475440976635e-06,
+      "loss": 0.3974,
+      "step": 2626
+    },
+    {
+      "epoch": 2.1867369589345174,
+      "grad_norm": 0.3298386335372925,
+      "learning_rate": 2.0744170039876255e-06,
+      "loss": 0.3923,
+      "step": 2627
+    },
+    {
+      "epoch": 2.1875693673695893,
+      "grad_norm": 0.295673131942749,
+      "learning_rate": 2.0704892110512458e-06,
+      "loss": 0.3811,
+      "step": 2628
+    },
+    {
+      "epoch": 2.1884017758046617,
+      "grad_norm": 0.3085336983203888,
+      "learning_rate": 2.0665641689767902e-06,
+      "loss": 0.392,
+      "step": 2629
+    },
+    {
+      "epoch": 2.1892341842397336,
+      "grad_norm": 0.3050670921802521,
+      "learning_rate": 2.0626418814499428e-06,
+      "loss": 0.3935,
+      "step": 2630
+    },
+    {
+      "epoch": 2.190066592674806,
+      "grad_norm": 0.32354554533958435,
+      "learning_rate": 2.0587223521537996e-06,
+      "loss": 0.4429,
+      "step": 2631
+    },
+    {
+      "epoch": 2.190899001109878,
+      "grad_norm": 0.3790428638458252,
+      "learning_rate": 2.0548055847688676e-06,
+      "loss": 0.3709,
+      "step": 2632
+    },
+    {
+      "epoch": 2.19173140954495,
+      "grad_norm": 0.2934742271900177,
+      "learning_rate": 2.0508915829730595e-06,
+      "loss": 0.3677,
+      "step": 2633
+    },
+    {
+      "epoch": 2.192563817980022,
+      "grad_norm": 0.30265501141548157,
+      "learning_rate": 2.046980350441694e-06,
+      "loss": 0.4166,
+      "step": 2634
+    },
+    {
+      "epoch": 2.1933962264150946,
+      "grad_norm": 0.3148394525051117,
+      "learning_rate": 2.0430718908474813e-06,
+      "loss": 0.3848,
+      "step": 2635
+    },
+    {
+      "epoch": 2.1942286348501665,
+      "grad_norm": 0.35043779015541077,
+      "learning_rate": 2.0391662078605383e-06,
+      "loss": 0.4055,
+      "step": 2636
+    },
+    {
+      "epoch": 2.1950610432852384,
+      "grad_norm": 0.27816712856292725,
+      "learning_rate": 2.0352633051483705e-06,
+      "loss": 0.3301,
+      "step": 2637
+    },
+    {
+      "epoch": 2.195893451720311,
+      "grad_norm": 0.3395947813987732,
+      "learning_rate": 2.0313631863758677e-06,
+      "loss": 0.4272,
+      "step": 2638
+    },
+    {
+      "epoch": 2.1967258601553827,
+      "grad_norm": 0.3466813564300537,
+      "learning_rate": 2.02746585520531e-06,
+      "loss": 0.3971,
+      "step": 2639
+    },
+    {
+      "epoch": 2.197558268590455,
+      "grad_norm": 0.32729557156562805,
+      "learning_rate": 2.0235713152963627e-06,
+      "loss": 0.3828,
+      "step": 2640
+    },
+    {
+      "epoch": 2.198390677025527,
+      "grad_norm": 0.35937994718551636,
+      "learning_rate": 2.019679570306068e-06,
+      "loss": 0.3956,
+      "step": 2641
+    },
+    {
+      "epoch": 2.1992230854605994,
+      "grad_norm": 0.2958609163761139,
+      "learning_rate": 2.0157906238888376e-06,
+      "loss": 0.3495,
+      "step": 2642
+    },
+    {
+      "epoch": 2.2000554938956713,
+      "grad_norm": 0.31037241220474243,
+      "learning_rate": 2.0119044796964614e-06,
+      "loss": 0.3748,
+      "step": 2643
+    },
+    {
+      "epoch": 2.2008879023307437,
+      "grad_norm": 0.3517056107521057,
+      "learning_rate": 2.008021141378102e-06,
+      "loss": 0.4095,
+      "step": 2644
+    },
+    {
+      "epoch": 2.2017203107658156,
+      "grad_norm": 0.3355526030063629,
+      "learning_rate": 2.0041406125802764e-06,
+      "loss": 0.3765,
+      "step": 2645
+    },
+    {
+      "epoch": 2.202552719200888,
+      "grad_norm": 0.32547831535339355,
+      "learning_rate": 2.0002628969468713e-06,
+      "loss": 0.4366,
+      "step": 2646
+    },
+    {
+      "epoch": 2.20338512763596,
+      "grad_norm": 0.2848454415798187,
+      "learning_rate": 1.9963879981191288e-06,
+      "loss": 0.3593,
+      "step": 2647
+    },
+    {
+      "epoch": 2.2042175360710323,
+      "grad_norm": 0.3068901598453522,
+      "learning_rate": 1.9925159197356475e-06,
+      "loss": 0.38,
+      "step": 2648
+    },
+    {
+      "epoch": 2.2050499445061043,
+      "grad_norm": 0.33143675327301025,
+      "learning_rate": 1.9886466654323765e-06,
+      "loss": 0.4175,
+      "step": 2649
+    },
+    {
+      "epoch": 2.2058823529411766,
+      "grad_norm": 0.323045551776886,
+      "learning_rate": 1.9847802388426137e-06,
+      "loss": 0.4157,
+      "step": 2650
+    },
+    {
+      "epoch": 2.2067147613762486,
+      "grad_norm": 0.27590107917785645,
+      "learning_rate": 1.9809166435970006e-06,
+      "loss": 0.3653,
+      "step": 2651
+    },
+    {
+      "epoch": 2.207547169811321,
+      "grad_norm": 0.3093126714229584,
+      "learning_rate": 1.9770558833235215e-06,
+      "loss": 0.4101,
+      "step": 2652
+    },
+    {
+      "epoch": 2.208379578246393,
+      "grad_norm": 0.32659175992012024,
+      "learning_rate": 1.973197961647498e-06,
+      "loss": 0.4096,
+      "step": 2653
+    },
+    {
+      "epoch": 2.2092119866814652,
+      "grad_norm": 0.29490897059440613,
+      "learning_rate": 1.969342882191585e-06,
+      "loss": 0.3598,
+      "step": 2654
+    },
+    {
+      "epoch": 2.210044395116537,
+      "grad_norm": 0.2893614172935486,
+      "learning_rate": 1.9654906485757707e-06,
+      "loss": 0.3784,
+      "step": 2655
+    },
+    {
+      "epoch": 2.210876803551609,
+      "grad_norm": 0.2968060374259949,
+      "learning_rate": 1.9616412644173697e-06,
+      "loss": 0.3807,
+      "step": 2656
+    },
+    {
+      "epoch": 2.2117092119866815,
+      "grad_norm": 0.2902612090110779,
+      "learning_rate": 1.957794733331021e-06,
+      "loss": 0.3769,
+      "step": 2657
+    },
+    {
+      "epoch": 2.212541620421754,
+      "grad_norm": 0.30103838443756104,
+      "learning_rate": 1.9539510589286848e-06,
+      "loss": 0.3738,
+      "step": 2658
+    },
+    {
+      "epoch": 2.2133740288568258,
+      "grad_norm": 0.3171117901802063,
+      "learning_rate": 1.950110244819638e-06,
+      "loss": 0.3899,
+      "step": 2659
+    },
+    {
+      "epoch": 2.2142064372918977,
+      "grad_norm": 0.32826194167137146,
+      "learning_rate": 1.9462722946104727e-06,
+      "loss": 0.3894,
+      "step": 2660
+    },
+    {
+      "epoch": 2.21503884572697,
+      "grad_norm": 0.32006773352622986,
+      "learning_rate": 1.942437211905092e-06,
+      "loss": 0.3982,
+      "step": 2661
+    },
+    {
+      "epoch": 2.215871254162042,
+      "grad_norm": 0.33344799280166626,
+      "learning_rate": 1.9386050003047047e-06,
+      "loss": 0.3706,
+      "step": 2662
+    },
+    {
+      "epoch": 2.2167036625971144,
+      "grad_norm": 0.30667856335639954,
+      "learning_rate": 1.9347756634078273e-06,
+      "loss": 0.4086,
+      "step": 2663
+    },
+    {
+      "epoch": 2.2175360710321863,
+      "grad_norm": 0.31673869490623474,
+      "learning_rate": 1.93094920481027e-06,
+      "loss": 0.4003,
+      "step": 2664
+    },
+    {
+      "epoch": 2.2183684794672587,
+      "grad_norm": 0.3328154981136322,
+      "learning_rate": 1.9271256281051443e-06,
+      "loss": 0.4012,
+      "step": 2665
+    },
+    {
+      "epoch": 2.2192008879023306,
+      "grad_norm": 0.32742953300476074,
+      "learning_rate": 1.92330493688286e-06,
+      "loss": 0.4188,
+      "step": 2666
+    },
+    {
+      "epoch": 2.220033296337403,
+      "grad_norm": 0.27765265107154846,
+      "learning_rate": 1.9194871347311115e-06,
+      "loss": 0.39,
+      "step": 2667
+    },
+    {
+      "epoch": 2.220865704772475,
+      "grad_norm": 0.28916940093040466,
+      "learning_rate": 1.91567222523488e-06,
+      "loss": 0.389,
+      "step": 2668
+    },
+    {
+      "epoch": 2.2216981132075473,
+      "grad_norm": 0.2962568402290344,
+      "learning_rate": 1.9118602119764325e-06,
+      "loss": 0.3644,
+      "step": 2669
+    },
+    {
+      "epoch": 2.222530521642619,
+      "grad_norm": 0.33371245861053467,
+      "learning_rate": 1.90805109853532e-06,
+      "loss": 0.4107,
+      "step": 2670
+    },
+    {
+      "epoch": 2.2233629300776916,
+      "grad_norm": 0.3214453160762787,
+      "learning_rate": 1.9042448884883618e-06,
+      "loss": 0.4116,
+      "step": 2671
+    },
+    {
+      "epoch": 2.2241953385127635,
+      "grad_norm": 0.2842710614204407,
+      "learning_rate": 1.9004415854096586e-06,
+      "loss": 0.3673,
+      "step": 2672
+    },
+    {
+      "epoch": 2.225027746947836,
+      "grad_norm": 0.3369787633419037,
+      "learning_rate": 1.8966411928705757e-06,
+      "loss": 0.3995,
+      "step": 2673
+    },
+    {
+      "epoch": 2.225860155382908,
+      "grad_norm": 0.3225659132003784,
+      "learning_rate": 1.8928437144397538e-06,
+      "loss": 0.403,
+      "step": 2674
+    },
+    {
+      "epoch": 2.22669256381798,
+      "grad_norm": 0.32059574127197266,
+      "learning_rate": 1.8890491536830863e-06,
+      "loss": 0.356,
+      "step": 2675
+    },
+    {
+      "epoch": 2.227524972253052,
+      "grad_norm": 0.3181704580783844,
+      "learning_rate": 1.8852575141637347e-06,
+      "loss": 0.4074,
+      "step": 2676
+    },
+    {
+      "epoch": 2.2283573806881245,
+      "grad_norm": 0.317874938249588,
+      "learning_rate": 1.8814687994421138e-06,
+      "loss": 0.3938,
+      "step": 2677
+    },
+    {
+      "epoch": 2.2291897891231964,
+      "grad_norm": 0.3188163638114929,
+      "learning_rate": 1.8776830130758939e-06,
+      "loss": 0.3644,
+      "step": 2678
+    },
+    {
+      "epoch": 2.2300221975582684,
+      "grad_norm": 0.3100307285785675,
+      "learning_rate": 1.873900158619994e-06,
+      "loss": 0.4062,
+      "step": 2679
+    },
+    {
+      "epoch": 2.2308546059933407,
+      "grad_norm": 0.3106209635734558,
+      "learning_rate": 1.8701202396265815e-06,
+      "loss": 0.3857,
+      "step": 2680
+    },
+    {
+      "epoch": 2.231687014428413,
+      "grad_norm": 0.35905230045318604,
+      "learning_rate": 1.866343259645066e-06,
+      "loss": 0.4164,
+      "step": 2681
+    },
+    {
+      "epoch": 2.232519422863485,
+      "grad_norm": 0.28830498456954956,
+      "learning_rate": 1.8625692222220977e-06,
+      "loss": 0.3477,
+      "step": 2682
+    },
+    {
+      "epoch": 2.233351831298557,
+      "grad_norm": 0.30920305848121643,
+      "learning_rate": 1.8587981309015635e-06,
+      "loss": 0.4061,
+      "step": 2683
+    },
+    {
+      "epoch": 2.2341842397336293,
+      "grad_norm": 0.3114646375179291,
+      "learning_rate": 1.8550299892245854e-06,
+      "loss": 0.3915,
+      "step": 2684
+    },
+    {
+      "epoch": 2.2350166481687013,
+      "grad_norm": 0.3108902871608734,
+      "learning_rate": 1.851264800729513e-06,
+      "loss": 0.3879,
+      "step": 2685
+    },
+    {
+      "epoch": 2.2358490566037736,
+      "grad_norm": 0.2933717370033264,
+      "learning_rate": 1.8475025689519256e-06,
+      "loss": 0.3962,
+      "step": 2686
+    },
+    {
+      "epoch": 2.2366814650388456,
+      "grad_norm": 0.29493871331214905,
+      "learning_rate": 1.8437432974246238e-06,
+      "loss": 0.3765,
+      "step": 2687
+    },
+    {
+      "epoch": 2.237513873473918,
+      "grad_norm": 0.31428149342536926,
+      "learning_rate": 1.8399869896776296e-06,
+      "loss": 0.4303,
+      "step": 2688
+    },
+    {
+      "epoch": 2.23834628190899,
+      "grad_norm": 0.30307847261428833,
+      "learning_rate": 1.8362336492381832e-06,
+      "loss": 0.3772,
+      "step": 2689
+    },
+    {
+      "epoch": 2.2391786903440623,
+      "grad_norm": 0.31176915764808655,
+      "learning_rate": 1.8324832796307323e-06,
+      "loss": 0.3836,
+      "step": 2690
+    },
+    {
+      "epoch": 2.240011098779134,
+      "grad_norm": 0.2853330075740814,
+      "learning_rate": 1.8287358843769448e-06,
+      "loss": 0.3783,
+      "step": 2691
+    },
+    {
+      "epoch": 2.2408435072142066,
+      "grad_norm": 0.28568682074546814,
+      "learning_rate": 1.8249914669956886e-06,
+      "loss": 0.392,
+      "step": 2692
+    },
+    {
+      "epoch": 2.2416759156492785,
+      "grad_norm": 0.2875567674636841,
+      "learning_rate": 1.8212500310030385e-06,
+      "loss": 0.4076,
+      "step": 2693
+    },
+    {
+      "epoch": 2.242508324084351,
+      "grad_norm": 0.28917601704597473,
+      "learning_rate": 1.8175115799122656e-06,
+      "loss": 0.3889,
+      "step": 2694
+    },
+    {
+      "epoch": 2.243340732519423,
+      "grad_norm": 0.29513514041900635,
+      "learning_rate": 1.8137761172338404e-06,
+      "loss": 0.4111,
+      "step": 2695
+    },
+    {
+      "epoch": 2.244173140954495,
+      "grad_norm": 0.2941649854183197,
+      "learning_rate": 1.810043646475431e-06,
+      "loss": 0.3975,
+      "step": 2696
+    },
+    {
+      "epoch": 2.245005549389567,
+      "grad_norm": 0.28224435448646545,
+      "learning_rate": 1.8063141711418941e-06,
+      "loss": 0.3946,
+      "step": 2697
+    },
+    {
+      "epoch": 2.2458379578246395,
+      "grad_norm": 0.286742627620697,
+      "learning_rate": 1.8025876947352677e-06,
+      "loss": 0.3954,
+      "step": 2698
+    },
+    {
+      "epoch": 2.2466703662597114,
+      "grad_norm": 0.3206160366535187,
+      "learning_rate": 1.7988642207547784e-06,
+      "loss": 0.4243,
+      "step": 2699
+    },
+    {
+      "epoch": 2.2475027746947838,
+      "grad_norm": 0.3006853759288788,
+      "learning_rate": 1.795143752696839e-06,
+      "loss": 0.3759,
+      "step": 2700
+    },
+    {
+      "epoch": 2.2483351831298557,
+      "grad_norm": 0.29394397139549255,
+      "learning_rate": 1.7914262940550292e-06,
+      "loss": 0.389,
+      "step": 2701
+    },
+    {
+      "epoch": 2.2491675915649276,
+      "grad_norm": 0.3121004104614258,
+      "learning_rate": 1.7877118483201095e-06,
+      "loss": 0.3977,
+      "step": 2702
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.3045893609523773,
+      "learning_rate": 1.784000418980007e-06,
+      "loss": 0.4071,
+      "step": 2703
+    },
+    {
+      "epoch": 2.2508324084350724,
+      "grad_norm": 0.30945533514022827,
+      "learning_rate": 1.7802920095198246e-06,
+      "loss": 0.3923,
+      "step": 2704
+    },
+    {
+      "epoch": 2.2516648168701443,
+      "grad_norm": 0.28781136870384216,
+      "learning_rate": 1.7765866234218187e-06,
+      "loss": 0.3648,
+      "step": 2705
+    },
+    {
+      "epoch": 2.2524972253052162,
+      "grad_norm": 0.32264959812164307,
+      "learning_rate": 1.7728842641654125e-06,
+      "loss": 0.4221,
+      "step": 2706
+    },
+    {
+      "epoch": 2.2533296337402886,
+      "grad_norm": 0.30960628390312195,
+      "learning_rate": 1.7691849352271872e-06,
+      "loss": 0.3859,
+      "step": 2707
+    },
+    {
+      "epoch": 2.2541620421753605,
+      "grad_norm": 0.3129541873931885,
+      "learning_rate": 1.7654886400808774e-06,
+      "loss": 0.3869,
+      "step": 2708
+    },
+    {
+      "epoch": 2.254994450610433,
+      "grad_norm": 0.2916148900985718,
+      "learning_rate": 1.7617953821973682e-06,
+      "loss": 0.363,
+      "step": 2709
+    },
+    {
+      "epoch": 2.255826859045505,
+      "grad_norm": 0.3174515664577484,
+      "learning_rate": 1.758105165044694e-06,
+      "loss": 0.4119,
+      "step": 2710
+    },
+    {
+      "epoch": 2.256659267480577,
+      "grad_norm": 0.28080281615257263,
+      "learning_rate": 1.7544179920880333e-06,
+      "loss": 0.3623,
+      "step": 2711
+    },
+    {
+      "epoch": 2.257491675915649,
+      "grad_norm": 0.31006425619125366,
+      "learning_rate": 1.7507338667897062e-06,
+      "loss": 0.4584,
+      "step": 2712
+    },
+    {
+      "epoch": 2.2583240843507215,
+      "grad_norm": 0.3159677982330322,
+      "learning_rate": 1.7470527926091702e-06,
+      "loss": 0.3642,
+      "step": 2713
+    },
+    {
+      "epoch": 2.2591564927857934,
+      "grad_norm": 0.34636712074279785,
+      "learning_rate": 1.7433747730030188e-06,
+      "loss": 0.3917,
+      "step": 2714
+    },
+    {
+      "epoch": 2.259988901220866,
+      "grad_norm": 0.28924471139907837,
+      "learning_rate": 1.7396998114249786e-06,
+      "loss": 0.3665,
+      "step": 2715
+    },
+    {
+      "epoch": 2.2608213096559377,
+      "grad_norm": 0.3206512928009033,
+      "learning_rate": 1.7360279113258977e-06,
+      "loss": 0.3851,
+      "step": 2716
+    },
+    {
+      "epoch": 2.26165371809101,
+      "grad_norm": 0.30951130390167236,
+      "learning_rate": 1.7323590761537595e-06,
+      "loss": 0.4265,
+      "step": 2717
+    },
+    {
+      "epoch": 2.262486126526082,
+      "grad_norm": 0.318645179271698,
+      "learning_rate": 1.7286933093536634e-06,
+      "loss": 0.3968,
+      "step": 2718
+    },
+    {
+      "epoch": 2.2633185349611544,
+      "grad_norm": 0.31277957558631897,
+      "learning_rate": 1.7250306143678292e-06,
+      "loss": 0.3947,
+      "step": 2719
+    },
+    {
+      "epoch": 2.2641509433962264,
+      "grad_norm": 0.30135253071784973,
+      "learning_rate": 1.7213709946355879e-06,
+      "loss": 0.3715,
+      "step": 2720
+    },
+    {
+      "epoch": 2.2649833518312987,
+      "grad_norm": 0.3124295771121979,
+      "learning_rate": 1.7177144535933903e-06,
+      "loss": 0.4376,
+      "step": 2721
+    },
+    {
+      "epoch": 2.2658157602663707,
+      "grad_norm": 0.30148327350616455,
+      "learning_rate": 1.7140609946747915e-06,
+      "loss": 0.382,
+      "step": 2722
+    },
+    {
+      "epoch": 2.266648168701443,
+      "grad_norm": 0.32914572954177856,
+      "learning_rate": 1.7104106213104554e-06,
+      "loss": 0.4407,
+      "step": 2723
+    },
+    {
+      "epoch": 2.267480577136515,
+      "grad_norm": 0.3174681067466736,
+      "learning_rate": 1.7067633369281422e-06,
+      "loss": 0.3585,
+      "step": 2724
+    },
+    {
+      "epoch": 2.268312985571587,
+      "grad_norm": 0.29542532563209534,
+      "learning_rate": 1.7031191449527162e-06,
+      "loss": 0.3803,
+      "step": 2725
+    },
+    {
+      "epoch": 2.2691453940066593,
+      "grad_norm": 0.2985629439353943,
+      "learning_rate": 1.699478048806143e-06,
+      "loss": 0.3619,
+      "step": 2726
+    },
+    {
+      "epoch": 2.2699778024417316,
+      "grad_norm": 0.3362717628479004,
+      "learning_rate": 1.6958400519074696e-06,
+      "loss": 0.3884,
+      "step": 2727
+    },
+    {
+      "epoch": 2.2708102108768036,
+      "grad_norm": 0.33345919847488403,
+      "learning_rate": 1.6922051576728415e-06,
+      "loss": 0.4431,
+      "step": 2728
+    },
+    {
+      "epoch": 2.2716426193118755,
+      "grad_norm": 0.2820887267589569,
+      "learning_rate": 1.6885733695154855e-06,
+      "loss": 0.319,
+      "step": 2729
+    },
+    {
+      "epoch": 2.272475027746948,
+      "grad_norm": 0.2951180636882782,
+      "learning_rate": 1.6849446908457201e-06,
+      "loss": 0.3953,
+      "step": 2730
+    },
+    {
+      "epoch": 2.27330743618202,
+      "grad_norm": 0.2896723747253418,
+      "learning_rate": 1.6813191250709326e-06,
+      "loss": 0.4086,
+      "step": 2731
+    },
+    {
+      "epoch": 2.274139844617092,
+      "grad_norm": 0.29083332419395447,
+      "learning_rate": 1.6776966755955941e-06,
+      "loss": 0.3748,
+      "step": 2732
+    },
+    {
+      "epoch": 2.274972253052164,
+      "grad_norm": 0.3114146888256073,
+      "learning_rate": 1.674077345821249e-06,
+      "loss": 0.4199,
+      "step": 2733
+    },
+    {
+      "epoch": 2.2758046614872365,
+      "grad_norm": 0.311087965965271,
+      "learning_rate": 1.6704611391465103e-06,
+      "loss": 0.4003,
+      "step": 2734
+    },
+    {
+      "epoch": 2.2766370699223084,
+      "grad_norm": 0.3135244846343994,
+      "learning_rate": 1.6668480589670604e-06,
+      "loss": 0.3827,
+      "step": 2735
+    },
+    {
+      "epoch": 2.277469478357381,
+      "grad_norm": 0.3102346956729889,
+      "learning_rate": 1.6632381086756439e-06,
+      "loss": 0.3987,
+      "step": 2736
+    },
+    {
+      "epoch": 2.2783018867924527,
+      "grad_norm": 0.2921448349952698,
+      "learning_rate": 1.6596312916620677e-06,
+      "loss": 0.3694,
+      "step": 2737
+    },
+    {
+      "epoch": 2.279134295227525,
+      "grad_norm": 0.31638967990875244,
+      "learning_rate": 1.6560276113131968e-06,
+      "loss": 0.4139,
+      "step": 2738
+    },
+    {
+      "epoch": 2.279966703662597,
+      "grad_norm": 0.29766252636909485,
+      "learning_rate": 1.6524270710129491e-06,
+      "loss": 0.3582,
+      "step": 2739
+    },
+    {
+      "epoch": 2.2807991120976694,
+      "grad_norm": 0.308046817779541,
+      "learning_rate": 1.6488296741422955e-06,
+      "loss": 0.3882,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2816315205327413,
+      "grad_norm": 0.30890682339668274,
+      "learning_rate": 1.6452354240792561e-06,
+      "loss": 0.4078,
+      "step": 2741
+    },
+    {
+      "epoch": 2.2824639289678137,
+      "grad_norm": 0.31466445326805115,
+      "learning_rate": 1.64164432419889e-06,
+      "loss": 0.3866,
+      "step": 2742
+    },
+    {
+      "epoch": 2.2832963374028856,
+      "grad_norm": 0.3218589425086975,
+      "learning_rate": 1.6380563778733078e-06,
+      "loss": 0.3788,
+      "step": 2743
+    },
+    {
+      "epoch": 2.284128745837958,
+      "grad_norm": 0.3170068860054016,
+      "learning_rate": 1.6344715884716517e-06,
+      "loss": 0.3912,
+      "step": 2744
+    },
+    {
+      "epoch": 2.28496115427303,
+      "grad_norm": 0.28382614254951477,
+      "learning_rate": 1.630889959360104e-06,
+      "loss": 0.3923,
+      "step": 2745
+    },
+    {
+      "epoch": 2.2857935627081023,
+      "grad_norm": 0.3070010840892792,
+      "learning_rate": 1.627311493901872e-06,
+      "loss": 0.4414,
+      "step": 2746
+    },
+    {
+      "epoch": 2.2866259711431742,
+      "grad_norm": 0.2942955493927002,
+      "learning_rate": 1.6237361954572023e-06,
+      "loss": 0.3476,
+      "step": 2747
+    },
+    {
+      "epoch": 2.287458379578246,
+      "grad_norm": 0.2962301969528198,
+      "learning_rate": 1.6201640673833613e-06,
+      "loss": 0.3807,
+      "step": 2748
+    },
+    {
+      "epoch": 2.2882907880133185,
+      "grad_norm": 0.3063211441040039,
+      "learning_rate": 1.6165951130346408e-06,
+      "loss": 0.4005,
+      "step": 2749
+    },
+    {
+      "epoch": 2.289123196448391,
+      "grad_norm": 0.30910301208496094,
+      "learning_rate": 1.6130293357623473e-06,
+      "loss": 0.4213,
+      "step": 2750
+    },
+    {
+      "epoch": 2.289955604883463,
+      "grad_norm": 0.3249872922897339,
+      "learning_rate": 1.6094667389148128e-06,
+      "loss": 0.3776,
+      "step": 2751
+    },
+    {
+      "epoch": 2.2907880133185348,
+      "grad_norm": 0.3110400140285492,
+      "learning_rate": 1.605907325837378e-06,
+      "loss": 0.4065,
+      "step": 2752
+    },
+    {
+      "epoch": 2.291620421753607,
+      "grad_norm": 0.28765764832496643,
+      "learning_rate": 1.6023510998723906e-06,
+      "loss": 0.3952,
+      "step": 2753
+    },
+    {
+      "epoch": 2.292452830188679,
+      "grad_norm": 0.3131961524486542,
+      "learning_rate": 1.598798064359211e-06,
+      "loss": 0.4005,
+      "step": 2754
+    },
+    {
+      "epoch": 2.2932852386237514,
+      "grad_norm": 0.30940961837768555,
+      "learning_rate": 1.5952482226342003e-06,
+      "loss": 0.3693,
+      "step": 2755
+    },
+    {
+      "epoch": 2.2941176470588234,
+      "grad_norm": 0.32620513439178467,
+      "learning_rate": 1.5917015780307265e-06,
+      "loss": 0.418,
+      "step": 2756
+    },
+    {
+      "epoch": 2.2949500554938957,
+      "grad_norm": 0.2872724235057831,
+      "learning_rate": 1.5881581338791462e-06,
+      "loss": 0.343,
+      "step": 2757
+    },
+    {
+      "epoch": 2.2957824639289677,
+      "grad_norm": 0.28437381982803345,
+      "learning_rate": 1.5846178935068173e-06,
+      "loss": 0.3797,
+      "step": 2758
+    },
+    {
+      "epoch": 2.29661487236404,
+      "grad_norm": 0.29751530289649963,
+      "learning_rate": 1.5810808602380872e-06,
+      "loss": 0.4177,
+      "step": 2759
+    },
+    {
+      "epoch": 2.297447280799112,
+      "grad_norm": 0.29856476187705994,
+      "learning_rate": 1.5775470373942926e-06,
+      "loss": 0.3655,
+      "step": 2760
+    },
+    {
+      "epoch": 2.2982796892341844,
+      "grad_norm": 0.3321475684642792,
+      "learning_rate": 1.5740164282937548e-06,
+      "loss": 0.4332,
+      "step": 2761
+    },
+    {
+      "epoch": 2.2991120976692563,
+      "grad_norm": 0.2865569591522217,
+      "learning_rate": 1.5704890362517772e-06,
+      "loss": 0.3488,
+      "step": 2762
+    },
+    {
+      "epoch": 2.2999445061043287,
+      "grad_norm": 0.30639970302581787,
+      "learning_rate": 1.5669648645806428e-06,
+      "loss": 0.3751,
+      "step": 2763
+    },
+    {
+      "epoch": 2.3007769145394006,
+      "grad_norm": 0.3253311812877655,
+      "learning_rate": 1.5634439165896103e-06,
+      "loss": 0.3768,
+      "step": 2764
+    },
+    {
+      "epoch": 2.301609322974473,
+      "grad_norm": 0.33678269386291504,
+      "learning_rate": 1.5599261955849126e-06,
+      "loss": 0.3722,
+      "step": 2765
+    },
+    {
+      "epoch": 2.302441731409545,
+      "grad_norm": 0.32921579480171204,
+      "learning_rate": 1.5564117048697503e-06,
+      "loss": 0.425,
+      "step": 2766
+    },
+    {
+      "epoch": 2.3032741398446173,
+      "grad_norm": 0.29400959610939026,
+      "learning_rate": 1.5529004477442921e-06,
+      "loss": 0.3497,
+      "step": 2767
+    },
+    {
+      "epoch": 2.304106548279689,
+      "grad_norm": 0.30234500765800476,
+      "learning_rate": 1.5493924275056699e-06,
+      "loss": 0.3947,
+      "step": 2768
+    },
+    {
+      "epoch": 2.3049389567147616,
+      "grad_norm": 0.29784688353538513,
+      "learning_rate": 1.5458876474479757e-06,
+      "loss": 0.3688,
+      "step": 2769
+    },
+    {
+      "epoch": 2.3057713651498335,
+      "grad_norm": 0.3415769636631012,
+      "learning_rate": 1.5423861108622601e-06,
+      "loss": 0.4432,
+      "step": 2770
+    },
+    {
+      "epoch": 2.3066037735849054,
+      "grad_norm": 0.2958844304084778,
+      "learning_rate": 1.5388878210365283e-06,
+      "loss": 0.3377,
+      "step": 2771
+    },
+    {
+      "epoch": 2.307436182019978,
+      "grad_norm": 0.3629142940044403,
+      "learning_rate": 1.5353927812557306e-06,
+      "loss": 0.4083,
+      "step": 2772
+    },
+    {
+      "epoch": 2.30826859045505,
+      "grad_norm": 0.3380175530910492,
+      "learning_rate": 1.5319009948017765e-06,
+      "loss": 0.3929,
+      "step": 2773
+    },
+    {
+      "epoch": 2.309100998890122,
+      "grad_norm": 0.2930530607700348,
+      "learning_rate": 1.528412464953512e-06,
+      "loss": 0.389,
+      "step": 2774
+    },
+    {
+      "epoch": 2.309933407325194,
+      "grad_norm": 0.3223347067832947,
+      "learning_rate": 1.5249271949867294e-06,
+      "loss": 0.4326,
+      "step": 2775
+    },
+    {
+      "epoch": 2.3107658157602664,
+      "grad_norm": 0.2680168151855469,
+      "learning_rate": 1.5214451881741544e-06,
+      "loss": 0.3597,
+      "step": 2776
+    },
+    {
+      "epoch": 2.3115982241953383,
+      "grad_norm": 0.29601550102233887,
+      "learning_rate": 1.5179664477854556e-06,
+      "loss": 0.3948,
+      "step": 2777
+    },
+    {
+      "epoch": 2.3124306326304107,
+      "grad_norm": 0.31549885869026184,
+      "learning_rate": 1.5144909770872324e-06,
+      "loss": 0.4028,
+      "step": 2778
+    },
+    {
+      "epoch": 2.3132630410654826,
+      "grad_norm": 0.3089504837989807,
+      "learning_rate": 1.5110187793430086e-06,
+      "loss": 0.3954,
+      "step": 2779
+    },
+    {
+      "epoch": 2.314095449500555,
+      "grad_norm": 0.3053048551082611,
+      "learning_rate": 1.5075498578132398e-06,
+      "loss": 0.4161,
+      "step": 2780
+    },
+    {
+      "epoch": 2.314927857935627,
+      "grad_norm": 0.3325631320476532,
+      "learning_rate": 1.504084215755306e-06,
+      "loss": 0.4186,
+      "step": 2781
+    },
+    {
+      "epoch": 2.3157602663706993,
+      "grad_norm": 0.2945425510406494,
+      "learning_rate": 1.5006218564235058e-06,
+      "loss": 0.3561,
+      "step": 2782
+    },
+    {
+      "epoch": 2.3165926748057712,
+      "grad_norm": 0.35173118114471436,
+      "learning_rate": 1.4971627830690533e-06,
+      "loss": 0.4315,
+      "step": 2783
+    },
+    {
+      "epoch": 2.3174250832408436,
+      "grad_norm": 0.326571524143219,
+      "learning_rate": 1.4937069989400782e-06,
+      "loss": 0.3789,
+      "step": 2784
+    },
+    {
+      "epoch": 2.3182574916759155,
+      "grad_norm": 0.32909318804740906,
+      "learning_rate": 1.4902545072816266e-06,
+      "loss": 0.4114,
+      "step": 2785
+    },
+    {
+      "epoch": 2.319089900110988,
+      "grad_norm": 0.3058091998100281,
+      "learning_rate": 1.4868053113356446e-06,
+      "loss": 0.3674,
+      "step": 2786
+    },
+    {
+      "epoch": 2.31992230854606,
+      "grad_norm": 0.31880733370780945,
+      "learning_rate": 1.483359414340989e-06,
+      "loss": 0.4133,
+      "step": 2787
+    },
+    {
+      "epoch": 2.3207547169811322,
+      "grad_norm": 0.28259992599487305,
+      "learning_rate": 1.4799168195334174e-06,
+      "loss": 0.3619,
+      "step": 2788
+    },
+    {
+      "epoch": 2.321587125416204,
+      "grad_norm": 0.33174195885658264,
+      "learning_rate": 1.4764775301455859e-06,
+      "loss": 0.41,
+      "step": 2789
+    },
+    {
+      "epoch": 2.3224195338512765,
+      "grad_norm": 0.3177911043167114,
+      "learning_rate": 1.4730415494070482e-06,
+      "loss": 0.3861,
+      "step": 2790
+    },
+    {
+      "epoch": 2.3232519422863485,
+      "grad_norm": 0.2958433926105499,
+      "learning_rate": 1.4696088805442505e-06,
+      "loss": 0.3628,
+      "step": 2791
+    },
+    {
+      "epoch": 2.324084350721421,
+      "grad_norm": 0.2952271103858948,
+      "learning_rate": 1.466179526780529e-06,
+      "loss": 0.3991,
+      "step": 2792
+    },
+    {
+      "epoch": 2.3249167591564928,
+      "grad_norm": 0.3214179575443268,
+      "learning_rate": 1.4627534913361064e-06,
+      "loss": 0.409,
+      "step": 2793
+    },
+    {
+      "epoch": 2.3257491675915647,
+      "grad_norm": 0.31064051389694214,
+      "learning_rate": 1.4593307774280895e-06,
+      "loss": 0.4192,
+      "step": 2794
+    },
+    {
+      "epoch": 2.326581576026637,
+      "grad_norm": 0.30104267597198486,
+      "learning_rate": 1.4559113882704683e-06,
+      "loss": 0.3648,
+      "step": 2795
+    },
+    {
+      "epoch": 2.3274139844617094,
+      "grad_norm": 0.3028804659843445,
+      "learning_rate": 1.4524953270741077e-06,
+      "loss": 0.3694,
+      "step": 2796
+    },
+    {
+      "epoch": 2.3282463928967814,
+      "grad_norm": 0.309151828289032,
+      "learning_rate": 1.4490825970467493e-06,
+      "loss": 0.4093,
+      "step": 2797
+    },
+    {
+      "epoch": 2.3290788013318533,
+      "grad_norm": 0.31857770681381226,
+      "learning_rate": 1.4456732013930064e-06,
+      "loss": 0.3984,
+      "step": 2798
+    },
+    {
+      "epoch": 2.3299112097669257,
+      "grad_norm": 0.3101446032524109,
+      "learning_rate": 1.442267143314361e-06,
+      "loss": 0.3722,
+      "step": 2799
+    },
+    {
+      "epoch": 2.3307436182019976,
+      "grad_norm": 0.29778382182121277,
+      "learning_rate": 1.4388644260091617e-06,
+      "loss": 0.3865,
+      "step": 2800
+    },
+    {
+      "epoch": 2.33157602663707,
+      "grad_norm": 0.309423565864563,
+      "learning_rate": 1.435465052672621e-06,
+      "loss": 0.3819,
+      "step": 2801
+    },
+    {
+      "epoch": 2.332408435072142,
+      "grad_norm": 0.301540732383728,
+      "learning_rate": 1.432069026496805e-06,
+      "loss": 0.3653,
+      "step": 2802
+    },
+    {
+      "epoch": 2.3332408435072143,
+      "grad_norm": 0.28494688868522644,
+      "learning_rate": 1.4286763506706474e-06,
+      "loss": 0.3809,
+      "step": 2803
+    },
+    {
+      "epoch": 2.334073251942286,
+      "grad_norm": 0.2828214764595032,
+      "learning_rate": 1.425287028379929e-06,
+      "loss": 0.4076,
+      "step": 2804
+    },
+    {
+      "epoch": 2.3349056603773586,
+      "grad_norm": 0.3035745620727539,
+      "learning_rate": 1.4219010628072806e-06,
+      "loss": 0.4237,
+      "step": 2805
+    },
+    {
+      "epoch": 2.3357380688124305,
+      "grad_norm": 0.30221620202064514,
+      "learning_rate": 1.418518457132182e-06,
+      "loss": 0.4185,
+      "step": 2806
+    },
+    {
+      "epoch": 2.336570477247503,
+      "grad_norm": 0.29032158851623535,
+      "learning_rate": 1.4151392145309634e-06,
+      "loss": 0.4039,
+      "step": 2807
+    },
+    {
+      "epoch": 2.337402885682575,
+      "grad_norm": 0.3030306398868561,
+      "learning_rate": 1.4117633381767925e-06,
+      "loss": 0.4108,
+      "step": 2808
+    },
+    {
+      "epoch": 2.338235294117647,
+      "grad_norm": 0.3300309181213379,
+      "learning_rate": 1.4083908312396727e-06,
+      "loss": 0.4123,
+      "step": 2809
+    },
+    {
+      "epoch": 2.339067702552719,
+      "grad_norm": 0.3179803490638733,
+      "learning_rate": 1.4050216968864477e-06,
+      "loss": 0.3796,
+      "step": 2810
+    },
+    {
+      "epoch": 2.3399001109877915,
+      "grad_norm": 0.28045690059661865,
+      "learning_rate": 1.401655938280798e-06,
+      "loss": 0.368,
+      "step": 2811
+    },
+    {
+      "epoch": 2.3407325194228634,
+      "grad_norm": 0.32688331604003906,
+      "learning_rate": 1.3982935585832253e-06,
+      "loss": 0.3847,
+      "step": 2812
+    },
+    {
+      "epoch": 2.341564927857936,
+      "grad_norm": 0.29030436277389526,
+      "learning_rate": 1.3949345609510645e-06,
+      "loss": 0.3711,
+      "step": 2813
+    },
+    {
+      "epoch": 2.3423973362930077,
+      "grad_norm": 0.33373093605041504,
+      "learning_rate": 1.3915789485384718e-06,
+      "loss": 0.4009,
+      "step": 2814
+    },
+    {
+      "epoch": 2.34322974472808,
+      "grad_norm": 0.3265412151813507,
+      "learning_rate": 1.3882267244964304e-06,
+      "loss": 0.4302,
+      "step": 2815
+    },
+    {
+      "epoch": 2.344062153163152,
+      "grad_norm": 0.30916792154312134,
+      "learning_rate": 1.3848778919727324e-06,
+      "loss": 0.392,
+      "step": 2816
+    },
+    {
+      "epoch": 2.344894561598224,
+      "grad_norm": 0.30912598967552185,
+      "learning_rate": 1.3815324541119924e-06,
+      "loss": 0.3968,
+      "step": 2817
+    },
+    {
+      "epoch": 2.3457269700332963,
+      "grad_norm": 0.31527265906333923,
+      "learning_rate": 1.3781904140556352e-06,
+      "loss": 0.3742,
+      "step": 2818
+    },
+    {
+      "epoch": 2.3465593784683687,
+      "grad_norm": 0.2920280694961548,
+      "learning_rate": 1.3748517749418944e-06,
+      "loss": 0.3646,
+      "step": 2819
+    },
+    {
+      "epoch": 2.3473917869034406,
+      "grad_norm": 0.3410710394382477,
+      "learning_rate": 1.3715165399058106e-06,
+      "loss": 0.426,
+      "step": 2820
+    },
+    {
+      "epoch": 2.3482241953385126,
+      "grad_norm": 0.3250875771045685,
+      "learning_rate": 1.368184712079228e-06,
+      "loss": 0.417,
+      "step": 2821
+    },
+    {
+      "epoch": 2.349056603773585,
+      "grad_norm": 0.2889362573623657,
+      "learning_rate": 1.3648562945907916e-06,
+      "loss": 0.3865,
+      "step": 2822
+    },
+    {
+      "epoch": 2.349889012208657,
+      "grad_norm": 0.3278926610946655,
+      "learning_rate": 1.3615312905659434e-06,
+      "loss": 0.4317,
+      "step": 2823
+    },
+    {
+      "epoch": 2.3507214206437292,
+      "grad_norm": 0.3064905107021332,
+      "learning_rate": 1.3582097031269208e-06,
+      "loss": 0.3656,
+      "step": 2824
+    },
+    {
+      "epoch": 2.351553829078801,
+      "grad_norm": 0.31251177191734314,
+      "learning_rate": 1.3548915353927516e-06,
+      "loss": 0.415,
+      "step": 2825
+    },
+    {
+      "epoch": 2.3523862375138735,
+      "grad_norm": 0.3331948220729828,
+      "learning_rate": 1.3515767904792548e-06,
+      "loss": 0.3624,
+      "step": 2826
+    },
+    {
+      "epoch": 2.3532186459489455,
+      "grad_norm": 0.31363222002983093,
+      "learning_rate": 1.3482654714990323e-06,
+      "loss": 0.3953,
+      "step": 2827
+    },
+    {
+      "epoch": 2.354051054384018,
+      "grad_norm": 0.30317604541778564,
+      "learning_rate": 1.3449575815614719e-06,
+      "loss": 0.4045,
+      "step": 2828
+    },
+    {
+      "epoch": 2.35488346281909,
+      "grad_norm": 0.30388328433036804,
+      "learning_rate": 1.3416531237727398e-06,
+      "loss": 0.4031,
+      "step": 2829
+    },
+    {
+      "epoch": 2.355715871254162,
+      "grad_norm": 0.30564767122268677,
+      "learning_rate": 1.338352101235781e-06,
+      "loss": 0.3676,
+      "step": 2830
+    },
+    {
+      "epoch": 2.356548279689234,
+      "grad_norm": 0.3132300078868866,
+      "learning_rate": 1.3350545170503087e-06,
+      "loss": 0.4245,
+      "step": 2831
+    },
+    {
+      "epoch": 2.3573806881243065,
+      "grad_norm": 0.3112890124320984,
+      "learning_rate": 1.3317603743128177e-06,
+      "loss": 0.3984,
+      "step": 2832
+    },
+    {
+      "epoch": 2.3582130965593784,
+      "grad_norm": 0.32551440596580505,
+      "learning_rate": 1.3284696761165634e-06,
+      "loss": 0.4194,
+      "step": 2833
+    },
+    {
+      "epoch": 2.3590455049944508,
+      "grad_norm": 0.3218403160572052,
+      "learning_rate": 1.3251824255515704e-06,
+      "loss": 0.4249,
+      "step": 2834
+    },
+    {
+      "epoch": 2.3598779134295227,
+      "grad_norm": 0.2741534113883972,
+      "learning_rate": 1.3218986257046217e-06,
+      "loss": 0.3283,
+      "step": 2835
+    },
+    {
+      "epoch": 2.360710321864595,
+      "grad_norm": 0.33001041412353516,
+      "learning_rate": 1.3186182796592634e-06,
+      "loss": 0.4112,
+      "step": 2836
+    },
+    {
+      "epoch": 2.361542730299667,
+      "grad_norm": 0.29912424087524414,
+      "learning_rate": 1.3153413904958024e-06,
+      "loss": 0.3664,
+      "step": 2837
+    },
+    {
+      "epoch": 2.3623751387347394,
+      "grad_norm": 0.2997702360153198,
+      "learning_rate": 1.3120679612912896e-06,
+      "loss": 0.3769,
+      "step": 2838
+    },
+    {
+      "epoch": 2.3632075471698113,
+      "grad_norm": 0.295955628156662,
+      "learning_rate": 1.308797995119534e-06,
+      "loss": 0.3931,
+      "step": 2839
+    },
+    {
+      "epoch": 2.3640399556048832,
+      "grad_norm": 0.3053940236568451,
+      "learning_rate": 1.30553149505109e-06,
+      "loss": 0.3741,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3648723640399556,
+      "grad_norm": 0.3083863854408264,
+      "learning_rate": 1.302268464153263e-06,
+      "loss": 0.4192,
+      "step": 2841
+    },
+    {
+      "epoch": 2.365704772475028,
+      "grad_norm": 0.3068256378173828,
+      "learning_rate": 1.2990089054900918e-06,
+      "loss": 0.4053,
+      "step": 2842
+    },
+    {
+      "epoch": 2.3665371809101,
+      "grad_norm": 0.3323042392730713,
+      "learning_rate": 1.2957528221223591e-06,
+      "loss": 0.44,
+      "step": 2843
+    },
+    {
+      "epoch": 2.367369589345172,
+      "grad_norm": 0.29590079188346863,
+      "learning_rate": 1.2925002171075846e-06,
+      "loss": 0.3591,
+      "step": 2844
+    },
+    {
+      "epoch": 2.368201997780244,
+      "grad_norm": 0.3084840476512909,
+      "learning_rate": 1.2892510935000252e-06,
+      "loss": 0.4094,
+      "step": 2845
+    },
+    {
+      "epoch": 2.369034406215316,
+      "grad_norm": 0.2761549651622772,
+      "learning_rate": 1.2860054543506595e-06,
+      "loss": 0.349,
+      "step": 2846
+    },
+    {
+      "epoch": 2.3698668146503885,
+      "grad_norm": 0.30774176120758057,
+      "learning_rate": 1.2827633027072017e-06,
+      "loss": 0.4483,
+      "step": 2847
+    },
+    {
+      "epoch": 2.3706992230854604,
+      "grad_norm": 0.2745443284511566,
+      "learning_rate": 1.2795246416140895e-06,
+      "loss": 0.3591,
+      "step": 2848
+    },
+    {
+      "epoch": 2.371531631520533,
+      "grad_norm": 0.3255792558193207,
+      "learning_rate": 1.2762894741124814e-06,
+      "loss": 0.4007,
+      "step": 2849
+    },
+    {
+      "epoch": 2.3723640399556047,
+      "grad_norm": 0.34070661664009094,
+      "learning_rate": 1.273057803240257e-06,
+      "loss": 0.4215,
+      "step": 2850
+    },
+    {
+      "epoch": 2.373196448390677,
+      "grad_norm": 0.31039872765541077,
+      "learning_rate": 1.2698296320320113e-06,
+      "loss": 0.3986,
+      "step": 2851
+    },
+    {
+      "epoch": 2.374028856825749,
+      "grad_norm": 0.2800928056240082,
+      "learning_rate": 1.2666049635190535e-06,
+      "loss": 0.3828,
+      "step": 2852
+    },
+    {
+      "epoch": 2.3748612652608214,
+      "grad_norm": 0.28755927085876465,
+      "learning_rate": 1.2633838007294048e-06,
+      "loss": 0.3997,
+      "step": 2853
+    },
+    {
+      "epoch": 2.3756936736958933,
+      "grad_norm": 0.28583160042762756,
+      "learning_rate": 1.260166146687793e-06,
+      "loss": 0.3762,
+      "step": 2854
+    },
+    {
+      "epoch": 2.3765260821309657,
+      "grad_norm": 0.33797693252563477,
+      "learning_rate": 1.2569520044156509e-06,
+      "loss": 0.415,
+      "step": 2855
+    },
+    {
+      "epoch": 2.3773584905660377,
+      "grad_norm": 0.31852585077285767,
+      "learning_rate": 1.2537413769311163e-06,
+      "loss": 0.4078,
+      "step": 2856
+    },
+    {
+      "epoch": 2.37819089900111,
+      "grad_norm": 0.3200073838233948,
+      "learning_rate": 1.25053426724902e-06,
+      "loss": 0.3974,
+      "step": 2857
+    },
+    {
+      "epoch": 2.379023307436182,
+      "grad_norm": 0.3143557608127594,
+      "learning_rate": 1.247330678380899e-06,
+      "loss": 0.3691,
+      "step": 2858
+    },
+    {
+      "epoch": 2.3798557158712543,
+      "grad_norm": 0.30389317870140076,
+      "learning_rate": 1.2441306133349785e-06,
+      "loss": 0.4205,
+      "step": 2859
+    },
+    {
+      "epoch": 2.3806881243063263,
+      "grad_norm": 0.2985035181045532,
+      "learning_rate": 1.2409340751161753e-06,
+      "loss": 0.3629,
+      "step": 2860
+    },
+    {
+      "epoch": 2.3815205327413986,
+      "grad_norm": 0.3365674316883087,
+      "learning_rate": 1.2377410667260914e-06,
+      "loss": 0.4372,
+      "step": 2861
+    },
+    {
+      "epoch": 2.3823529411764706,
+      "grad_norm": 0.31339481472969055,
+      "learning_rate": 1.2345515911630223e-06,
+      "loss": 0.4141,
+      "step": 2862
+    },
+    {
+      "epoch": 2.3831853496115425,
+      "grad_norm": 0.3070381283760071,
+      "learning_rate": 1.2313656514219408e-06,
+      "loss": 0.3372,
+      "step": 2863
+    },
+    {
+      "epoch": 2.384017758046615,
+      "grad_norm": 0.2941051423549652,
+      "learning_rate": 1.2281832504944967e-06,
+      "loss": 0.383,
+      "step": 2864
+    },
+    {
+      "epoch": 2.3848501664816872,
+      "grad_norm": 0.2965616285800934,
+      "learning_rate": 1.2250043913690235e-06,
+      "loss": 0.4028,
+      "step": 2865
+    },
+    {
+      "epoch": 2.385682574916759,
+      "grad_norm": 0.30054527521133423,
+      "learning_rate": 1.2218290770305218e-06,
+      "loss": 0.3639,
+      "step": 2866
+    },
+    {
+      "epoch": 2.386514983351831,
+      "grad_norm": 0.3061928451061249,
+      "learning_rate": 1.2186573104606735e-06,
+      "loss": 0.4202,
+      "step": 2867
+    },
+    {
+      "epoch": 2.3873473917869035,
+      "grad_norm": 0.29752233624458313,
+      "learning_rate": 1.2154890946378178e-06,
+      "loss": 0.3646,
+      "step": 2868
+    },
+    {
+      "epoch": 2.3881798002219754,
+      "grad_norm": 0.3129211664199829,
+      "learning_rate": 1.2123244325369665e-06,
+      "loss": 0.4046,
+      "step": 2869
+    },
+    {
+      "epoch": 2.3890122086570478,
+      "grad_norm": 0.3117455244064331,
+      "learning_rate": 1.2091633271297916e-06,
+      "loss": 0.4251,
+      "step": 2870
+    },
+    {
+      "epoch": 2.3898446170921197,
+      "grad_norm": 0.2949092984199524,
+      "learning_rate": 1.20600578138463e-06,
+      "loss": 0.3705,
+      "step": 2871
+    },
+    {
+      "epoch": 2.390677025527192,
+      "grad_norm": 0.31923535466194153,
+      "learning_rate": 1.2028517982664683e-06,
+      "loss": 0.4063,
+      "step": 2872
+    },
+    {
+      "epoch": 2.391509433962264,
+      "grad_norm": 0.30522382259368896,
+      "learning_rate": 1.1997013807369535e-06,
+      "loss": 0.3575,
+      "step": 2873
+    },
+    {
+      "epoch": 2.3923418423973364,
+      "grad_norm": 0.29187336564064026,
+      "learning_rate": 1.196554531754383e-06,
+      "loss": 0.3778,
+      "step": 2874
+    },
+    {
+      "epoch": 2.3931742508324083,
+      "grad_norm": 0.3098593056201935,
+      "learning_rate": 1.193411254273703e-06,
+      "loss": 0.4229,
+      "step": 2875
+    },
+    {
+      "epoch": 2.3940066592674807,
+      "grad_norm": 0.2713955342769623,
+      "learning_rate": 1.1902715512465057e-06,
+      "loss": 0.3677,
+      "step": 2876
+    },
+    {
+      "epoch": 2.3948390677025526,
+      "grad_norm": 0.26880118250846863,
+      "learning_rate": 1.1871354256210277e-06,
+      "loss": 0.4,
+      "step": 2877
+    },
+    {
+      "epoch": 2.395671476137625,
+      "grad_norm": 0.31260430812835693,
+      "learning_rate": 1.1840028803421455e-06,
+      "loss": 0.4367,
+      "step": 2878
+    },
+    {
+      "epoch": 2.396503884572697,
+      "grad_norm": 0.28427496552467346,
+      "learning_rate": 1.1808739183513745e-06,
+      "loss": 0.3538,
+      "step": 2879
+    },
+    {
+      "epoch": 2.3973362930077693,
+      "grad_norm": 0.32038092613220215,
+      "learning_rate": 1.1777485425868639e-06,
+      "loss": 0.368,
+      "step": 2880
+    },
+    {
+      "epoch": 2.398168701442841,
+      "grad_norm": 0.32951870560646057,
+      "learning_rate": 1.1746267559833973e-06,
+      "loss": 0.3891,
+      "step": 2881
+    },
+    {
+      "epoch": 2.3990011098779136,
+      "grad_norm": 0.2811376452445984,
+      "learning_rate": 1.1715085614723881e-06,
+      "loss": 0.375,
+      "step": 2882
+    },
+    {
+      "epoch": 2.3998335183129855,
+      "grad_norm": 0.29977062344551086,
+      "learning_rate": 1.1683939619818708e-06,
+      "loss": 0.4002,
+      "step": 2883
+    },
+    {
+      "epoch": 2.400665926748058,
+      "grad_norm": 0.303155779838562,
+      "learning_rate": 1.1652829604365135e-06,
+      "loss": 0.3454,
+      "step": 2884
+    },
+    {
+      "epoch": 2.40149833518313,
+      "grad_norm": 0.3123762309551239,
+      "learning_rate": 1.1621755597575996e-06,
+      "loss": 0.4271,
+      "step": 2885
+    },
+    {
+      "epoch": 2.4023307436182018,
+      "grad_norm": 0.28050366044044495,
+      "learning_rate": 1.1590717628630337e-06,
+      "loss": 0.3879,
+      "step": 2886
+    },
+    {
+      "epoch": 2.403163152053274,
+      "grad_norm": 0.2736967206001282,
+      "learning_rate": 1.155971572667332e-06,
+      "loss": 0.3857,
+      "step": 2887
+    },
+    {
+      "epoch": 2.4039955604883465,
+      "grad_norm": 0.3153398334980011,
+      "learning_rate": 1.1528749920816319e-06,
+      "loss": 0.391,
+      "step": 2888
+    },
+    {
+      "epoch": 2.4048279689234184,
+      "grad_norm": 0.3044677972793579,
+      "learning_rate": 1.1497820240136753e-06,
+      "loss": 0.4164,
+      "step": 2889
+    },
+    {
+      "epoch": 2.4056603773584904,
+      "grad_norm": 0.3012680411338806,
+      "learning_rate": 1.1466926713678117e-06,
+      "loss": 0.382,
+      "step": 2890
+    },
+    {
+      "epoch": 2.4064927857935627,
+      "grad_norm": 0.32723140716552734,
+      "learning_rate": 1.143606937044997e-06,
+      "loss": 0.4242,
+      "step": 2891
+    },
+    {
+      "epoch": 2.4073251942286347,
+      "grad_norm": 0.3037180006504059,
+      "learning_rate": 1.140524823942793e-06,
+      "loss": 0.4052,
+      "step": 2892
+    },
+    {
+      "epoch": 2.408157602663707,
+      "grad_norm": 0.300297349691391,
+      "learning_rate": 1.137446334955357e-06,
+      "loss": 0.3693,
+      "step": 2893
+    },
+    {
+      "epoch": 2.408990011098779,
+      "grad_norm": 0.30542272329330444,
+      "learning_rate": 1.1343714729734424e-06,
+      "loss": 0.3466,
+      "step": 2894
+    },
+    {
+      "epoch": 2.4098224195338513,
+      "grad_norm": 0.32545021176338196,
+      "learning_rate": 1.1313002408843986e-06,
+      "loss": 0.3899,
+      "step": 2895
+    },
+    {
+      "epoch": 2.4106548279689233,
+      "grad_norm": 0.30949637293815613,
+      "learning_rate": 1.1282326415721657e-06,
+      "loss": 0.4122,
+      "step": 2896
+    },
+    {
+      "epoch": 2.4114872364039956,
+      "grad_norm": 0.27907073497772217,
+      "learning_rate": 1.1251686779172772e-06,
+      "loss": 0.3984,
+      "step": 2897
+    },
+    {
+      "epoch": 2.4123196448390676,
+      "grad_norm": 0.3177613914012909,
+      "learning_rate": 1.122108352796844e-06,
+      "loss": 0.3756,
+      "step": 2898
+    },
+    {
+      "epoch": 2.41315205327414,
+      "grad_norm": 0.31491708755493164,
+      "learning_rate": 1.119051669084567e-06,
+      "loss": 0.4117,
+      "step": 2899
+    },
+    {
+      "epoch": 2.413984461709212,
+      "grad_norm": 0.2828831374645233,
+      "learning_rate": 1.1159986296507259e-06,
+      "loss": 0.3907,
+      "step": 2900
+    },
+    {
+      "epoch": 2.4148168701442843,
+      "grad_norm": 0.281389057636261,
+      "learning_rate": 1.112949237362177e-06,
+      "loss": 0.3957,
+      "step": 2901
+    },
+    {
+      "epoch": 2.415649278579356,
+      "grad_norm": 0.29541143774986267,
+      "learning_rate": 1.1099034950823539e-06,
+      "loss": 0.3692,
+      "step": 2902
+    },
+    {
+      "epoch": 2.4164816870144286,
+      "grad_norm": 0.3205885887145996,
+      "learning_rate": 1.1068614056712624e-06,
+      "loss": 0.4481,
+      "step": 2903
+    },
+    {
+      "epoch": 2.4173140954495005,
+      "grad_norm": 0.3070029020309448,
+      "learning_rate": 1.103822971985477e-06,
+      "loss": 0.3735,
+      "step": 2904
+    },
+    {
+      "epoch": 2.418146503884573,
+      "grad_norm": 0.31543856859207153,
+      "learning_rate": 1.1007881968781403e-06,
+      "loss": 0.3967,
+      "step": 2905
+    },
+    {
+      "epoch": 2.418978912319645,
+      "grad_norm": 0.29121309518814087,
+      "learning_rate": 1.0977570831989593e-06,
+      "loss": 0.3626,
+      "step": 2906
+    },
+    {
+      "epoch": 2.419811320754717,
+      "grad_norm": 0.3155393600463867,
+      "learning_rate": 1.0947296337942026e-06,
+      "loss": 0.4054,
+      "step": 2907
+    },
+    {
+      "epoch": 2.420643729189789,
+      "grad_norm": 0.3240245282649994,
+      "learning_rate": 1.091705851506698e-06,
+      "loss": 0.4109,
+      "step": 2908
+    },
+    {
+      "epoch": 2.421476137624861,
+      "grad_norm": 0.27017274498939514,
+      "learning_rate": 1.088685739175831e-06,
+      "loss": 0.3569,
+      "step": 2909
+    },
+    {
+      "epoch": 2.4223085460599334,
+      "grad_norm": 0.3277089595794678,
+      "learning_rate": 1.085669299637539e-06,
+      "loss": 0.4304,
+      "step": 2910
+    },
+    {
+      "epoch": 2.4231409544950058,
+      "grad_norm": 0.30966663360595703,
+      "learning_rate": 1.0826565357243125e-06,
+      "loss": 0.3984,
+      "step": 2911
+    },
+    {
+      "epoch": 2.4239733629300777,
+      "grad_norm": 0.28824105858802795,
+      "learning_rate": 1.0796474502651893e-06,
+      "loss": 0.3827,
+      "step": 2912
+    },
+    {
+      "epoch": 2.4248057713651496,
+      "grad_norm": 0.29764944314956665,
+      "learning_rate": 1.0766420460857507e-06,
+      "loss": 0.3925,
+      "step": 2913
+    },
+    {
+      "epoch": 2.425638179800222,
+      "grad_norm": 0.31288978457450867,
+      "learning_rate": 1.0736403260081279e-06,
+      "loss": 0.3844,
+      "step": 2914
+    },
+    {
+      "epoch": 2.426470588235294,
+      "grad_norm": 0.2968754172325134,
+      "learning_rate": 1.070642292850987e-06,
+      "loss": 0.3646,
+      "step": 2915
+    },
+    {
+      "epoch": 2.4273029966703663,
+      "grad_norm": 0.3165760636329651,
+      "learning_rate": 1.067647949429534e-06,
+      "loss": 0.4002,
+      "step": 2916
+    },
+    {
+      "epoch": 2.4281354051054382,
+      "grad_norm": 0.2920931875705719,
+      "learning_rate": 1.0646572985555071e-06,
+      "loss": 0.3664,
+      "step": 2917
+    },
+    {
+      "epoch": 2.4289678135405106,
+      "grad_norm": 0.2827328145503998,
+      "learning_rate": 1.0616703430371833e-06,
+      "loss": 0.4066,
+      "step": 2918
+    },
+    {
+      "epoch": 2.4298002219755825,
+      "grad_norm": 0.29017403721809387,
+      "learning_rate": 1.0586870856793657e-06,
+      "loss": 0.3744,
+      "step": 2919
+    },
+    {
+      "epoch": 2.430632630410655,
+      "grad_norm": 0.3129844665527344,
+      "learning_rate": 1.0557075292833836e-06,
+      "loss": 0.4184,
+      "step": 2920
+    },
+    {
+      "epoch": 2.431465038845727,
+      "grad_norm": 0.3073255121707916,
+      "learning_rate": 1.052731676647092e-06,
+      "loss": 0.3929,
+      "step": 2921
+    },
+    {
+      "epoch": 2.432297447280799,
+      "grad_norm": 0.2809605300426483,
+      "learning_rate": 1.049759530564871e-06,
+      "loss": 0.3918,
+      "step": 2922
+    },
+    {
+      "epoch": 2.433129855715871,
+      "grad_norm": 0.29906994104385376,
+      "learning_rate": 1.0467910938276182e-06,
+      "loss": 0.4187,
+      "step": 2923
+    },
+    {
+      "epoch": 2.4339622641509435,
+      "grad_norm": 0.2949870228767395,
+      "learning_rate": 1.0438263692227452e-06,
+      "loss": 0.3776,
+      "step": 2924
+    },
+    {
+      "epoch": 2.4347946725860155,
+      "grad_norm": 0.33523961901664734,
+      "learning_rate": 1.0408653595341812e-06,
+      "loss": 0.4238,
+      "step": 2925
+    },
+    {
+      "epoch": 2.435627081021088,
+      "grad_norm": 0.3024636507034302,
+      "learning_rate": 1.0379080675423664e-06,
+      "loss": 0.4127,
+      "step": 2926
+    },
+    {
+      "epoch": 2.4364594894561598,
+      "grad_norm": 0.2905980944633484,
+      "learning_rate": 1.0349544960242496e-06,
+      "loss": 0.3608,
+      "step": 2927
+    },
+    {
+      "epoch": 2.437291897891232,
+      "grad_norm": 0.3159032464027405,
+      "learning_rate": 1.0320046477532864e-06,
+      "loss": 0.4126,
+      "step": 2928
+    },
+    {
+      "epoch": 2.438124306326304,
+      "grad_norm": 0.3154861629009247,
+      "learning_rate": 1.0290585254994356e-06,
+      "loss": 0.4027,
+      "step": 2929
+    },
+    {
+      "epoch": 2.4389567147613764,
+      "grad_norm": 0.29416272044181824,
+      "learning_rate": 1.0261161320291586e-06,
+      "loss": 0.4003,
+      "step": 2930
+    },
+    {
+      "epoch": 2.4397891231964484,
+      "grad_norm": 0.30298373103141785,
+      "learning_rate": 1.0231774701054126e-06,
+      "loss": 0.3521,
+      "step": 2931
+    },
+    {
+      "epoch": 2.4406215316315203,
+      "grad_norm": 0.3236266076564789,
+      "learning_rate": 1.020242542487654e-06,
+      "loss": 0.4003,
+      "step": 2932
+    },
+    {
+      "epoch": 2.4414539400665927,
+      "grad_norm": 0.2998720407485962,
+      "learning_rate": 1.017311351931831e-06,
+      "loss": 0.3764,
+      "step": 2933
+    },
+    {
+      "epoch": 2.442286348501665,
+      "grad_norm": 0.36174139380455017,
+      "learning_rate": 1.0143839011903822e-06,
+      "loss": 0.3811,
+      "step": 2934
+    },
+    {
+      "epoch": 2.443118756936737,
+      "grad_norm": 0.306686133146286,
+      "learning_rate": 1.0114601930122363e-06,
+      "loss": 0.3726,
+      "step": 2935
+    },
+    {
+      "epoch": 2.443951165371809,
+      "grad_norm": 0.3181160092353821,
+      "learning_rate": 1.0085402301428055e-06,
+      "loss": 0.4274,
+      "step": 2936
+    },
+    {
+      "epoch": 2.4447835738068813,
+      "grad_norm": 0.3226320743560791,
+      "learning_rate": 1.005624015323986e-06,
+      "loss": 0.4226,
+      "step": 2937
+    },
+    {
+      "epoch": 2.445615982241953,
+      "grad_norm": 0.28478971123695374,
+      "learning_rate": 1.0027115512941549e-06,
+      "loss": 0.3913,
+      "step": 2938
+    },
+    {
+      "epoch": 2.4464483906770256,
+      "grad_norm": 0.2990976572036743,
+      "learning_rate": 9.998028407881672e-07,
+      "loss": 0.3943,
+      "step": 2939
+    },
+    {
+      "epoch": 2.4472807991120975,
+      "grad_norm": 0.326755166053772,
+      "learning_rate": 9.96897886537353e-07,
+      "loss": 0.4393,
+      "step": 2940
+    },
+    {
+      "epoch": 2.44811320754717,
+      "grad_norm": 0.29832005500793457,
+      "learning_rate": 9.939966912695143e-07,
+      "loss": 0.3817,
+      "step": 2941
+    },
+    {
+      "epoch": 2.448945615982242,
+      "grad_norm": 0.3191179037094116,
+      "learning_rate": 9.910992577089269e-07,
+      "loss": 0.3915,
+      "step": 2942
+    },
+    {
+      "epoch": 2.449778024417314,
+      "grad_norm": 0.29789450764656067,
+      "learning_rate": 9.882055885763264e-07,
+      "loss": 0.4053,
+      "step": 2943
+    },
+    {
+      "epoch": 2.450610432852386,
+      "grad_norm": 0.3055591285228729,
+      "learning_rate": 9.853156865889234e-07,
+      "loss": 0.4068,
+      "step": 2944
+    },
+    {
+      "epoch": 2.4514428412874585,
+      "grad_norm": 0.2884024381637573,
+      "learning_rate": 9.824295544603863e-07,
+      "loss": 0.3908,
+      "step": 2945
+    },
+    {
+      "epoch": 2.4522752497225304,
+      "grad_norm": 0.29266828298568726,
+      "learning_rate": 9.795471949008411e-07,
+      "loss": 0.3763,
+      "step": 2946
+    },
+    {
+      "epoch": 2.453107658157603,
+      "grad_norm": 0.28186294436454773,
+      "learning_rate": 9.766686106168744e-07,
+      "loss": 0.3785,
+      "step": 2947
+    },
+    {
+      "epoch": 2.4539400665926747,
+      "grad_norm": 0.2927301228046417,
+      "learning_rate": 9.73793804311529e-07,
+      "loss": 0.4081,
+      "step": 2948
+    },
+    {
+      "epoch": 2.454772475027747,
+      "grad_norm": 0.30842170119285583,
+      "learning_rate": 9.70922778684299e-07,
+      "loss": 0.4011,
+      "step": 2949
+    },
+    {
+      "epoch": 2.455604883462819,
+      "grad_norm": 0.30753591656684875,
+      "learning_rate": 9.680555364311251e-07,
+      "loss": 0.391,
+      "step": 2950
+    },
+    {
+      "epoch": 2.4564372918978914,
+      "grad_norm": 0.3024488091468811,
+      "learning_rate": 9.651920802443971e-07,
+      "loss": 0.3796,
+      "step": 2951
+    },
+    {
+      "epoch": 2.4572697003329633,
+      "grad_norm": 0.29086968302726746,
+      "learning_rate": 9.623324128129557e-07,
+      "loss": 0.3864,
+      "step": 2952
+    },
+    {
+      "epoch": 2.4581021087680357,
+      "grad_norm": 0.28945299983024597,
+      "learning_rate": 9.594765368220737e-07,
+      "loss": 0.3656,
+      "step": 2953
+    },
+    {
+      "epoch": 2.4589345172031076,
+      "grad_norm": 0.30900830030441284,
+      "learning_rate": 9.56624454953471e-07,
+      "loss": 0.4122,
+      "step": 2954
+    },
+    {
+      "epoch": 2.4597669256381796,
+      "grad_norm": 0.3098362982273102,
+      "learning_rate": 9.537761698853016e-07,
+      "loss": 0.3975,
+      "step": 2955
+    },
+    {
+      "epoch": 2.460599334073252,
+      "grad_norm": 0.2963133454322815,
+      "learning_rate": 9.509316842921551e-07,
+      "loss": 0.3775,
+      "step": 2956
+    },
+    {
+      "epoch": 2.4614317425083243,
+      "grad_norm": 0.28471919894218445,
+      "learning_rate": 9.480910008450534e-07,
+      "loss": 0.3706,
+      "step": 2957
+    },
+    {
+      "epoch": 2.4622641509433962,
+      "grad_norm": 0.32631343603134155,
+      "learning_rate": 9.452541222114481e-07,
+      "loss": 0.4478,
+      "step": 2958
+    },
+    {
+      "epoch": 2.463096559378468,
+      "grad_norm": 0.2862439751625061,
+      "learning_rate": 9.424210510552179e-07,
+      "loss": 0.35,
+      "step": 2959
+    },
+    {
+      "epoch": 2.4639289678135405,
+      "grad_norm": 0.2939037084579468,
+      "learning_rate": 9.395917900366663e-07,
+      "loss": 0.4093,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4647613762486125,
+      "grad_norm": 0.27758315205574036,
+      "learning_rate": 9.36766341812519e-07,
+      "loss": 0.3851,
+      "step": 2961
+    },
+    {
+      "epoch": 2.465593784683685,
+      "grad_norm": 0.3188072443008423,
+      "learning_rate": 9.33944709035921e-07,
+      "loss": 0.4501,
+      "step": 2962
+    },
+    {
+      "epoch": 2.4664261931187568,
+      "grad_norm": 0.29086801409721375,
+      "learning_rate": 9.31126894356435e-07,
+      "loss": 0.386,
+      "step": 2963
+    },
+    {
+      "epoch": 2.467258601553829,
+      "grad_norm": 0.30861911177635193,
+      "learning_rate": 9.283129004200381e-07,
+      "loss": 0.4013,
+      "step": 2964
+    },
+    {
+      "epoch": 2.468091009988901,
+      "grad_norm": 0.31104356050491333,
+      "learning_rate": 9.255027298691205e-07,
+      "loss": 0.3772,
+      "step": 2965
+    },
+    {
+      "epoch": 2.4689234184239734,
+      "grad_norm": 0.3085804879665375,
+      "learning_rate": 9.226963853424815e-07,
+      "loss": 0.3887,
+      "step": 2966
+    },
+    {
+      "epoch": 2.4697558268590454,
+      "grad_norm": 0.3194611072540283,
+      "learning_rate": 9.198938694753268e-07,
+      "loss": 0.4214,
+      "step": 2967
+    },
+    {
+      "epoch": 2.4705882352941178,
+      "grad_norm": 0.30297356843948364,
+      "learning_rate": 9.170951848992693e-07,
+      "loss": 0.4159,
+      "step": 2968
+    },
+    {
+      "epoch": 2.4714206437291897,
+      "grad_norm": 0.3000932037830353,
+      "learning_rate": 9.143003342423212e-07,
+      "loss": 0.4052,
+      "step": 2969
+    },
+    {
+      "epoch": 2.472253052164262,
+      "grad_norm": 0.300144225358963,
+      "learning_rate": 9.115093201288977e-07,
+      "loss": 0.4005,
+      "step": 2970
+    },
+    {
+      "epoch": 2.473085460599334,
+      "grad_norm": 0.304669588804245,
+      "learning_rate": 9.0872214517981e-07,
+      "loss": 0.3844,
+      "step": 2971
+    },
+    {
+      "epoch": 2.4739178690344064,
+      "grad_norm": 0.2950827181339264,
+      "learning_rate": 9.059388120122626e-07,
+      "loss": 0.3914,
+      "step": 2972
+    },
+    {
+      "epoch": 2.4747502774694783,
+      "grad_norm": 0.2657037079334259,
+      "learning_rate": 9.031593232398539e-07,
+      "loss": 0.3532,
+      "step": 2973
+    },
+    {
+      "epoch": 2.4755826859045507,
+      "grad_norm": 0.3045920133590698,
+      "learning_rate": 9.003836814725742e-07,
+      "loss": 0.4007,
+      "step": 2974
+    },
+    {
+      "epoch": 2.4764150943396226,
+      "grad_norm": 0.295762300491333,
+      "learning_rate": 8.976118893168006e-07,
+      "loss": 0.3683,
+      "step": 2975
+    },
+    {
+      "epoch": 2.477247502774695,
+      "grad_norm": 0.29629120230674744,
+      "learning_rate": 8.94843949375292e-07,
+      "loss": 0.4194,
+      "step": 2976
+    },
+    {
+      "epoch": 2.478079911209767,
+      "grad_norm": 0.28778138756752014,
+      "learning_rate": 8.920798642471918e-07,
+      "loss": 0.3705,
+      "step": 2977
+    },
+    {
+      "epoch": 2.478912319644839,
+      "grad_norm": 0.32166433334350586,
+      "learning_rate": 8.893196365280282e-07,
+      "loss": 0.4143,
+      "step": 2978
+    },
+    {
+      "epoch": 2.479744728079911,
+      "grad_norm": 0.301952987909317,
+      "learning_rate": 8.865632688097004e-07,
+      "loss": 0.381,
+      "step": 2979
+    },
+    {
+      "epoch": 2.4805771365149836,
+      "grad_norm": 0.2990775406360626,
+      "learning_rate": 8.83810763680486e-07,
+      "loss": 0.3671,
+      "step": 2980
+    },
+    {
+      "epoch": 2.4814095449500555,
+      "grad_norm": 0.31099602580070496,
+      "learning_rate": 8.810621237250355e-07,
+      "loss": 0.4344,
+      "step": 2981
+    },
+    {
+      "epoch": 2.4822419533851274,
+      "grad_norm": 0.2816608250141144,
+      "learning_rate": 8.783173515243725e-07,
+      "loss": 0.3796,
+      "step": 2982
+    },
+    {
+      "epoch": 2.4830743618202,
+      "grad_norm": 0.27356284856796265,
+      "learning_rate": 8.755764496558838e-07,
+      "loss": 0.3578,
+      "step": 2983
+    },
+    {
+      "epoch": 2.4839067702552717,
+      "grad_norm": 0.32336634397506714,
+      "learning_rate": 8.728394206933239e-07,
+      "loss": 0.4262,
+      "step": 2984
+    },
+    {
+      "epoch": 2.484739178690344,
+      "grad_norm": 0.3039683699607849,
+      "learning_rate": 8.701062672068122e-07,
+      "loss": 0.3967,
+      "step": 2985
+    },
+    {
+      "epoch": 2.485571587125416,
+      "grad_norm": 0.2810039520263672,
+      "learning_rate": 8.673769917628272e-07,
+      "loss": 0.3698,
+      "step": 2986
+    },
+    {
+      "epoch": 2.4864039955604884,
+      "grad_norm": 0.2974560260772705,
+      "learning_rate": 8.646515969242065e-07,
+      "loss": 0.3925,
+      "step": 2987
+    },
+    {
+      "epoch": 2.4872364039955603,
+      "grad_norm": 0.3146367371082306,
+      "learning_rate": 8.619300852501427e-07,
+      "loss": 0.425,
+      "step": 2988
+    },
+    {
+      "epoch": 2.4880688124306327,
+      "grad_norm": 0.29236578941345215,
+      "learning_rate": 8.592124592961843e-07,
+      "loss": 0.3563,
+      "step": 2989
+    },
+    {
+      "epoch": 2.4889012208657046,
+      "grad_norm": 0.29718172550201416,
+      "learning_rate": 8.56498721614229e-07,
+      "loss": 0.3998,
+      "step": 2990
+    },
+    {
+      "epoch": 2.489733629300777,
+      "grad_norm": 0.313275009393692,
+      "learning_rate": 8.537888747525236e-07,
+      "loss": 0.4364,
+      "step": 2991
+    },
+    {
+      "epoch": 2.490566037735849,
+      "grad_norm": 0.30545952916145325,
+      "learning_rate": 8.51082921255662e-07,
+      "loss": 0.4099,
+      "step": 2992
+    },
+    {
+      "epoch": 2.4913984461709213,
+      "grad_norm": 0.33663320541381836,
+      "learning_rate": 8.483808636645824e-07,
+      "loss": 0.4288,
+      "step": 2993
+    },
+    {
+      "epoch": 2.4922308546059933,
+      "grad_norm": 0.3012736141681671,
+      "learning_rate": 8.456827045165638e-07,
+      "loss": 0.3524,
+      "step": 2994
+    },
+    {
+      "epoch": 2.4930632630410656,
+      "grad_norm": 0.31417447328567505,
+      "learning_rate": 8.429884463452248e-07,
+      "loss": 0.3816,
+      "step": 2995
+    },
+    {
+      "epoch": 2.4938956714761376,
+      "grad_norm": 0.31972286105155945,
+      "learning_rate": 8.402980916805215e-07,
+      "loss": 0.4074,
+      "step": 2996
+    },
+    {
+      "epoch": 2.49472807991121,
+      "grad_norm": 0.33134639263153076,
+      "learning_rate": 8.376116430487441e-07,
+      "loss": 0.4231,
+      "step": 2997
+    },
+    {
+      "epoch": 2.495560488346282,
+      "grad_norm": 0.28740981221199036,
+      "learning_rate": 8.349291029725126e-07,
+      "loss": 0.369,
+      "step": 2998
+    },
+    {
+      "epoch": 2.4963928967813542,
+      "grad_norm": 0.29424378275871277,
+      "learning_rate": 8.322504739707821e-07,
+      "loss": 0.3857,
+      "step": 2999
+    },
+    {
+      "epoch": 2.497225305216426,
+      "grad_norm": 0.275423526763916,
+      "learning_rate": 8.295757585588304e-07,
+      "loss": 0.3734,
+      "step": 3000
+    },
+    {
+      "epoch": 2.498057713651498,
+      "grad_norm": 0.3008045256137848,
+      "learning_rate": 8.269049592482648e-07,
+      "loss": 0.3864,
+      "step": 3001
+    },
+    {
+      "epoch": 2.4988901220865705,
+      "grad_norm": 0.2958872318267822,
+      "learning_rate": 8.242380785470088e-07,
+      "loss": 0.3839,
+      "step": 3002
+    },
+    {
+      "epoch": 2.499722530521643,
+      "grad_norm": 0.31030040979385376,
+      "learning_rate": 8.215751189593107e-07,
+      "loss": 0.4039,
+      "step": 3003
+    },
+    {
+      "epoch": 2.5005549389567148,
+      "grad_norm": 0.3011883497238159,
+      "learning_rate": 8.189160829857396e-07,
+      "loss": 0.3923,
+      "step": 3004
+    },
+    {
+      "epoch": 2.5013873473917867,
+      "grad_norm": 0.314978688955307,
+      "learning_rate": 8.16260973123173e-07,
+      "loss": 0.3885,
+      "step": 3005
+    },
+    {
+      "epoch": 2.502219755826859,
+      "grad_norm": 0.31006646156311035,
+      "learning_rate": 8.136097918648073e-07,
+      "loss": 0.3841,
+      "step": 3006
+    },
+    {
+      "epoch": 2.5030521642619314,
+      "grad_norm": 0.3044472336769104,
+      "learning_rate": 8.109625417001465e-07,
+      "loss": 0.4087,
+      "step": 3007
+    },
+    {
+      "epoch": 2.5038845726970034,
+      "grad_norm": 0.3212951421737671,
+      "learning_rate": 8.08319225115009e-07,
+      "loss": 0.3969,
+      "step": 3008
+    },
+    {
+      "epoch": 2.5047169811320753,
+      "grad_norm": 0.3018619120121002,
+      "learning_rate": 8.056798445915115e-07,
+      "loss": 0.3801,
+      "step": 3009
+    },
+    {
+      "epoch": 2.5055493895671477,
+      "grad_norm": 0.3130761384963989,
+      "learning_rate": 8.030444026080791e-07,
+      "loss": 0.3594,
+      "step": 3010
+    },
+    {
+      "epoch": 2.5063817980022196,
+      "grad_norm": 0.30970242619514465,
+      "learning_rate": 8.004129016394374e-07,
+      "loss": 0.4117,
+      "step": 3011
+    },
+    {
+      "epoch": 2.507214206437292,
+      "grad_norm": 0.2875807583332062,
+      "learning_rate": 7.977853441566152e-07,
+      "loss": 0.3824,
+      "step": 3012
+    },
+    {
+      "epoch": 2.508046614872364,
+      "grad_norm": 0.3063216805458069,
+      "learning_rate": 7.951617326269318e-07,
+      "loss": 0.3899,
+      "step": 3013
+    },
+    {
+      "epoch": 2.5088790233074363,
+      "grad_norm": 0.2871801257133484,
+      "learning_rate": 7.925420695140052e-07,
+      "loss": 0.3954,
+      "step": 3014
+    },
+    {
+      "epoch": 2.509711431742508,
+      "grad_norm": 0.30276262760162354,
+      "learning_rate": 7.899263572777454e-07,
+      "loss": 0.4196,
+      "step": 3015
+    },
+    {
+      "epoch": 2.5105438401775806,
+      "grad_norm": 0.3048497140407562,
+      "learning_rate": 7.873145983743513e-07,
+      "loss": 0.3925,
+      "step": 3016
+    },
+    {
+      "epoch": 2.5113762486126525,
+      "grad_norm": 0.3085763454437256,
+      "learning_rate": 7.847067952563103e-07,
+      "loss": 0.3928,
+      "step": 3017
+    },
+    {
+      "epoch": 2.512208657047725,
+      "grad_norm": 0.3082400858402252,
+      "learning_rate": 7.821029503723959e-07,
+      "loss": 0.3913,
+      "step": 3018
+    },
+    {
+      "epoch": 2.513041065482797,
+      "grad_norm": 0.2877948582172394,
+      "learning_rate": 7.795030661676633e-07,
+      "loss": 0.3754,
+      "step": 3019
+    },
+    {
+      "epoch": 2.513873473917869,
+      "grad_norm": 0.29405730962753296,
+      "learning_rate": 7.769071450834498e-07,
+      "loss": 0.3884,
+      "step": 3020
+    },
+    {
+      "epoch": 2.514705882352941,
+      "grad_norm": 0.31952765583992004,
+      "learning_rate": 7.743151895573703e-07,
+      "loss": 0.4085,
+      "step": 3021
+    },
+    {
+      "epoch": 2.5155382907880135,
+      "grad_norm": 0.3090550899505615,
+      "learning_rate": 7.717272020233169e-07,
+      "loss": 0.4127,
+      "step": 3022
+    },
+    {
+      "epoch": 2.5163706992230854,
+      "grad_norm": 0.27301451563835144,
+      "learning_rate": 7.691431849114561e-07,
+      "loss": 0.3753,
+      "step": 3023
+    },
+    {
+      "epoch": 2.5172031076581574,
+      "grad_norm": 0.31073275208473206,
+      "learning_rate": 7.665631406482216e-07,
+      "loss": 0.4172,
+      "step": 3024
+    },
+    {
+      "epoch": 2.5180355160932297,
+      "grad_norm": 0.32717081904411316,
+      "learning_rate": 7.639870716563236e-07,
+      "loss": 0.4001,
+      "step": 3025
+    },
+    {
+      "epoch": 2.518867924528302,
+      "grad_norm": 0.27241167426109314,
+      "learning_rate": 7.614149803547354e-07,
+      "loss": 0.3609,
+      "step": 3026
+    },
+    {
+      "epoch": 2.519700332963374,
+      "grad_norm": 0.31278035044670105,
+      "learning_rate": 7.588468691586964e-07,
+      "loss": 0.3819,
+      "step": 3027
+    },
+    {
+      "epoch": 2.520532741398446,
+      "grad_norm": 0.3114195764064789,
+      "learning_rate": 7.562827404797046e-07,
+      "loss": 0.3831,
+      "step": 3028
+    },
+    {
+      "epoch": 2.5213651498335183,
+      "grad_norm": 0.29425179958343506,
+      "learning_rate": 7.537225967255252e-07,
+      "loss": 0.3664,
+      "step": 3029
+    },
+    {
+      "epoch": 2.5221975582685907,
+      "grad_norm": 0.2959001958370209,
+      "learning_rate": 7.511664403001778e-07,
+      "loss": 0.3783,
+      "step": 3030
+    },
+    {
+      "epoch": 2.5230299667036626,
+      "grad_norm": 0.3024890720844269,
+      "learning_rate": 7.486142736039364e-07,
+      "loss": 0.3908,
+      "step": 3031
+    },
+    {
+      "epoch": 2.5238623751387346,
+      "grad_norm": 0.3172726333141327,
+      "learning_rate": 7.460660990333307e-07,
+      "loss": 0.422,
+      "step": 3032
+    },
+    {
+      "epoch": 2.524694783573807,
+      "grad_norm": 0.26689743995666504,
+      "learning_rate": 7.435219189811404e-07,
+      "loss": 0.381,
+      "step": 3033
+    },
+    {
+      "epoch": 2.525527192008879,
+      "grad_norm": 0.29678985476493835,
+      "learning_rate": 7.409817358363986e-07,
+      "loss": 0.3954,
+      "step": 3034
+    },
+    {
+      "epoch": 2.5263596004439512,
+      "grad_norm": 0.2859072983264923,
+      "learning_rate": 7.38445551984378e-07,
+      "loss": 0.3582,
+      "step": 3035
+    },
+    {
+      "epoch": 2.527192008879023,
+      "grad_norm": 0.29791417717933655,
+      "learning_rate": 7.359133698066012e-07,
+      "loss": 0.4127,
+      "step": 3036
+    },
+    {
+      "epoch": 2.5280244173140956,
+      "grad_norm": 0.27158403396606445,
+      "learning_rate": 7.333851916808298e-07,
+      "loss": 0.3757,
+      "step": 3037
+    },
+    {
+      "epoch": 2.5288568257491675,
+      "grad_norm": 0.2889906167984009,
+      "learning_rate": 7.308610199810717e-07,
+      "loss": 0.367,
+      "step": 3038
+    },
+    {
+      "epoch": 2.52968923418424,
+      "grad_norm": 0.31212955713272095,
+      "learning_rate": 7.28340857077564e-07,
+      "loss": 0.4116,
+      "step": 3039
+    },
+    {
+      "epoch": 2.530521642619312,
+      "grad_norm": 0.30906566977500916,
+      "learning_rate": 7.258247053367856e-07,
+      "loss": 0.4001,
+      "step": 3040
+    },
+    {
+      "epoch": 2.531354051054384,
+      "grad_norm": 0.27870288491249084,
+      "learning_rate": 7.233125671214469e-07,
+      "loss": 0.3846,
+      "step": 3041
+    },
+    {
+      "epoch": 2.532186459489456,
+      "grad_norm": 0.2938634157180786,
+      "learning_rate": 7.208044447904893e-07,
+      "loss": 0.4034,
+      "step": 3042
+    },
+    {
+      "epoch": 2.5330188679245285,
+      "grad_norm": 0.2890358865261078,
+      "learning_rate": 7.183003406990841e-07,
+      "loss": 0.3691,
+      "step": 3043
+    },
+    {
+      "epoch": 2.5338512763596004,
+      "grad_norm": 0.28360605239868164,
+      "learning_rate": 7.158002571986283e-07,
+      "loss": 0.3825,
+      "step": 3044
+    },
+    {
+      "epoch": 2.5346836847946728,
+      "grad_norm": 0.277585506439209,
+      "learning_rate": 7.133041966367443e-07,
+      "loss": 0.3718,
+      "step": 3045
+    },
+    {
+      "epoch": 2.5355160932297447,
+      "grad_norm": 0.2901941239833832,
+      "learning_rate": 7.108121613572771e-07,
+      "loss": 0.3666,
+      "step": 3046
+    },
+    {
+      "epoch": 2.5363485016648166,
+      "grad_norm": 0.2988186776638031,
+      "learning_rate": 7.083241537002905e-07,
+      "loss": 0.3939,
+      "step": 3047
+    },
+    {
+      "epoch": 2.537180910099889,
+      "grad_norm": 0.31158897280693054,
+      "learning_rate": 7.058401760020689e-07,
+      "loss": 0.4082,
+      "step": 3048
+    },
+    {
+      "epoch": 2.5380133185349614,
+      "grad_norm": 0.29564598202705383,
+      "learning_rate": 7.033602305951104e-07,
+      "loss": 0.3702,
+      "step": 3049
+    },
+    {
+      "epoch": 2.5388457269700333,
+      "grad_norm": 0.29291555285453796,
+      "learning_rate": 7.008843198081239e-07,
+      "loss": 0.4032,
+      "step": 3050
+    },
+    {
+      "epoch": 2.5396781354051052,
+      "grad_norm": 0.29521968960762024,
+      "learning_rate": 6.984124459660374e-07,
+      "loss": 0.3727,
+      "step": 3051
+    },
+    {
+      "epoch": 2.5405105438401776,
+      "grad_norm": 0.29539427161216736,
+      "learning_rate": 6.95944611389982e-07,
+      "loss": 0.4056,
+      "step": 3052
+    },
+    {
+      "epoch": 2.54134295227525,
+      "grad_norm": 0.3013669550418854,
+      "learning_rate": 6.934808183972986e-07,
+      "loss": 0.377,
+      "step": 3053
+    },
+    {
+      "epoch": 2.542175360710322,
+      "grad_norm": 0.3182709217071533,
+      "learning_rate": 6.910210693015285e-07,
+      "loss": 0.4025,
+      "step": 3054
+    },
+    {
+      "epoch": 2.543007769145394,
+      "grad_norm": 0.3235478401184082,
+      "learning_rate": 6.885653664124226e-07,
+      "loss": 0.4023,
+      "step": 3055
+    },
+    {
+      "epoch": 2.543840177580466,
+      "grad_norm": 0.28089457750320435,
+      "learning_rate": 6.861137120359296e-07,
+      "loss": 0.3754,
+      "step": 3056
+    },
+    {
+      "epoch": 2.544672586015538,
+      "grad_norm": 0.3240266740322113,
+      "learning_rate": 6.836661084741924e-07,
+      "loss": 0.4258,
+      "step": 3057
+    },
+    {
+      "epoch": 2.5455049944506105,
+      "grad_norm": 0.30531975626945496,
+      "learning_rate": 6.812225580255549e-07,
+      "loss": 0.4008,
+      "step": 3058
+    },
+    {
+      "epoch": 2.5463374028856824,
+      "grad_norm": 0.2742519676685333,
+      "learning_rate": 6.787830629845549e-07,
+      "loss": 0.3597,
+      "step": 3059
+    },
+    {
+      "epoch": 2.547169811320755,
+      "grad_norm": 0.3020766079425812,
+      "learning_rate": 6.763476256419215e-07,
+      "loss": 0.3942,
+      "step": 3060
+    },
+    {
+      "epoch": 2.5480022197558267,
+      "grad_norm": 0.314591646194458,
+      "learning_rate": 6.739162482845707e-07,
+      "loss": 0.3769,
+      "step": 3061
+    },
+    {
+      "epoch": 2.548834628190899,
+      "grad_norm": 0.308002233505249,
+      "learning_rate": 6.714889331956087e-07,
+      "loss": 0.3942,
+      "step": 3062
+    },
+    {
+      "epoch": 2.549667036625971,
+      "grad_norm": 0.300611287355423,
+      "learning_rate": 6.690656826543285e-07,
+      "loss": 0.4121,
+      "step": 3063
+    },
+    {
+      "epoch": 2.5504994450610434,
+      "grad_norm": 0.27304670214653015,
+      "learning_rate": 6.666464989362054e-07,
+      "loss": 0.3678,
+      "step": 3064
+    },
+    {
+      "epoch": 2.5513318534961154,
+      "grad_norm": 0.2949070632457733,
+      "learning_rate": 6.642313843128922e-07,
+      "loss": 0.4077,
+      "step": 3065
+    },
+    {
+      "epoch": 2.5521642619311877,
+      "grad_norm": 0.2929401695728302,
+      "learning_rate": 6.618203410522262e-07,
+      "loss": 0.376,
+      "step": 3066
+    },
+    {
+      "epoch": 2.5529966703662597,
+      "grad_norm": 0.30417951941490173,
+      "learning_rate": 6.594133714182178e-07,
+      "loss": 0.4095,
+      "step": 3067
+    },
+    {
+      "epoch": 2.553829078801332,
+      "grad_norm": 0.32251882553100586,
+      "learning_rate": 6.570104776710551e-07,
+      "loss": 0.4429,
+      "step": 3068
+    },
+    {
+      "epoch": 2.554661487236404,
+      "grad_norm": 0.2788321077823639,
+      "learning_rate": 6.546116620670961e-07,
+      "loss": 0.3727,
+      "step": 3069
+    },
+    {
+      "epoch": 2.555493895671476,
+      "grad_norm": 0.3155818581581116,
+      "learning_rate": 6.522169268588713e-07,
+      "loss": 0.3847,
+      "step": 3070
+    },
+    {
+      "epoch": 2.5563263041065483,
+      "grad_norm": 0.3051152527332306,
+      "learning_rate": 6.49826274295079e-07,
+      "loss": 0.3391,
+      "step": 3071
+    },
+    {
+      "epoch": 2.5571587125416206,
+      "grad_norm": 0.2933105230331421,
+      "learning_rate": 6.474397066205834e-07,
+      "loss": 0.3977,
+      "step": 3072
+    },
+    {
+      "epoch": 2.5579911209766926,
+      "grad_norm": 0.2970825135707855,
+      "learning_rate": 6.450572260764137e-07,
+      "loss": 0.4147,
+      "step": 3073
+    },
+    {
+      "epoch": 2.5588235294117645,
+      "grad_norm": 0.31705281138420105,
+      "learning_rate": 6.42678834899761e-07,
+      "loss": 0.4048,
+      "step": 3074
+    },
+    {
+      "epoch": 2.559655937846837,
+      "grad_norm": 0.28795450925827026,
+      "learning_rate": 6.403045353239757e-07,
+      "loss": 0.4011,
+      "step": 3075
+    },
+    {
+      "epoch": 2.5604883462819092,
+      "grad_norm": 0.3047225773334503,
+      "learning_rate": 6.379343295785673e-07,
+      "loss": 0.3888,
+      "step": 3076
+    },
+    {
+      "epoch": 2.561320754716981,
+      "grad_norm": 0.29493415355682373,
+      "learning_rate": 6.355682198892005e-07,
+      "loss": 0.3831,
+      "step": 3077
+    },
+    {
+      "epoch": 2.562153163152053,
+      "grad_norm": 0.29540005326271057,
+      "learning_rate": 6.33206208477693e-07,
+      "loss": 0.4203,
+      "step": 3078
+    },
+    {
+      "epoch": 2.5629855715871255,
+      "grad_norm": 0.266559898853302,
+      "learning_rate": 6.308482975620161e-07,
+      "loss": 0.3354,
+      "step": 3079
+    },
+    {
+      "epoch": 2.5638179800221974,
+      "grad_norm": 0.29652097821235657,
+      "learning_rate": 6.284944893562872e-07,
+      "loss": 0.4215,
+      "step": 3080
+    },
+    {
+      "epoch": 2.56465038845727,
+      "grad_norm": 0.3205231726169586,
+      "learning_rate": 6.261447860707753e-07,
+      "loss": 0.4226,
+      "step": 3081
+    },
+    {
+      "epoch": 2.5654827968923417,
+      "grad_norm": 0.3019580841064453,
+      "learning_rate": 6.23799189911894e-07,
+      "loss": 0.3837,
+      "step": 3082
+    },
+    {
+      "epoch": 2.566315205327414,
+      "grad_norm": 0.26572704315185547,
+      "learning_rate": 6.214577030821967e-07,
+      "loss": 0.3495,
+      "step": 3083
+    },
+    {
+      "epoch": 2.567147613762486,
+      "grad_norm": 0.2902132272720337,
+      "learning_rate": 6.191203277803798e-07,
+      "loss": 0.4022,
+      "step": 3084
+    },
+    {
+      "epoch": 2.5679800221975584,
+      "grad_norm": 0.2867758572101593,
+      "learning_rate": 6.167870662012831e-07,
+      "loss": 0.3833,
+      "step": 3085
+    },
+    {
+      "epoch": 2.5688124306326303,
+      "grad_norm": 0.27622854709625244,
+      "learning_rate": 6.144579205358786e-07,
+      "loss": 0.3983,
+      "step": 3086
+    },
+    {
+      "epoch": 2.5696448390677027,
+      "grad_norm": 0.2973408102989197,
+      "learning_rate": 6.121328929712739e-07,
+      "loss": 0.3968,
+      "step": 3087
+    },
+    {
+      "epoch": 2.5704772475027746,
+      "grad_norm": 0.3048017919063568,
+      "learning_rate": 6.098119856907103e-07,
+      "loss": 0.431,
+      "step": 3088
+    },
+    {
+      "epoch": 2.571309655937847,
+      "grad_norm": 0.29575997591018677,
+      "learning_rate": 6.074952008735624e-07,
+      "loss": 0.3685,
+      "step": 3089
+    },
+    {
+      "epoch": 2.572142064372919,
+      "grad_norm": 0.3209153413772583,
+      "learning_rate": 6.051825406953316e-07,
+      "loss": 0.4429,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5729744728079913,
+      "grad_norm": 0.30033037066459656,
+      "learning_rate": 6.02874007327644e-07,
+      "loss": 0.364,
+      "step": 3091
+    },
+    {
+      "epoch": 2.5738068812430632,
+      "grad_norm": 0.2988393306732178,
+      "learning_rate": 6.005696029382535e-07,
+      "loss": 0.3802,
+      "step": 3092
+    },
+    {
+      "epoch": 2.574639289678135,
+      "grad_norm": 0.3061862289905548,
+      "learning_rate": 5.982693296910386e-07,
+      "loss": 0.3916,
+      "step": 3093
+    },
+    {
+      "epoch": 2.5754716981132075,
+      "grad_norm": 0.31209561228752136,
+      "learning_rate": 5.959731897459936e-07,
+      "loss": 0.4099,
+      "step": 3094
+    },
+    {
+      "epoch": 2.57630410654828,
+      "grad_norm": 0.3136289417743683,
+      "learning_rate": 5.93681185259235e-07,
+      "loss": 0.3519,
+      "step": 3095
+    },
+    {
+      "epoch": 2.577136514983352,
+      "grad_norm": 0.29192236065864563,
+      "learning_rate": 5.91393318382995e-07,
+      "loss": 0.3679,
+      "step": 3096
+    },
+    {
+      "epoch": 2.5779689234184238,
+      "grad_norm": 0.3013075590133667,
+      "learning_rate": 5.891095912656208e-07,
+      "loss": 0.471,
+      "step": 3097
+    },
+    {
+      "epoch": 2.578801331853496,
+      "grad_norm": 0.2735782265663147,
+      "learning_rate": 5.86830006051572e-07,
+      "loss": 0.3486,
+      "step": 3098
+    },
+    {
+      "epoch": 2.5796337402885685,
+      "grad_norm": 0.28365886211395264,
+      "learning_rate": 5.845545648814188e-07,
+      "loss": 0.3943,
+      "step": 3099
+    },
+    {
+      "epoch": 2.5804661487236404,
+      "grad_norm": 0.262623131275177,
+      "learning_rate": 5.822832698918413e-07,
+      "loss": 0.3549,
+      "step": 3100
+    },
+    {
+      "epoch": 2.5812985571587124,
+      "grad_norm": 0.3032148778438568,
+      "learning_rate": 5.800161232156238e-07,
+      "loss": 0.3379,
+      "step": 3101
+    },
+    {
+      "epoch": 2.5821309655937847,
+      "grad_norm": 0.3096363842487335,
+      "learning_rate": 5.777531269816577e-07,
+      "loss": 0.4437,
+      "step": 3102
+    },
+    {
+      "epoch": 2.5829633740288567,
+      "grad_norm": 0.2988356649875641,
+      "learning_rate": 5.754942833149363e-07,
+      "loss": 0.3767,
+      "step": 3103
+    },
+    {
+      "epoch": 2.583795782463929,
+      "grad_norm": 0.29221031069755554,
+      "learning_rate": 5.732395943365526e-07,
+      "loss": 0.3716,
+      "step": 3104
+    },
+    {
+      "epoch": 2.584628190899001,
+      "grad_norm": 0.2928762137889862,
+      "learning_rate": 5.709890621636993e-07,
+      "loss": 0.3776,
+      "step": 3105
+    },
+    {
+      "epoch": 2.5854605993340734,
+      "grad_norm": 0.29212868213653564,
+      "learning_rate": 5.687426889096659e-07,
+      "loss": 0.3987,
+      "step": 3106
+    },
+    {
+      "epoch": 2.5862930077691453,
+      "grad_norm": 0.34992143511772156,
+      "learning_rate": 5.665004766838356e-07,
+      "loss": 0.4072,
+      "step": 3107
+    },
+    {
+      "epoch": 2.5871254162042177,
+      "grad_norm": 0.3027849495410919,
+      "learning_rate": 5.642624275916852e-07,
+      "loss": 0.3612,
+      "step": 3108
+    },
+    {
+      "epoch": 2.5879578246392896,
+      "grad_norm": 0.29607823491096497,
+      "learning_rate": 5.620285437347834e-07,
+      "loss": 0.4058,
+      "step": 3109
+    },
+    {
+      "epoch": 2.588790233074362,
+      "grad_norm": 0.28273841738700867,
+      "learning_rate": 5.597988272107824e-07,
+      "loss": 0.3771,
+      "step": 3110
+    },
+    {
+      "epoch": 2.589622641509434,
+      "grad_norm": 0.293728232383728,
+      "learning_rate": 5.575732801134287e-07,
+      "loss": 0.3977,
+      "step": 3111
+    },
+    {
+      "epoch": 2.5904550499445063,
+      "grad_norm": 0.28861936926841736,
+      "learning_rate": 5.553519045325501e-07,
+      "loss": 0.3839,
+      "step": 3112
+    },
+    {
+      "epoch": 2.591287458379578,
+      "grad_norm": 0.2736756503582001,
+      "learning_rate": 5.531347025540546e-07,
+      "loss": 0.374,
+      "step": 3113
+    },
+    {
+      "epoch": 2.5921198668146506,
+      "grad_norm": 0.2778788208961487,
+      "learning_rate": 5.509216762599339e-07,
+      "loss": 0.3654,
+      "step": 3114
+    },
+    {
+      "epoch": 2.5929522752497225,
+      "grad_norm": 0.29448696970939636,
+      "learning_rate": 5.487128277282605e-07,
+      "loss": 0.3952,
+      "step": 3115
+    },
+    {
+      "epoch": 2.5937846836847944,
+      "grad_norm": 0.31573858857154846,
+      "learning_rate": 5.465081590331817e-07,
+      "loss": 0.4203,
+      "step": 3116
+    },
+    {
+      "epoch": 2.594617092119867,
+      "grad_norm": 0.2849540114402771,
+      "learning_rate": 5.443076722449186e-07,
+      "loss": 0.3638,
+      "step": 3117
+    },
+    {
+      "epoch": 2.595449500554939,
+      "grad_norm": 0.2975523769855499,
+      "learning_rate": 5.421113694297664e-07,
+      "loss": 0.3668,
+      "step": 3118
+    },
+    {
+      "epoch": 2.596281908990011,
+      "grad_norm": 0.3016306459903717,
+      "learning_rate": 5.399192526500946e-07,
+      "loss": 0.4132,
+      "step": 3119
+    },
+    {
+      "epoch": 2.597114317425083,
+      "grad_norm": 0.2861490845680237,
+      "learning_rate": 5.377313239643367e-07,
+      "loss": 0.3821,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5979467258601554,
+      "grad_norm": 0.2761737108230591,
+      "learning_rate": 5.355475854269964e-07,
+      "loss": 0.3899,
+      "step": 3121
+    },
+    {
+      "epoch": 2.5987791342952278,
+      "grad_norm": 0.2985495328903198,
+      "learning_rate": 5.333680390886426e-07,
+      "loss": 0.4112,
+      "step": 3122
+    },
+    {
+      "epoch": 2.5996115427302997,
+      "grad_norm": 0.2759629786014557,
+      "learning_rate": 5.311926869959094e-07,
+      "loss": 0.3573,
+      "step": 3123
+    },
+    {
+      "epoch": 2.6004439511653716,
+      "grad_norm": 0.343628466129303,
+      "learning_rate": 5.290215311914881e-07,
+      "loss": 0.4135,
+      "step": 3124
+    },
+    {
+      "epoch": 2.601276359600444,
+      "grad_norm": 0.2960171103477478,
+      "learning_rate": 5.268545737141323e-07,
+      "loss": 0.3888,
+      "step": 3125
+    },
+    {
+      "epoch": 2.602108768035516,
+      "grad_norm": 0.277442991733551,
+      "learning_rate": 5.246918165986537e-07,
+      "loss": 0.3336,
+      "step": 3126
+    },
+    {
+      "epoch": 2.6029411764705883,
+      "grad_norm": 0.31210342049598694,
+      "learning_rate": 5.225332618759193e-07,
+      "loss": 0.4181,
+      "step": 3127
+    },
+    {
+      "epoch": 2.6037735849056602,
+      "grad_norm": 0.27689164876937866,
+      "learning_rate": 5.203789115728486e-07,
+      "loss": 0.3818,
+      "step": 3128
+    },
+    {
+      "epoch": 2.6046059933407326,
+      "grad_norm": 0.28218722343444824,
+      "learning_rate": 5.182287677124159e-07,
+      "loss": 0.3801,
+      "step": 3129
+    },
+    {
+      "epoch": 2.6054384017758045,
+      "grad_norm": 0.29998162388801575,
+      "learning_rate": 5.160828323136424e-07,
+      "loss": 0.4012,
+      "step": 3130
+    },
+    {
+      "epoch": 2.606270810210877,
+      "grad_norm": 0.2883782684803009,
+      "learning_rate": 5.139411073916001e-07,
+      "loss": 0.3856,
+      "step": 3131
+    },
+    {
+      "epoch": 2.607103218645949,
+      "grad_norm": 0.3021693825721741,
+      "learning_rate": 5.118035949574057e-07,
+      "loss": 0.4013,
+      "step": 3132
+    },
+    {
+      "epoch": 2.6079356270810212,
+      "grad_norm": 0.26575303077697754,
+      "learning_rate": 5.096702970182204e-07,
+      "loss": 0.3247,
+      "step": 3133
+    },
+    {
+      "epoch": 2.608768035516093,
+      "grad_norm": 0.3065766394138336,
+      "learning_rate": 5.075412155772492e-07,
+      "loss": 0.382,
+      "step": 3134
+    },
+    {
+      "epoch": 2.6096004439511655,
+      "grad_norm": 0.30673807859420776,
+      "learning_rate": 5.054163526337364e-07,
+      "loss": 0.4343,
+      "step": 3135
+    },
+    {
+      "epoch": 2.6104328523862375,
+      "grad_norm": 0.3258250951766968,
+      "learning_rate": 5.032957101829661e-07,
+      "loss": 0.3875,
+      "step": 3136
+    },
+    {
+      "epoch": 2.61126526082131,
+      "grad_norm": 0.2834281623363495,
+      "learning_rate": 5.011792902162572e-07,
+      "loss": 0.379,
+      "step": 3137
+    },
+    {
+      "epoch": 2.6120976692563818,
+      "grad_norm": 0.3125614523887634,
+      "learning_rate": 4.990670947209675e-07,
+      "loss": 0.3715,
+      "step": 3138
+    },
+    {
+      "epoch": 2.6129300776914537,
+      "grad_norm": 0.29795554280281067,
+      "learning_rate": 4.969591256804824e-07,
+      "loss": 0.402,
+      "step": 3139
+    },
+    {
+      "epoch": 2.613762486126526,
+      "grad_norm": 0.30222252011299133,
+      "learning_rate": 4.948553850742238e-07,
+      "loss": 0.392,
+      "step": 3140
+    },
+    {
+      "epoch": 2.6145948945615984,
+      "grad_norm": 0.30855488777160645,
+      "learning_rate": 4.927558748776412e-07,
+      "loss": 0.4307,
+      "step": 3141
+    },
+    {
+      "epoch": 2.6154273029966704,
+      "grad_norm": 0.27991849184036255,
+      "learning_rate": 4.906605970622114e-07,
+      "loss": 0.3593,
+      "step": 3142
+    },
+    {
+      "epoch": 2.6162597114317423,
+      "grad_norm": 0.269772469997406,
+      "learning_rate": 4.885695535954361e-07,
+      "loss": 0.3599,
+      "step": 3143
+    },
+    {
+      "epoch": 2.6170921198668147,
+      "grad_norm": 0.2982839345932007,
+      "learning_rate": 4.8648274644084e-07,
+      "loss": 0.4084,
+      "step": 3144
+    },
+    {
+      "epoch": 2.617924528301887,
+      "grad_norm": 0.30452045798301697,
+      "learning_rate": 4.844001775579766e-07,
+      "loss": 0.4204,
+      "step": 3145
+    },
+    {
+      "epoch": 2.618756936736959,
+      "grad_norm": 0.28936830163002014,
+      "learning_rate": 4.8232184890241e-07,
+      "loss": 0.377,
+      "step": 3146
+    },
+    {
+      "epoch": 2.619589345172031,
+      "grad_norm": 0.29566147923469543,
+      "learning_rate": 4.802477624257285e-07,
+      "loss": 0.399,
+      "step": 3147
+    },
+    {
+      "epoch": 2.6204217536071033,
+      "grad_norm": 0.29598045349121094,
+      "learning_rate": 4.781779200755354e-07,
+      "loss": 0.3818,
+      "step": 3148
+    },
+    {
+      "epoch": 2.621254162042175,
+      "grad_norm": 0.2791507840156555,
+      "learning_rate": 4.7611232379545124e-07,
+      "loss": 0.3618,
+      "step": 3149
+    },
+    {
+      "epoch": 2.6220865704772476,
+      "grad_norm": 0.28765299916267395,
+      "learning_rate": 4.740509755251038e-07,
+      "loss": 0.3683,
+      "step": 3150
+    },
+    {
+      "epoch": 2.6229189789123195,
+      "grad_norm": 0.2902831435203552,
+      "learning_rate": 4.71993877200137e-07,
+      "loss": 0.3989,
+      "step": 3151
+    },
+    {
+      "epoch": 2.623751387347392,
+      "grad_norm": 0.29683712124824524,
+      "learning_rate": 4.6994103075220175e-07,
+      "loss": 0.4098,
+      "step": 3152
+    },
+    {
+      "epoch": 2.624583795782464,
+      "grad_norm": 0.28899383544921875,
+      "learning_rate": 4.678924381089567e-07,
+      "loss": 0.3437,
+      "step": 3153
+    },
+    {
+      "epoch": 2.625416204217536,
+      "grad_norm": 0.29837122559547424,
+      "learning_rate": 4.658481011940663e-07,
+      "loss": 0.4139,
+      "step": 3154
+    },
+    {
+      "epoch": 2.626248612652608,
+      "grad_norm": 0.31046080589294434,
+      "learning_rate": 4.63808021927199e-07,
+      "loss": 0.3861,
+      "step": 3155
+    },
+    {
+      "epoch": 2.6270810210876805,
+      "grad_norm": 0.3166535198688507,
+      "learning_rate": 4.617722022240245e-07,
+      "loss": 0.4264,
+      "step": 3156
+    },
+    {
+      "epoch": 2.6279134295227524,
+      "grad_norm": 0.3086240589618683,
+      "learning_rate": 4.597406439962138e-07,
+      "loss": 0.4188,
+      "step": 3157
+    },
+    {
+      "epoch": 2.628745837957825,
+      "grad_norm": 0.29972097277641296,
+      "learning_rate": 4.5771334915143516e-07,
+      "loss": 0.3938,
+      "step": 3158
+    },
+    {
+      "epoch": 2.6295782463928967,
+      "grad_norm": 0.3148014545440674,
+      "learning_rate": 4.5569031959335374e-07,
+      "loss": 0.402,
+      "step": 3159
+    },
+    {
+      "epoch": 2.630410654827969,
+      "grad_norm": 0.31971946358680725,
+      "learning_rate": 4.536715572216299e-07,
+      "loss": 0.3992,
+      "step": 3160
+    },
+    {
+      "epoch": 2.631243063263041,
+      "grad_norm": 0.3092584013938904,
+      "learning_rate": 4.5165706393191676e-07,
+      "loss": 0.4067,
+      "step": 3161
+    },
+    {
+      "epoch": 2.632075471698113,
+      "grad_norm": 0.306518018245697,
+      "learning_rate": 4.496468416158595e-07,
+      "loss": 0.3765,
+      "step": 3162
+    },
+    {
+      "epoch": 2.6329078801331853,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.4764089216109144e-07,
+      "loss": 0.4048,
+      "step": 3163
+    },
+    {
+      "epoch": 2.6337402885682577,
+      "grad_norm": 0.2979956269264221,
+      "learning_rate": 4.456392174512347e-07,
+      "loss": 0.3945,
+      "step": 3164
+    },
+    {
+      "epoch": 2.6345726970033296,
+      "grad_norm": 0.29669734835624695,
+      "learning_rate": 4.4364181936589536e-07,
+      "loss": 0.417,
+      "step": 3165
+    },
+    {
+      "epoch": 2.6354051054384016,
+      "grad_norm": 0.28595003485679626,
+      "learning_rate": 4.4164869978066684e-07,
+      "loss": 0.3818,
+      "step": 3166
+    },
+    {
+      "epoch": 2.636237513873474,
+      "grad_norm": 0.2954268455505371,
+      "learning_rate": 4.3965986056712316e-07,
+      "loss": 0.3785,
+      "step": 3167
+    },
+    {
+      "epoch": 2.6370699223085463,
+      "grad_norm": 0.3013925850391388,
+      "learning_rate": 4.376753035928194e-07,
+      "loss": 0.4055,
+      "step": 3168
+    },
+    {
+      "epoch": 2.6379023307436182,
+      "grad_norm": 0.2962302565574646,
+      "learning_rate": 4.3569503072128703e-07,
+      "loss": 0.4024,
+      "step": 3169
+    },
+    {
+      "epoch": 2.63873473917869,
+      "grad_norm": 0.2931113839149475,
+      "learning_rate": 4.3371904381203976e-07,
+      "loss": 0.3846,
+      "step": 3170
+    },
+    {
+      "epoch": 2.6395671476137625,
+      "grad_norm": 0.290157675743103,
+      "learning_rate": 4.3174734472056334e-07,
+      "loss": 0.3722,
+      "step": 3171
+    },
+    {
+      "epoch": 2.6403995560488345,
+      "grad_norm": 0.3102055490016937,
+      "learning_rate": 4.2977993529831675e-07,
+      "loss": 0.3862,
+      "step": 3172
+    },
+    {
+      "epoch": 2.641231964483907,
+      "grad_norm": 0.30981799960136414,
+      "learning_rate": 4.278168173927322e-07,
+      "loss": 0.3891,
+      "step": 3173
+    },
+    {
+      "epoch": 2.6420643729189788,
+      "grad_norm": 0.28204596042633057,
+      "learning_rate": 4.258579928472106e-07,
+      "loss": 0.36,
+      "step": 3174
+    },
+    {
+      "epoch": 2.642896781354051,
+      "grad_norm": 0.30792534351348877,
+      "learning_rate": 4.2390346350112634e-07,
+      "loss": 0.4047,
+      "step": 3175
+    },
+    {
+      "epoch": 2.643729189789123,
+      "grad_norm": 0.2940234839916229,
+      "learning_rate": 4.219532311898128e-07,
+      "loss": 0.4073,
+      "step": 3176
+    },
+    {
+      "epoch": 2.6445615982241955,
+      "grad_norm": 0.27562248706817627,
+      "learning_rate": 4.200072977445735e-07,
+      "loss": 0.4024,
+      "step": 3177
+    },
+    {
+      "epoch": 2.6453940066592674,
+      "grad_norm": 0.2981874942779541,
+      "learning_rate": 4.180656649926745e-07,
+      "loss": 0.413,
+      "step": 3178
+    },
+    {
+      "epoch": 2.6462264150943398,
+      "grad_norm": 0.31285783648490906,
+      "learning_rate": 4.161283347573425e-07,
+      "loss": 0.4159,
+      "step": 3179
+    },
+    {
+      "epoch": 2.6470588235294117,
+      "grad_norm": 0.2978651523590088,
+      "learning_rate": 4.141953088577644e-07,
+      "loss": 0.3846,
+      "step": 3180
+    },
+    {
+      "epoch": 2.647891231964484,
+      "grad_norm": 0.28265270590782166,
+      "learning_rate": 4.12266589109086e-07,
+      "loss": 0.3768,
+      "step": 3181
+    },
+    {
+      "epoch": 2.648723640399556,
+      "grad_norm": 0.2694186270236969,
+      "learning_rate": 4.103421773224081e-07,
+      "loss": 0.3781,
+      "step": 3182
+    },
+    {
+      "epoch": 2.6495560488346284,
+      "grad_norm": 0.2963635325431824,
+      "learning_rate": 4.0842207530478793e-07,
+      "loss": 0.4111,
+      "step": 3183
+    },
+    {
+      "epoch": 2.6503884572697003,
+      "grad_norm": 0.29516974091529846,
+      "learning_rate": 4.0650628485923385e-07,
+      "loss": 0.3699,
+      "step": 3184
+    },
+    {
+      "epoch": 2.6512208657047722,
+      "grad_norm": 0.3196733295917511,
+      "learning_rate": 4.0459480778470786e-07,
+      "loss": 0.4343,
+      "step": 3185
+    },
+    {
+      "epoch": 2.6520532741398446,
+      "grad_norm": 0.28815674781799316,
+      "learning_rate": 4.026876458761192e-07,
+      "loss": 0.3925,
+      "step": 3186
+    },
+    {
+      "epoch": 2.652885682574917,
+      "grad_norm": 0.30133241415023804,
+      "learning_rate": 4.0078480092432705e-07,
+      "loss": 0.3972,
+      "step": 3187
+    },
+    {
+      "epoch": 2.653718091009989,
+      "grad_norm": 0.3248240351676941,
+      "learning_rate": 3.9888627471613595e-07,
+      "loss": 0.405,
+      "step": 3188
+    },
+    {
+      "epoch": 2.654550499445061,
+      "grad_norm": 0.3046268820762634,
+      "learning_rate": 3.969920690342954e-07,
+      "loss": 0.3948,
+      "step": 3189
+    },
+    {
+      "epoch": 2.655382907880133,
+      "grad_norm": 0.29888278245925903,
+      "learning_rate": 3.9510218565749823e-07,
+      "loss": 0.3884,
+      "step": 3190
+    },
+    {
+      "epoch": 2.6562153163152056,
+      "grad_norm": 0.2877199351787567,
+      "learning_rate": 3.9321662636037537e-07,
+      "loss": 0.3705,
+      "step": 3191
+    },
+    {
+      "epoch": 2.6570477247502775,
+      "grad_norm": 0.2959778606891632,
+      "learning_rate": 3.913353929135033e-07,
+      "loss": 0.3945,
+      "step": 3192
+    },
+    {
+      "epoch": 2.6578801331853494,
+      "grad_norm": 0.27556687593460083,
+      "learning_rate": 3.8945848708339173e-07,
+      "loss": 0.369,
+      "step": 3193
+    },
+    {
+      "epoch": 2.658712541620422,
+      "grad_norm": 0.27974632382392883,
+      "learning_rate": 3.8758591063248864e-07,
+      "loss": 0.3902,
+      "step": 3194
+    },
+    {
+      "epoch": 2.6595449500554937,
+      "grad_norm": 0.2980170249938965,
+      "learning_rate": 3.8571766531917466e-07,
+      "loss": 0.4005,
+      "step": 3195
+    },
+    {
+      "epoch": 2.660377358490566,
+      "grad_norm": 0.26741084456443787,
+      "learning_rate": 3.838537528977659e-07,
+      "loss": 0.3446,
+      "step": 3196
+    },
+    {
+      "epoch": 2.661209766925638,
+      "grad_norm": 0.2738521993160248,
+      "learning_rate": 3.8199417511851023e-07,
+      "loss": 0.3572,
+      "step": 3197
+    },
+    {
+      "epoch": 2.6620421753607104,
+      "grad_norm": 0.30660688877105713,
+      "learning_rate": 3.8013893372758125e-07,
+      "loss": 0.398,
+      "step": 3198
+    },
+    {
+      "epoch": 2.6628745837957823,
+      "grad_norm": 0.31190434098243713,
+      "learning_rate": 3.782880304670833e-07,
+      "loss": 0.3662,
+      "step": 3199
+    },
+    {
+      "epoch": 2.6637069922308547,
+      "grad_norm": 0.27871236205101013,
+      "learning_rate": 3.7644146707504826e-07,
+      "loss": 0.3878,
+      "step": 3200
+    },
+    {
+      "epoch": 2.6645394006659266,
+      "grad_norm": 0.28115060925483704,
+      "learning_rate": 3.7459924528543247e-07,
+      "loss": 0.3932,
+      "step": 3201
+    },
+    {
+      "epoch": 2.665371809100999,
+      "grad_norm": 0.29334890842437744,
+      "learning_rate": 3.727613668281116e-07,
+      "loss": 0.4139,
+      "step": 3202
+    },
+    {
+      "epoch": 2.666204217536071,
+      "grad_norm": 0.30630093812942505,
+      "learning_rate": 3.709278334288874e-07,
+      "loss": 0.4072,
+      "step": 3203
+    },
+    {
+      "epoch": 2.6670366259711433,
+      "grad_norm": 0.2759256958961487,
+      "learning_rate": 3.6909864680947815e-07,
+      "loss": 0.396,
+      "step": 3204
+    },
+    {
+      "epoch": 2.6678690344062153,
+      "grad_norm": 0.289121150970459,
+      "learning_rate": 3.672738086875255e-07,
+      "loss": 0.393,
+      "step": 3205
+    },
+    {
+      "epoch": 2.6687014428412876,
+      "grad_norm": 0.3110848069190979,
+      "learning_rate": 3.6545332077658146e-07,
+      "loss": 0.4241,
+      "step": 3206
+    },
+    {
+      "epoch": 2.6695338512763596,
+      "grad_norm": 0.2876022160053253,
+      "learning_rate": 3.63637184786117e-07,
+      "loss": 0.3783,
+      "step": 3207
+    },
+    {
+      "epoch": 2.6703662597114315,
+      "grad_norm": 0.28069189190864563,
+      "learning_rate": 3.618254024215156e-07,
+      "loss": 0.4027,
+      "step": 3208
+    },
+    {
+      "epoch": 2.671198668146504,
+      "grad_norm": 0.2795030474662781,
+      "learning_rate": 3.6001797538407214e-07,
+      "loss": 0.3889,
+      "step": 3209
+    },
+    {
+      "epoch": 2.6720310765815762,
+      "grad_norm": 0.28693515062332153,
+      "learning_rate": 3.582149053709932e-07,
+      "loss": 0.3846,
+      "step": 3210
+    },
+    {
+      "epoch": 2.672863485016648,
+      "grad_norm": 0.30839934945106506,
+      "learning_rate": 3.564161940753924e-07,
+      "loss": 0.4299,
+      "step": 3211
+    },
+    {
+      "epoch": 2.67369589345172,
+      "grad_norm": 0.2880555987358093,
+      "learning_rate": 3.5462184318629134e-07,
+      "loss": 0.3687,
+      "step": 3212
+    },
+    {
+      "epoch": 2.6745283018867925,
+      "grad_norm": 0.3026648759841919,
+      "learning_rate": 3.528318543886172e-07,
+      "loss": 0.3653,
+      "step": 3213
+    },
+    {
+      "epoch": 2.675360710321865,
+      "grad_norm": 0.2983366847038269,
+      "learning_rate": 3.510462293632e-07,
+      "loss": 0.3878,
+      "step": 3214
+    },
+    {
+      "epoch": 2.6761931187569368,
+      "grad_norm": 0.2791326344013214,
+      "learning_rate": 3.4926496978677393e-07,
+      "loss": 0.3921,
+      "step": 3215
+    },
+    {
+      "epoch": 2.6770255271920087,
+      "grad_norm": 0.29134318232536316,
+      "learning_rate": 3.4748807733197223e-07,
+      "loss": 0.3876,
+      "step": 3216
+    },
+    {
+      "epoch": 2.677857935627081,
+      "grad_norm": 0.3067108392715454,
+      "learning_rate": 3.457155536673279e-07,
+      "loss": 0.3933,
+      "step": 3217
+    },
+    {
+      "epoch": 2.678690344062153,
+      "grad_norm": 0.31518760323524475,
+      "learning_rate": 3.439474004572724e-07,
+      "loss": 0.4012,
+      "step": 3218
+    },
+    {
+      "epoch": 2.6795227524972254,
+      "grad_norm": 0.27954304218292236,
+      "learning_rate": 3.4218361936213195e-07,
+      "loss": 0.3536,
+      "step": 3219
+    },
+    {
+      "epoch": 2.6803551609322973,
+      "grad_norm": 0.30983665585517883,
+      "learning_rate": 3.4042421203812904e-07,
+      "loss": 0.4255,
+      "step": 3220
+    },
+    {
+      "epoch": 2.6811875693673697,
+      "grad_norm": 0.28379014134407043,
+      "learning_rate": 3.386691801373754e-07,
+      "loss": 0.3885,
+      "step": 3221
+    },
+    {
+      "epoch": 2.6820199778024416,
+      "grad_norm": 0.2883853316307068,
+      "learning_rate": 3.369185253078794e-07,
+      "loss": 0.3971,
+      "step": 3222
+    },
+    {
+      "epoch": 2.682852386237514,
+      "grad_norm": 0.28160157799720764,
+      "learning_rate": 3.3517224919353555e-07,
+      "loss": 0.3796,
+      "step": 3223
+    },
+    {
+      "epoch": 2.683684794672586,
+      "grad_norm": 0.2900620102882385,
+      "learning_rate": 3.334303534341277e-07,
+      "loss": 0.4238,
+      "step": 3224
+    },
+    {
+      "epoch": 2.6845172031076583,
+      "grad_norm": 0.2966312766075134,
+      "learning_rate": 3.3169283966532517e-07,
+      "loss": 0.3779,
+      "step": 3225
+    },
+    {
+      "epoch": 2.68534961154273,
+      "grad_norm": 0.27945834398269653,
+      "learning_rate": 3.2995970951868574e-07,
+      "loss": 0.3731,
+      "step": 3226
+    },
+    {
+      "epoch": 2.6861820199778026,
+      "grad_norm": 0.31320464611053467,
+      "learning_rate": 3.2823096462164915e-07,
+      "loss": 0.3988,
+      "step": 3227
+    },
+    {
+      "epoch": 2.6870144284128745,
+      "grad_norm": 0.28876248002052307,
+      "learning_rate": 3.265066065975353e-07,
+      "loss": 0.3668,
+      "step": 3228
+    },
+    {
+      "epoch": 2.687846836847947,
+      "grad_norm": 0.2871847152709961,
+      "learning_rate": 3.2478663706554724e-07,
+      "loss": 0.4239,
+      "step": 3229
+    },
+    {
+      "epoch": 2.688679245283019,
+      "grad_norm": 0.31527018547058105,
+      "learning_rate": 3.2307105764076694e-07,
+      "loss": 0.3833,
+      "step": 3230
+    },
+    {
+      "epoch": 2.6895116537180908,
+      "grad_norm": 0.31052833795547485,
+      "learning_rate": 3.213598699341547e-07,
+      "loss": 0.3889,
+      "step": 3231
+    },
+    {
+      "epoch": 2.690344062153163,
+      "grad_norm": 0.28673428297042847,
+      "learning_rate": 3.1965307555254343e-07,
+      "loss": 0.4006,
+      "step": 3232
+    },
+    {
+      "epoch": 2.6911764705882355,
+      "grad_norm": 0.28148752450942993,
+      "learning_rate": 3.1795067609864395e-07,
+      "loss": 0.3643,
+      "step": 3233
+    },
+    {
+      "epoch": 2.6920088790233074,
+      "grad_norm": 0.28988078236579895,
+      "learning_rate": 3.162526731710386e-07,
+      "loss": 0.4032,
+      "step": 3234
+    },
+    {
+      "epoch": 2.6928412874583794,
+      "grad_norm": 0.2785470485687256,
+      "learning_rate": 3.14559068364183e-07,
+      "loss": 0.3915,
+      "step": 3235
+    },
+    {
+      "epoch": 2.6936736958934517,
+      "grad_norm": 0.27616074681282043,
+      "learning_rate": 3.1286986326840076e-07,
+      "loss": 0.3812,
+      "step": 3236
+    },
+    {
+      "epoch": 2.694506104328524,
+      "grad_norm": 0.28551867604255676,
+      "learning_rate": 3.1118505946988506e-07,
+      "loss": 0.3644,
+      "step": 3237
+    },
+    {
+      "epoch": 2.695338512763596,
+      "grad_norm": 0.31371888518333435,
+      "learning_rate": 3.095046585506967e-07,
+      "loss": 0.4134,
+      "step": 3238
+    },
+    {
+      "epoch": 2.696170921198668,
+      "grad_norm": 0.299771785736084,
+      "learning_rate": 3.0782866208876163e-07,
+      "loss": 0.3827,
+      "step": 3239
+    },
+    {
+      "epoch": 2.6970033296337403,
+      "grad_norm": 0.29559215903282166,
+      "learning_rate": 3.0615707165786937e-07,
+      "loss": 0.385,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6978357380688123,
+      "grad_norm": 0.27681657671928406,
+      "learning_rate": 3.044898888276726e-07,
+      "loss": 0.3581,
+      "step": 3241
+    },
+    {
+      "epoch": 2.6986681465038846,
+      "grad_norm": 0.29248616099357605,
+      "learning_rate": 3.0282711516368524e-07,
+      "loss": 0.3735,
+      "step": 3242
+    },
+    {
+      "epoch": 2.6995005549389566,
+      "grad_norm": 0.3053610026836395,
+      "learning_rate": 3.011687522272816e-07,
+      "loss": 0.4158,
+      "step": 3243
+    },
+    {
+      "epoch": 2.700332963374029,
+      "grad_norm": 0.31310683488845825,
+      "learning_rate": 2.995148015756927e-07,
+      "loss": 0.3995,
+      "step": 3244
+    },
+    {
+      "epoch": 2.701165371809101,
+      "grad_norm": 0.30295878648757935,
+      "learning_rate": 2.978652647620073e-07,
+      "loss": 0.4164,
+      "step": 3245
+    },
+    {
+      "epoch": 2.7019977802441733,
+      "grad_norm": 0.2850441336631775,
+      "learning_rate": 2.962201433351697e-07,
+      "loss": 0.3935,
+      "step": 3246
+    },
+    {
+      "epoch": 2.702830188679245,
+      "grad_norm": 0.2884097397327423,
+      "learning_rate": 2.9457943883997696e-07,
+      "loss": 0.3741,
+      "step": 3247
+    },
+    {
+      "epoch": 2.7036625971143176,
+      "grad_norm": 0.2926424741744995,
+      "learning_rate": 2.929431528170801e-07,
+      "loss": 0.3759,
+      "step": 3248
+    },
+    {
+      "epoch": 2.7044950055493895,
+      "grad_norm": 0.29279258847236633,
+      "learning_rate": 2.91311286802981e-07,
+      "loss": 0.3754,
+      "step": 3249
+    },
+    {
+      "epoch": 2.705327413984462,
+      "grad_norm": 0.2717744708061218,
+      "learning_rate": 2.8968384233002855e-07,
+      "loss": 0.3732,
+      "step": 3250
+    },
+    {
+      "epoch": 2.706159822419534,
+      "grad_norm": 0.29794079065322876,
+      "learning_rate": 2.8806082092642186e-07,
+      "loss": 0.3902,
+      "step": 3251
+    },
+    {
+      "epoch": 2.706992230854606,
+      "grad_norm": 0.31436416506767273,
+      "learning_rate": 2.8644222411620793e-07,
+      "loss": 0.4225,
+      "step": 3252
+    },
+    {
+      "epoch": 2.707824639289678,
+      "grad_norm": 0.2860269546508789,
+      "learning_rate": 2.848280534192777e-07,
+      "loss": 0.3869,
+      "step": 3253
+    },
+    {
+      "epoch": 2.70865704772475,
+      "grad_norm": 0.2989305257797241,
+      "learning_rate": 2.832183103513636e-07,
+      "loss": 0.3976,
+      "step": 3254
+    },
+    {
+      "epoch": 2.7094894561598224,
+      "grad_norm": 0.30432286858558655,
+      "learning_rate": 2.816129964240433e-07,
+      "loss": 0.4112,
+      "step": 3255
+    },
+    {
+      "epoch": 2.7103218645948948,
+      "grad_norm": 0.3015022575855255,
+      "learning_rate": 2.800121131447353e-07,
+      "loss": 0.404,
+      "step": 3256
+    },
+    {
+      "epoch": 2.7111542730299667,
+      "grad_norm": 0.2788195013999939,
+      "learning_rate": 2.784156620166983e-07,
+      "loss": 0.3442,
+      "step": 3257
+    },
+    {
+      "epoch": 2.7119866814650386,
+      "grad_norm": 0.32228487730026245,
+      "learning_rate": 2.7682364453902487e-07,
+      "loss": 0.4665,
+      "step": 3258
+    },
+    {
+      "epoch": 2.712819089900111,
+      "grad_norm": 0.2844350039958954,
+      "learning_rate": 2.7523606220664854e-07,
+      "loss": 0.3602,
+      "step": 3259
+    },
+    {
+      "epoch": 2.7136514983351834,
+      "grad_norm": 0.270535409450531,
+      "learning_rate": 2.736529165103385e-07,
+      "loss": 0.3578,
+      "step": 3260
+    },
+    {
+      "epoch": 2.7144839067702553,
+      "grad_norm": 0.31310030817985535,
+      "learning_rate": 2.7207420893669455e-07,
+      "loss": 0.3897,
+      "step": 3261
+    },
+    {
+      "epoch": 2.7153163152053272,
+      "grad_norm": 0.3238847851753235,
+      "learning_rate": 2.704999409681508e-07,
+      "loss": 0.4193,
+      "step": 3262
+    },
+    {
+      "epoch": 2.7161487236403996,
+      "grad_norm": 0.2968219220638275,
+      "learning_rate": 2.6893011408297196e-07,
+      "loss": 0.3856,
+      "step": 3263
+    },
+    {
+      "epoch": 2.7169811320754715,
+      "grad_norm": 0.2701689600944519,
+      "learning_rate": 2.6736472975525564e-07,
+      "loss": 0.3865,
+      "step": 3264
+    },
+    {
+      "epoch": 2.717813540510544,
+      "grad_norm": 0.30333271622657776,
+      "learning_rate": 2.65803789454922e-07,
+      "loss": 0.4149,
+      "step": 3265
+    },
+    {
+      "epoch": 2.718645948945616,
+      "grad_norm": 0.2998022735118866,
+      "learning_rate": 2.6424729464772316e-07,
+      "loss": 0.3999,
+      "step": 3266
+    },
+    {
+      "epoch": 2.719478357380688,
+      "grad_norm": 0.2897208034992218,
+      "learning_rate": 2.626952467952343e-07,
+      "loss": 0.388,
+      "step": 3267
+    },
+    {
+      "epoch": 2.72031076581576,
+      "grad_norm": 0.27508842945098877,
+      "learning_rate": 2.611476473548552e-07,
+      "loss": 0.3711,
+      "step": 3268
+    },
+    {
+      "epoch": 2.7211431742508325,
+      "grad_norm": 0.30373603105545044,
+      "learning_rate": 2.596044977798101e-07,
+      "loss": 0.4133,
+      "step": 3269
+    },
+    {
+      "epoch": 2.7219755826859044,
+      "grad_norm": 0.30435818433761597,
+      "learning_rate": 2.5806579951914214e-07,
+      "loss": 0.411,
+      "step": 3270
+    },
+    {
+      "epoch": 2.722807991120977,
+      "grad_norm": 0.28073209524154663,
+      "learning_rate": 2.5653155401771655e-07,
+      "loss": 0.3691,
+      "step": 3271
+    },
+    {
+      "epoch": 2.7236403995560488,
+      "grad_norm": 0.330555260181427,
+      "learning_rate": 2.550017627162166e-07,
+      "loss": 0.4357,
+      "step": 3272
+    },
+    {
+      "epoch": 2.724472807991121,
+      "grad_norm": 0.25710800290107727,
+      "learning_rate": 2.534764270511431e-07,
+      "loss": 0.3366,
+      "step": 3273
+    },
+    {
+      "epoch": 2.725305216426193,
+      "grad_norm": 0.29701754450798035,
+      "learning_rate": 2.5195554845481306e-07,
+      "loss": 0.4171,
+      "step": 3274
+    },
+    {
+      "epoch": 2.7261376248612654,
+      "grad_norm": 0.2765384316444397,
+      "learning_rate": 2.5043912835535867e-07,
+      "loss": 0.3471,
+      "step": 3275
+    },
+    {
+      "epoch": 2.7269700332963374,
+      "grad_norm": 0.29196688532829285,
+      "learning_rate": 2.4892716817672304e-07,
+      "loss": 0.4051,
+      "step": 3276
+    },
+    {
+      "epoch": 2.7278024417314093,
+      "grad_norm": 0.2899174690246582,
+      "learning_rate": 2.474196693386649e-07,
+      "loss": 0.3756,
+      "step": 3277
+    },
+    {
+      "epoch": 2.7286348501664817,
+      "grad_norm": 0.2992348074913025,
+      "learning_rate": 2.45916633256752e-07,
+      "loss": 0.4266,
+      "step": 3278
+    },
+    {
+      "epoch": 2.729467258601554,
+      "grad_norm": 0.29235145449638367,
+      "learning_rate": 2.4441806134236137e-07,
+      "loss": 0.4102,
+      "step": 3279
+    },
+    {
+      "epoch": 2.730299667036626,
+      "grad_norm": 0.26745110750198364,
+      "learning_rate": 2.4292395500267796e-07,
+      "loss": 0.3314,
+      "step": 3280
+    },
+    {
+      "epoch": 2.731132075471698,
+      "grad_norm": 0.3117216229438782,
+      "learning_rate": 2.4143431564069344e-07,
+      "loss": 0.4081,
+      "step": 3281
+    },
+    {
+      "epoch": 2.7319644839067703,
+      "grad_norm": 0.2923487424850464,
+      "learning_rate": 2.39949144655206e-07,
+      "loss": 0.3971,
+      "step": 3282
+    },
+    {
+      "epoch": 2.7327968923418426,
+      "grad_norm": 0.2777581810951233,
+      "learning_rate": 2.38468443440818e-07,
+      "loss": 0.3751,
+      "step": 3283
+    },
+    {
+      "epoch": 2.7336293007769146,
+      "grad_norm": 0.30919012427330017,
+      "learning_rate": 2.3699221338793155e-07,
+      "loss": 0.4077,
+      "step": 3284
+    },
+    {
+      "epoch": 2.7344617092119865,
+      "grad_norm": 0.2907434105873108,
+      "learning_rate": 2.355204558827534e-07,
+      "loss": 0.3853,
+      "step": 3285
+    },
+    {
+      "epoch": 2.735294117647059,
+      "grad_norm": 0.2914058566093445,
+      "learning_rate": 2.340531723072914e-07,
+      "loss": 0.3711,
+      "step": 3286
+    },
+    {
+      "epoch": 2.736126526082131,
+      "grad_norm": 0.2743302583694458,
+      "learning_rate": 2.3259036403934843e-07,
+      "loss": 0.4138,
+      "step": 3287
+    },
+    {
+      "epoch": 2.736958934517203,
+      "grad_norm": 0.30281293392181396,
+      "learning_rate": 2.3113203245252734e-07,
+      "loss": 0.3994,
+      "step": 3288
+    },
+    {
+      "epoch": 2.737791342952275,
+      "grad_norm": 0.303541898727417,
+      "learning_rate": 2.2967817891622724e-07,
+      "loss": 0.3588,
+      "step": 3289
+    },
+    {
+      "epoch": 2.7386237513873475,
+      "grad_norm": 0.31531253457069397,
+      "learning_rate": 2.2822880479564325e-07,
+      "loss": 0.3762,
+      "step": 3290
+    },
+    {
+      "epoch": 2.7394561598224194,
+      "grad_norm": 0.29251205921173096,
+      "learning_rate": 2.2678391145176115e-07,
+      "loss": 0.3847,
+      "step": 3291
+    },
+    {
+      "epoch": 2.740288568257492,
+      "grad_norm": 0.2942062318325043,
+      "learning_rate": 2.2534350024136232e-07,
+      "loss": 0.3904,
+      "step": 3292
+    },
+    {
+      "epoch": 2.7411209766925637,
+      "grad_norm": 0.30196094512939453,
+      "learning_rate": 2.2390757251701756e-07,
+      "loss": 0.3586,
+      "step": 3293
+    },
+    {
+      "epoch": 2.741953385127636,
+      "grad_norm": 0.3154551386833191,
+      "learning_rate": 2.224761296270883e-07,
+      "loss": 0.4193,
+      "step": 3294
+    },
+    {
+      "epoch": 2.742785793562708,
+      "grad_norm": 0.2840886116027832,
+      "learning_rate": 2.2104917291572435e-07,
+      "loss": 0.3629,
+      "step": 3295
+    },
+    {
+      "epoch": 2.7436182019977804,
+      "grad_norm": 0.28531089425086975,
+      "learning_rate": 2.196267037228633e-07,
+      "loss": 0.389,
+      "step": 3296
+    },
+    {
+      "epoch": 2.7444506104328523,
+      "grad_norm": 0.2992643713951111,
+      "learning_rate": 2.1820872338422838e-07,
+      "loss": 0.3948,
+      "step": 3297
+    },
+    {
+      "epoch": 2.7452830188679247,
+      "grad_norm": 0.296223908662796,
+      "learning_rate": 2.1679523323132835e-07,
+      "loss": 0.3902,
+      "step": 3298
+    },
+    {
+      "epoch": 2.7461154273029966,
+      "grad_norm": 0.270108699798584,
+      "learning_rate": 2.153862345914548e-07,
+      "loss": 0.3622,
+      "step": 3299
+    },
+    {
+      "epoch": 2.7469478357380686,
+      "grad_norm": 0.2876308262348175,
+      "learning_rate": 2.139817287876822e-07,
+      "loss": 0.3826,
+      "step": 3300
+    },
+    {
+      "epoch": 2.747780244173141,
+      "grad_norm": 0.28309187293052673,
+      "learning_rate": 2.125817171388672e-07,
+      "loss": 0.3575,
+      "step": 3301
+    },
+    {
+      "epoch": 2.7486126526082133,
+      "grad_norm": 0.31576332449913025,
+      "learning_rate": 2.111862009596427e-07,
+      "loss": 0.425,
+      "step": 3302
+    },
+    {
+      "epoch": 2.7494450610432852,
+      "grad_norm": 0.3037078082561493,
+      "learning_rate": 2.097951815604249e-07,
+      "loss": 0.3768,
+      "step": 3303
+    },
+    {
+      "epoch": 2.750277469478357,
+      "grad_norm": 0.3079027831554413,
+      "learning_rate": 2.0840866024740502e-07,
+      "loss": 0.3602,
+      "step": 3304
+    },
+    {
+      "epoch": 2.7511098779134295,
+      "grad_norm": 0.3146269917488098,
+      "learning_rate": 2.070266383225511e-07,
+      "loss": 0.4562,
+      "step": 3305
+    },
+    {
+      "epoch": 2.751942286348502,
+      "grad_norm": 0.2855137288570404,
+      "learning_rate": 2.0564911708360447e-07,
+      "loss": 0.3737,
+      "step": 3306
+    },
+    {
+      "epoch": 2.752774694783574,
+      "grad_norm": 0.30360037088394165,
+      "learning_rate": 2.0427609782408265e-07,
+      "loss": 0.4307,
+      "step": 3307
+    },
+    {
+      "epoch": 2.7536071032186458,
+      "grad_norm": 0.25891706347465515,
+      "learning_rate": 2.029075818332754e-07,
+      "loss": 0.3776,
+      "step": 3308
+    },
+    {
+      "epoch": 2.754439511653718,
+      "grad_norm": 0.2823725938796997,
+      "learning_rate": 2.0154357039624317e-07,
+      "loss": 0.3631,
+      "step": 3309
+    },
+    {
+      "epoch": 2.75527192008879,
+      "grad_norm": 0.30981922149658203,
+      "learning_rate": 2.0018406479381525e-07,
+      "loss": 0.4291,
+      "step": 3310
+    },
+    {
+      "epoch": 2.7561043285238624,
+      "grad_norm": 0.29691869020462036,
+      "learning_rate": 1.9882906630259158e-07,
+      "loss": 0.3768,
+      "step": 3311
+    },
+    {
+      "epoch": 2.7569367369589344,
+      "grad_norm": 0.28396308422088623,
+      "learning_rate": 1.9747857619494105e-07,
+      "loss": 0.3652,
+      "step": 3312
+    },
+    {
+      "epoch": 2.7577691453940067,
+      "grad_norm": 0.2920913100242615,
+      "learning_rate": 1.961325957389959e-07,
+      "loss": 0.3836,
+      "step": 3313
+    },
+    {
+      "epoch": 2.7586015538290787,
+      "grad_norm": 0.2953694760799408,
+      "learning_rate": 1.9479112619865513e-07,
+      "loss": 0.3953,
+      "step": 3314
+    },
+    {
+      "epoch": 2.759433962264151,
+      "grad_norm": 0.3114762604236603,
+      "learning_rate": 1.934541688335828e-07,
+      "loss": 0.4324,
+      "step": 3315
+    },
+    {
+      "epoch": 2.760266370699223,
+      "grad_norm": 0.28087711334228516,
+      "learning_rate": 1.9212172489920632e-07,
+      "loss": 0.3553,
+      "step": 3316
+    },
+    {
+      "epoch": 2.7610987791342954,
+      "grad_norm": 0.2810649871826172,
+      "learning_rate": 1.9079379564671207e-07,
+      "loss": 0.3584,
+      "step": 3317
+    },
+    {
+      "epoch": 2.7619311875693673,
+      "grad_norm": 0.3019134998321533,
+      "learning_rate": 1.8947038232304981e-07,
+      "loss": 0.4394,
+      "step": 3318
+    },
+    {
+      "epoch": 2.7627635960044397,
+      "grad_norm": 0.2816515266895294,
+      "learning_rate": 1.8815148617092772e-07,
+      "loss": 0.3543,
+      "step": 3319
+    },
+    {
+      "epoch": 2.7635960044395116,
+      "grad_norm": 0.30382615327835083,
+      "learning_rate": 1.8683710842881174e-07,
+      "loss": 0.4034,
+      "step": 3320
+    },
+    {
+      "epoch": 2.764428412874584,
+      "grad_norm": 0.29469168186187744,
+      "learning_rate": 1.8552725033092635e-07,
+      "loss": 0.3681,
+      "step": 3321
+    },
+    {
+      "epoch": 2.765260821309656,
+      "grad_norm": 0.3065103590488434,
+      "learning_rate": 1.8422191310725147e-07,
+      "loss": 0.4048,
+      "step": 3322
+    },
+    {
+      "epoch": 2.766093229744728,
+      "grad_norm": 0.29562950134277344,
+      "learning_rate": 1.8292109798352054e-07,
+      "loss": 0.3898,
+      "step": 3323
+    },
+    {
+      "epoch": 2.7669256381798,
+      "grad_norm": 0.28922104835510254,
+      "learning_rate": 1.816248061812226e-07,
+      "loss": 0.3753,
+      "step": 3324
+    },
+    {
+      "epoch": 2.7677580466148726,
+      "grad_norm": 0.2866446375846863,
+      "learning_rate": 1.8033303891759835e-07,
+      "loss": 0.3665,
+      "step": 3325
+    },
+    {
+      "epoch": 2.7685904550499445,
+      "grad_norm": 0.3014658987522125,
+      "learning_rate": 1.7904579740563921e-07,
+      "loss": 0.3785,
+      "step": 3326
+    },
+    {
+      "epoch": 2.7694228634850164,
+      "grad_norm": 0.28989580273628235,
+      "learning_rate": 1.7776308285408826e-07,
+      "loss": 0.4271,
+      "step": 3327
+    },
+    {
+      "epoch": 2.770255271920089,
+      "grad_norm": 0.27226322889328003,
+      "learning_rate": 1.7648489646743648e-07,
+      "loss": 0.3668,
+      "step": 3328
+    },
+    {
+      "epoch": 2.771087680355161,
+      "grad_norm": 0.286642462015152,
+      "learning_rate": 1.752112394459232e-07,
+      "loss": 0.4169,
+      "step": 3329
+    },
+    {
+      "epoch": 2.771920088790233,
+      "grad_norm": 0.2984205186367035,
+      "learning_rate": 1.7394211298553508e-07,
+      "loss": 0.3907,
+      "step": 3330
+    },
+    {
+      "epoch": 2.772752497225305,
+      "grad_norm": 0.26750481128692627,
+      "learning_rate": 1.726775182780044e-07,
+      "loss": 0.3586,
+      "step": 3331
+    },
+    {
+      "epoch": 2.7735849056603774,
+      "grad_norm": 0.2971529960632324,
+      "learning_rate": 1.7141745651080565e-07,
+      "loss": 0.4357,
+      "step": 3332
+    },
+    {
+      "epoch": 2.7744173140954493,
+      "grad_norm": 0.2746163308620453,
+      "learning_rate": 1.7016192886716132e-07,
+      "loss": 0.3554,
+      "step": 3333
+    },
+    {
+      "epoch": 2.7752497225305217,
+      "grad_norm": 0.2701033651828766,
+      "learning_rate": 1.689109365260333e-07,
+      "loss": 0.392,
+      "step": 3334
+    },
+    {
+      "epoch": 2.7760821309655936,
+      "grad_norm": 0.3108593225479126,
+      "learning_rate": 1.676644806621247e-07,
+      "loss": 0.38,
+      "step": 3335
+    },
+    {
+      "epoch": 2.776914539400666,
+      "grad_norm": 0.28853678703308105,
+      "learning_rate": 1.664225624458793e-07,
+      "loss": 0.381,
+      "step": 3336
+    },
+    {
+      "epoch": 2.777746947835738,
+      "grad_norm": 0.30430543422698975,
+      "learning_rate": 1.651851830434803e-07,
+      "loss": 0.4189,
+      "step": 3337
+    },
+    {
+      "epoch": 2.7785793562708103,
+      "grad_norm": 0.2828444838523865,
+      "learning_rate": 1.6395234361684943e-07,
+      "loss": 0.3827,
+      "step": 3338
+    },
+    {
+      "epoch": 2.7794117647058822,
+      "grad_norm": 0.2710029184818268,
+      "learning_rate": 1.6272404532364337e-07,
+      "loss": 0.4074,
+      "step": 3339
+    },
+    {
+      "epoch": 2.7802441731409546,
+      "grad_norm": 0.2796766757965088,
+      "learning_rate": 1.615002893172557e-07,
+      "loss": 0.3915,
+      "step": 3340
+    },
+    {
+      "epoch": 2.7810765815760266,
+      "grad_norm": 0.31362029910087585,
+      "learning_rate": 1.6028107674681547e-07,
+      "loss": 0.395,
+      "step": 3341
+    },
+    {
+      "epoch": 2.781908990011099,
+      "grad_norm": 0.30837786197662354,
+      "learning_rate": 1.5906640875718525e-07,
+      "loss": 0.3694,
+      "step": 3342
+    },
+    {
+      "epoch": 2.782741398446171,
+      "grad_norm": 0.290594220161438,
+      "learning_rate": 1.5785628648895767e-07,
+      "loss": 0.3946,
+      "step": 3343
+    },
+    {
+      "epoch": 2.7835738068812432,
+      "grad_norm": 0.28468722105026245,
+      "learning_rate": 1.5665071107845987e-07,
+      "loss": 0.3844,
+      "step": 3344
+    },
+    {
+      "epoch": 2.784406215316315,
+      "grad_norm": 0.2934969365596771,
+      "learning_rate": 1.5544968365774792e-07,
+      "loss": 0.3824,
+      "step": 3345
+    },
+    {
+      "epoch": 2.785238623751387,
+      "grad_norm": 0.295932799577713,
+      "learning_rate": 1.542532053546081e-07,
+      "loss": 0.4102,
+      "step": 3346
+    },
+    {
+      "epoch": 2.7860710321864595,
+      "grad_norm": 0.292585551738739,
+      "learning_rate": 1.5306127729255382e-07,
+      "loss": 0.3556,
+      "step": 3347
+    },
+    {
+      "epoch": 2.786903440621532,
+      "grad_norm": 0.30441945791244507,
+      "learning_rate": 1.5187390059082706e-07,
+      "loss": 0.3754,
+      "step": 3348
+    },
+    {
+      "epoch": 2.7877358490566038,
+      "grad_norm": 0.2834872305393219,
+      "learning_rate": 1.5069107636439484e-07,
+      "loss": 0.4026,
+      "step": 3349
+    },
+    {
+      "epoch": 2.7885682574916757,
+      "grad_norm": 0.29688331484794617,
+      "learning_rate": 1.4951280572394977e-07,
+      "loss": 0.3684,
+      "step": 3350
+    },
+    {
+      "epoch": 2.789400665926748,
+      "grad_norm": 0.28374260663986206,
+      "learning_rate": 1.483390897759096e-07,
+      "loss": 0.4047,
+      "step": 3351
+    },
+    {
+      "epoch": 2.7902330743618204,
+      "grad_norm": 0.30708324909210205,
+      "learning_rate": 1.4716992962241272e-07,
+      "loss": 0.3917,
+      "step": 3352
+    },
+    {
+      "epoch": 2.7910654827968924,
+      "grad_norm": 0.2824838161468506,
+      "learning_rate": 1.4600532636132147e-07,
+      "loss": 0.3869,
+      "step": 3353
+    },
+    {
+      "epoch": 2.7918978912319643,
+      "grad_norm": 0.294177383184433,
+      "learning_rate": 1.4484528108621942e-07,
+      "loss": 0.4117,
+      "step": 3354
+    },
+    {
+      "epoch": 2.7927302996670367,
+      "grad_norm": 0.3030126988887787,
+      "learning_rate": 1.4368979488640855e-07,
+      "loss": 0.3446,
+      "step": 3355
+    },
+    {
+      "epoch": 2.7935627081021086,
+      "grad_norm": 0.2987135350704193,
+      "learning_rate": 1.4253886884691148e-07,
+      "loss": 0.4387,
+      "step": 3356
+    },
+    {
+      "epoch": 2.794395116537181,
+      "grad_norm": 0.28978827595710754,
+      "learning_rate": 1.4139250404846704e-07,
+      "loss": 0.384,
+      "step": 3357
+    },
+    {
+      "epoch": 2.795227524972253,
+      "grad_norm": 0.28679752349853516,
+      "learning_rate": 1.4025070156753196e-07,
+      "loss": 0.346,
+      "step": 3358
+    },
+    {
+      "epoch": 2.7960599334073253,
+      "grad_norm": 0.341983437538147,
+      "learning_rate": 1.391134624762791e-07,
+      "loss": 0.4132,
+      "step": 3359
+    },
+    {
+      "epoch": 2.796892341842397,
+      "grad_norm": 0.3002639412879944,
+      "learning_rate": 1.3798078784259594e-07,
+      "loss": 0.399,
+      "step": 3360
+    },
+    {
+      "epoch": 2.7977247502774696,
+      "grad_norm": 0.3027034103870392,
+      "learning_rate": 1.368526787300839e-07,
+      "loss": 0.4074,
+      "step": 3361
+    },
+    {
+      "epoch": 2.7985571587125415,
+      "grad_norm": 0.2931459844112396,
+      "learning_rate": 1.3572913619805616e-07,
+      "loss": 0.374,
+      "step": 3362
+    },
+    {
+      "epoch": 2.799389567147614,
+      "grad_norm": 0.2804553806781769,
+      "learning_rate": 1.3461016130153993e-07,
+      "loss": 0.3715,
+      "step": 3363
+    },
+    {
+      "epoch": 2.800221975582686,
+      "grad_norm": 0.2980172038078308,
+      "learning_rate": 1.3349575509127244e-07,
+      "loss": 0.371,
+      "step": 3364
+    },
+    {
+      "epoch": 2.801054384017758,
+      "grad_norm": 0.28683868050575256,
+      "learning_rate": 1.3238591861369943e-07,
+      "loss": 0.3872,
+      "step": 3365
+    },
+    {
+      "epoch": 2.80188679245283,
+      "grad_norm": 0.28891491889953613,
+      "learning_rate": 1.3128065291097724e-07,
+      "loss": 0.3925,
+      "step": 3366
+    },
+    {
+      "epoch": 2.8027192008879025,
+      "grad_norm": 0.3039287030696869,
+      "learning_rate": 1.3017995902097013e-07,
+      "loss": 0.4261,
+      "step": 3367
+    },
+    {
+      "epoch": 2.8035516093229744,
+      "grad_norm": 0.30227887630462646,
+      "learning_rate": 1.290838379772491e-07,
+      "loss": 0.3702,
+      "step": 3368
+    },
+    {
+      "epoch": 2.8043840177580464,
+      "grad_norm": 0.3222786784172058,
+      "learning_rate": 1.2799229080909026e-07,
+      "loss": 0.3801,
+      "step": 3369
+    },
+    {
+      "epoch": 2.8052164261931187,
+      "grad_norm": 0.2965812683105469,
+      "learning_rate": 1.2690531854147537e-07,
+      "loss": 0.3763,
+      "step": 3370
+    },
+    {
+      "epoch": 2.806048834628191,
+      "grad_norm": 0.2899739146232605,
+      "learning_rate": 1.2582292219509184e-07,
+      "loss": 0.3972,
+      "step": 3371
+    },
+    {
+      "epoch": 2.806881243063263,
+      "grad_norm": 0.3159641921520233,
+      "learning_rate": 1.2474510278632779e-07,
+      "loss": 0.4052,
+      "step": 3372
+    },
+    {
+      "epoch": 2.807713651498335,
+      "grad_norm": 0.30823588371276855,
+      "learning_rate": 1.2367186132727415e-07,
+      "loss": 0.3822,
+      "step": 3373
+    },
+    {
+      "epoch": 2.8085460599334073,
+      "grad_norm": 0.286836177110672,
+      "learning_rate": 1.2260319882572425e-07,
+      "loss": 0.3836,
+      "step": 3374
+    },
+    {
+      "epoch": 2.8093784683684797,
+      "grad_norm": 0.2870038151741028,
+      "learning_rate": 1.2153911628517036e-07,
+      "loss": 0.3896,
+      "step": 3375
+    },
+    {
+      "epoch": 2.8102108768035516,
+      "grad_norm": 0.3150666058063507,
+      "learning_rate": 1.2047961470480485e-07,
+      "loss": 0.4282,
+      "step": 3376
+    },
+    {
+      "epoch": 2.8110432852386236,
+      "grad_norm": 0.28661397099494934,
+      "learning_rate": 1.1942469507951803e-07,
+      "loss": 0.3712,
+      "step": 3377
+    },
+    {
+      "epoch": 2.811875693673696,
+      "grad_norm": 0.29480233788490295,
+      "learning_rate": 1.1837435839989808e-07,
+      "loss": 0.416,
+      "step": 3378
+    },
+    {
+      "epoch": 2.812708102108768,
+      "grad_norm": 0.2871578335762024,
+      "learning_rate": 1.1732860565222936e-07,
+      "loss": 0.3905,
+      "step": 3379
+    },
+    {
+      "epoch": 2.8135405105438402,
+      "grad_norm": 0.27761033177375793,
+      "learning_rate": 1.162874378184925e-07,
+      "loss": 0.3864,
+      "step": 3380
+    },
+    {
+      "epoch": 2.814372918978912,
+      "grad_norm": 0.29687756299972534,
+      "learning_rate": 1.1525085587636209e-07,
+      "loss": 0.3494,
+      "step": 3381
+    },
+    {
+      "epoch": 2.8152053274139845,
+      "grad_norm": 0.2945989668369293,
+      "learning_rate": 1.1421886079920619e-07,
+      "loss": 0.3968,
+      "step": 3382
+    },
+    {
+      "epoch": 2.8160377358490565,
+      "grad_norm": 0.2789284884929657,
+      "learning_rate": 1.1319145355608684e-07,
+      "loss": 0.381,
+      "step": 3383
+    },
+    {
+      "epoch": 2.816870144284129,
+      "grad_norm": 0.3032738268375397,
+      "learning_rate": 1.1216863511175736e-07,
+      "loss": 0.3858,
+      "step": 3384
+    },
+    {
+      "epoch": 2.817702552719201,
+      "grad_norm": 0.30183014273643494,
+      "learning_rate": 1.111504064266622e-07,
+      "loss": 0.398,
+      "step": 3385
+    },
+    {
+      "epoch": 2.818534961154273,
+      "grad_norm": 0.29048189520835876,
+      "learning_rate": 1.1013676845693544e-07,
+      "loss": 0.4148,
+      "step": 3386
+    },
+    {
+      "epoch": 2.819367369589345,
+      "grad_norm": 0.2871493101119995,
+      "learning_rate": 1.0912772215440182e-07,
+      "loss": 0.3881,
+      "step": 3387
+    },
+    {
+      "epoch": 2.8201997780244175,
+      "grad_norm": 0.30180609226226807,
+      "learning_rate": 1.0812326846657228e-07,
+      "loss": 0.3678,
+      "step": 3388
+    },
+    {
+      "epoch": 2.8210321864594894,
+      "grad_norm": 0.2673914134502411,
+      "learning_rate": 1.0712340833664737e-07,
+      "loss": 0.3572,
+      "step": 3389
+    },
+    {
+      "epoch": 2.8218645948945618,
+      "grad_norm": 0.28488829731941223,
+      "learning_rate": 1.0612814270351324e-07,
+      "loss": 0.4244,
+      "step": 3390
+    },
+    {
+      "epoch": 2.8226970033296337,
+      "grad_norm": 0.2882390320301056,
+      "learning_rate": 1.0513747250174123e-07,
+      "loss": 0.3718,
+      "step": 3391
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 0.30586186051368713,
+      "learning_rate": 1.0415139866158774e-07,
+      "loss": 0.3935,
+      "step": 3392
+    },
+    {
+      "epoch": 2.824361820199778,
+      "grad_norm": 0.296487033367157,
+      "learning_rate": 1.0316992210899435e-07,
+      "loss": 0.38,
+      "step": 3393
+    },
+    {
+      "epoch": 2.8251942286348504,
+      "grad_norm": 0.26386237144470215,
+      "learning_rate": 1.0219304376558492e-07,
+      "loss": 0.3801,
+      "step": 3394
+    },
+    {
+      "epoch": 2.8260266370699223,
+      "grad_norm": 0.26951250433921814,
+      "learning_rate": 1.0122076454866347e-07,
+      "loss": 0.3868,
+      "step": 3395
+    },
+    {
+      "epoch": 2.8268590455049942,
+      "grad_norm": 0.30076080560684204,
+      "learning_rate": 1.0025308537121859e-07,
+      "loss": 0.4074,
+      "step": 3396
+    },
+    {
+      "epoch": 2.8276914539400666,
+      "grad_norm": 0.3036608397960663,
+      "learning_rate": 9.929000714191838e-08,
+      "loss": 0.4016,
+      "step": 3397
+    },
+    {
+      "epoch": 2.828523862375139,
+      "grad_norm": 0.2856789231300354,
+      "learning_rate": 9.833153076510893e-08,
+      "loss": 0.3614,
+      "step": 3398
+    },
+    {
+      "epoch": 2.829356270810211,
+      "grad_norm": 0.3074534237384796,
+      "learning_rate": 9.737765714081748e-08,
+      "loss": 0.3782,
+      "step": 3399
+    },
+    {
+      "epoch": 2.830188679245283,
+      "grad_norm": 0.2953753173351288,
+      "learning_rate": 9.642838716474645e-08,
+      "loss": 0.4025,
+      "step": 3400
+    },
+    {
+      "epoch": 2.831021087680355,
+      "grad_norm": 0.2885286808013916,
+      "learning_rate": 9.548372172827946e-08,
+      "loss": 0.3511,
+      "step": 3401
+    },
+    {
+      "epoch": 2.831853496115427,
+      "grad_norm": 0.2874261736869812,
+      "learning_rate": 9.454366171847196e-08,
+      "loss": 0.3967,
+      "step": 3402
+    },
+    {
+      "epoch": 2.8326859045504995,
+      "grad_norm": 0.29668954014778137,
+      "learning_rate": 9.360820801805726e-08,
+      "loss": 0.4091,
+      "step": 3403
+    },
+    {
+      "epoch": 2.8335183129855714,
+      "grad_norm": 0.28590044379234314,
+      "learning_rate": 9.267736150544271e-08,
+      "loss": 0.3797,
+      "step": 3404
+    },
+    {
+      "epoch": 2.834350721420644,
+      "grad_norm": 0.31228768825531006,
+      "learning_rate": 9.175112305470913e-08,
+      "loss": 0.4138,
+      "step": 3405
+    },
+    {
+      "epoch": 2.8351831298557157,
+      "grad_norm": 0.29430899024009705,
+      "learning_rate": 9.082949353561187e-08,
+      "loss": 0.3866,
+      "step": 3406
+    },
+    {
+      "epoch": 2.836015538290788,
+      "grad_norm": 0.30940887331962585,
+      "learning_rate": 8.991247381357593e-08,
+      "loss": 0.4006,
+      "step": 3407
+    },
+    {
+      "epoch": 2.83684794672586,
+      "grad_norm": 0.29085415601730347,
+      "learning_rate": 8.900006474969913e-08,
+      "loss": 0.3581,
+      "step": 3408
+    },
+    {
+      "epoch": 2.8376803551609324,
+      "grad_norm": 0.29941922426223755,
+      "learning_rate": 8.809226720075059e-08,
+      "loss": 0.4113,
+      "step": 3409
+    },
+    {
+      "epoch": 2.8385127635960044,
+      "grad_norm": 0.2955285906791687,
+      "learning_rate": 8.718908201916676e-08,
+      "loss": 0.3789,
+      "step": 3410
+    },
+    {
+      "epoch": 2.8393451720310767,
+      "grad_norm": 0.2934470772743225,
+      "learning_rate": 8.629051005305478e-08,
+      "loss": 0.3859,
+      "step": 3411
+    },
+    {
+      "epoch": 2.8401775804661487,
+      "grad_norm": 0.28304851055145264,
+      "learning_rate": 8.539655214618969e-08,
+      "loss": 0.3857,
+      "step": 3412
+    },
+    {
+      "epoch": 2.841009988901221,
+      "grad_norm": 0.2950180172920227,
+      "learning_rate": 8.450720913801336e-08,
+      "loss": 0.3564,
+      "step": 3413
+    },
+    {
+      "epoch": 2.841842397336293,
+      "grad_norm": 0.2903018593788147,
+      "learning_rate": 8.362248186363441e-08,
+      "loss": 0.41,
+      "step": 3414
+    },
+    {
+      "epoch": 2.842674805771365,
+      "grad_norm": 0.3056175708770752,
+      "learning_rate": 8.274237115382777e-08,
+      "loss": 0.381,
+      "step": 3415
+    },
+    {
+      "epoch": 2.8435072142064373,
+      "grad_norm": 0.2846352756023407,
+      "learning_rate": 8.186687783503289e-08,
+      "loss": 0.376,
+      "step": 3416
+    },
+    {
+      "epoch": 2.8443396226415096,
+      "grad_norm": 0.29959842562675476,
+      "learning_rate": 8.09960027293516e-08,
+      "loss": 0.395,
+      "step": 3417
+    },
+    {
+      "epoch": 2.8451720310765816,
+      "grad_norm": 0.30339470505714417,
+      "learning_rate": 8.012974665455308e-08,
+      "loss": 0.3928,
+      "step": 3418
+    },
+    {
+      "epoch": 2.8460044395116535,
+      "grad_norm": 0.3100747764110565,
+      "learning_rate": 7.926811042406557e-08,
+      "loss": 0.4121,
+      "step": 3419
+    },
+    {
+      "epoch": 2.846836847946726,
+      "grad_norm": 0.28325793147087097,
+      "learning_rate": 7.841109484698184e-08,
+      "loss": 0.4077,
+      "step": 3420
+    },
+    {
+      "epoch": 2.8476692563817982,
+      "grad_norm": 0.30448758602142334,
+      "learning_rate": 7.755870072805316e-08,
+      "loss": 0.3853,
+      "step": 3421
+    },
+    {
+      "epoch": 2.84850166481687,
+      "grad_norm": 0.3250058591365814,
+      "learning_rate": 7.67109288676926e-08,
+      "loss": 0.4308,
+      "step": 3422
+    },
+    {
+      "epoch": 2.849334073251942,
+      "grad_norm": 0.27683284878730774,
+      "learning_rate": 7.586778006197337e-08,
+      "loss": 0.3575,
+      "step": 3423
+    },
+    {
+      "epoch": 2.8501664816870145,
+      "grad_norm": 0.30187785625457764,
+      "learning_rate": 7.50292551026277e-08,
+      "loss": 0.4145,
+      "step": 3424
+    },
+    {
+      "epoch": 2.8509988901220864,
+      "grad_norm": 0.2923542261123657,
+      "learning_rate": 7.419535477704354e-08,
+      "loss": 0.3981,
+      "step": 3425
+    },
+    {
+      "epoch": 2.851831298557159,
+      "grad_norm": 0.3036320209503174,
+      "learning_rate": 7.336607986826839e-08,
+      "loss": 0.3879,
+      "step": 3426
+    },
+    {
+      "epoch": 2.8526637069922307,
+      "grad_norm": 0.29370933771133423,
+      "learning_rate": 7.254143115500711e-08,
+      "loss": 0.3971,
+      "step": 3427
+    },
+    {
+      "epoch": 2.853496115427303,
+      "grad_norm": 0.307032972574234,
+      "learning_rate": 7.17214094116181e-08,
+      "loss": 0.398,
+      "step": 3428
+    },
+    {
+      "epoch": 2.854328523862375,
+      "grad_norm": 0.3023212254047394,
+      "learning_rate": 7.090601540811648e-08,
+      "loss": 0.3883,
+      "step": 3429
+    },
+    {
+      "epoch": 2.8551609322974474,
+      "grad_norm": 0.2827087938785553,
+      "learning_rate": 7.009524991017091e-08,
+      "loss": 0.3788,
+      "step": 3430
+    },
+    {
+      "epoch": 2.8559933407325193,
+      "grad_norm": 0.28021904826164246,
+      "learning_rate": 6.928911367910573e-08,
+      "loss": 0.3843,
+      "step": 3431
+    },
+    {
+      "epoch": 2.8568257491675917,
+      "grad_norm": 0.2957780063152313,
+      "learning_rate": 6.848760747189598e-08,
+      "loss": 0.3939,
+      "step": 3432
+    },
+    {
+      "epoch": 2.8576581576026636,
+      "grad_norm": 0.307271808385849,
+      "learning_rate": 6.769073204117016e-08,
+      "loss": 0.4193,
+      "step": 3433
+    },
+    {
+      "epoch": 2.858490566037736,
+      "grad_norm": 0.3014625012874603,
+      "learning_rate": 6.689848813520805e-08,
+      "loss": 0.3848,
+      "step": 3434
+    },
+    {
+      "epoch": 2.859322974472808,
+      "grad_norm": 0.2933945059776306,
+      "learning_rate": 6.611087649794124e-08,
+      "loss": 0.3986,
+      "step": 3435
+    },
+    {
+      "epoch": 2.8601553829078803,
+      "grad_norm": 0.29639217257499695,
+      "learning_rate": 6.532789786895033e-08,
+      "loss": 0.3729,
+      "step": 3436
+    },
+    {
+      "epoch": 2.8609877913429522,
+      "grad_norm": 0.31975582242012024,
+      "learning_rate": 6.454955298346555e-08,
+      "loss": 0.4172,
+      "step": 3437
+    },
+    {
+      "epoch": 2.861820199778024,
+      "grad_norm": 0.2899722754955292,
+      "learning_rate": 6.377584257236724e-08,
+      "loss": 0.3822,
+      "step": 3438
+    },
+    {
+      "epoch": 2.8626526082130965,
+      "grad_norm": 0.27522486448287964,
+      "learning_rate": 6.300676736218258e-08,
+      "loss": 0.3855,
+      "step": 3439
+    },
+    {
+      "epoch": 2.863485016648169,
+      "grad_norm": 0.2691723704338074,
+      "learning_rate": 6.224232807508667e-08,
+      "loss": 0.3849,
+      "step": 3440
+    },
+    {
+      "epoch": 2.864317425083241,
+      "grad_norm": 0.2858082056045532,
+      "learning_rate": 6.148252542890198e-08,
+      "loss": 0.4041,
+      "step": 3441
+    },
+    {
+      "epoch": 2.8651498335183128,
+      "grad_norm": 0.27565300464630127,
+      "learning_rate": 6.072736013709557e-08,
+      "loss": 0.3919,
+      "step": 3442
+    },
+    {
+      "epoch": 2.865982241953385,
+      "grad_norm": 0.2915458679199219,
+      "learning_rate": 5.997683290878131e-08,
+      "loss": 0.4133,
+      "step": 3443
+    },
+    {
+      "epoch": 2.8668146503884575,
+      "grad_norm": 0.27250799536705017,
+      "learning_rate": 5.923094444871713e-08,
+      "loss": 0.3583,
+      "step": 3444
+    },
+    {
+      "epoch": 2.8676470588235294,
+      "grad_norm": 0.28215593099594116,
+      "learning_rate": 5.848969545730554e-08,
+      "loss": 0.4154,
+      "step": 3445
+    },
+    {
+      "epoch": 2.8684794672586014,
+      "grad_norm": 0.2830628752708435,
+      "learning_rate": 5.775308663059309e-08,
+      "loss": 0.385,
+      "step": 3446
+    },
+    {
+      "epoch": 2.8693118756936737,
+      "grad_norm": 0.2738587558269501,
+      "learning_rate": 5.702111866026705e-08,
+      "loss": 0.373,
+      "step": 3447
+    },
+    {
+      "epoch": 2.870144284128746,
+      "grad_norm": 0.28816235065460205,
+      "learning_rate": 5.629379223365872e-08,
+      "loss": 0.3963,
+      "step": 3448
+    },
+    {
+      "epoch": 2.870976692563818,
+      "grad_norm": 0.3096372187137604,
+      "learning_rate": 5.557110803374066e-08,
+      "loss": 0.4023,
+      "step": 3449
+    },
+    {
+      "epoch": 2.87180910099889,
+      "grad_norm": 0.29053497314453125,
+      "learning_rate": 5.485306673912616e-08,
+      "loss": 0.401,
+      "step": 3450
+    },
+    {
+      "epoch": 2.8726415094339623,
+      "grad_norm": 0.29939737915992737,
+      "learning_rate": 5.413966902406753e-08,
+      "loss": 0.3722,
+      "step": 3451
+    },
+    {
+      "epoch": 2.8734739178690343,
+      "grad_norm": 0.285515159368515,
+      "learning_rate": 5.343091555845781e-08,
+      "loss": 0.3878,
+      "step": 3452
+    },
+    {
+      "epoch": 2.8743063263041067,
+      "grad_norm": 0.28648898005485535,
+      "learning_rate": 5.272680700783073e-08,
+      "loss": 0.3773,
+      "step": 3453
+    },
+    {
+      "epoch": 2.8751387347391786,
+      "grad_norm": 0.275084525346756,
+      "learning_rate": 5.2027344033354077e-08,
+      "loss": 0.3646,
+      "step": 3454
+    },
+    {
+      "epoch": 2.875971143174251,
+      "grad_norm": 0.2844296395778656,
+      "learning_rate": 5.1332527291837465e-08,
+      "loss": 0.3902,
+      "step": 3455
+    },
+    {
+      "epoch": 2.876803551609323,
+      "grad_norm": 0.32090917229652405,
+      "learning_rate": 5.06423574357251e-08,
+      "loss": 0.3932,
+      "step": 3456
+    },
+    {
+      "epoch": 2.8776359600443953,
+      "grad_norm": 0.29726335406303406,
+      "learning_rate": 4.9956835113099676e-08,
+      "loss": 0.3991,
+      "step": 3457
+    },
+    {
+      "epoch": 2.878468368479467,
+      "grad_norm": 0.30311334133148193,
+      "learning_rate": 4.927596096767795e-08,
+      "loss": 0.385,
+      "step": 3458
+    },
+    {
+      "epoch": 2.8793007769145396,
+      "grad_norm": 0.3128688633441925,
+      "learning_rate": 4.8599735638812373e-08,
+      "loss": 0.4243,
+      "step": 3459
+    },
+    {
+      "epoch": 2.8801331853496115,
+      "grad_norm": 0.2947447896003723,
+      "learning_rate": 4.7928159761490566e-08,
+      "loss": 0.4102,
+      "step": 3460
+    },
+    {
+      "epoch": 2.8809655937846834,
+      "grad_norm": 0.2862451374530792,
+      "learning_rate": 4.7261233966334196e-08,
+      "loss": 0.3962,
+      "step": 3461
+    },
+    {
+      "epoch": 2.881798002219756,
+      "grad_norm": 0.2894607186317444,
+      "learning_rate": 4.659895887959787e-08,
+      "loss": 0.3894,
+      "step": 3462
+    },
+    {
+      "epoch": 2.882630410654828,
+      "grad_norm": 0.2788543701171875,
+      "learning_rate": 4.594133512317023e-08,
+      "loss": 0.3739,
+      "step": 3463
+    },
+    {
+      "epoch": 2.8834628190899,
+      "grad_norm": 0.289505273103714,
+      "learning_rate": 4.528836331457065e-08,
+      "loss": 0.4062,
+      "step": 3464
+    },
+    {
+      "epoch": 2.884295227524972,
+      "grad_norm": 0.26619216799736023,
+      "learning_rate": 4.4640044066951994e-08,
+      "loss": 0.3453,
+      "step": 3465
+    },
+    {
+      "epoch": 2.8851276359600444,
+      "grad_norm": 0.29299643635749817,
+      "learning_rate": 4.399637798909673e-08,
+      "loss": 0.4205,
+      "step": 3466
+    },
+    {
+      "epoch": 2.8859600443951168,
+      "grad_norm": 0.2827985882759094,
+      "learning_rate": 4.335736568541915e-08,
+      "loss": 0.4005,
+      "step": 3467
+    },
+    {
+      "epoch": 2.8867924528301887,
+      "grad_norm": 0.2879890203475952,
+      "learning_rate": 4.272300775596205e-08,
+      "loss": 0.3817,
+      "step": 3468
+    },
+    {
+      "epoch": 2.8876248612652606,
+      "grad_norm": 0.30056655406951904,
+      "learning_rate": 4.2093304796399504e-08,
+      "loss": 0.372,
+      "step": 3469
+    },
+    {
+      "epoch": 2.888457269700333,
+      "grad_norm": 0.285110205411911,
+      "learning_rate": 4.146825739803295e-08,
+      "loss": 0.3608,
+      "step": 3470
+    },
+    {
+      "epoch": 2.8892896781354054,
+      "grad_norm": 0.29550155997276306,
+      "learning_rate": 4.084786614779346e-08,
+      "loss": 0.4042,
+      "step": 3471
+    },
+    {
+      "epoch": 2.8901220865704773,
+      "grad_norm": 0.3029278516769409,
+      "learning_rate": 4.023213162823947e-08,
+      "loss": 0.4232,
+      "step": 3472
+    },
+    {
+      "epoch": 2.8909544950055492,
+      "grad_norm": 0.2764630913734436,
+      "learning_rate": 3.962105441755515e-08,
+      "loss": 0.3862,
+      "step": 3473
+    },
+    {
+      "epoch": 2.8917869034406216,
+      "grad_norm": 0.2890421748161316,
+      "learning_rate": 3.9014635089554274e-08,
+      "loss": 0.3944,
+      "step": 3474
+    },
+    {
+      "epoch": 2.8926193118756935,
+      "grad_norm": 0.3060329258441925,
+      "learning_rate": 3.841287421367412e-08,
+      "loss": 0.3911,
+      "step": 3475
+    },
+    {
+      "epoch": 2.893451720310766,
+      "grad_norm": 0.29384008049964905,
+      "learning_rate": 3.781577235497935e-08,
+      "loss": 0.3783,
+      "step": 3476
+    },
+    {
+      "epoch": 2.894284128745838,
+      "grad_norm": 0.2772958278656006,
+      "learning_rate": 3.7223330074158126e-08,
+      "loss": 0.3298,
+      "step": 3477
+    },
+    {
+      "epoch": 2.89511653718091,
+      "grad_norm": 0.32114148139953613,
+      "learning_rate": 3.663554792752544e-08,
+      "loss": 0.4557,
+      "step": 3478
+    },
+    {
+      "epoch": 2.895948945615982,
+      "grad_norm": 0.27607429027557373,
+      "learning_rate": 3.605242646701812e-08,
+      "loss": 0.3632,
+      "step": 3479
+    },
+    {
+      "epoch": 2.8967813540510545,
+      "grad_norm": 0.3078801929950714,
+      "learning_rate": 3.547396624019817e-08,
+      "loss": 0.4103,
+      "step": 3480
+    },
+    {
+      "epoch": 2.8976137624861265,
+      "grad_norm": 0.2985548675060272,
+      "learning_rate": 3.490016779024885e-08,
+      "loss": 0.3895,
+      "step": 3481
+    },
+    {
+      "epoch": 2.898446170921199,
+      "grad_norm": 0.3042149245738983,
+      "learning_rate": 3.4331031655976955e-08,
+      "loss": 0.3638,
+      "step": 3482
+    },
+    {
+      "epoch": 2.8992785793562708,
+      "grad_norm": 0.28391167521476746,
+      "learning_rate": 3.3766558371812754e-08,
+      "loss": 0.4081,
+      "step": 3483
+    },
+    {
+      "epoch": 2.9001109877913427,
+      "grad_norm": 0.29962408542633057,
+      "learning_rate": 3.320674846780503e-08,
+      "loss": 0.3843,
+      "step": 3484
+    },
+    {
+      "epoch": 2.900943396226415,
+      "grad_norm": 0.27982330322265625,
+      "learning_rate": 3.265160246962607e-08,
+      "loss": 0.3801,
+      "step": 3485
+    },
+    {
+      "epoch": 2.9017758046614874,
+      "grad_norm": 0.30797380208969116,
+      "learning_rate": 3.210112089856721e-08,
+      "loss": 0.4107,
+      "step": 3486
+    },
+    {
+      "epoch": 2.9026082130965594,
+      "grad_norm": 0.30911773443222046,
+      "learning_rate": 3.155530427153997e-08,
+      "loss": 0.4328,
+      "step": 3487
+    },
+    {
+      "epoch": 2.9034406215316313,
+      "grad_norm": 0.2736690044403076,
+      "learning_rate": 3.1014153101076026e-08,
+      "loss": 0.3458,
+      "step": 3488
+    },
+    {
+      "epoch": 2.9042730299667037,
+      "grad_norm": 0.28448286652565,
+      "learning_rate": 3.0477667895326133e-08,
+      "loss": 0.3888,
+      "step": 3489
+    },
+    {
+      "epoch": 2.905105438401776,
+      "grad_norm": 0.29954880475997925,
+      "learning_rate": 2.994584915805898e-08,
+      "loss": 0.3806,
+      "step": 3490
+    },
+    {
+      "epoch": 2.905937846836848,
+      "grad_norm": 0.27898576855659485,
+      "learning_rate": 2.9418697388661766e-08,
+      "loss": 0.3819,
+      "step": 3491
+    },
+    {
+      "epoch": 2.90677025527192,
+      "grad_norm": 0.305177241563797,
+      "learning_rate": 2.889621308213908e-08,
+      "loss": 0.3835,
+      "step": 3492
+    },
+    {
+      "epoch": 2.9076026637069923,
+      "grad_norm": 0.29762381315231323,
+      "learning_rate": 2.8378396729113466e-08,
+      "loss": 0.3718,
+      "step": 3493
+    },
+    {
+      "epoch": 2.9084350721420646,
+      "grad_norm": 0.29369136691093445,
+      "learning_rate": 2.7865248815822087e-08,
+      "loss": 0.3917,
+      "step": 3494
+    },
+    {
+      "epoch": 2.9092674805771366,
+      "grad_norm": 0.29049572348594666,
+      "learning_rate": 2.7356769824121166e-08,
+      "loss": 0.4114,
+      "step": 3495
+    },
+    {
+      "epoch": 2.9100998890122085,
+      "grad_norm": 0.27511221170425415,
+      "learning_rate": 2.6852960231480985e-08,
+      "loss": 0.3677,
+      "step": 3496
+    },
+    {
+      "epoch": 2.910932297447281,
+      "grad_norm": 0.28010135889053345,
+      "learning_rate": 2.635382051098756e-08,
+      "loss": 0.406,
+      "step": 3497
+    },
+    {
+      "epoch": 2.911764705882353,
+      "grad_norm": 0.28749653697013855,
+      "learning_rate": 2.585935113134208e-08,
+      "loss": 0.3838,
+      "step": 3498
+    },
+    {
+      "epoch": 2.912597114317425,
+      "grad_norm": 0.2730650305747986,
+      "learning_rate": 2.5369552556859243e-08,
+      "loss": 0.3455,
+      "step": 3499
+    },
+    {
+      "epoch": 2.913429522752497,
+      "grad_norm": 0.2961804270744324,
+      "learning_rate": 2.4884425247468924e-08,
+      "loss": 0.3869,
+      "step": 3500
+    },
+    {
+      "epoch": 2.9142619311875695,
+      "grad_norm": 0.29125046730041504,
+      "learning_rate": 2.44039696587145e-08,
+      "loss": 0.383,
+      "step": 3501
+    },
+    {
+      "epoch": 2.9150943396226414,
+      "grad_norm": 0.3115370273590088,
+      "learning_rate": 2.392818624175175e-08,
+      "loss": 0.3976,
+      "step": 3502
+    },
+    {
+      "epoch": 2.915926748057714,
+      "grad_norm": 0.3151339590549469,
+      "learning_rate": 2.345707544334941e-08,
+      "loss": 0.3907,
+      "step": 3503
+    },
+    {
+      "epoch": 2.9167591564927857,
+      "grad_norm": 0.29095959663391113,
+      "learning_rate": 2.2990637705889717e-08,
+      "loss": 0.3903,
+      "step": 3504
+    },
+    {
+      "epoch": 2.917591564927858,
+      "grad_norm": 0.30376264452934265,
+      "learning_rate": 2.2528873467365098e-08,
+      "loss": 0.4254,
+      "step": 3505
+    },
+    {
+      "epoch": 2.91842397336293,
+      "grad_norm": 0.28083696961402893,
+      "learning_rate": 2.2071783161379812e-08,
+      "loss": 0.3617,
+      "step": 3506
+    },
+    {
+      "epoch": 2.919256381798002,
+      "grad_norm": 0.3202035129070282,
+      "learning_rate": 2.1619367217150522e-08,
+      "loss": 0.4465,
+      "step": 3507
+    },
+    {
+      "epoch": 2.9200887902330743,
+      "grad_norm": 0.2700662612915039,
+      "learning_rate": 2.1171626059503514e-08,
+      "loss": 0.3814,
+      "step": 3508
+    },
+    {
+      "epoch": 2.9209211986681467,
+      "grad_norm": 0.2775675058364868,
+      "learning_rate": 2.0728560108875807e-08,
+      "loss": 0.3732,
+      "step": 3509
+    },
+    {
+      "epoch": 2.9217536071032186,
+      "grad_norm": 0.2907252311706543,
+      "learning_rate": 2.0290169781313483e-08,
+      "loss": 0.418,
+      "step": 3510
+    },
+    {
+      "epoch": 2.9225860155382906,
+      "grad_norm": 0.27951502799987793,
+      "learning_rate": 1.985645548847337e-08,
+      "loss": 0.3932,
+      "step": 3511
+    },
+    {
+      "epoch": 2.923418423973363,
+      "grad_norm": 0.2931711673736572,
+      "learning_rate": 1.9427417637619685e-08,
+      "loss": 0.3873,
+      "step": 3512
+    },
+    {
+      "epoch": 2.9242508324084353,
+      "grad_norm": 0.30858153104782104,
+      "learning_rate": 1.9003056631627935e-08,
+      "loss": 0.4157,
+      "step": 3513
+    },
+    {
+      "epoch": 2.9250832408435072,
+      "grad_norm": 0.2937333285808563,
+      "learning_rate": 1.8583372868979933e-08,
+      "loss": 0.3687,
+      "step": 3514
+    },
+    {
+      "epoch": 2.925915649278579,
+      "grad_norm": 0.2885035574436188,
+      "learning_rate": 1.8168366743765432e-08,
+      "loss": 0.3595,
+      "step": 3515
+    },
+    {
+      "epoch": 2.9267480577136515,
+      "grad_norm": 0.2732231914997101,
+      "learning_rate": 1.775803864568326e-08,
+      "loss": 0.3982,
+      "step": 3516
+    },
+    {
+      "epoch": 2.927580466148724,
+      "grad_norm": 0.29774099588394165,
+      "learning_rate": 1.7352388960038548e-08,
+      "loss": 0.4037,
+      "step": 3517
+    },
+    {
+      "epoch": 2.928412874583796,
+      "grad_norm": 0.304595410823822,
+      "learning_rate": 1.695141806774325e-08,
+      "loss": 0.3923,
+      "step": 3518
+    },
+    {
+      "epoch": 2.9292452830188678,
+      "grad_norm": 0.27520307898521423,
+      "learning_rate": 1.6555126345316197e-08,
+      "loss": 0.3416,
+      "step": 3519
+    },
+    {
+      "epoch": 2.93007769145394,
+      "grad_norm": 0.30991899967193604,
+      "learning_rate": 1.6163514164882486e-08,
+      "loss": 0.4287,
+      "step": 3520
+    },
+    {
+      "epoch": 2.930910099889012,
+      "grad_norm": 0.29716676473617554,
+      "learning_rate": 1.577658189417186e-08,
+      "loss": 0.3619,
+      "step": 3521
+    },
+    {
+      "epoch": 2.9317425083240845,
+      "grad_norm": 0.2918042540550232,
+      "learning_rate": 1.539432989652201e-08,
+      "loss": 0.3875,
+      "step": 3522
+    },
+    {
+      "epoch": 2.9325749167591564,
+      "grad_norm": 0.27701038122177124,
+      "learning_rate": 1.5016758530873033e-08,
+      "loss": 0.3748,
+      "step": 3523
+    },
+    {
+      "epoch": 2.9334073251942288,
+      "grad_norm": 0.31016066670417786,
+      "learning_rate": 1.4643868151771323e-08,
+      "loss": 0.394,
+      "step": 3524
+    },
+    {
+      "epoch": 2.9342397336293007,
+      "grad_norm": 0.29906758666038513,
+      "learning_rate": 1.4275659109367346e-08,
+      "loss": 0.4065,
+      "step": 3525
+    },
+    {
+      "epoch": 2.935072142064373,
+      "grad_norm": 0.31481271982192993,
+      "learning_rate": 1.3912131749416746e-08,
+      "loss": 0.3975,
+      "step": 3526
+    },
+    {
+      "epoch": 2.935904550499445,
+      "grad_norm": 0.2872684597969055,
+      "learning_rate": 1.3553286413277022e-08,
+      "loss": 0.3697,
+      "step": 3527
+    },
+    {
+      "epoch": 2.9367369589345174,
+      "grad_norm": 0.29366058111190796,
+      "learning_rate": 1.3199123437910855e-08,
+      "loss": 0.3647,
+      "step": 3528
+    },
+    {
+      "epoch": 2.9375693673695893,
+      "grad_norm": 0.29561111330986023,
+      "learning_rate": 1.2849643155882773e-08,
+      "loss": 0.3693,
+      "step": 3529
+    },
+    {
+      "epoch": 2.938401775804661,
+      "grad_norm": 0.2844444811344147,
+      "learning_rate": 1.2504845895361384e-08,
+      "loss": 0.3812,
+      "step": 3530
+    },
+    {
+      "epoch": 2.9392341842397336,
+      "grad_norm": 0.26833999156951904,
+      "learning_rate": 1.2164731980117694e-08,
+      "loss": 0.4002,
+      "step": 3531
+    },
+    {
+      "epoch": 2.940066592674806,
+      "grad_norm": 0.2657759487628937,
+      "learning_rate": 1.1829301729524567e-08,
+      "loss": 0.3429,
+      "step": 3532
+    },
+    {
+      "epoch": 2.940899001109878,
+      "grad_norm": 0.31585893034935,
+      "learning_rate": 1.1498555458555604e-08,
+      "loss": 0.3983,
+      "step": 3533
+    },
+    {
+      "epoch": 2.94173140954495,
+      "grad_norm": 0.3071114420890808,
+      "learning_rate": 1.1172493477789037e-08,
+      "loss": 0.4385,
+      "step": 3534
+    },
+    {
+      "epoch": 2.942563817980022,
+      "grad_norm": 0.2719644010066986,
+      "learning_rate": 1.085111609340217e-08,
+      "loss": 0.3399,
+      "step": 3535
+    },
+    {
+      "epoch": 2.9433962264150946,
+      "grad_norm": 0.2977891266345978,
+      "learning_rate": 1.0534423607173604e-08,
+      "loss": 0.38,
+      "step": 3536
+    },
+    {
+      "epoch": 2.9442286348501665,
+      "grad_norm": 0.2826799154281616,
+      "learning_rate": 1.022241631648324e-08,
+      "loss": 0.3787,
+      "step": 3537
+    },
+    {
+      "epoch": 2.9450610432852384,
+      "grad_norm": 0.2901372015476227,
+      "learning_rate": 9.915094514311719e-09,
+      "loss": 0.4123,
+      "step": 3538
+    },
+    {
+      "epoch": 2.945893451720311,
+      "grad_norm": 0.2899583578109741,
+      "learning_rate": 9.612458489239308e-09,
+      "loss": 0.4093,
+      "step": 3539
+    },
+    {
+      "epoch": 2.946725860155383,
+      "grad_norm": 0.27518314123153687,
+      "learning_rate": 9.314508525446464e-09,
+      "loss": 0.3605,
+      "step": 3540
+    },
+    {
+      "epoch": 2.947558268590455,
+      "grad_norm": 0.29123416543006897,
+      "learning_rate": 9.021244902713833e-09,
+      "loss": 0.3942,
+      "step": 3541
+    },
+    {
+      "epoch": 2.948390677025527,
+      "grad_norm": 0.2904079258441925,
+      "learning_rate": 8.732667896421131e-09,
+      "loss": 0.3999,
+      "step": 3542
+    },
+    {
+      "epoch": 2.9492230854605994,
+      "grad_norm": 0.29001057147979736,
+      "learning_rate": 8.448777777546601e-09,
+      "loss": 0.3751,
+      "step": 3543
+    },
+    {
+      "epoch": 2.9500554938956713,
+      "grad_norm": 0.2952643632888794,
+      "learning_rate": 8.169574812668668e-09,
+      "loss": 0.3758,
+      "step": 3544
+    },
+    {
+      "epoch": 2.9508879023307437,
+      "grad_norm": 0.3037968873977661,
+      "learning_rate": 7.895059263963168e-09,
+      "loss": 0.4062,
+      "step": 3545
+    },
+    {
+      "epoch": 2.9517203107658156,
+      "grad_norm": 0.29707083106040955,
+      "learning_rate": 7.625231389205567e-09,
+      "loss": 0.3736,
+      "step": 3546
+    },
+    {
+      "epoch": 2.952552719200888,
+      "grad_norm": 0.3003142178058624,
+      "learning_rate": 7.360091441768746e-09,
+      "loss": 0.3837,
+      "step": 3547
+    },
+    {
+      "epoch": 2.95338512763596,
+      "grad_norm": 0.2987581789493561,
+      "learning_rate": 7.099639670623548e-09,
+      "loss": 0.3834,
+      "step": 3548
+    },
+    {
+      "epoch": 2.9542175360710323,
+      "grad_norm": 0.28041496872901917,
+      "learning_rate": 6.8438763203393375e-09,
+      "loss": 0.366,
+      "step": 3549
+    },
+    {
+      "epoch": 2.9550499445061043,
+      "grad_norm": 0.28142842650413513,
+      "learning_rate": 6.59280163108178e-09,
+      "loss": 0.4162,
+      "step": 3550
+    },
+    {
+      "epoch": 2.9558823529411766,
+      "grad_norm": 0.2837742865085602,
+      "learning_rate": 6.346415838614506e-09,
+      "loss": 0.3371,
+      "step": 3551
+    },
+    {
+      "epoch": 2.9567147613762486,
+      "grad_norm": 0.29957345128059387,
+      "learning_rate": 6.104719174298557e-09,
+      "loss": 0.4113,
+      "step": 3552
+    },
+    {
+      "epoch": 2.9575471698113205,
+      "grad_norm": 0.2940880358219147,
+      "learning_rate": 5.867711865090719e-09,
+      "loss": 0.3893,
+      "step": 3553
+    },
+    {
+      "epoch": 2.958379578246393,
+      "grad_norm": 0.29609382152557373,
+      "learning_rate": 5.635394133545191e-09,
+      "loss": 0.3688,
+      "step": 3554
+    },
+    {
+      "epoch": 2.9592119866814652,
+      "grad_norm": 0.29784825444221497,
+      "learning_rate": 5.40776619781247e-09,
+      "loss": 0.4102,
+      "step": 3555
+    },
+    {
+      "epoch": 2.960044395116537,
+      "grad_norm": 0.2848165035247803,
+      "learning_rate": 5.184828271639353e-09,
+      "loss": 0.3718,
+      "step": 3556
+    },
+    {
+      "epoch": 2.960876803551609,
+      "grad_norm": 0.30922722816467285,
+      "learning_rate": 4.966580564368384e-09,
+      "loss": 0.3542,
+      "step": 3557
+    },
+    {
+      "epoch": 2.9617092119866815,
+      "grad_norm": 0.2894100546836853,
+      "learning_rate": 4.7530232809378515e-09,
+      "loss": 0.4281,
+      "step": 3558
+    },
+    {
+      "epoch": 2.962541620421754,
+      "grad_norm": 0.2721484899520874,
+      "learning_rate": 4.54415662188179e-09,
+      "loss": 0.392,
+      "step": 3559
+    },
+    {
+      "epoch": 2.9633740288568258,
+      "grad_norm": 0.30090397596359253,
+      "learning_rate": 4.339980783329423e-09,
+      "loss": 0.3977,
+      "step": 3560
+    },
+    {
+      "epoch": 2.9642064372918977,
+      "grad_norm": 0.29261451959609985,
+      "learning_rate": 4.140495957006274e-09,
+      "loss": 0.3847,
+      "step": 3561
+    },
+    {
+      "epoch": 2.96503884572697,
+      "grad_norm": 0.2850942015647888,
+      "learning_rate": 3.945702330230839e-09,
+      "loss": 0.4031,
+      "step": 3562
+    },
+    {
+      "epoch": 2.9658712541620424,
+      "grad_norm": 0.2756821811199188,
+      "learning_rate": 3.755600085918465e-09,
+      "loss": 0.364,
+      "step": 3563
+    },
+    {
+      "epoch": 2.9667036625971144,
+      "grad_norm": 0.31473076343536377,
+      "learning_rate": 3.5701894025791383e-09,
+      "loss": 0.446,
+      "step": 3564
+    },
+    {
+      "epoch": 2.9675360710321863,
+      "grad_norm": 0.29244500398635864,
+      "learning_rate": 3.3894704543152578e-09,
+      "loss": 0.4213,
+      "step": 3565
+    },
+    {
+      "epoch": 2.9683684794672587,
+      "grad_norm": 0.29834458231925964,
+      "learning_rate": 3.213443410826078e-09,
+      "loss": 0.3975,
+      "step": 3566
+    },
+    {
+      "epoch": 2.9692008879023306,
+      "grad_norm": 0.29561495780944824,
+      "learning_rate": 3.0421084374038234e-09,
+      "loss": 0.3765,
+      "step": 3567
+    },
+    {
+      "epoch": 2.970033296337403,
+      "grad_norm": 0.28357571363449097,
+      "learning_rate": 2.875465694935353e-09,
+      "loss": 0.4125,
+      "step": 3568
+    },
+    {
+      "epoch": 2.970865704772475,
+      "grad_norm": 0.3278956413269043,
+      "learning_rate": 2.7135153399004967e-09,
+      "loss": 0.4449,
+      "step": 3569
+    },
+    {
+      "epoch": 2.9716981132075473,
+      "grad_norm": 0.2880827784538269,
+      "learning_rate": 2.5562575243737176e-09,
+      "loss": 0.364,
+      "step": 3570
+    },
+    {
+      "epoch": 2.972530521642619,
+      "grad_norm": 0.30118292570114136,
+      "learning_rate": 2.4036923960230053e-09,
+      "loss": 0.3864,
+      "step": 3571
+    },
+    {
+      "epoch": 2.9733629300776916,
+      "grad_norm": 0.30025699734687805,
+      "learning_rate": 2.255820098109873e-09,
+      "loss": 0.3575,
+      "step": 3572
+    },
+    {
+      "epoch": 2.9741953385127635,
+      "grad_norm": 0.31271275877952576,
+      "learning_rate": 2.112640769488805e-09,
+      "loss": 0.424,
+      "step": 3573
+    },
+    {
+      "epoch": 2.975027746947836,
+      "grad_norm": 0.28177496790885925,
+      "learning_rate": 1.974154544607254e-09,
+      "loss": 0.3734,
+      "step": 3574
+    },
+    {
+      "epoch": 2.975860155382908,
+      "grad_norm": 0.2948905825614929,
+      "learning_rate": 1.8403615535067531e-09,
+      "loss": 0.4235,
+      "step": 3575
+    },
+    {
+      "epoch": 2.9766925638179798,
+      "grad_norm": 0.28703781962394714,
+      "learning_rate": 1.71126192182125e-09,
+      "loss": 0.3976,
+      "step": 3576
+    },
+    {
+      "epoch": 2.977524972253052,
+      "grad_norm": 0.3083464205265045,
+      "learning_rate": 1.586855770777107e-09,
+      "loss": 0.4025,
+      "step": 3577
+    },
+    {
+      "epoch": 2.9783573806881245,
+      "grad_norm": 0.30387553572654724,
+      "learning_rate": 1.4671432171947663e-09,
+      "loss": 0.4421,
+      "step": 3578
+    },
+    {
+      "epoch": 2.9791897891231964,
+      "grad_norm": 0.2780333161354065,
+      "learning_rate": 1.3521243734854195e-09,
+      "loss": 0.3695,
+      "step": 3579
+    },
+    {
+      "epoch": 2.9800221975582684,
+      "grad_norm": 0.31834548711776733,
+      "learning_rate": 1.2417993476543377e-09,
+      "loss": 0.4009,
+      "step": 3580
+    },
+    {
+      "epoch": 2.9808546059933407,
+      "grad_norm": 0.27020248770713806,
+      "learning_rate": 1.136168243298097e-09,
+      "loss": 0.3783,
+      "step": 3581
+    },
+    {
+      "epoch": 2.981687014428413,
+      "grad_norm": 0.29213130474090576,
+      "learning_rate": 1.0352311596067976e-09,
+      "loss": 0.398,
+      "step": 3582
+    },
+    {
+      "epoch": 2.982519422863485,
+      "grad_norm": 0.307743102312088,
+      "learning_rate": 9.389881913618448e-10,
+      "loss": 0.4089,
+      "step": 3583
+    },
+    {
+      "epoch": 2.983351831298557,
+      "grad_norm": 0.31198549270629883,
+      "learning_rate": 8.474394289376131e-10,
+      "loss": 0.4056,
+      "step": 3584
+    },
+    {
+      "epoch": 2.9841842397336293,
+      "grad_norm": 0.30002060532569885,
+      "learning_rate": 7.605849582986713e-10,
+      "loss": 0.3865,
+      "step": 3585
+    },
+    {
+      "epoch": 2.9850166481687017,
+      "grad_norm": 0.28173932433128357,
+      "learning_rate": 6.784248610042232e-10,
+      "loss": 0.3695,
+      "step": 3586
+    },
+    {
+      "epoch": 2.9858490566037736,
+      "grad_norm": 0.29308822751045227,
+      "learning_rate": 6.009592142036669e-10,
+      "loss": 0.3928,
+      "step": 3587
+    },
+    {
+      "epoch": 2.9866814650388456,
+      "grad_norm": 0.317353218793869,
+      "learning_rate": 5.281880906382597e-10,
+      "loss": 0.4158,
+      "step": 3588
+    },
+    {
+      "epoch": 2.987513873473918,
+      "grad_norm": 0.30219030380249023,
+      "learning_rate": 4.6011155864111865e-10,
+      "loss": 0.3933,
+      "step": 3589
+    },
+    {
+      "epoch": 2.98834628190899,
+      "grad_norm": 0.28345680236816406,
+      "learning_rate": 3.967296821383304e-10,
+      "loss": 0.3719,
+      "step": 3590
+    },
+    {
+      "epoch": 2.9891786903440623,
+      "grad_norm": 0.2939116358757019,
+      "learning_rate": 3.380425206461757e-10,
+      "loss": 0.3956,
+      "step": 3591
+    },
+    {
+      "epoch": 2.990011098779134,
+      "grad_norm": 0.30421409010887146,
+      "learning_rate": 2.8405012927223975e-10,
+      "loss": 0.384,
+      "step": 3592
+    },
+    {
+      "epoch": 2.9908435072142066,
+      "grad_norm": 0.2770386338233948,
+      "learning_rate": 2.3475255871707737e-10,
+      "loss": 0.3904,
+      "step": 3593
+    },
+    {
+      "epoch": 2.9916759156492785,
+      "grad_norm": 0.28260740637779236,
+      "learning_rate": 1.901498552714376e-10,
+      "loss": 0.3735,
+      "step": 3594
+    },
+    {
+      "epoch": 2.992508324084351,
+      "grad_norm": 0.29308584332466125,
+      "learning_rate": 1.5024206081848401e-10,
+      "loss": 0.4103,
+      "step": 3595
+    },
+    {
+      "epoch": 2.993340732519423,
+      "grad_norm": 0.3127232491970062,
+      "learning_rate": 1.1502921283212953e-10,
+      "loss": 0.442,
+      "step": 3596
+    },
+    {
+      "epoch": 2.994173140954495,
+      "grad_norm": 0.3048544228076935,
+      "learning_rate": 8.451134437814646e-11,
+      "loss": 0.3834,
+      "step": 3597
+    },
+    {
+      "epoch": 2.995005549389567,
+      "grad_norm": 0.2796577513217926,
+      "learning_rate": 5.86884841130564e-11,
+      "loss": 0.3338,
+      "step": 3598
+    },
+    {
+      "epoch": 2.995837957824639,
+      "grad_norm": 0.29299935698509216,
+      "learning_rate": 3.7560656284685305e-11,
+      "loss": 0.418,
+      "step": 3599
+    },
+    {
+      "epoch": 2.9966703662597114,
+      "grad_norm": 0.28302428126335144,
+      "learning_rate": 2.1127880733273764e-11,
+      "loss": 0.3776,
+      "step": 3600
+    },
+    {
+      "epoch": 2.9975027746947838,
+      "grad_norm": 0.2750588357448578,
+      "learning_rate": 9.390172888701366e-12,
+      "loss": 0.3898,
+      "step": 3601
+    },
+    {
+      "epoch": 2.9983351831298557,
+      "grad_norm": 0.30292966961860657,
+      "learning_rate": 2.347543773262295e-12,
+      "loss": 0.4144,
+      "step": 3602
+    },
+    {
+      "epoch": 2.9991675915649276,
+      "grad_norm": 0.29896464943885803,
+      "learning_rate": 0.0,
+      "loss": 0.3686,
+      "step": 3603
+    },
+    {
+      "epoch": 2.9991675915649276,
+      "step": 3603,
+      "total_flos": 4646503280312320.0,
+      "train_loss": 0.438159415813211,
+      "train_runtime": 72439.8933,
+      "train_samples_per_second": 4.776,
+      "train_steps_per_second": 0.05
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3603,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4646503280312320.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}